The World According To Basketry An Ethno

Ail
Willeke
Wendrich
T h e W o r l d
A c c o r d i n g t o
B a s k e t r y
A n E t h n o - a r c h a e o l o g i c a
Interpretation of
Basketry P r o d u c t i o n
in E g y p t
T h e W o r l d A c c o r d i n g to B a s k e t r y
T h e W o r l d According to Basketry
An Ethno-archaeological Interpretation of Basketry Production

in Egypt
PROEFSCHRIFT
ter verkrijging van

de graad van Doctor aan de Universiteit Leiden,
op gezag van de Rector Magnificus Dr. W . A . Wagenaar,
hoogleraar in de faculteit der Sociale Wetenschappen,
volgens besluit van het College voor Promoties
te verdedigen op donderdag 25 november 1999
te klokke 14.15 urn-
door
Willemina Zwanida Wendrich
geboren te Haarlem
in 1961
Promotor: Prof.Dr. J.F. Borghouts
Co-promotor: Dr. P. van dcVclde
Referent: Dr. L. Bender J0rgensen
Vitenskapsmuscct, Trondhcim, Noorwegcn
Ovcrige leden: Prof.Dr. C.C. Bakcls
Prof.Dr. S. Bollcma
Dr. D. Mcijcr
Prof.Dr. H.D. Schneider
Dr. R. van Walscm

"If you are careful," Garp wrote, "if you use
good ingredients, and you don't take any
shortcuts, then you can usually cook something
very good. (...) With writing, I find, you can
have all the right ingredients, give plenty of
time and care, and still get nothing. Also true
of love."
John Irving, 1978 p. 176

Promolor: Prof.Dr. J.F. Borghouts
Co-promotor: Dr. P. van deVelde
Referent: Dr. L. Bender J0rgensen
Vitenskapsmusect, Trondhcim, Noorwcgcn
Overigc leden: Prof.Dr. C.C. Bakcls
Prof.Dr. S. Botlema
Dr. D. Mcijcr
Prof.Dr. H.D. Schneider
Dr. R. van Walscm

"If you are careful," Garp wrote, "if you use
good ingredients, and you don't take any
shortcuts, then you can usually cook something
very good. (...) With writing, I find, you can
have all the right ingredients, give plenty of
time and care, and still get nothing. Also true
of love."
John Irving, 1978 p. 176

CONTENTS
Preface and Acknowledgements
Part O n e
Methods and Materials
Chapter 1 Introduction
Chapter 2 The fabric of archaeology
2.1 Introduction 5
2.2 Pigeons and Pots 10
2.2.7 Classification 11
2.2.2 Terminology 15
2.3 Comparing Apples and Oranges?
2.3.1 Analogy 16
2.3.2 Ethno-archaeology 18
2.3.3 Meaning and Metaphor 20
2.5 The Fabric 21
Chapter 3 O n the classification of basketry

3.1 Introduction 24
3.2 Purpose of classification 28
3.3 Defining the boundaries
3.3.1 Survey of boundaries 31
3.3.2 Summary 34
3.3.3 Discussion 35
3.4 Research Questions and Criteria 37
3.4.1 Lehmann 38
3.4.2 Vogt 39
3.4.3 Mason, Adovasio and McGregor 40
3.4.4 Lewi Gourhan and Balfet 44
3.4.5 Seiler-Baldinger and Keppel 46
3.4.6 Emery, Larsen and North 48
3.4.7 Crowfoot, Lucas and Forbes 52
3.4.8 Gourlay 52
3.5 Conclusions and Consequences 55
The World According to Basketry
Chapter 4 Field Work Methods

4.1 Introduction 57
4.2 Recording Basketry Objects
4.2.1 Technology
4.2.1.1 The Basic Structure 58
4.2.1.2 Beginning and End of the Basic Structure 59
4.2.1.3 Incorporation of R a w Material 60
4.2.1.4 Construction of the Shape 62
4.2.1.5 Decoration 63
4.2.1.6 Additional Features 65
4.2.2 Size and Shape 66
4.2.3 Identifying the R a w Material 66
4.2.4 Wear, D a m a g e and Repair 67
4.2.5 Context 68
4.3 Recording Basketry Production 68
4.3.1 Production phases 69
4.3.2 Timing 70
4.3.3 Context 70
4.4 Audiovisual means
4.4.1 Drawings 71
4.4.2 Photographs 72
4.4.3 Video 73
4.5 Problems and Pitfalls 74
Chapter 5 Basketry Analysis

5.1 Analysing the Basic Structure 81
5.1.1 N u m b e r of Systems 81
5.1.2 Orientation of Systems 88
5.1.3 Activity of the Systems 90
5.1.4 The Composition of a System 91
5.1.4.1 R a w material 91
5.1.4.2 Element size, shape and flexibility 91
5.1.4.3 Preparation of elements 81
5.1.4.4 O n e element / Set of elements 92
5.1.4.5 Element members (number/orientation) 93
5.1.4.6 Element mobility 95
5.1.5 System Interaction 97
5.2 Analy sing the Basket as a Structure 99
Chapter 6 And Action

6.1 Video and Archaeology 101
6.2 Process Analysis 103
6.3 Action Analysis 104
6.4 Highliights in the Video 106
Contents 1X
Chapter 7 Sources along the River Nile

7.1 Selection of Sources 109
7.2 Amarna
7.2.1 Historical Context 112
7.2.2 Excavation History of the Workmen's Village 116
7.2.3 Description of the Workmen's Village 117
7.2.4 Population of the Workmen's Village 120
7.3 Middle Egypt 122
7.4 Qasr Ibrim 124
7.4.1 Historical Context 125
7.4.2 Excavation History of Qasr Ibrim 127
7.4.3 A Meroitic and Ballana Periods House 131
7.4.4 Cultural Background of the Population 134
7.5 New Nubia 137
Part T w o
Static Aspects of Basketry
Chapter 8 Leaves and Branches

8.1 Palm Species (Palmae) 145
8.2 Grass Species (Graminae) 147
8.3 Sedges (Cyperaceae) 150
8.4 Rushes (Juncaceae) 151
8.5 Plant Fibres 152
Chapter 9 Basketry Techniques, a Classification

9.1 Classification Criteria 153
9.1.1 One System Techniques 155
9.1.2 Two Systems Techniques 157
9.1.3 Three Systems Techniques 166
9.2 Occurrence in Egypt 168
9.3 Empty Classes 170
Chapter 10 Basketry from Amarna and Middle Egypt 175

10.1 One-system techniques 176
(knotless netting, grommet, knotted netting)
10.2 Two-systems techniques
10.2.1 Passive and active system, one direction 179
(sewn plaits, coiling, wrapping, knotting)
x The World According to Basketry
10.2.2 Passive and one active system, two directions 190

(weaving with one strand, weaving with
two strands, twining, pierced)
10.2.3 T w o active systems, one direction 203
(rope making)
10.2.4 T w o active systems, two directions 203
(plaited strips, knotting)
10.3 Three-systems techniques 206
(binding)
Chapter 11 Basketry from Qasr Ibrim and New Nubia

11.1 One-system techniques 207
(grommet, knotted netting)
11.2 Two-systems techniques
11.2.1 Passive and active system, one direction 208
(sewn plaits, coiling, wrapping)
11.2.2 Passive and active system, two directions 221
(weaving with one strand, weaving with
two strands, twining, waling).
11.2.3 T w o active systems, one direction 236
(rope making)
11.2.4 T w o active systems, rwo directions 237
(continuous plaiting, plaited strips, knotting)
11.3 Three-systems techniques
11.3.1 T w o passive and one active systems, 1 direction 244
(coiled inlay)
11.3.2 T w o passive and one active systems, 2 directions 246
(binding)
11.3.3 O n e passive and two active systems, 2 directions 247
(plaiting with a core)
Chapter 12 The Status of the Static

12.1 Static Statistics 249
12.2 The Mat, the Bag and the Basket
72.2.7 Sleeping and Sitting 251
12.2.2 Storing and Covering 254
12.2.3 Carrying and Transporting 258
72.2.4 Sleeping and Sitting 2 261
12.3 Decoration 263
12.4 Regionality, Tradition and Change over Time 264
Contents
Part Three
Dynamic Aspects of Basketry
Chapter 13 The Introduction of Time 269
Chapter 14 Climbing the Tree

14.1 Obtaining and Preparing R a w Materials 273
14.2 Palm Species (Palmae) 274
14.3 Grass Species (Graminae) 282
14.4 Sedges (Cyperaceae) 284
14.5 Rushes (Juncaceae) 286
14.6 Plant Fibres 287
Chapter 15 Making Mats, Bags and Baskets

15.1 knotless netting 289
15.2 grommets 291
15.3 knotted netting 292
15.4 sewn plaits, plaiting and string 295
15.5 coiling 303
15.6 wrapping 309
15.7 knotting 311
15.8 weaving (1 strand) 312
15.9 weaving (2 strands) 317
15.10 twining 318
15.11 waling 323
15.12 pierced basketry 324
15.13 rope making 325
15.14 continuous plaiting 326
15.15 plaited strips 326
15.16 knotting 327
15.17 coiled inlay 328
15.18 looping 328
15.19 binding 330
15.20 plaiting around a core 330
Chapter 16 The Choreography of Work

16.1 TheMaqtaf
76.7.7 Working Order 331
16.1.2 Production Time 333
16.1.3 Working Rhythm 334
16.1.4 The Body as Instrument 341
16.1.5 Workspace 249
xii The World According to Basketry
16.2 The Sabat

76.2. J Working Rhythm 351
16.3 The Shauwer
16.4 The Masballa
16.5 The Serirbil Habl
Chapter 17 The Dance of the Dynamics

17.1 Production Process 389
17.2 Skill 390
17.3 Tradition, Learning and Innovations 391
17.4 Feedback to the Archaeology 392
Part F o u r
Basket Makers
Chapter 18 Producers at Present

15.1 Mohammed and Nabawiyya, sewn plaits basket makers 397
15.2 Rauwhayya and Khadidja, coiled basket makers 400
15.3 Amrit, twined basket maker 404
15.4 Shahed and Karima, mat weavers 404
15.5 Professionality 405
15.6 Gender 406
Contents
Chapter 19 Producers in the Past

19.1 Amarna Basket Makers 409
19.2 Qasr Ibrim Basket Makers 410
19.3 Some New Kingdom Basketry Terms 411
Chapter 20 Summary and Conclusion

20.1 The Textual, the Audio-visual and the Tactile 419
20.2 Classification and Terminology 420
20.3 Systems, Activity and Actions 420
20.4 Context and Meaning 422
20.5 Continuity and Change 425
Appendices
Appendix A Fibre Identification Criteria 429

by O. Brinkkemper with contributions by E. van der Heijden.
Appendix B Questionnaire ethno-archaeological survey 441
Appendix C Video Concordance and Index 449
Appendix D Glossary 461
Bibliography 473
Index 483
Nederlandse Samenvatting
Curriculum Vitae
Stellingen
PREFACE A N D A C K N O W L E D G E M E N T S
S o m e critical readers might think the title "the world according to basketry" a
delusion of grandeur of a modest kind of artefact and a sign of megalomania, on
the part of the author, w h o claims to pronounce upon the world by studying only
basketry, while even interdisciplinary research projects do not succeed in giving a
full interpretation of Egyptian past.
W h e n I started the study of basketry I was only mildly interested in baskets
while m y main interest was Egyptian religious text and practices. Gradually the
baskets took over with surprising force and stamina. I had not realised this before,
but the world is truly riddled with baskets. They are everywhere in large numbers.
Once you have lost your innocence, you cannot stop noticing them.
The person I have to thank for m y involvement is m y dear friend Gillian
Vogelsang-Eastwood. I was interested in archaeology and fascinated by her work
as a textile specialist. She cunningly averted m y attention, because next to her
textiles in A m a r n a there was this pressing quantity of basketry. She brought m e in
contact with the A m a r n a team. T o Barry K e m p and many of the A m a r n a team
members, I o w e a great deal. The stimulating discussions and the fact that
everybody was doing a P h D quietly guided m e into the track of which I have n o w
reached the end. Discussions with Paul Nicholson, Delwen Samuel, P a m Rose and
Margaret Serpico were the start of stimulating friendships.
In m y work I have been influenced and inspired by many. The dilemma of
acknowledgements is that there are always people w h o do not get mentioned,
although at one m o m e n t or another they have been, or still are important in m y
professional or personal life. With apologies to those w h o m I have forgotten, I
would like to thank Magdi A n w a r Abdeen for all his help over the years. M a n y a
pleasant afternoon was spent with Asim Suleiman, w h o translated Magdi's Arabic
notes into English. Boyce Driskell, w h o was m y predecessor at Qasr Ibrim and
also worked extensively with Nubian basket makers, gave m e all his notes. I
cannot say h o w much I value this grand gesture. I wish the academic world would
count more like him among its ranks.
For three months I was given the opportunity to learn fibre identification,
thanks to Ferry B o u m a n of the laboratory of plant taxonomy of the University of
Amsterdam, followed by a practice on Scanning Electron Microscopy taught by
Bill Cooke at the University of Manchester Institute for Science and Technology.
Finally, due to a lack of time, I did only the fibre identification in the field, while
most of the identificatory work was done by Edwin van de Heijden and Otto
Brinkkemper. They have written the first appendix.
xvi The W o r l d According to Basketry
Other colleagues that have contributed to the work at hand, mostly without
even realizing it, are John Peter Wild and Freek Colombijn, the first a textile
specialist, the second an anthropologist. Through m y appointment at the Centre of
Non-Western Studies (as it was then still called), I have met m a n y students of
different disciplines, which has strongly influenced m y work. Through the C N W S
I also got into contact with the department of ethno-cinematography. Metje Postma
and especially Dirk Nijland have taught m e all I know about video, but more
importantly, about watching. Carinda Strangio and Ciska van Beek have had an
important contribution in the editing of the video.
I would like to thank Ben Haring for giving m e his notes on ancient Egyptian
basketry terms, Jacques van der Vliet for sending literature on Coptic monastic
basket makers and Klaas W o r p for discussing Greek basketry terminology.
T o understand basketry, meant that I had to take lessons of basket makers,
both in the Netherlands and in Egypt. I want to thank m y teachers, Tonnie,
M o h a m m e d , Rawhayya, U m m Ali, Khadidja and Sophie. Inspirational was also
m y contact with the board members of the foundation Wilg & M a n d , Tutein
Nolthenius, Arie 't Hoog and especially Mia Pot-Van Regteren Altena, in whose
beautiful garden w e have had the most delightful and exciting conversations.
In Egypt I have had the opportunity to discuss m y work with Fayza Haikal
and Nessim Henein. I would also like to thank the Supreme Council of Antiquities
and their succesive directors, professor A b d el-Halim Nur el-Din and professor
Gaballa Ali Gaballa for their ongoing support of m y work in Egypt. H a m a m ' s
good care of papers and cars allowed m y mind to wander towards less practical
matters.
Kate Trott has checked the English in an early stage, and later patiently acted
as m y personal grammar book via e-mail. M y parents never said: "why don't you
do something useful", and never complained that for four years I spent the only
month of summer leave from Cairo glued to their computer.
Finally, I want to thank Hans Barnard, m y partner for more than half of m y
life, for drawing and planning and cooking and teasing and so many other things I
cannot even think where to begin.
PART ONE
Methods and Materials

Figure 1-1 Egypt and Nubia
CHAPTER ONE
INTRODUCTION
The world according to basketry is an introduction to basketry analysis, focusing

on ancient Egyptian basketry production and its parallels in the present, as well as
ancient and m o d e m basket makers.1 Even though it is impossible to interpret
ancient society from the study of just one artefact group, the title expresses exactly
what this book is about: concentrating on one aspect of the material culture in
order to draw conclusions on a m u c h wider scope.
T h e reason for using basketry analysis to pronounce upon society is partly a
prosaic one: it's there. At a considerable number of excavations in Egypt basketry
is found. Recording material without any reason other than that it is present is,
however, of little use and tends to result in archaeological bureaucracy.
The potential of basketry analysis has long been underestimated. Basketry
w a s an important artefact in the daily life of ordinary people. During excavations
at Qasr Ibrim (Egyptian Nubia), where the preservational circumstances are
exceptionally good, the quantity of basketry artefacts came second, right behind
pottery. Assuming that a large amount of basketry never was deposited, because
discarded baskets were probably used as fuel, w e could even maintain that the
number of basketry containers will have approximated the number of pots.
T h e large quantity of basketry artefacts at Qasr Ibrim shed light on the
information that is missed at m a n y other sites. Since the preservation of organic
material is dependent on either anaerobic circumstances (be it extremely dry,
extremely cold, or waterlogged), only specific sites yield basketry. Although it is
quite obvious that baskets and mats were used in every household in Egypt, on
m a n y sites this is not reflected in the archaeological record.
But even w h e n basketry was found in excavations, the attention was often
focused on other artefacts, as Trigger remarks while mentioning criticism of the
cultural historical approach: "Pottery and lithic material were studied m u c h more
carefully than was surviving evidence of basketry" (Trigger 1989, 276).
T h e baskets which form the basis of this book, were found during
excavations in the workmen's village of Tell e l - A m a m a , in Middle Egypt, and at
Qasr Ibrim, in Egyptian Nubia. Both excavations were funded by the Egypt
Exploration Society. The A m a r n a project was directed by Barry K e m p , and the
'Basketry is defined as a class of artefacts made out of vegetable fibres of limited

length or with a shape which is specific to the raw material. Basketry thus defined
comprises various objects such as: baskets and bags; food covers, mats, awnings and wind
shields; brushes and brooms; sandals, hats and belts (cf. Wendrich 1991c, 4).
2 The W o r l d According to Basketry
excavations at Qasr Ibrim by Mark Horton. The basketry from Amarna and Qasr
Ibrim was recorded on site by the author between 1987 and 1992. In some
instances reference is made to basketry found in other excavations.2 The
publication of the catalogue and interpretation of the A m a r n a and Qasr Ibrim
basketry will appear elsewhere as part of a study on function, meaning, use, re-use
and discard. The present work concentrates on basketry production. A large
section is dedicated to the method of analysing basketry technology and the use of
ethno-archaeology in an attempt to identify the ancient basketry producers.
The one hour video tape which is part of this book includes sequences of the
production of five basketry types. The first sequence, which is the longest, shows
the production of a sewn plaits basket, m a d e in Middle Egypt. It follows the
production from beginning to end. T w o short sequences, one filmed in Middle
Egypt, the other in Egyptian Nubia, show the start of making coiled basketry. The
last two sequences show the finishing off of two types of matting (sack matting
m a d e in Middle Egypt and furniture matting made in Egyptian Nubia).3
The video can be watched separately from the book, merely as a record of
the production process. The voice over is designed to elucidate the questions that
arise while watching.4 The power of using video, however, lies in its relation to
the text, as illustration, addition and replacement with its o w n narrative. Watching
the actions of basket makers is preferable over reading lengthy and perhaps boring
descriptions. O n the other hand, the sequences which have been fully written out in
Chapter 17, draw attention to aspects in the video, which are easily missed even
after repeated viewings. In order to link the text and the video, a time code is
included in the right bottom corner of the image, which functions as a reference to
remarks m a d e in the text. Furthermore, at intervals a code is displayed in the left
top corner which indicates the aspects highlighted at that particular moment. The
main value, however, is the analysis of the filmed material, an introduction to
which is included in Chapter 6.
The video displays the world according to basketry: the baskets are the
protagonists, the basket makers are the supporting actors (cf. D e France 1989, 41).
Nevertheless, the book focuses not only on the baskets, but also on the production
process and the basket makers. The use and function of basketry are considered
briefly, because many of the basket makers produce for their o w n use only. S o m e
2These are the excavations at Abu Sha'ar, directed from 1987 to 1993 by Steven
Sidebotham (University of Delaware) and the ongoing excavations at Berenike, directed by
Steven Sidebotham and the author (University of Delaware, Leiden University).
3In Appendix C section 1 these sequences are listed together with the timer codes
where they can be found in the tape.
4The edited version of the video has been shown individually with only original sound
to 30 persons. Their questions were noted down and used to design the text of the voice
over.
1 Introduction 3
remarks are made in the conclusion in respect to the meaning of basketry in

Egyptian society.
It has been maintained that there is a great continuity in the Egyptian basket
making tradition (Blackman 1927, 304 ). In the course of this study it appeared
that, indeed, there is a clear continuity in basket making traditions (plural), but that
the opposite is true also: there have been m a n y changes over time, for instance in
the technology and the use of raw materials. Still, it is possible to draw some clear
lines from 1350 B C to the present in Middle Egypt and from the third century A D
to the present in Egyptian Nubia. There also appears to be a clear difference
between the (Middle) Egyptian and the Nubian basketry traditions, to the extent
that the main conclusion of this book is that the regional traditions over time are
stronger than similarities between regions at a given time. In other words: present
day basketry from Middle Egypt has a closer resemblance to basketry from the
A m a r n a region during the N e w Kingdom than to basketry from present day N e w
Nubia. In the following chapters these similarities will be defined using formal
criteria.
Studying archaeological basketry in its context gave information on the
technology. This is a static kind of knowledge, represented by the finished product.
S o m e installations, tools, raw materials or half products were found, which gave
an indication of the earlier stages of the production process. Mostly, however, the
archaeological context reflected a late stage in the life of a basket: discarded on the
refuse d u m p , or re-used on roofs or in pits.
In order to understand the production as a process, considerable time was
spent with present day basket makers. In the outset, the basketry of A m a r n a was
compared with basket making in the villages in the immediate vicinity of the
ancient site, namely the villages of el-Till, el-Hagg Kandil and el-Amariyya. For a
parallel with the Qasr Ibrim basketry a study was made of Nubian basketry in the
area where the population was resettled after the A s w a n High D a m was built
(around K o m O m b o , north of Aswan). The implicit supposition that there was a
continuity and that using ethno-archaeology was valid as a direct historical
approach, was abandoned. Instead, the comparison between then and n o w was
used in a more flexible way, which allowed for conclusions on a larger scope.
This resulted, a m o n g other things, in the urge to record the dynamics of
basketry production by using moving images. With the help of video the aspects of
time and movement could be introduced into the analysis of basketry production.
This leads to an understanding of the physical demands of the process and the time
involved. Perhaps the most important aspect of the ethno-archaeological work,
however, was on a m u c h more basic level: archaeological basketry, stored in a
computer file, suddenly came alive w h e n I realised they once had owners and
creators. This is not an impressionist addition to the archaeological material, but
the creation of n e w and different questions, with which higher demands are put on
the study of the archaeological material.
The book is divided in four parts, of which the first five chapters are best
read as an introduction to the method used in the rest of the book. These
introductory chapters can be considered answers to concise questions:
W h y is it important to study basketry? (Chapter 1, introduction)
H o w have archaeology and ethno-archaeology been applied? (Chapter 2, method)
W h o has been writing about basketry? (Chapter 3, classification)
What method was used in collecting the information? (Chapter 4, field work)
H o w is basketry systemized? (Chapter 5, basketry analysis)
W h e n do time and movement get involved? (Chapter 6, video analysis)
Where was field work done? (Chapter 7, introduction of sources).
The second part of the book (Chapters 8 to 12) presents a model of the
basketry in Egypt and Nubia, both in the past and at present. This part generalizes
the basketry techniques, by studying ancient and modern baskets according to a
fixed body of criteria. In Chapter 9 twenty different Egyptian basketry techniques
are defined on the basis of a basketry classification, which forms the basis of a
presentation of the basketry techniques occurring in Amarna, Qasr Ibrim, present
day Middle Egypt and Egyptian Nubia.
The third part of the book (Chapters 13 to 18) concentrates on the production
process. In this section the twenty basketry techniques which were defined in
Chapter 9 are brought to life, by studying present day basket makers and referring
this information back to the archaeological material. Chapter 17 focuses on the
analysis of the video images in order to determine the actions and movements of
basket makers. In this part the emphasis shifts from the general (classification of
basketry techniques) to the particular (the movements of an individual basket
maker).
In part four (Chapters 19 to 20) the individual approach of Chapter 17, which
concentrated on a particular (modern) basket maker, is reflected onto the ancient
basket makers. From the particular a new generalization is made, trying to
visualize the ancient basket maker.
Specialists' knowledge is a prerequisite to analyse artefacts, because
sweeping statements about something that has not been studied in detail, are of
little value beyond the obvious. Specialists run the risk of getting completely
submerged in their specialism. In order to prevent this, and to succeed in finding
the wider scope of the world according to basketry, communication with
colleagues is of vital importance. Therefore, the presentation has to be clear and
open to criticism. I have attempted to keep these standards throughout the book, by
making m y presuppositions and assumptions explicit. I also tried to use a m i n i m u m
of specialist jargon, although this could not always be avoided. The terminology
has been explained in the text and is also listed in the glossary (Appendix D ) and
index.
CHAPTER T W O
THE FABRIC OF ARCHAEOLOGY
2.1 Introduction
Basketry is m a d e of interweaving strands that form a fabric. O n e strand keeps the
next in place and creates a system of interconnections, which in a larger frame also
encompasses the life of the basket makers and the society in which they live.
Likewise, archaeology is built of different interlacing strands. Fieldwork,
description and analysis are part of one fabric, as well as colleagues, students and
friends. T h e fabric of archaeology is complicated and rests on a long tradition of
methodological discussions. S o m e strands are deemed important to account for
what has been done in the pages that follow.
Western scientific research has been concerned with absolute values in
finding the truth, and reflecting reality. The eerie phenomena in the world had to
be explained: falling objects, thunder and lightning, the change of the seasons. T h e
concept of truth is deeply rooted in our society. In our legal system witnesses are
asked to v o w that they will tell the truth. Jewish-Christian-Islamic traditions, the
religions of the book, are concerned with absolute values: not only the truth in
religion, but also the notion that there is one true religion. Likewise, it is a western
idiosyncrasy to try to define one true scientific method, reflecting one true reality.
In ancient Egypt the existence of several parallel truths was apparently acceptable.
The god Horus was a powerful sky god and a weak child. Images of the afterworld
differ widely, without any indication that they were adhered to by groups, which
considered other explanations and images as heresy or superstition.
In the era of the Enlightenment scientific research w a s contrasted with the
superstitious explanations of previous periods. The Western scientists have been
concerned with constructing theories to pronounce upon reality in physics, in
mathematics, but also in history, anthropology and archaeology. Science was not
only thought to reflect reality, but through understanding also geared to m a k e the
world manageable, to provide control. Reality w a s thought to be absolute and
existing outside and independent of those w h o study and interpret it. B y running
tests, comparing the consequences of a theory with empirical data, it could be
determined to what level theories were reflecting the truth. This method became
well established in science, to the extent that it w a s claimed to be true science.
Truth, however, is a complicated concept and implicitly w e accept this: though a
witness is compelled to tell the truth, still judge or jury are needed to decide.
Perception is highly personal: ask two people to describe a situation to the
best of their knowledge and two different accounts will be given. T w o realities
have c o m e into being, which do not exist in duplicate worlds, but in this one world
and we have to find a way to live with them. Perception is linked to cultural
background, past experiences, and present interests, to n a m e a few aspects.
Gradually it is accepted that the perception of reality is subjective and that it is
ruled by preconceptions such as the belief in truth, in progress, in creation or in
reason. These preconceptions are not fixed, but they fluctuate with time and (sub)
culture. Studying the history of science brings to light which preconceptions rule a
certain era (cf. Trigger 1989), but determining them in our o w n time and
surroundings is difficult. Therefore, it is extremely important to recognize that w e
are working under certain preconceptions and attempt to explicate these, while
realising that w e can easily fail in this endeavour.
The realization that research is not the objective action which positivist
science thought it to be has resulted in a multitude of n e w theories, methods and
research agendas.1 At present, archaeologists work with methods ranging from a
positivist to a relativist approach. The basic division of these approaches is the w a y
the relation between data and theory is perceived. In the positivist view empirical
data and theory are opposites, while in the relativist approach they are part of a
continuum. In both ends of the present research scale the relation between data and
theory is problematic.
In the positivist approach the two opposites, empirical data and theory, are
kept apart to ascertain objectivity. At some point, however, theory and data have
to be confronted in order to m a k e sense. If research is to tell us more than the
obvious, there is an immediate gap between theory and empirically retrievable
data.
The solutions to close the gap and still maintain 'objective science', range
from insisting on building a pyramid of generalizations, which in the end forms a
high level theory, to deducing middle-level theories from high level theories and
testing them against the low-level theories induced from the empirical data. Thus,
the division of theories in high-, middle-, and low-level is used to indicate levels of
inference, or empirical content, low-level theories having a more direct link with
the database than high-level theories.
Trigger uses an example with which the difference of high and low level
theories can be clarified (Trigger 1989, 23). The geologist Buckland (1784-1856)
found strata with gradually more developed fossil remains. The implication was
(and this can be considered a low-level theory) that geologically stratified finds
represent successive temporal phases. The high level theory, Bucklands'
explanation of the phenomenon, was that a series of disasters had ended the
existence of the species found in the strata. Each disaster was followed by the
creation of a slightly more developed species. The Darwinists, although working
with the same low-level theory (strata represent temporal phases) gave a
completely different explanation. The two explanations were due to different high-
'Shanks and Tilley 1987 and 1992; Tilley 1998; Hodder 1999; also Whitley 1998, 2).
2 The Fabric of Archaeology 7
level theories: one based on the biblical notions of creation and deluge, the other
on adaption and evolution.
Dunnell defines theory as "the principles by which explanation is achieved"
(1971, p. 32). The structure of archaeological science is thus as follows: "one
starts with a set of explicitly defined notions (theory) which are capable of being
organized according to some of the defined relations in a model for the solution of
a particular class of problems (method), which in turn is capable of being matched
with phenomena (technique) in order to produce a testable hypothesis capable of
being used as an explanation (prediction/control)." (Dunnell 1971, p. 41).
Dunnell uses the term model, in the sense of model of explanation, free from
empirical content. A hypothesis is "a proposed explanation for a specific set of
things or events" (Dunnell 1971, p. 37). Thus, he makes a strict division between
theory and data. Ideally the theory is formulated first and the testing of the theory
is performed by dressing the theoretical skeleton with empirical flesh.
Although Salmon does not agree with the order in which Dunnell sets up
scientific work, first formulating theory, then filling the theory with empirical
data, she equally separates theory and data. In her definition, theory is "a set of
interrelated, rather high-level principles or laws that can provide an explanatory
framework to accommodate a broad range of phenomena" (Salmon 1982, p. 140).
For archaeology these phenomena are patterns of connection between h u m a n
behaviour, material remains and the physical settings in which they are found.
Social-political and economic organization are aspects that are to be explained by
theory as well. The laws of archaeology which form the explanatory framework
that constitutes a theory are defined as "the regularities or empirical
generalizations that relate various aspects of material culture to one another, or
connect aspects of material culture with patterns of h u m a n behavior" (Salmon
1982, p. 20).
In archaeology the positivist approach of the 'New Archaeology' and
subsequent processual archaeology consider archaeology as a science and a means
to gain knowledge about past reality. The past can be reconstructed provided that
sufficient information can be retrieved. The central concept of processual
archaeology is that theories, for instance about the relation between material
culture and societies, have to be tested. In order to enable testing the complex
whole is divided into manageable portions. Middle Range Theories have been
developed to close the gap between theory and data (Binford 1977).
A similar attempt to c o m e to grips with the cluttered, complex whole that
makes up h u m a n society (past or present) is a deliberate simplification by defining
spheres of h u m a n activities as different systems that interact. Thus economy,
religion, subsistence and social interaction can be considered as subsystem of
society, each answering to different laws. They can be analysed and brought
together in the system that forms the whole (Clarke 19782).
The research questions of processual archaeology concentrate on long term
developments, functional and behavioural patterns, although in recent years on the
8 T h e W o r l d According to Basketry
basis of a processual approach social and cognitive theories have been developed
(Renfrew 1973, 1982).
Increasingly, however, the idea has eroded that data can be collected without
the influence of either subconscious preconceptions or conscious, well pronounced
theoretical structures. Observing is interpreting. Accepting that making a clear
division between theory and data is not possible, does imply that w e can only dis-
tinguish between results that are more or less influenced by our assumptions.
Probably w e are not even able to observe if w e do not start from certain
preconceptions (cf. Hodder 1999, 53, 92). In post-processual publications the term
interpretation indicates that our ideas about the past are firmly rooted in the
presence. Interpretation is specifically used to describe those phases of the research
process in which uncertainty exists (Tilley 1993). Since in the post-processual
approaches theory and data are no longer considered an opposition, the widely
accepted method of testing theory against data, such as employed by the
hypothetico-deductive method is not possible.
Several solutions have been proposed to solve this dilemma. A mild form of
empiricism is used by those w h o test their theories against 'close-to-data' data, i.e.
by using data as if they were 'pure' data. It is a practical solution that grants that
there is a fundamental problem, but on the other hand, prefers to use the well
developed and established positivist methods.
Other solutions are to test the theory on its internal coherency, rather than its
empiricist values; to follow an ethical line of reasoning and test theories by holding
the results against (ethical) values or social consequences. The most rigorous
consequence is that theories cannot be tested and that the best scholars can do is to
produce a construction, a narrative that is deeply rooted in our o w n society.
In the terminology used by archaeologists this change in approach is
expressed most vividly in the diminishing use of the word explanation and the
increasing importance of the term interpretation (cf. the contributions in Hodder
e.a. 1995). With the realization that data collecting is part of interpretation, the
attention has shifted to a more complex subject matter: from a mere functional or
behavioural interpretation to an integration of the influence of culture groups and
even individuals Hodder states that in order to widen the subject in this manner, a
theory of material culture is needed (Hodder 1982, 212).
In his later work Hodder develops this line of reasoning, focusing on the
meaning of material culture. H e considers material culture as language, a narrative
which includes rhetorical devices such as irony and metaphors, which can be easily
misinterpreted (e.g. Hodder 1995, 165). The approach to understand the meanings
of texts is hermeneutics, a term originally used specifically for a discipline within
theology dealing with the explanation of biblical texts.
The analogy of (material) culture with text is one that has taken great hold in
archaeology and the social sciences (Shanks 1996, 126). Material culture is thought
to be a form of communication, a social production, m a d e up of relations such as
parallelism, opposition, linearity, equivalence and inversion. "Material culture is
polysemous, located along open systems of signified-signifiers" (Shanks and Tilley

1987, 117, cf. 103).
Hermeneutics in social theory is a dialectic process. "According to the
hermeneutic circle, the meaning of a part derives from its relationship to a whole,
while the whole is understood from the relationship with the parts." (Hodder 1999,
32-33). The terms 'part' and 'whole' can be used at m a n y different levels. The
part is the basketry and the whole is the site, or the part is the site and the whole is
a region. W h a t matters is the relation between the parts and whole. During the
hermeneutic process information is gathered from a set of prejudgements, which
enable us to see and observe, rather than just look.
F r o m scattered information a narrative is formed, which helps is finding n e w
information, by allowing changes and augmentations. The procedure is judged by
internal coherency and 'fitting' of all available information (Hodder 1999, 33).
Past reality is a present realization of the past, a fluent entity, which
constantly has to be adapted to n e w insights. The narrative is ours to create and
this leaves us to consider at least the ambiguity of texts in which w e report our
finds and interpretations, an ambiguity that has to be communicated to colleagues.
T o speak with Shanks and Hodder: "To argue a relativism which maintains
objectivity is socially constructed is to argue simply for relationality" (Shanks and
Hodder 1995, 19). Thus, interpretation is closely linked to its original meaning in
h u m a n communication: translating and relating.
The metaphor of material culture as language is in some respects an unhappy
one. N o doubt material culture carries messages and communications, but these
are non-verbal. W h a t w e , archaeologists, try to do is bring to a conscious level,
that what never was supposed to be formulated in words. Hodder calls it
"deciphering a system of signs" (Hodder 1999, 32), but a system of signs has been
designed to be deciphered, while material culture for the most part has not.
Chances are that w e read what was never written.
O u r study and experience give us a role in society to interpret for others
(colleagues, the public). W e can do this in different guises, either as the great
masters of interpretation, or as humble translators w h o indicate exactly which
terms are unknown to us and where the interpretation might have gone wrong.
Needless to say the latter approach is perhaps less exciting, but such self-
reflexivity, although vulnerable, enables discussion and cooperation.
The narrative of archaeology is the structure w e weave around our perception
of the material remains of the past. The w a y that narrative permeates the
archaeological work from excavation to publication, makes one wonder: is
narrative in fact not the post-modern form of the empiricist's high-level theory?
Both are the all-encompassing result, which forms the crown on the scholarly
research. The fundamental difference, however, is that the narrative has been built
by interweaving data and theory, rather than by confronting 'independently'
developed theory with theory-free data.
W h e r e Shanks states ominously that "Elements of the methodological
hegemony of processual archaeology have been identified as positivist; and in
philosophy and social theory a positivist is not a good thing to be" (Shanks 1992,
25), Hodder poses that "It could be argued that the different theoretical
perspectives in archaeology are not contradictory but complementary. For
example, the overall division between processual and post processual archaeologies
can be interpreted in terms of different objects of study." (Hodder 1999, 12).
While the first quotation tends towards rigidity, the second is too flexible in
smoothing out some fundamental differences.
In this light, w e should re-evaluate the careful approach of recording data, as
advocated by processual archaeology. Employing formal analysis has been given
great incentive and although the method of formal analysis has a different status
and function than in a positivist approach, its value for a post-processual approach
is perhaps even greater. W h e n w e abandon the claim that w e can m a k e an
objective representation of the past, two persons involved in archaeological
research no longer have the past as stable c o m m o n ground. The subject of their
research is related to the present, and c o m m o n interpretations of the past are based
on intersubjectivity. W h e n the communication between scholars is a central
concept in the research method, it is of utmost importance to present one's work in
a manner that is both self-reflexive and open to the criticism of others. The well-
developed recording methods and formal analysis of processual archaeology can be
adapted to serve the purpose of open communication.
In the following paragraphs I will survey several issues that are of direct
importance in respect to the analysis of archaeological basketry. Section 2.2
surveys classification and terminology, while section 2.3 concentrates on analogy
and ethno-archaeology. The end of this chapter, section 2.4, gives a summarizing
account of the method used in the second, third and fourth part of this book.
2.2 Pigeons and Pots

Travelling through Egypt, the landscape is adorned at regular intervals with
towering white shapes, which are dwarfing the date palm trees. These are pigeon
towers, built completely out of pots. The openings of the pots are facing inwards.
Through a number of openings the pigeons can enter the tower and build their nest
in one of the pots, until the day they are served on the dinner table, their little
bodies stuffed with wheat.
Each pigeon has its o w n pot, which is a good way of maintaining the pigeon
population. Pigeonholing in archaeology, however, is a method that has great
weaknesses and has been used too often. Observation presents us with a chaos of
information, which is processed in a constant action of ordering. O n e w a y of
ordering is making a classification, which tells us where things belong in relation
to each other according to specific criteria. This is something else than
pigeonholing, ordering things for the sake of ordering. In Section 2.2.1 examples
of both classification and pigeonholing are given, based on an inventory of a
hypothetical kitchen area.
Descriptions and designations are based on classification and consequently

terminology is given some consideration in this paragraph too.
2.2.1 Classification
In the second part of this book, classification is used as part of context-dependent
formal analysis of recorded data. This seems to be a contradiction in terms, since
formal analysis, as applied by positivist scientists, consists of filling in a context-
independent frame with data. By using formal analysis as a heuristic, rather than
explanatory procedure its advantages can be maintained, while its disadvantage, no
surplus value of information beyond the theory from which the analytical frame is
derived deductively, is avoided. Although formal analysis is a rigid procedure, its
application as a tool to formulate imaginative hypotheses solves this drawback. The
most important gain of formal analysis is clarity and precision, rendering the re-
sults specifically open to discussion.
The application of formal analysis as a tool gives no guarantee that the results
will make sense. Classification cannot be applied to 'let the material speak for
itself. Instead, classification is used as a means to check hypothetical answers to
explicit research questions, to generate new questions and build the narrative
simultaneously.
In this respect explicating all assumptions is important, including the
'common sense' decisions made continuously while doing scholarly work. By
always explicitly accounting for presuppositions, keeping a check on the entire
procedure is possible. Involving the many aspects, which are of influence on a
given situation, in the analysis, results in a black box, a mechanism of which the
outcome may be predictable, but the working is unclear. Classification can be used
to make a well-considered simplification, which is not a representation of
'reality', but a construction by the scholar, a point in the line of reasoning.
Although I disagree with Dunnell (1971) that first a theory has to be
formulated, before the analytical structure can be filled in with data, I find his
discussion on classification extremely useful. Dunnell distinguished grouping and
classification, in which the first is a division of an existing corpus of objects in
groups. The aspects that are important in grouping are dictated by the composition
of the group-as-it-happens-to-be. If a new object is added, it is easily possible that
it does not fit any of the existing groups and a new 'class' or 'type' has to be
created on the spot.
In contrast to grouping, classification is a division into classes according to
certain aspects judged relevant for a specific research question. It can be
maintained that there are as many classifications as there are hypothetical answers.
Dunnell discerns hierarchical and paradigmatical classifications (Dunnell
1971, 69-86). The paradigmatical classification procedure involves the selection of
relevant aspects, which form the dimensions of the classification. Every dimension
has several features, or as Dunnell calls them: attributes. Every combination of the
attributes of the different dimensions forms a class. Although the selection of
relevant dimensions and attributes is inspired by k n o w n objects, and specific

research questions, the classes are created formally. They are filled with objects
only after the formulation of the entire classification.
A s an example of a paradigmatical classification of objects, Figure 2-2 shows
a classification of shape, function, and material of objects found in a hypothetical
kitchen area. All three dimensions have most attributes. Shape is specified as
cylindrical, globular and flat. Three functions are considered relevant: things to
put other things in, on or under. There are two material attributes: vegetable and
mineral. The number of classes that have thus been created is 3 x 3 x 2 = 18.
These 18 classes are assumed to represent a relevant reflection of the research
question, which will have been something like "what is the relation between form,
function and material of objects found in the kitchen area"?
cylindrical globular flat Form
< ^ ^ ' /
j ^ r 1>^
&
tf# 1 4/_
w
/ ? /
n y u y
—
//\ V -
2 / 5 ' 8
/
12 X. — j5_ y
/ 7
3 J 6 y — 9 ; / f
under
)
C
O
+—»
o
c
3
Figure 2-1 Paradigmatical classification of objects in a hypothetical kitchen area,
classified according to three dimensions: material (mineral / vegetable), form
(cylindrical / globular / flat) and function (to put things in / on / under). The
18 classes are occupied by the following objects:
l=baskets, 2 = pot stands, 3 = covers, 6 = stoppers, 8 = mats, 9 == awnings,
10 = vases. ll=pedestals, 13 = pots, 17 = plates/coasters, 18 = lids.
The classes 4, 5, 7, 12, 14, 15 and 16 are empty.
W h e n designating the objects to the classes, it appears that some classes are
filled, others are empty. It is predictable that classes 7 and 16 are empty, because
it is not possible to put something in a flat object, be it a mat (made of a vegetable
material) or a pottery coaster (mineral). A similar explanation occurs for the
empty classes 5 and 14, because spherical baskets and pots are not suitable for
putting things on. At this stage, the classification appears futile, since everybody
knows that coasters and mats are flat and that baskets and pots are supposed to
contain things.
1. In
a Cylindrical 2. On
3. Under
4. In
I. Vegetable b Globular 5. On
6. Under
7. In
c Flat 8. On
9. Under
| Kitchen objects
10. In
d Cylindrical 11 . On
12. Under
13 . In
II. Mineral e. Globular 14 . On
15 . Under
16. In
f. Flat 17. On
18 . Under
Figure 2-2 A hierarchical classification of objects found in a hypothetical kitchen area

according to the same criteria as used in Figure 2-1.
Above, it was stated that classification also has a heuristic value. Here the
classification of the kitchen inventory reveals a surprising phenomenon: there are
no globular baskets (class 4: "globular objects to put things in, made of vegetable
materials"). The heuristic value of classification lies in the need to explain these
unexpected phenomena. The objective of using classification as an analytical tool is
not to devise a magical method, in the sense that it produces results by itself, but
to use c o m m o n sense in a systematic and explicit manner.
Classifications can also be used hierarchically by giving some criteria more
weight than other. The result is a system of classes and subclasses. Using the same
dimensions as in the example cited above, the first level is material divided in
vegetable and mineral (see Figure 2-2). O n the second level a division is m a d e into
cylindrical, globular and flat and the third level divides these subgroups into the
three functional categories in, on and under. Making a division in function is
equally possible, while creating subclasses according to shape and material.
Most published classifications, however, are not m a d e according to
dimensions, with several attributes. Figure 2-3 is an example of such a non-
dimensional hierarchical classification. Every class has different subclasses. The
coarse red pottery (B.l.b) is divided into wheel thrown and hand made. Because
all the white pottery is wheel thrown, this technical feature does not occur in the
classification of the white pottery. Because all red hand m a d e pottery (B.l.b.ii) is
globular no division of shape is incorporated. The shape is classified at two
different levels: for the vegetable objects on the second level (classes A.l and
A.2), for the white pottery objects on the third level (classes B.2.a, b and c). Flat
white plates (B.2.c) are not used to put things in, so the classification of the
function of these plates only list that they are used to put things on or under
(B.2.c.i and ii).
a. in
Cylindrical b. on
c. under
A. Vegetable
a. stitches
2. Flat b. beads
I Kitchen c. colour Ii. wedding
objects
ii. funeral
I a. fine
1. Red b. coan i. wheel
I ii. hand
B. Mineral
a. globular
2. White b. cylindrical
c. Flat Ii. O n
Un
ii. Under
Figure 2-3 A hierarchical non-dimensional method, the result of which is not a
classification (Figures 2-1 and 2-2), but grouping according to different
criteria at each level and in each 'branch'. This is merely pigeonholing.
When a white hand made pot is found, however, this classification is

outdated. A n e w class has to be added between level 2.(white pottery) and level a.
b. and c. (forms of the white wheel thrown pottery). Furthermore, there are no
empty classes. The globular baskets, which did not occur in the hypothetical
kitchen area and resulted in an empty class in Figure 2-1, do not feature in the
classification of Figure 2-3. This type of classification, which is in fact grouping,
has lost all heuristic power and is nothing more than a set of pigeon holes neatly
arranging a number of objects. It gives an average of what occurs in a specific
context. It does not enable a comparison to what occurs in a different context,
because each assemblage requires its o w n grouping. Criticism of classification (for
instance Shanks and Hodder 1995, 9), refers mainly to the shortcomings of this
method of 'classification'.
A s said before, classification is assumed to be a relevant reflection of a
research question. In a dimensional classification, be it paradigmatical or
hierarchical, the relevant aspects are used to build classes. Differences in opinion
or developments in insight on what is relevant can be incorporated and the strategy
can be adjusted anytime. Several classifications can be used alongside each other,
to highlight specific aspects such as technique, shape, form or function in more
detail.
2.2.2 Terminology
The most general term for basket in Arabic is sella, but w h e n I used this term
while making an inventory of household basketry, trying to get the families to
show m e all the baskets they used in the house, the term sella produced just one
type of basket. I had to ask for all basketry and matting types specifically. Clearly,
the terms referred to very distinct objects. O n the other hand, using the term that
indicated the material (khus, palm leaf) brought out all locally produced baskets
(cf. Section 4.5, p. 75). Studying present basket makers brings h o m e that the
terminology and the implicit classification on which it is based, differs from ours.
Working in Egyptian archaeology is studying a culture and language, which are
removed from our o w n in both time and region.
The implicit classifications underlying the terminology are hard to discover
even w h e n there are people available w h o can be asked. Apart from this, the
persons w h o m a k e or use the baskets often have a more specialised terminology
than those, w h o speak the same language, but are relative outsiders. It makes you
wonder h o w well the tomb decorators of pharaonic Egypt knew the basketry
terms, which they quote in the 'captions' of the tomb reliefs with daily life scenes.
Furthermore, everyday object names are referring to different aspects, such
as function, technique, material or origin, e.g. 'Egg-basket', 'Coiled basketry',
'wicker work' and 'Nubian basketry'. Developing a consistent terminology is often
considered a primary task for those w h o study material culture and is directly
connected to classification. The existence of several basketry terminologies is a
reflection of different classifications (see Chapter 3). It is therefore not surprising
16 The World According to Basketry
that those terminologies are usually incompatible, in spite of their internal

coherency. Furthermore, there is the horror of terminologies designed to
encompass every possible type and subtype. Without the handbook in which the
types are defined and drawn, such terminologies are completely incomprehensible.
Using classification as an analytical tool, rather than a set of pigeon holes to
organize material culture, creates a problem and a liberation at the same time. The
problem is of course that almost every question requires a separate classification,
which could result in a large number of internally consistent terminologies. The
liberation is that we can stop attempting to create an all-embracing classification
cum terminology. It is preferable to stick to a number of generally known terms
and make ad hoc specifications, definitions and drawings, whenever needed.
2.3 Comparing Apples and Oranges?

A n archaeo-botanist, who finds a small fragment of an apple core, has a reference
collection to which he can compare the unknown fragment and identify it. If he has
seen many fragments of apple cores, then he 'recognizes' the fragment, without
the need of comparison.
Without thinking, purely based on experience, a pottery specialist picks up a
sherd and puts it on the stack of spouted vessels. From known equivalents, a sherd
with a hole under the rim is described as part of a spouted vessel, although the
spout itself has disappeared. This is an implicit analogy with complete vessels
made of the same fabric, with a spout under the rim.
Analogy is a comparison of an entity, with a better-known entity to come to
an interpretation. The entities that are compared range from aspects of single
objects to complete cultures. A widespread form of analogous reasoning in
archaeology is based on a comparison of archaeological and ethnographic
information. Ethno-archaeology is a method to make this comparison, based on
research which focuses on the material expression or reflection of culture.
Metaphors form a relation between material culture and the symbolic world.
In analogous reasoning one should be critical of what is compared with what.
W h e n apples are compared with oranges, the outcome will differ from a
comparison of apples with apples. It depends on the research question if the first
or the latter is useful, but seemingly far-fetched analogies (apples and oranges) are
not necessarily useless.
2.3.1 Analogy
Using analogy for heuristic purposes is a means of generating new research
questions by comparing a known with a lesser known situation. Referring to a
present day situation as an 'inspiration' for archaeological research, is implicitly
using the heuristic powers of analogy.
Explaining archaeological phenomena is often done by using analogies. A
pattern of post holes is reconstructed as a silo, house or corral, based on examples
2 T h e Fabric of Archaeology 17
known from other sites or from the present. Traces of use are linked to human
behaviour by comparing the patterns or residues of present behaviour.
Analogy, however, can go seriously wrong for several reasons. W h e n the
aspects that are compared bear no relevance to the interpretation, analogy goes
limp. For instance, if there are red spouted pots and red bowls then the analogy of
the colour of the fabric of a single sherd does not help to distinguish the shape.
Although the use of analogy is widespread in archaeology, the validity is
subject to debate a m o n g archaeologists. Implicit use of analogy muddles the
argumentation. Naive use of analogy m a y lead to farfetched conclusions: a Fatimid
green glazed incense burner that looks conspicuously like an orange squeezer m a y
provoke the idea that orange juice was part of the diet in 10th century A D Cairo.
Hodder gives an overview of criticism of misleading and amateurish use of
analogies (Hodder 1982, 11-27).
Rather than rejecting analogous reasoning altogether, or restricting its use
drastically, a critical and explicit use makes analogy an important tool for
interpreting material culture (Wylie 1985). By discerning two kinds of analogy,
formal and relational, the conditions of using analogy become m u c h clearer.
Formal analogy compares aspects of objects or situations. Formal analogy
looks at the existence of similarities and the strength of the analogy, the number of
aspects that are similar. In case of a strong analogy, other aspects are supposed to
be similar as well
Relational analogy, according to Hodder, focuses on the natural or cultural
relations between aspects of the analogy. Relational analogy not only looks for the
existence and strength of the analogy but also for the nature and cause. The
questions that Hodder asks of an analogy thus are:
are there similarities between A and B
h o w m a n y similarities are there between A and B
what are these similarities like
what caused these similarities (Hodder 1982)
However, the opposition of formal and relational analogies is artificial at best.
Formal analogies are usually followed by explanations, which focus on causality
and in that respect c o m e close to relational analogies which focus on causal
relations.
T o m y understanding it is more important to highlight relations other than
causality, by making an inventory of similarities and dissimilarities. In such a use
of analogy the subject, object, purpose and relevance should be defined. The
subject is the archaeological situation that has to be interpreted. The object is the
analogous situation, for instance another part of the site, another excavation or a
present day culture. The task of the researcher is to define w h y the object is
thought analogous to the subject.
Classification can be used as a means to m a k e analogy more controllable. In
a dimensional classification both the similarities and dissimilarities can be m a d e
visible. In order not to use formal criteria doggedly, defining the relevance of
those criteria to the research question is important. Determining the relevance is
not an objective exercise, because it involves judgement and selection. Making

such decisions is an internal argumentation, which, w h e n m a d e explicit, can be
criticized by colleagues.
The variation in archaeological research questions is great. S o m e examples
of archaeological research using analogy are the appearance, function and meaning
of objects, the lay out of work shops, the formation of archaeological sites,
subsistence strategies or social structures and relations.
2.3.2 Ethno-archaeology
Comparing archaeological and ethnographical data is a form of analogy frequently
used by archaeologists. The discussion on analogous reasoning based on comparing
the present and the past has focused on the object (which aspects are compared and
what are the preconceptions connected to it) and the purpose (cf. Hodder 1982, 40-
46; Ochsenschlager 1998, 103). The critical discussions centre on the value and
validity of ethno-archaeology and the preconception (either implicit or explicit) that
govern the analogous thinking.
In the discussion of the object of ethno-archaeology, the reasons given of
w h y a comparison between past and present should be considered possible are:
• There is a continuity of tradition and thus it is valid to study the present, to
understand the past as long as this is done in the same area (direct historical
approach). This approach centres on a study of a region in past and present.
• People in present and past exploit similar material resources and their actions,
w a y of life and material culture are determined by those circumstances. Here, the
researcher will concentrate on the properties of the material resources and their
effect on h u m a n behaviour.
• People living in the same natural conditions are compared, because they have to
cope with the same problems. This is also a behavioural approach, concentrating
on the manner in which humans cope with natural circumstances as cold, drought,
etcetera. This approach, sees h u m a n behaviour primarily as a reaction on the
surroundings.
• There are behavioural laws that are valid for humanity as a whole (cross-cultural
laws). Thus it is useful even to study ideas, for instance, about family, society,
religion and ethics, within two or more societies.
Based on arguments such as these, processual archaeology has m a d e
extensive use of ethno-archaeology, in two main functions:first,it has a heuristic
purpose. The validity of the hypotheses suggested by ethno-archeology has to be
screened by testing the hypothesis against archaeological data (Binford 1972, 33).
Secondly, ethno-archaeological research is used for testing theories about the past,
on the basis of an analogy of the residue of specific actions, which are thought to
be the same in the past and at present (Binford 1967, 1980).
Ethno-archaeology focuses on processes instead of artefacts. It brings the

past alive and people, rather than objects, back into view. In this context the
dichotomy of static and dynamic is often introduced (Binford 1977).
The processual use of ethno-archaeology centres around behavioural and
functional interpretations. Gould, for instance, argues that ethno-archaeology is a
systematic exploration of present h u m a n behaviour, from which material
correlations are inferred, providing a basis for ideas about past h u m a n behaviour
(Gould 1978, 292). Hodder, on the other hand, stresses that arguments of
relevance should not just be functional or behavioural. By applying a relational
analogy, conclusions on a m u c h wider scope can be drawn (Hodder 1982).
There is a difference, between ethnographic analogy and ethno-archaeology.
The former uses existing descriptions made by ethnographers or anthropologists,
the latter is based on field work in present society with a specific archaeological
focus. Because interpretation is part of even the first stage of basic research, doing
ones o w n ethno-archaeological fieldwork is highly preferable to using existing
ethnographies.
The material remains of h u m a n actions are all that is left for the
archaeologists. Furthermore, these material remains are not simply the residues of
'frozen' actions, but the result of a wide variety of processes and motives. The
processes, such as production, use, re-use, deposition and post-deposition are
understood better by studying them at present. Objects get lost, are thrown away,
are taken from the world of the living in a different world by burying them with
the dead or storing them in a special place. Studies of the motives of present day
informants m a y give less tangible results, but are of importance in understanding
that things might not be what they seem.
The major problem of ethno-archaeology is to what extent the present can be
used to interpret the past. A n analogy which just puts the two side by side, without
making an attempt to interpret the one in the light of the other, is not very
powerful. O n the other hand intertwining past and present is easily done too
quickly and too simply. A n example of such an intertwining of present and past by
ethno-archaeological research, is found in the work of McGregor. At some point
the author, w h o writes about archaic life ways of the lower Pecos, Texas, is
referring to the past, in a description which is based on the present:
" W o m e n , with few exceptions, are the basketmaker (...). Most basket makers
depend upon two or three culturally preferred materials for the majority of
their work. Although the weaving of baskets is not done by specialists, there
must be superior basket makers." ( M c Gregor 1992, 10, emphasis added).
In this paragraph M c Gregor has applied analogy descriptive ("Women are the
basketmaker"; "not done by specialists" and "superior basketmaker") and
explanatory ("culturally preferred materials"). Is this a good illustration of the
wide scope of analogy as advocated by Hodder? M c Gregor's use of ethnographic
information on American Indian peoples is a direct historical approach bridging
5000 years. Without hesitation the present situation is transplanted onto the past. In
this interpretation there is no firm link to the archaeology. The result is an
impressionistic rendering of ancient basket makers which is a possible

interpretation, but by no means the only one, as Ochsenschlager has shown
(Ochsenschlager 1998, 116-118).
The present work on Egyptian basketry production started on the one hand
from the archaeological baskets and the context in which they were found, on the
other hand, the present day basket makers have been studied in order to understand
the present production process. These were not two independent research projects.
They were intertwined from the beginning, because m y interest in the modern
basket makers was born from the ancient baskets.
O n the basis of the results of archaeology and ethno-archaeology the
similarities and differences of four groups of basketry have been determined
(Chapters 10 and 11). These four groups were the ancient Middle Egyptian
basketry (Amarna), the ancient Nubian basketry (Qasr Ibrim), the present Middle
Egyptian basketry and the present Nubian basketry. The focus of the research was
to understand the ancient production processes, to identify the basket makers (i.e.
to determine their gender, economical and social position), and to find their
objective for making the baskets.
Understanding the ancient production processes involved understanding the
properties of the raw materials used and the strength and skill needed to work
these materials. For these aspects I have taken lessons from the basket makers.
Parts two and three of the book focus on static and dynamic aspects of
basketry. This arrangement is not based on Binford's division, but stems from the
involvement of time and movement through the use of video in the study and
interpretation of ancient and present basket makers.
2.3.3 Meaning and Metaphor

Traces in the earth are perhaps the result of practical subsistence strategies or m a y
have a symbolic meaning. Ethno-archaeology has been concentrating mainly on
matter, but is learning from anthropology that sometimes there is more to an
object than appears on first sight, sometimes less.
Tilley quotes some examples of the rich meaning of material culture of two
kinds of basketry. The bilum net bag from N e w Guinea is utilitarian, decorative,
but has also a symbolic and social meaning. It is connected to child birth,
protection, but also acquired a new cultural value as a neo-traditional symbol of
independent Papua N e w Guinea (Tilley 1998, 62-68). The basketry of the Yekuana
(south Venezuela) has a strong symbolic and powerful social meaning. "The
process of basket-making is one of a constant metamorphosis of reality into a
comprehensible, coherent and legible order" (Tilley 1998, 68). Textile production
also carries a rich and important symbolic meaning, which is expressed in the
weaving and decoration patterns (Geirnaert-Martin 1992). O n the other hand, the
main motive for a recent spectacular change in Iraqi pottery is an attempt to attract
customers (Ochsenschlager 1998, 116-118).
Technological processes are used metaphorically also. The creation and

production of objects are often linked to ideas of spiritual or ancestral involvement
in the production process (Trigger 1998, 57). In ancient Egypt, the act of knotting
had a strong magical side to it. Knotting stood for protection as well as hampering
(in child birth, but also for the dead, cf. Wendrich 1988). Both Hodder (1982) and
Ochsenschlager (1998) give ethno-archaeological examples that contain warnings
for oversimplification while interpreting material culture.
Hodder focuses on the symbolic meaning of material culture in an attempt to
formulate a theory of material culture (Hodder 1982). H e states that material
symbols are value laden; that symbols should be understood by reference to the
particular historical situation; that visual images are ambiguous and multifocal, in
contrast to spoken text; that meaning can be composite and can change over time;
that material culture can be used actively to have an emotive effect; that material
culture is not only functional, but often refers to an ideal, symbolic, world; that
symbolic meaning m a y be subconscious; that the c o m m o n scholarly language
might not be sufficient to express the emotional values of material culture; that
material symbols are important in ideological and social strategies. Hodder lists
these nine aspects of a theory of material culture, without proceeding actually to
formulate the relation between them and without a clear link to archaeology. In his
opinion, the key to the interpretation of material culture is the underlying
structure, but it is a key of w a x that Hodder presents, because he does not
elaborate on h o w to define this underlying structure.
Seventeen years later he has refrained from attempts to formulate such an all-
encompassing theory of material culture, but points instead at the pluralism and
multivocality of archaeology (Hodder 1999, 159). This is combined with some
very d o w n to earth examples of h o w the major post-processual themes work out in
the field.
2.6 T h e Fabric
Perhaps the positivist approach can be compared with the fabric of a textile: the
warp is formed by empirical observations, which are held together by the weft of
theory. W a r p and weft are two separate systems which interact systematically and
are beaten to a tight fabric with the beating rod of method. Or, as Dunnell
describes the procedure: "one starts with a set of explicitly defined notions (theory)
which are capable of being organized according to some of the defined relations in
a model for the solution of a particular class of problems (method), which in turn
is capable of being matched with phenomena (technique) in order to produce a
testable hypothesis capable of being used as an explanation (prediction/control)."
(Dunnell 1971, p.41).
The underlying work, however, is better compared with the fabric of
basketry: a complex interaction of oddly shaped leaves and branches. The
observations on present day Egyptian basket makers and ancient Egyptian
archaeological baskets have to be interwoven with care, in order for the fabric to
hold. Or, as Shanks has formulated it: "The archaeologist can only weave
connections that establish insights and plausibilities and are as m u c h about the
present as the past" (Shanks 1996, 128).
The method used is a hermeneutic spiral: a constant confrontation in which
theory and observations are not considered opposites, but as part of the same
continuous process (Hodder 1992,188-193, 239). The spiral represents the process
of learning. In principle nothing makes the spiral end, in practice the deadline for
publication does.
'Testing' theories developed in the course of the spiral, is problematic in this
set-up. For one thing the theory is built on top of its o w n earlier stages. Secondly,
since a confrontation of theory and observation is not possible, because they are
considered part of the same continuum, the positivist procedure of testing theory
against data is not possible. A s has been indicated in section 2.1 there are a
number of alternatives in testing theories, which have to be checked for suitability.
Testing theories on their internal coherency is useful w h e n a theory is
formulated deductively. However, spiral theory building is a mixture of induction
and deduction. This does not present a problem w h e n judging theories from their
social implications. This method of testing is linked to the important and extensive
subject of ethical implications of archaeology. Even a seemingly 'innocent' subject
as basketry m a y have political and ethical implications: the social position of
basket makers can be presented in several ways.
A practical solution is accepting that observation and theory cannot be
parted, while acting as if they can. A s long as it is kept in mind that this is a
makeshift measure, it is possible to check ones construction of reality by referring
to the data base as if it were an independent entity. Avoiding this self-deceitful
solution would be preferable, however.
In post-processual archaeology the construction of a narrative, built up by
linking small elements to a meaningful whole is presented to the outside world.
Therefore, the central demand is that the work has to be presented with m a x i m u m
openness to critique. This can be achieved by using explicit presuppositions and
use formal analysis wherever relevant. A s indicated in section 2.2 classification
and analogy are the most important tools used in basketry analysis. Via a flexible
application of classification correspondences and differences are being found.
In Chapter 9 classification is used to analyse four groups of basketry (ancient
Middle Egyptian basketry, the ancient Nubian basketry, the present Middle
Egyptian basketry and the present Nubian basketry). T o check the degree of
correspondence and difference, a fifth group is involved which is supposed to have
no historical links with the Egyptian material and uses different raw materials.
Following this, the use and properties of raw materials are considered.
Analogy between ethno-archaeological material and archaeology is
determined by following four steps. The degree of similarity is determined.
Secondly, attention is paid to dissimilarities. Next the relevance of both similarities
and dissimilarities to the interpretation of the object or residue is determined. After
2 T h e Fabric of Archaeology 23
having benefited from analogy, the last step is putting it aside, in order not to be
limited by an interpretation that is mainly focused by the analogy.
T h e example of the lower Pesos basket makers in section 2.3.2 shows that
without additional information, such as Hodders' relational reasoning, the results
give but a possible variant. Analogy thus applied forms even a threat to
interpretation, first because it gives one-sided information, secondly because it
blocks the search for alternatives.
In the following chapters an attempt has been m a d e to avoid this trap, by
making increasingly bold analogies. Part II gives a 'static' image of basketry
production by comparing ancient and modern basketry techniques and raw
materials on the basis of well-defined criteria. Part III focuses on the 'dynamic'
aspects of the production process by presenting the time involved in the present
day process and the movements of the basket maker. Part IV tries to identify the
ancient basket makers and their motives, by making a comparison with the present
day basket makers.
Thus this is a book on basketry production, with two interlinked starting
points: archaeological basketry and present day basket makers. Comparisons are
m a d e between basketry remains from the N e w Kingdom (thirteenth century B C )
and the Ballana-culture (third to sixth centuries A D ) ; between Middle Egyptian and
Nubian basketry; between modern and ancient basketry. These are gaps that have
to be bridged. Perhaps the biggest gap, however, is that between the researcher, a
west European highly educated female, without first hand experience of having to
m a k e a living by producing a craft, and the informants, mostly barely educated
Egyptian and Nubian m e n and w o m e n , w h o have been living in the same village
for their entire life and use their craft to survive or to improve their lives.
Mason 1904 Lehmann 1907
(English) i (German)
Vogt 1937
(German)
Lucas 1926'5' (English)

Lucas 1948-19593"1
Biihler-Opp<:nheim 1948
(German)
V
Gourlay 1981
(French)
'!
ii
Crowfoot 1954
(English)
''
Lucas and Harris 1962
(English)
ii
Forbes 1964
(English)
I
L . ---» Emery 1961 Seiler-Baldinger 1973
(English (German; 1977 English)
if i."
North 1990 Larsen 1986 Keppel 1984
(English) (English) (Dutch)
Adovasio 1977 Leroi-Gourhan 1943

(English) (French)
i
McGregor 1992 B.ilfet 1952
(English) <French)
1 A
• 1
1
I
Figure 3-1 Influence of major publications of basketry classification on each other.
A is known to B = B is developed out of A
B is influenced by A E E E E B is using method of A
CHAPTER THREE
CLASSIFICATION
3.1 Introduction
In section 2.3 a kitchen inventory was used to illustrate the nature of classification.
In Chapter 3 a survey is m a d e of basketry classifications used in literature on
archaeological and ethnographic basketry from different parts of the world.
A n introduction to the literature on basketry classification gives a firm basis
for the analysis of Egyptian basketry. The literature that will be discussed in this
section includes publications of Egyptian archaeological material and world wide
classificatory systems. Within the field of basketry different subjects are
emphasized, due to the specific interests of the authors. T o be able to compare
these different classificatory systems, four aspects are highlighted. T h efirstaspect
on which the classifications are compared is the purpose of classification (section
3.2). In section 3.3 a comparison is m a d e of the definitions given of basketry.
Furthermore the opinions of the different authors on the relation between textiles
and basketry are presented. Section 3.4 gives a brief outline of the different
classificatory systems, by concentrating on the research question and the criteria
used. T h e drawings and photographs, which form an important part in any work
on basketry, could not be included. In the last section some concluding remarks
will be m a d e , resulting in an outline of the use of classification in Part II.
The terminologies used in books and articles differ considerably, not only
because they are in French, G e r m a n and English, but more importantly because
they reflect different classificatory principles. It is possible to discern a number of
'traditions', by identifying the extent to which the work of one author is source
material for the others (see Figure 3-1).
The purpose of each of these books and articles differs, as does the subject
matter as far as region and period are concerned. The publication of M a s o n (1904)
describes American Indian ethnographic basketry. Adovasio (1977) and M c G r e g o r
(1992) deal with American Indian archaeological basketry. L e h m a n n (1907) has
developed a general classification, inspired mainly by material of the Dresden
M u s e u m s , from South-East Asia, Africa and South-America. T h e objective of
Leroi-Gourhan (1943) is to give a description of world technology, including the
technology of making baskets. His work is extended and improved by Balfet
(1952), w h o studied basketry from all over the world from the collections of six
ethnographic m u s e u m s .
Biihler and Biihler-Oppenheim (1948) base their classification on the

ethnographic collection of Ikle-Huber in the M u s e u m of ethnology in Basel. Seiler-
Baldinger's (1973) update and reorganization of the work of Biihler and Biihler-
Oppenheim, amounts in what can be considered a standard work on textile
techniques. Keppel (1984) uses Seiler-Baldinger's classification in her work on the
basketry of the Mentawai Islands (Indonesia). Emery (1961) has written a classifi-
cation of world textile techniques, which has been internationally accepted as a
standard. Larsen (1986) elaborates on her work, concentrating on the subject of
interlacing. A brief introduction to basketry techniques is included in a recent work
on conservation of plant material by North (1990).
Vogt (1937) is a publication of a collection of stone age basketry and textiles
of the Schweizerisches Landesmuseum. Crowfoot's (1954) world history of
basketry technology is based on archaeological evidence. Her work is used, in
places almost literally, by Forbes (1964), w h o includes a short section on textiles,
basketry and mats in his Ancient technology, in which he also uses the work of
Buhler and Buhler-Oppenheim. T w o editions of Lucas's Ancient Egyptian
Materials and Industries should be considered here. The third edition (Lucas 1959)
concentrates on raw materials and hardly mentions any techniques. The fourth
edition (Lucas 1962), which has been revised by Harris, uses Crowfoot's
classification of techniques (Crowfoot 1954).
Crowfoot, Forbes and Lucas (1959) are the main source of information for
Gourlay (1981) in his publication of the archaeological basketry from Deir el-
Medina (Egypt, N e w Kingdom period).
World-Wide Regional
Ethnographic Lehmann (1907) Mason (1904)

Leroi-Gourhan (1943) Keppel (1984)
Buhler and Buhler-Oppenheim (1948)
Balfet (1952)
Emery (1961)
Seiler-Baldinger (1973)
Larsen (1986)
North (1990)
Archaeological Crowfoot (1954) Vogt (1937)

Forbes (1964) Lucas (1926)
Lucas and Harris (1962)
Adovasio (1977)
Gourlay (1981)
McGregor (1992)
Table 3-2 Division of publications in those dealing with ethnographic and

archaeological basketry, on a regional as well as a world-wide scale.
3 Classification 27
In Table 3-2 the publications have been divided into four groups to clarify
their subject matter. The largest part of the literature, eight of the eighteen
publications, has been based on information on ethnographic basketry of different
continents, with the objective to give a survey of basketry techniques on a world
wide scale. T w o publications, M a s o n and Keppel, concentrate on ethnographic
basketry from one region. Although Mason's book covers a large area, it deals
with basketry from the American continent only, and, more specifically, American
Indian ethnographic basketry, with some attention to archaeological material.
Keppel describes basketry from a specific Indonesian region.
The same division in world wide and regional can be m a d e in respect to the
work done on archaeological basketry. The publications of Crowfoot and Forbes
both describe a history of basketry technology, which is meant to be world wide.
Different regions are subject of the work of Vogt, Lucas, Lucas and Harris,
Adovasio, Gourlay and McGregor. The work of Adovasio focuses on ancient
American basketry, while McGregor describes archaeological basketry from
specifically the Lower Pecos area (Texas). Vogt's work deals with archaeological
material from Switzerland.
There are three publications on ancient Egyptian basketry, Lucas, Lucas and
Harris, and Gourlay. Lucas mainly lists archaeological basketry, found in Egypt.
In the later edition, revised by Harris, there is more attention for the different
techniques. The publication of archaeological basketry, which can be considered
the most closely related to material from the workmen's village at Amarna, is
Gourlay's catalogue of the finds from the workmen's village at Deir el-Medina.
In the next sections the above-mentioned publications will be compared in
more detail. Because of the difference in terminology used, all classifications will
be 'translated' into a basic description, using a limited number of terms'. Apart
from the wide range of terms that are direct quotations or representations of those
used by other authors, the following expressions will be used:
• Technique is used as a general term for the structure of the basket. The
technique can be recorded by studying a basket or fragment of a basket.
• Technology is used to indicate the production method of the technique. This
includes the use of tools and the working order. Technology can be recorded by
observing modern basket makers, or (partly) inferred from the technique.
• Strand is used to indicate the smallest unit with which a basket is made, be this a
string, leaflet, twig or tree.
• Element is a strand, or number of strands worked up together. W h e n an element
consists of more than one strand, the element has several members.
• Set of elements are a number of strands, which have the same function in a
technique. This function can be active ox passive. Thus, a basketry fabric can be
m a d e from one long element or several parallel elements.
More definitions can be found in Chapter 4 and the glossary. The terms are
explained and illustrated in Wendrich 1991c.
• Passive and active indicate the function of elements. Passive elements make up
the body of a technique, while active elements hold the technique together.
Together, the active elements form the active system, which causes the coherency
of the technique. Without the active system, all the passive elements (the passive
system), would just be a number of loose strands.
• Basic structure is the w a y in which a technique is made up of several strands.
The basic structure consists of a number of systems. The systems each consist of
an element, or set of elements, that have the same role in the technique. The
interaction of the systems results in a coherent structure that makes up the
technique.
3.2 Purpose of classification

In the earlier works on basketry the reason for making a basketry classification is
often not made explicit. O n e should, of course, be aware that searching for an
explicit purpose is looking with late-twentieth century eyes to work that has been
done from the beginning of this century onwards. By reading between the lines, a
reason can always be detected. They are presented here in chronological order.
M a s o n sets out to formulate criteria to identify the source of production of
American Indian basketry. In his o w n words: "In basket-making there are several
characteristics to be observed which will enable one to classify the objects and to
refer them to their several tribal manufacturers." (Mason 1904, 190).
For L e h m a n n the purpose is to create an order. H e does not discuss the gain
of such an orderly arrangement extensively, but he remarks: "Schon u m die
einzelnen Geflechtsarten systematisch ordnen zu konnen und nicht spater mit
anderen ahnlichen Arbeiten zu kollidieren, ist eine systematische Ubersicht iiber
samtliche existierenden Geflechtsarten unbedingt erforderlich." (Lehmann 1907,
p. 1) Lehmann is convinced that a classification is imperative, because all attempts
of studying basketry and making a classification clash with work previously done,
if a complete system is lacking. Reading between the lines, it m a y be concluded
that Lehmann sets out to classify basketry to create the possibility of comparison
between different people's work.
The presentation of the collection of stone age basketry and textiles at the
Schweizerisches Landesmuseum forces Vogt to make models to explain the
technical differences between the objects. H e presents these 'models' (drawings of
the technical construction) in a specific order, but Vogt does not m e a n to m a k e a
classificatory system. For that he refers to Lehmann's system, which he mentions
and declines as too complicated and yet not detailed enough for his purpose.
Leroi Gourhan uses basketry from the Musee de I'Homme and his o w n
observations during field work as the basis for the description of basketry
technology. This is part of an attempt to create a complete technological
description of the world. "Aucun chercheur ne pourrait decrire l'activite des
h o m m e s en tous les temps et dans tous les pays, mais les grandes classifications se
font, bien avant qu'une science soit completement exploitee." (Leroi Gourhan
3 Classification 29
1971, 14). Thus Leroi Gourhan implies that a good classification contains
knowledge of which scholars are still unaware.
T h e purpose that Balfet has in mind is not m a d e explicit, but can be inferred
from what is written in her book: "Si quelques formes bien etudiees sont connues
avec toutes leurs variantes dans certaines aires geographiques, il est encore
impossible de se faire une idee des rapports de certains types entre eux, et le
defaut de terminologie rend difficiles les descriptions et incertaine 1'interpretation
des donnees publiees." (Balfet 1952, 295). The purpose of classification is, in her
view, to find correlations on a worldwide scale between types of basketry, rather
than just concentrating on describing a certain region in detail, and to develop a
terminology that makes it possible to communicate on basketry.
Partly basing herself on Vogt, Crowfoot does not give a purpose. In her
introduction to the section called "Types of basketry", she writes: "Before
discussing these finds it is necessary to describe in some detail the main techniques
in use." (Crowfoot 1954, 415). Although she does not m a k e explicit w h y this is
necessary, she does give a clue to what she is really interested in, w h e n describing
basketry techniques, namely the evolution of techniques: "The evolutionary story
of these techniques is still full of gaps" (Crowfoot 1954, 414).
Emery's purpose for classification is creating a consistent terminology. In her
o w n words: "The bewildering inconsistencies and incongruities (...) led m e to
undertake a detailed and systematic investigation of the essential characteristics of
ancient and primitive fabrics." (Emery 1980, xi).
Both Lucas' and Forbes' objective is comparable to that of Crowfoot: to find
the evolution of techniques, Forbes worldwide and Lucas in Egypt only.
Seiler-Baldinger does not give an introduction to her classificatory system. In
a preface Buhler, whose work forms the basis of Seiler-Baldinger's system (Buhler
and Buhler-Oppenheim 1948), gives two reasons for making a classification. First,
classification provides the basis for a textile terminology, which was lacking
before. The terminology as it was used until the publication of Seiler-Baldinger's
book, is unclear, especially internationally. A classificatory system is necessary, to
create a consistent terminology. Secondly, Buhler considers the production process
as a source for detecting a development of technology. In his introduction to
Seiler-Baldinger's book he writes: "Die Gliederung nach Anfertigungsprozessen
hat den grossen Vorteil, dass m a n die einzelnen Verfahren nach ihrer technischen
Entwicklungshohe einstufen kann." (Buhler in Seiler-Baldinger 1973, X X V I ) .
Adovasio considers basketry classification analogous to the taxonomy of
plants and animals. H e expresses this as follows: "The entire procedure reduces
the assemblage to progressively smaller units of increasingly greater taxonomic
resolution or precision." (Adovasio 1977, p.l). The objective of making a
classification is thus to create an order, but nowhere Adovasio specifies or hints to
what purpose.
Gourlay notes that Lucas has provided a sketch of the evolution of basketry
techniques in Egypt and he clearly states that he does not set out to do the same.
Gourlay's purpose is to identify the date and place of the different techniques by
studying them in detail, and putting them in an order according to k n o w n and
newly created criteria: "II conviendra dorenavant de mettre en fiches, selon des
criteres definis dans ce present travail si possible, sinon en creant selon les besoins
des criteres nouveaux rentrant facilements dans la classification ici employee.
Seule une mise en fiches systematiques, repondant a des normes bien strides,
permettra une datation et une localisation de tout objet de sparterie avec un
m a x i m u m de certitudes." (Gourlay 1981, vii-viii). Since he publishes basketry
from one site, his material is not really suited for this purpose. H e probably means
that by ordering basketry according to clear criteria, the descriptions can be
compared with the work of others, which makes it possible to compare the data
and in the end develop a set of criteria against which every n e w find can be dated
and placed. Interpreted this way, Gourlay's objective is close to that of Balfet,
Emery and Seiler-Baldinger.
Larsen is convinced that classification is necessary, and that his classification
fills a gap. H e does not indicate, however, what the purpose of his classification
is. "A sweeping view and survey of interlacing in all its forms is needed" and:
"No classification has appeared that consistently and logically relates the weaves of
textiles to other interlacings. Our classification is proposed as a further step toward
the comprehension of all interlacings as a single entity, which can be described by
one system of classification" (Larsen 1986, 10).
In a chapter on basketry techniques, North explains the importance of this
subject for conservators: "Familiarity with construction methods is important to
accurately document artifact construction and to understand the mechanical
strengths and weaknesses of artifacts." (North 1990, 87). Her only objective is to
give a simple, clear survey of techniques, in an order that is easy to grasp.
The question of the purpose of basketry classification thus receives a wide
variety of answers, which range from none, to a heuristic application of
classification.2 In a nutshell the purposes as found above are:
• N o clear purpose (Larsen, North)
• Creating an order (Lehmann, Adovasio)
• Identifying the source of production (Mason)
• Determining the evolution of techniques (Crowfoot, Lucas, Forbes)
• Determining the level of development of a technique (Buhler / Seiler-
Baldinger).
• M a k e communication on basketry possible, by creating a consistent
terminology (Balfet, Emery, Buhler/Seiler-Baldinger, Gourlay)
• Explain technical differences (Vogt)
• Give a technological description of the world (Leroi Gourhan)
Perhaps Leroi Gourhan's remark that great classifications can describe h u m a n
action and contain knowledge of which scholars are still unaware (Leroi Gourhan
1971, 14), can be interpreted as the heuristic power of classification. O n the other
"The purpose of classification within the scope of the underlying work, will be put
forward in section 3.5.
3 Classification 31
hand, this remark could also be interpreted as a somewhat naive belief in 'letting
the material speak for itself.
3.3 Defining the Boundaries

Definitions of basketry are often concerned with defining the boundaries between
basketry and textiles. Both aspects, definitions and defining the boundaries, are
surveyed in the work of the different authors, introduced at the beginning of this
chapter. Their statements on the subject are presented in order of publication date.
3.3.1 Survey of boundaries

M a s o n defines a basket as "A vessel or receptacle in textile material; a technic
product resembling this" and basketry as: "A general term including (1) basket
making, the process or art; (2) basket work, the technic or stitches, any textile
motive resembling work in baskets; (3) basket ware, a collection of finished
products."(Mason 1904, p. 193). H e sees basketry as a variety of textiles, but
considers the use of a loom and the rigidity of the material as the discriminating
factors between basketry and cloth: "Basketry is differentiated from (...) loom
products, not only by the material, which is usually less rigid, but by the
workmanship, which is done by machinery. (...) N o wide gulf separates the
different varieties of textiles, however, beginning with such coarse products as
brush fences and fish weirs and ending with the finest lace and needlework."
(Mason 1904, p. 184)
L e h m a n n states that giving a definition of basketry with a limited amount of
words is impossible, but nevertheless attempts to do so: "Flechten nennt m a n eine
ohne oder mit Instrument, das jedoch nicht zur Versteifung des Streifenendes
(Fadenendes) dient, vorgenommene Manipulation, die bezweckt, aus einem
Material von geringer Breitenausdehnung und unbeschrankter Langenausdehnung
durch Verschrankung entweder ein an Breitenausdehnung grosseres und somit
festeres oder ein Gebilde zu erhalten, dessen Breitenausdehnung von der Willkur
des Manipulanten abhangt." (1907 p 2.) A definition of basketry is not easy; that
m u c h m a y be clear. L e h m a n n proposes that basketry is made with or without
instruments, but in any case without looms. It is made from material with limited
width and unlimited length. This material is interlaced to make either bigger and
stronger material, or shapes for which the size is determined by the producer.
L e h m a n n does not concern himself with discerning textiles and basketry. H e
maintains that, although the boundaries between textiles and basketry are not easily
drawn, in practice it will be clear what is basketry and what is a textile.
Vogt presents stone age textiles and basketry in two separate parts.
Therefore, mentioning the criteria that rule this division is important to him. Vogt
distinguishes two actions, rather than two artefact groups: "Als Hauptunterschied
zwischen Flechterei und Weberei kann der Umstand gelten, dass bei der letzteren
wenigstens ein Teil der Faden mit einer gemeinsamen Vorrichtung zur Bildung
eines Faches versehen sein muss." (Vogt 1937, p. 2). Weaving requires a loom
that has a device to lift a part of the warp in one movement. Thus the
sophistication of the loom is what divides textiles from basketry, for example mat
weaving, which might be done on a simple loom, is classed as basketry, because
each strand of the weft is inserted by hand.
The third edition of Lucas's work on ancient Egyptian basketry (1948-1959)
mentions that basketry is m a d e without the use of any kind of machinery, whereas
weaving requires tools, such as looms and a specific preparation of the fibres
(spinning).
Leroi Gourhan uses the same word for basketry and textiles (tissus). The
difference between the two groups is a practical one, m a d e by c o m m o n sense and
based on shape, use and appearance (Leroi Gourhan 1971, p.268). In technique,
however, Leroi Gourhan wants to incorporate m a n y similarities in his
classification: "L'essentiel de ce qu'on peut dire des tissus (vannerie et etoffes) est
c o m m u n a toutes les techniques qui ont pour but l'enchevetrement de deux nappes
et la distinction apparait c o m m e secondaire." (Leroi Gourhan 1971, p. 272). This
approach can be criticized, because although the technical structure of some
baskets and textiles are similar, the manner in which this structure was m a d e - the
technology - often is not.
Balfet uses the same definition of basketry and textiles as Leroi Gourhan:
"Vannerie est l'assemblage a la main de fibres relativement rigides et d'assez gros
calibre pour en faire des surfaces continues, le plus souvent des recipients." and:
"le tissage est l'assemblage sur un metier a tisser de fibres fines pour fabriquer une
surface plane." (Leroi Gourhan 1971, p. 268) (Balfet 1952, p. 260) According to
these two definitions the difference is determined by the flexibility and size of the
strands, and the shape or function of the object (container versus flat surface). The
use of words such as "le plus souvent" indicates that the limit between basketry
and textiles is arbitrary, something which Balfet also states explicitly.
The title of Crowfoot's article "Textiles, basketry and mats", implies a
relation between these three groups: "Basketry and mats are commonly
distinguished from weaving, but it is often difficult to k n o w where to m a k e the
division" (Crowfoot 1964, p. 414). She points to the function (baskets are vessels)
and, with respect to the difference between mats and textiles, to the use of looms,
notes: "Only with the fuller development of the loom did the techniques draw
apart." (Crowfoot 1954, p. 414).
E m e r y gives ample attention to the manner in which basketry and textiles
("cloth" in her terminology) are separated. She lists four criteria from the work of
others (see p. 34) and states that none of these criteria can be used to m a k e a non-
arbitrary separation. Since Emery uses the structure of the techniques as
classificatory criterion (see section 3.4), she argues that a separation into two
technical groups is counter productive. She considers it: "a false assumption that
the terms basketry and cloth refer to mutually exclusive groups of fabrics. Both
terms are broadly generic and actually refer to fabric groupings that overlap in so
m a n y areas that neither group can be properly studied or understood without
reference to the other. (...) separation of the two groups is necessarily arbitrary.
The habit of treating them as disparate subjects for investigation has led unhappily
to the use of different terms for identical structures as well as frequent failure to
3 Classification 33
recognize and record structural identity or similarities in closely related fabrics that
happen to have been relegated to separate categories. " (Emery 19802, p.210). Not
only the four subjects that Emery mentions can be used as criteria to split the two
realms. The technical structure of basketry and textile shows fundamental
differences also, as will be illustrated below. Emery's endeavour to m a k e one all-
encompassing terminology focuses her attention on the similarities, rather than the
differences.
Three years later, Forbes writes: "Baskets and mats are n o w c o m m o n l y
distinguished from weaving, but w e cannot make this sharp division in earlier
periods of h u m a n history" (Forbes 1964, p. 178). His argument is the same as
Crowfoot's. Only after the development of the loom, textiles have m o v e d away
from basketry.
Just like Emery, Seiler-Baldinger does not m a k e a distinction between textiles
and basketry. She uses the term textile techniques3, which she divides into two
groups, the "primary" and the "advanced" techniques. The "advanced" techniques
require more or less complicated instruments, such as looms. This division is
thought relevant for detecting levels of development (see section 3.2). All basketry
techniques fall within the first group.
Adovasio does not c o m m e n t on the difference between basketry and textiles.
H e states that the term basketry applies to different kinds of items, such as
containers, matting, bags, fish traps, hats and cradles. H e does not give a clear
definition of basketry. A s a description of the general attitude towards basketry and
textiles he states: "Specifically, all forms of basketry are manually assembled or
w o v e n without a frame or loom. Being woven, they are technically a class or
variety of textile. Usually, however, that term is restricted to "cloth" fabrics with
continuous plane surfaces produced on or with the aid of some sort of auxiliary
apparatus" (Adovasio 1977, p. 1). H e is m u c h more precise in his definitions of
the three subclasses, twining, coiling and plaiting, (see section 3.4).
Gourlay does not give a definition of basketry, nor considers the differences
or similarities between textiles and basketry.
Larsen agrees with Emery, that cloth and basketry should not be considered
as different techniques. His classification is limited to a specific group of
techniques, referred to as interlacing, which he defines as: "A fabric structure
interworked so that each element passes over and under elements that cross its
path-without other engagements such as twisting or linking" (emphasis by Larsen).
A fabric structure is: "The system by which linear elements or fibers are
interworked or enmeshed" (Larsen 1986, p. 38).
North does not define basketry either. She implicitly includes basketry and
textiles in one group, as far as they are m a d e from plant materials. This m a y be
"Als Stoffe werden (...) samtliche Produkte textiler Techniken betrachtet, die aus
untereinander rein mechanisch verbundenen Grundbestandteilen (Faden oder Fadengruppen)
bestehen. Die Feinheit der Faden ist dabei von sekundarer Bedeutung." (Seiler-Baldinger
1973 p.4).
because the subject of the book she contributes to, is the conservation of plant
materials.
3.3.2 Summary
Summing up the views of the different authors, w e find that some authors
(Gourlay, North) do not give the problem of defining basketry any thought. Others
(Emery and Seiler-Baldinger) have decided not to make a distinction between
textiles and basketry, because they think it counterproductive in the light of their
purpose, whish is creating an all encompassing terminology of textile techniques.
All other authors mention a number of criteria according to which they made
a division between basketry and textiles. Emery lists four criteria which she has
found in the work of others (Emery 1980, p. 209), to which a fifth should be
added to outline the criteria as mentioned by the different authors:
1. Implementation
Lehmann indicates that for basketry no instruments or only simple instruments are
used. Both he and Mason insist that no looms are used for basketry, while Vogt,
Lucas, Crowfoot and Forbes maintain that simple looms are used for basketry (i.e.
matting) while textiles are made on complicated looms.
2. Preparation of elements
According to Emery it is often maintained that an important criterion to discern
textiles from basketry is that textiles are made of elements that have been spun or
plied into long yarns. She refutes this criterion, which is used only by Lucas, by
pointing at the use of silk, which is often used unspun for weavingfinetextiles.
3. Relative size and flexibility of the raw material

Mason and Balfet use this criterion when they say that basketry material is rigid, to
which Balfet adds that it is coarse as well. Lehmann also refers to the aspect of
raw material, remarking that basketry material is of limited width and unlimited
length ("Material von geringer Breitenausdehnung und unbeschrankter Langenaus-
dehnung"). In most basketry materials (e.g. leaflets, twigs and grasses) both width
and length are limited. Seiler Baldinger indicates that a material criterion is not of
primary importance, stating that: "Die Feinheit der Faden ist dabei von sekundarer
Bedeutung" (Seiler-Baldinger 1973 p.4). Leroi Gourhan points out that the division
between basketry and textiles is often made because of the 'appearance', which is
one of the c o m m o n sense criteria he decided not to use in his classification. The
criterion of relative size and flexibility is closely related to that of preparation of
the strands, since the properties of the raw material can be changed through
working them. A rigid twig, for instance, can be soaked and twisted into a flexible
string.
3 Classification 35
4. Shape and quality offinishedproduct

Balfet indicates that textiles are a flat surface, while baskets are containers. This
criterion is a mixture of shape and function. The shape is one of the ' c o m m o n
sense' criteria, which Leroi Gourhan describes as not being essential.
5. Use and function

Leroi Gourhan mentions the function also as a c o m m o n sense criterion to discern
basketry and textiles. Balfet mixes functional criteria with criteria of shape (see
under point 4). Crowfoot indicates that baskets are vessels and Adovasio lists a
large number of functions of baskets.
3.3.3 Discussion
E m e r y is right in arguing that the division between basketry and textiles has been
arbitrarily made, on the basis of an inconsistent use of a wide range of criteria.
Still, I consider it important to m a k e the distinction, because there are two aspects
in which basketry and textiles differ fundamentally, which are relevant, especially
in a publication which focusses on the production process. Analysing basketry as a
textile technique results in missing out those aspects which primarily determine the
basketry technique.
Insertion of raw materials

The first aspect to be considered is the relation between technique and raw
material. Figure 3-3 represents a textile technique k n o w n as 111 tabby or plain
weave. T h e same basic structure is visible in both Figures 3-4 and 3-5: passive
vertical elements ('warp') are crossed at right angles by active horizontal elements
('weft') in an under 1 I over 1 - shift 1 pattern. Even though the structures of
Figures 3-3, 3-4 and 3-5 are identical w h e n considered purely schematical, the two
basketry techniques differ fundamentally in the w a y the raw material has been
inserted: In Figure 3-4 a number of willow rods have been inserted and are
worked up simultaneously. In Figure 3-5 one willow rod is put in and worked up,
then the next one is inserted and so on.
s i s S i S X
s ™ m p
s i w l s s p
Figure 3-3 The tabby, or plain weave: weaving under one / over one
36
Figure 3-4 French randing: weaving under 1 / over 1 with a number of willow rods
simultaneously, which are all layed in with the thick side.
Figure 3-5 English randing: weaving under 1 / over 1 with one willow rod at a
time. The first willow rod is worked around until it reaches the starting
point. Then the next one is inserted, between the next two stakes, in
order to divide the thickness of the butts is evenly.
3 Classification 37
The willow rods have a limited length (about 1.50 m) and a specific shape
(the diameter varies from about 15 m m at the butt, to 1 m m at the top). In order to
m a k e a basket for which the sides are the same height everywhere, the willow rods
cannot be inserted randomly. The thick butts and the thin tops have to be evenly
distributed. T w o methods to do this are illustrated by figure 3-4 and 3-5.
Describing both basketry techniques as '1/1 tabby' would be missing the point.
Thus, the limited length and the specific shape of the plant parts determine
the technique. Regular insertion is necessary because the raw material has a
limited length and the shape determines in what manner the insertion is done.
Textiles, on the other hand, are m a d e from uniform materials of unlimited length.
Shape
Secondly, the intended shape of the end product determines the technique. Just as
with some textile techniques (for instance knitting), a basket is shaped
simultaneously with the construction of the basketry 'fabric'. The shaping of the
object, such as transition from base to side, is an intrinsic part of the technique.
The construction often needs strengthening at certain points, for instance, to
maintain the shape.
Start, insertion of new material, shaping and finish

The combination of shaping the object as an intrinsic part of the technique and the
requirements of the raw material results in four aspects which do not occur in
textile techniques, but are important in understanding basketry techniques. These
aspects are: the start, the insertion of n e w material, the shaping and the finishing
off of the basket. The start of the construction, usually the centre of a basket, is
the place where a number of elements are fastened from which the basket is built
up. During the entire production process new material has to be inserted, since the
raw material is of limited length. The number of strands has to be increased or
decreased, in order to m a k e a specific shape. Technical solutions are sought to
maintain the shape, for instance by strengthening the basket in places. Lastly, all
elements have to be fastened, usually at the rim. Often the fastening of the
elements at the rim also fulfills other requirements, such as fixing the shape.
Thus, it is judged important to distinguish basketry from textiles, because an
analysis of the production process needs focusing on the basketry-specific aspects
mentioned above. The main reason for Emery and Seiler-Baldinger not to split
basketry and textile techniques, is their interest in creating one terminology. This
is not the purpose of the present work, however. O n the contrary, in section 2.3.2
it was said specifically that an all encompassing terminology is not deemed
important, because the classification on which such a terminology would have to
be based, depends on changing research questions.
3.4 Research questions and criteria

In this section classificatory systems with similar underlying research questions
and criteria are presented together, rather than in order of their year of
publication. First Lehmann and Vogt will be considered separately. Then three
classifications of American Indian basketry will be analysed (Mason, Adovasio and
McGregor). Thirdly, the French tradition will be looked at (Leroi Gourhan and
Balfet). Buhler and Buhler-Oppenheim wil not be discussed separately, but the
continuation and expansion of their work by Seiler-Baldinger will be. A second
American tradition is formed by Emery and Larsen. Crowfoot, Lucas and Forbes
will be discussed together. Lastly, w e will look at the classificatory principles used
for ancient Egyptian basketry by Gourlay.
3.4.1 Lehmann
Lehmann classifies according to technique, which he calls Geflechtsart. H e is
particular about distinguishing this from Flechtart, the manner in which a
technique is made (in m y terminology the technology). Other than this focus on
technique it is not possible to find an explicit research question.
Lehmann is not at all interested in developing a basketry terminology derived
from his classification. In Lehmann's opinion the peculiarities of different
languages would only hamper international discussions on basketry. Therefore he
uses a numbering system instead of terminology. His system harbours a large
amount of techniques, which are partly classified according to formal criteria.
A m o n g these criteria are orientation and number of systems, weaving patterns and
connection of the systems.
Class I, for instance, contains basketry made from strands orientated in two
directions. This class has four sub-classes (Table 3-6).
A Geflechte mit Taftbindung tabby-weave
B Geflechte mit Koperbindung twill weave
C Geflechte mit Atlasbindung satin weave
D Zopf- und Bandgeflechte, sofern sienicht zu plaited straps, other than those
A und B gehoren under A and B
Table 3-6 Representation of Class I of the classification of Lehmann 1907
These four subclasses are divided into the same subsubclasses:

1. made from strands of the same width
2. made from strands of different widths.
A further subdivision (indicated by letters of the greek alphabet) is in most cases
made without explicit criteria. Class IAacx, for instance, is described as
"Gewohnliches Taftgeflecht" (normal tabby weave), while class IAaP is divided
into IAaPl "Geflechte, bei denen die Geflechsstreifen der einen Richtung aus
verschiedenen Streifen bestehen" and IAaP2 "Geflechte, bei denen alle Streifen aus
verschiedenen Streifen bestehen".
3 Classification 39
Basketry techniques are thus indicated with a letter and number code, most of
which are illustrated with drawings.
Indications such as "IAaP2" do not give a clue to which technique they are
referring, unless Lehmann's publication is at hand. T h e classification is thus hard
to use, unless the system is k n o w n by heart. M o r e importantly, the system is not a
classification in the sense in which I have chosen to use it. It is a clear example of
a non-dimensional hierarchical construction, which, as was shown in section 2.3,
is an indication of grouping. T h e sheer quantity of different techniques gives the
impression that his system is all-encompassing, but in fact it is not. Each n e w item
of basketry might cause the need to formulate n e w 'classes' and there are no empty
classes.
Despite its shortcomings Lehmann's classification is a valuable record of
examples of basketry techniques and he gives s o m e useful incentives for formal
criteria.
3.4.2 Vogt
Vogt k n o w s the w o r k of L e h m a n n , but judges his system on the one hand too
complicated and on the other hand not detailed enough to be of service (Vogt
1937, 5). H e maintains Lehmann's terminology by discriminating Geflechtarten,
the variety in techniques, and Flechtart, the technology, but he makes a m u c h
simpler division into seven classes (Table 3-7):
A Randparallele Geflechte straight angled plaiting
B Spiralwulstgeflechte coiling
C Geflechte mit Zwirnbindung twining
D Vliessgeflechte mit Zwirnbindung twining with inlay
E Netzgeflechte netting
F Geflechte mit Zopfbindung Neolithic braiding4
G Geflechte mit Leinwandbindung und Four-Directional Plaiting5

Diagonalstreifen
Table 3-7 Representation of the classification of Vogt 1937
4
Larsen calls this technique "Neolithic braiding" (Larsen 1986, 83).
5Term of Seiler-Baldinger (Seiler-Baldinger 1979, 30).
THE W O R L D ACCORDING TO BASKETRY
40
These classes are descriptive, and not based on criteria. S o m e are divided into
subclasses, or 'variants' are indicated. Again w e are dealing with a hierarchic non-
dimensional grouping, rather than with classification.
3.4.3 Mason, Adovasio and McGregor

M a s o n uses vague criteria for his basketry classification, although he does point
out the characteristics of basketry, which he considers important for classification:
"These characteristics are the material, the framework, the methods of weaving,
the coiling or sewing, the border, the decoration, the use, etc" (Mason 1904, 190).
O n the same page he proceeds by describing the typology as follows: "There are
two distinct types of technic in basketry, namely, (1) h a n d v w w z basketry, which is
built on a warp foundation, and (2) sewed or coiled basketry, which is built on a
foundation of rods, splints, or straws." (Table 3-8).
•
Kinds of W o v e n Basketry
A Checkerwork The warp and the weft having the same width,
thickness, and pliability.
B Diagonal or twilled basketry T w o or more weft strands over two or more

warp strands.
C Wickerwork Inflexible warp; slender, flexible weft.
D Wrapped weft, or single weft The weft strand is wrapped, or makes a bight
wrapped about the warp at each decussation, as in the
Mohave Kiho.
E Twined or wattled basketry Weft of two or more elements.

Kinds of Coiled Basketry
A Coiled work without foundation.
B Simple interlocking coils
C Single-rod foundation
D Two-rod foundation
E Rod and welt foundation
F Two-rod and splint foundation.
G Three-rod foundation
H Grass-coil foundation.
K Fuegian coiled basketry basketry specific for the Fuego tribe
Table 3-8 Representation of the classification of Mason, 1904

3 Classification 41
In the section on woven basketry he introduces criteria for classifying the coiled
basketry: "The sewed work goes by the n a m e of coiled basketry, and is classed
both by the foundation and the fastening. In addition to these technical methods on
the body, special ones are to be found in the border" (Mason 1904, 222). The
w o v e n basketry is divided into groups without clear criteria. Like L e h m a n n and
Vogt, M a s o n is grouping his material without classifying it (cf. Table 3-8).
L e h m a n n in his book, criticises Mason's work for being too limited, because
M a s o n he is only dealing with techniques that occur among the American Indians.
Adovasio, w h o writes also about (archaeological) basketry from American
Indians is explicit about the aspects used for classification: "A variety of attributes
have been and can be employed to classify basketry. Such diverse criteria as shape
of the object, rigidity or flexibility of the weave, and elements of decoration (to
n a m e but a few) have been used with widely varying degrees of success. I believe
that sub-classes should be defined exclusively by attributes of wall construction."
(Adovasio 1977, 1) Implicitly, however, he also involves technological aspects, for
instance the working position of the producer (Adovasio 1977, 15).
The techniques of making the wall construction, that is the technique in
which the side of a basket or the surface of a mat is made, is divided into three
mutually exclusive and taxonomically distinct sub-classes: twining, coiling and
plaiting. Adovasio's definitions of these three classes are: "Twining denotes a sub-
class of basket weaves manufactured by passing moving horizontal elements,
called wefts, around stationary vertical elements, called warps. In the twining
process the wefts are active while the warps are passive." (Adovasio 1977, 15)
and: "Coiling denotes a sub-class of basket weaves manufactured by sewing a
stationary horizontal element or set of elements, called the foundation, with
moving vertical elements, called stitches. The stitches are active, while the
foundation is passive." (Adovasio 1977, 53) and: "Plaiting denotes a sub-class of
basket weaves in which all elements are active. Single elements or sets of
elements, called strips, pass over and under each other at a more or less fixed
angle (about ninety degrees), without any other form of engagement" (Adovasio
1977, 99).
A basket is allocated to one of these classes and then described according to
different criteria for each group. A s with the other three systems w e have seen so
far, this is a taxonomic approach. Consequently the subclasses of twining, coiling
and plaiting cannot be compared (cf. Dunnell 1971, 76-84).
Twined basketry is subdivided by three criteria: the spacing of the weft rows,
with three attributes (open, close and a combination, cf. Table 3-9); the number,
arrangement and sequence of warps engaged at each weft crossing, with five
attributes, and the stitch slant with three attributes. A combination of all
possibilities results in 45 classes (3x5x3). Surprisingly, Adovasio only mentions 27
classes, although he claims to combine all of the construction attributes (Adovasio
1977, 20).
,
Twined Basketry
^ ~ ^ n o . of aspects: 1 1 2 3 i 4 5
characteristic: ~~^^^
A. Spacing of weft rows: open ; close open +

close
B. Number, arrangement simple j diagonal simple+ i cross wrapped

and sequence of warps diagonal | warp
engaged at each weft
crossing
C. Stitch slant S !Z S+Z i
Table 3-9 Representation of the classification of twined basketry by Adovasio, 1977
A similar phenomenon occurs in respect to coiled basketry. The three criteria, in

which the work of M a s o n can be recognised, are the spacing of the foundation (the
coil) in close, open and a combination of both (see Table 3-10 A.) Secondly, there
is the criterion of kind, number and arrangement of foundation elements. In fact
these are two criteria bundled together: 'kind of foundation', with three attributes
(rod, bundle, welt) and the criterium 'basic arrangement', with four attributes
(single element foundation, horizontal foundation, stacked foundation and bunched
foundation).
Coiled Basketry
^ ^ ^ n o . of aspects: 1 j 2 j 3 4
characteristic: —>,^
A. Spacing foundation: close j open ! open + close
B. Kind of foundation rod | bundle ; welt

elements
B' number and arrangement single j horizontal I stacked bunched

of foundation elements: element
C. Type of stitch simple | intricate j wrapping
Table 3-10 Representation of the classification of coiled basketry by Adovasio, 1977

The third criterion (Table 3-10 C ) is the type of stitch, which has three main
attributes. The first is the "simple stitch", subdivided in interlocking, non-
3 Classification 43
interlocking and split. The second is the "intricate stitch", the third is callled
"wrapping stitch".
A cross reference of these criteria results in 108 classes (3x3x4x3, not
counting the subdivision of the "simple stitch"). This time Adovasio explicitly
decides to present just a number of classes: "Combining the attributes of wall
construction can generate a very large number of coiling types. In fact more than a
hundred types of close coiling alone are theoretically possible using the kinds of
foundations and stitches listed above. The addition of rarer foundations increases
the potential number significantly. In lieu of enumerating all the types that can be
generated or even those that have been documented archeologically, I have m a d e a
representative selection of c o m m o n open and close types for illustration"
(Adovasio 1977, 62)
Plaiting is the most simple technique to classify, according to Adovasio. Only
the number, orientation and composition of plaiting elements are used as criteria.
The two classes are "simple-" and "twill-plaiting", which are further distinguished
by the "appropriate interval designators" (Adovasio 1977, 109), which form the
plait pattern. Furthermore, n e w classes can be m a d e whenever necessary. "It is
theoretically and mechanically possible to produce twill plaiting with an uneven
principal interval (such as 2/3 or 3/4). While I have observed few specimens with
these patterns, they do occur."
Adovasio's descriptions are clear and systematical, although hes is not fully
consistent. The technique, which is usually indicated by the term wickerwork, or
stake-and-strand basketry, is classed as a sub-class of plaiting. Wickerwork,
however, is an example of basketry in which there are clear active and passive
elements. According to his o w n definitions, Adovasio should have classed
wickerwork as twining. The reason that this he did not do this, is that Adovasio
implicitly added another criterium to the definition of twining: the horizontal,
active elements are worked up in pairs. The two twining strands are twisted around
each other, holding the passive vertical elements in place.
In a separate section, Adovasio describes other features than wall
construction, such as centres, rims and decoration. H e calls this the 'post-typing
analysis', indicating that the typology of basketry is m a d e solely on the basis of the
wall construction, while the understanding of the other features is secondary.
Although Adovasio formulates criteria, and makes a partly dimensional
hierarchical classification, he does not define consistently all classes that result
from his o w n criteria. H e does not feel the need to look at classes that remain
empty and resorts partly to grouping, adding n e w classes if n e w material comes
up. In contrast to Adovasio, I consider the centres and rims an intergral part of the
techniques.
M c G r e g o r follows Adovasio's system exactly by referring to his
classificatory system. In her classification of the basketry of the Lower Pecos
(Texas) she only includes the classes that actually occur.
44 The world According to Basketry
3.4.4 Leroi Gourhan and Balfet

In section 3.2 Leroi Gourhan's central interest w a s identified as being the
technology. H e emphasizes the importance of experiments in the present to better
understand the ancient technological process, and he involves the working position
of the producer and the orientation of the w o r k towards the producer. His
classification is a taxonomic system with three levels according to the criteria:
fundamental position of the strands (1. and 2.)
fundamental character of the interweaving (variable)
pattern of weaving (variable)
T h e system thus starts as a dimensional hierarchical classification. In the sub-
groups, howere there is a less clear use of criteria.
1. nappes superposees (strands arranged on top of each

other and fastened by a third strand)
a. character of interweaving
X. several patterns
2. nappes enchevetrees (interwoven strands)
Fundamental character of interweaving
a. vannerie a brins spiralis, etoffe a (coiling: the active element spirals

trame spiralee around a fixed element)
X several patterns
b. vannerie cordee, etoffe cordee (twining: a pair of active strands
holding the passive elements)
X several patterns
c. vannerie/etoffe tissee (weaving: active elements are woven
through passive elements)
X several patterns
Table 3-11 Representation of the classification of Leroi Gourhan, 1943
Balfet supplements and re-organizes Leroi Gourhan's w o r k from a

hierarchical classification into a dimensional paradigmatical classification. S o m e of
these are filled, others are empty. She combines properties of the montants
(warp), with properties of the brins (weft).
T h e dimension which contains properties of the warp, has five attributes,
which are defined by criteria of activity and orientation of the strands (see Table
3-12).
3 Classification 45
1 Passive and consisting of a number of parallel strands in one direction

2 Active and consisting of a number of parallel strands in one direction
3 Passive and consisting of a number of parallel strands in 2 or 3 directions
4 Passive and consisting of one strand which is spiraled
5 Active, and consisting of one strand
Table 3-12 Properties of the warp, Balfet 1952
These five groups of warp-possibilities are cross-linked with 13 weft-possibilities,

divided into three groups (Table 3-13). T h e first group is that of the linked
techniques (lie), which has five subgroups (A 1-5). T h e second group is m a d e up
of twined techniques (corde), with four subgroups (B 6-9). The third group is that
of w o v e n basketry (tisse) which is subdivided in four groups (C 10-13).
A. lie (linked techniques) /. route (wound)

2. cousu (sewn)
3. tourne (wrapped)
4. demi-clef (looped)
5. noue (knotted)
B. corde (twining techniques) 6. simple (simple)
7. croise (even twill)
8. serge (uneven twill)
9. tresse' (waling)
C. tisse (woven basketry) 10 Wile (plain, or tabby-weave)
11 croise (even twill)
12 serge (uneven twill)
13 satin (satin weave)
Table 3-13 Properties of the weft, Balfet 1952
T h e result is 5 x 13 = 65 combinations which form classes of which s o m e have

been filled, others are empty. Balfet introduces the techniques that are k n o w n to
exist, with schematic drawings. In the accompanying text she gives a description
of the different classes and indicates the variations that occur within the classes.
Although the properties of warp and weft are cross referred in a paradigmatical
classification, the classes have not been based consistently on formal criteria. The
warp and weft possibilities that Balfet includes are a selection for which the
criteria have not been m a d e explicit. Since her classification sets out to be a
supplement of Leroi Gourhans' work, w e might expect the criteria to be
technological, but nowhere in the text is this confirmed. She w a s very explicit
about rejecting the idea of classifying according to function or appearance (Balfet
1952, p. 261), but judged the latter important enough to introduce a 'side-line
classification' in which she m a d e a distinction between widely spaced basketry,
closely spaced basketry and basketry which is widely spaced in one direction,
closely spaced in the other.
Although this scheme is a 'real' classification, it certainly is not 'the'
basketry classification, as Balfet designed it to be. She does not put forward a
research question for this classification, but she does give the objective: "la
necessite de n o m m e r les types, et ce faisant de tenir compte de la terminologie
traditionnelle, sans toutefois sacrifier le besoin d'un cadre logique et coherent de
classification" (Balfet 1952, p. 260).
Because of her wish to take existing divisions and terminology into
consideration, she rejects Lehmann's system. Her most important criticism of
L e h m a n n is that his classification is not very useful, because of the formulas he
uses instead of developing a terminology. Her decision to develop her o w n
classification, rather than using for instance Mason's, is based on her criticism that
existing classifications are restricted to a specific geographical area and,
furthermore, that they are an orderly grouping rather than a proper classification.
Balfet's classification includes empty classes, but she does not utilize this
feature. She seems to be interested in the classification itself but not in its
consequences, because she does not ask herself w h y certain classes are filled and
others are empty.
3.4.5 Seiler-Baldinger and Keppel

In the book of Seiler-Baldinger, it is again Buhler w h o gives a clear account of the
basic assumptions of the systematization. The criteria are determined by
technology, and not technique, as is the case in Emery's book (1961). Buhler:
"Wie schon aus d e m Titel ersichtlich ist, geht unsere Systematik primar von den
Techniken, das heisst von den Herstellungsverfahren aus. Sie unterscheidet sich
damit grundsatzlich von der Gliederung Emery's, die in erster Linie auf den
Strukturen, auf der Beschaffenheit fertiger Textilien beruht." (in: Seiler-Baldinger
1973, p. X X V I ) A s shown in section 3.2 this is done in order to be able to
determine the level of development of different techniques.
Seiler-Baldinger's classification starts with a division into primary and
advanced techniques. The criterion for this opposition is the complexity of the
instruments (read: looms) to produce the fabric. Since all basketry techniques are
classed in the 'primary' section only these will be considered here. Weaving is
considered an "advanced technique", since a loom is needed, and thus does not
occur in Table 3-14.
3 Classification 47
•
A. Single Continuous Thread
a. mesh formation with threads of limited length
I linking Einhangen 6 sub-groups
11 looping Verschlingen 9 sub-groups
III knotting Verknolen 11 sub-groups
IV meshwork lace Maschenstoffspitzen 2 sub-groups
b. m e s h w o r k with single threads of unlimited length
I crocheting Hdkeln 4 sub-groups
II knitting Stricken 3 sub-groups
c. transition to plaiting
I knotted plaits Flechtknoten 2 sub-groups
B. Systems of Threads
a. half-plaiting
I splitting Durchstechen des einen Systems no sub-groups
II wrapping Wickeln 2 sub-groups 1
III binding Binden 4 sub-groups
IV coiling Wulsthalbflechten 7 sub-groups
V twining Zwirnbinden 4 sub-groups.
b. transitory forms to plaitir g and advanced techniques
c. real plaiting or braiding
I plaiting 2 directions Flechten in 2 Richtungen 5 sub-groups
II multi-directional Flechten in 3 und mehr 2 sub-groups

plaiting Richtungen
Table 3-14 Outline of the classification of Seiler-Baldinger, 1973 / 1977
The primary techniques are divided into two groups: fabrics m a d e from one
thread, and fabrics m a d e from a system of threads. These two classes are divided
into different sub-classes and sub-groups, ordered according to complexity.
N o real consideration is given to what the exact differences are between the
techniques classified in sub-groups. The 4 sub-groups of twining (BaV), for
instance, are discerned by: twining with two active strands around a passive
system, twining with two active strands around two passive systems, twining with
three or m o r e active strands, braided twining with three strands.
48 The W o r l d according to Basketry
When using the same criteria in a dimensional approach, one dimension

would contain the 3 attributes of the active strands (twining with two active
strands, twining with three or more active strands, braiding), and the other
dimension the 2 attributes of the passive strands (one passive system, two passive
systems). Cross-linking the two dimensions would result in 3x2 = 6 classes, rather
than the four classes that Seiler-Baldinger discerns. Furthermore, the criteria she
selects are too limited to describe all twining variations.
T w o things m a y be clear. Even though Seiler-Baldinger's system is an
attempt at classification, it is a hierarchical grouping. Secondly, the criteria chosen
are not the only possible ones. Other criteria would lead to a quite different
classification.
3.4.6 Emery, Larsen and North

In Emery's work it is the actual structural make-up of the fabrics that is the
classificatory feature: "The details of structure provide a sound factual basis for
more comprehensive description and, being determinable data, for comparative
studies and for classification" (Emery 1980, p. xi). The phrase "being
determinable data", probably refers to data that can be standardized. E m e r y
standardised techniques by simplifying them to structures and classifying those.
E m e r y does not clarify w h y looking just at structures is more "determinable" than
looking at other things.
Using the structure as the central feature in classification means that the
production method, the w a y these structures have c o m e into being, is not
considered. Emery is not consistent in this. She involved the edges of some
techniques, while for other techniques she does not look at the edges at all. Clear
examples are the techniques of 'oblique interlacing with one set of elements',
which have exactly the same structure as the 'plain weaves with two sets of
elements'6 The only thing that sets them apart is the production process, which is
exactly what Emery wants to avoid dealing with.
Table 3-15 shows the core of Emery's classification. The criteria she uses are
not made explicit. W e could interpret the criterion used at the R o m a n N u m b e r
level as 'physical properties of fibres' (having the possibility of getting stuck
together, or being woven). The criterion used at the level indicated by capital
letters consists in fact of two criteria: the number of elements, and the element
being used singly or as a set. The levels below that, are not grouped by
dimensions, but indicate a number of variations. Looking at her classification
makes clear than once again basketry is ordered by hierarchical non-dimensional
grouping.
Compare for instance the following photographs in the book (Emery 1961):
figure 74 and figure 90; figure 72 and figure 85; figure 75 and figure 115.
3 Classification 49
I. felted fibers
A. natural plant forms
B. agglomerated fibers
II. interworked elements
A. single element
1. linking 2 sub-groups
2. looping 3 sub-groups
3. interlooping 2 sub-groups
B. two single elements
1. linking stitches
2. looping stitches (coiling)
C. one set of elements
1. interlinking (plaiting)
2. oblique interlacing (braiding)
3. oblique twining, single and double
4. interknotting (macrame)
D. two or more sets of elements
1. interlacing warps and wefts
a. simple weaves
b. compound weaves
2. interacting elements
a. crossing and re-crossing
b. twining
3. wrapping wefts
Table 3-15 Concise presentation of Emery's classification (Emery 1961)
Larsen classifies according to a mixture of construction (technique) and

production (technology), although he insisted on dealing only with thefirstsubject.
"The classification (...) is concerned with the essential orders of interworking, not
with the methods of achieving them. "(Larsen 1986, p. 34)
Larsen gives the s a m e reason as Emery: a classification based on production
method breaks d o w n w h e n the production method cannot be determined.
According to Larsen this is often the case with archaeological material and
sometimes also with ethnographic material (Larsen 1986, p. 36).
The principles which form the basis for Larsens' system are: "count of
elements x orders of interlacing x modifiers (including materials and color effects)
x form (including size). It is this multiplication that produces variants numerous as
the stars, with many as yet untried" (Larsen 1986, 50).
1 count of elements 5 attributes

1 element A structure in which one strand circles about to
interwork with itself.7
2 elements T w o interworking strands
1 system of elements A structure in which all strands interwork in the same

manner
2 systems of elements T w o interweaving sets of elements
3 systems of elements Three interweaving sets of elements
2. orders of interlacing 6 attributes
over 1 under 1
over 2 under 2
over 3 under 3
over 1 under 3
over 1 under 4
patterned designs
3. materials and modifiers n u m e r o u s attributes
material size of the elements, count of elements per centimeter
(fineness), shape of the raw material, longitudinal
profile, basts, solidity, flexibility, twist, resilience
and grouping
structure density (open/close), balance, tension, take up, pitch.
color effects
4. form 4 attributes
single unit
linear
planar
3-dimensional
Table 3-16 The attributes of the four dimensions of Larsen's classification of interlacing
(Larsen 1986).
Larsen includes knots, but excludes knitting and crochet, because these
techniques involve looping, which is not the subject of a book on interlacing (Larsen 1986,
jy).
3 Classification 51
This suggests a dimensional classification, with four dimensions. The expression

"numerous as the stars", indicates, however, that not every technique has been
described.
The first dimension (count of elements) has five attributes, the second (orders
of interlacing, which stands for the plait pattern) has six and, the fourth (form) has
four attributes. This results in a classification with 120 classes (5 x 6 x 4), in
which the third dimension (materials and modifiers) has not yet been included,
because the number of attributes has not been given. The third dimension is the
difference in appearance of two objects, which are m a d e in the same technique.
The criteria involved here are innumerous in itself and divided into two sub-groups
focusing on the material as modifier, which involves size of the elements, fineness
(count of elements per centimetre), shape of the raw material, longitudinal profile,
basts, solidity, flexibility, twist, resilience and grouping. The second modifier is
the structure, involving the density (open/close), balance, tension, take up, pitch
and colour effects (Larsen 1986, 44).
In order to m a k e this potentially enormous classificatory scheme manageable,
Larsen makes two smaller classifications, one based on the count of elements and
order of interlacing, with 5 x 6 = 30 classes, of which 12 are empty. The second
classification is based on the count of elements and the form, with 6 x 4 = 24
classes, of which 9 are empty. In the first classification the dimension 'count of
elements' contains five attributes8, in the second there are suddenly six9.
A system that Larsen presents in a separate chapter is a hierarchical
dimensional classification. The first level (Roman numbers) refer to five varieties
in the count of elements10. At the next level Larsen discriminates the direction of
interlacing (single, linear, planar, 3 dimensional) and the structure (simple,
complex, compound). Every class is thus constructed by the same criteria, which
is typical for a dimensional classification. Larsen does not implement this system
consistently, because some classes are ruled by n e w criteria. The classification has
no empty classes. Larsen does not include technical combinations which he
supposes do not exist.
Although North was interested primarily in technology, the way in which
artefacts are made, she gave an overview of techniques mainly based on Emery,
considering the structure, rather than the method of making that structure.
Five kinds of 'count of elements': 1 element, 2 elements, 1 System Of Elements

(SOE), 2SOE, 3 S O E (Larsen 1986, 56)
Six kinds of 'counts of elements': 1 element and 2 elements are considered one
attribute, the other five are: 1SOE, 2SOE, 3SOE, 4SOE, 6 S O E (Larsen 1986, 66).
The five kinds of 'counts of elements' in the final classification are: 1E+2E,
1SOE, 2SOE, 3SOE, 4 or more SOE.
52 The World according to Basketry
3.4.7 Crowfoot, Lucas and Forbes

Crowfoot's classification is implicitly based on the technical structure. She is
influenced by Vogt, the third edition of Lucas and by Buhler and Buhler-
Oppenheim. Crowfoot discerns six major techniques, but uses no explicit criteria
to do so. She then proceeds by dividing these into sub-groups.
....
Coiling 3 stitch types 3 centre variations
Twining 5 methods of spacing
Wrapped work 1 type

Matting work 3 weaving patterns
Plaited work 1 type
Wicker work 1 type 4 centre variations
Table 3-17 Representation of the classification as put forward by Crowfoot (1954)
Coiling is divided into four sub-groups, by the manner in which the winding
element holds the coil. Furthermore she discerns three centre types. Twining
consists of five sub-groups, based on the spacing of the rows of twining. The class
of "wrapped work" contains only one type, while three types of matting work are
discerned by their weaving pattern. Furthermore, she lists one type of plaited work
and one type of wickerwork with four centre variations (Crowfoot 1954). It is
another example of grouping, rather than classifying material and the criteria are
not used consistently.
Even though Forbes gives a general classification of textile techniques, based
on Buhler-Oppenheim (Forbes 1964, 167, 177), he follows Crowfoot's
classification of basketry techniques exactly. H e does not intertwine the two
systems, which follow different criteria. Forbes seems to have no difficulty in
using an inconsistent combination of two inconsistent classifications.
3.4.8 Gourlay
Gourlay 's description of the basketry of Deir el-Medina is quite detailed with
proper attention for important feature such as centres, border and handles, as well
as decoration and form. His catalogue of basketry techniques, however, is not a
proper classification. Instead of ordering his material according to technical
criteria (which would seem the most logical approach for a catalogue of
techniques), he makes a main ordering according to the function of the basketry
objects. A s m a y be clear from Table 3-18 he divides the objects into brushes,
brooms, mats, sandals, baskets, etcetera.
F r o m this main division he starts a hierarchical grouping employing different
criteria for each group and even within one and the same group. T h e clearest
indication that w e are not dealing with a proper classification is class V , containing
3 Classification 53
miscellaneous objects. 'Class' and 'miscellaneous' are a contradiction in terms,

because a well formulated set of criteria encompasses all objects.
I Les balais de menage, (brushes, brooms)

Four types (A, B, Bl, and C ) discerned by technology, in which type B is
divided in B and Bl and type C is considered a variation of type A
II Les brosses et les pinceaux (paint brushes)

Seven types (A, B, A B , C, CI, D, Dl) discerned by technology (the way
the were made).
Ill Le garnissage ou le revetement de meuble (matting in furniture)
Five types of bed or chair matting (A-E). In this case the types are
determined by the weaving pattern. A separate section is devoted to one
type defined by material and technique.
IV Les cordes et les noeuds (strings and knots)
There is no classification of rope making techniques. Thirteen knots or
knot-constructions are listed.
V Divers
A Nine types of filling material for wigs (A, B, Bl, C, CI, C 2 , D, E ) ,
discerned by diferent methods of production
B T w o types of plaited palm leaf with unknown function (A, B), discerned by
the size of the palm leaf
C Five types of raw materials (A, A l , B, C, CI) discerned by different
rolling and folding methods
VI Les nattes (mats)
Three types of mats have been discerned on the basis of two different
techniques: A = twined, B = woven, C = twined (but with different
materials).
VII Les sacs, les resides et les filets

A Bags and nets are divided into eight groups: A, A B , A C , A D are varieties
of twined bags, group E are handles, F represent coarsely knotted nets, G
is a twined bag, but made of different materials than A - A D and H is a
plaited bag.
B There are two types of carrier nets, type A made of string, type B made of
palm leaf.
VIII Les sandales (sandals)

sandals made of rope: there are six types, of which B, C, D, E and F are
presented as a variation of type A. The differences are formed by technical
details
sandals made of leaves: there are eleven types and variants discerned
according to technical details:
A are plaited sandals, Al are plaited sandals with leather rims,
B are plaited sandals with sewn strenghtening rims,
C, CI and C 2 are plaited sandals with flexible wooden soles,
D are plaited sandals with sides,
E are coiled sandals,
F are sandals of parallel leaf strips,
then there is an unnumbered variety on sandal type A, with a curled toe
F, which is a varient on the previous type, with a straight toe
IX Les anneaux et les erseaux (rings)
classified according to use and technique
X La vannerie (basketry), in total there are 27 types
A, A2, A3, A4, A5, A6, B, B2, C, D, Dl, E, El, F, Fl. G, Gl, H, I,
represent variants of the centre of coiled baskets
J and Jl are twined centres of sieves with a coiled rim
K are sewn plaits baskets
L and LI are oval coiled baskets with different types of centre, while L2 is
a centre variant of a round coiled basket.
Q and Z have not been described
Table 3-18 Summary of the classification of Gourlay (1981).
The kind of information is unbalanced, because sometimes a class is defined

by a brief description and a drawing, while in other instances an extensive
disquisition on the production method of one object is given. Although these are of
great interest, they do not belong in a catalogue of techniques.
The section on basketry ('X') contains three catalogues, based on
respectively techniques, shape and decorations. Here w e are concerned with the
classification of techniques only and, therefore, the two other catalogues will not
be considered. The technical types A-I and L are all coiled baskets with different
centres, every type being sub-divided according to the variation that occurs in the
centre of the baskets. Type J is a coiled sieve with a twined base, type K is a sewn
plaits basket. The basketry is thus grouped according to centre construction, rather
3 Classification 55
than wall construction. In principle this could be a valid approach, if the purpose
and the criteria are made explicit. Does Gourlay propose that the difference in
centres has a specific meaning or relevance? N o , he does not and, furthermore, he
does not m a k e clear w h y and along which criteria the subdivisions are made. T w o
of his classes, 'Q' and 'Z' have not even been described in the catalogue.
3.5 Conclusions and consequences

After having made an inventory of the existing classifications of basketry, it is
clear that with exception of those of Balfet, and to a certain extent Larsen and
Adovasio, all classifications are in fact groupings in a hierarchical order. S o m e
authors use explicit criteria, others do not. I could not agree more with Adovasio
w h e n he writes: "It is interesting that lack of systematization and taxonomic
precision continued to characterize basketry research, despite the publication of an
updated version of Leroi Gourhan's classificatory scheme by Balfet in 1952"
(Adovasio 1977, 4) and I regret to see that even after this remark his o w n classifi-
cation still lacks the systematization that characterizes Balfet's work.
The reason for this is quite obvious, and has been suggested by the authors
themselves. Adovasio wrote: "Combining all attribute of wall construction can
generate a very large number of (...) types" (Adovasio 1977, 62). Larsen indicated
the same problem w h e n he remarked that it is the multiplication of a large number
of attributes "that produces variants numerous as the stars" (Larsen 1986, 50).
Apparently m a n y attributes are necessary to make a basketry classification. The
cross reference of all those attributes results in a number of classes, that is too
large to be of any use.
It is exactly in the definition of the use, or the purpose, that the power of a
classification can be determined. In the previous sections an inventory was m a d e of
the purpose, limits, research question and criteria for basketry classification as
used by different authors. There appears to be a great deal of variety, because all
classifications have their specific point of view. This is only clear, however, when
the purpose, research question and criteria of the classification are made explicit.
A good example can be found in the work of Emery and Seiler-Baldinger. The two
classificatory systems are both dealing with world wide textile techniques, but
from a different point of view, namely technique (structure) and technology
(production method). The two systems should not be considered as rivals, but as
supplementary.
There are at least three solutions to the problem of the myriad of possible
classes, all of which have been used by the authors whose work has been looked at
in the previous sections. The first solution is to make a paradigmatic classification,
but to eliminate the empty classes. This is what Larsen did w h e n he decided to
give examples of those techniques k n o w n to exist. Adovasio chose to make "a
representative selection of c o m m o n (...) types for illustration" (Adovasio 1977,
62), instead of presenting all theoretical possible techniques.
The second solution is to m a k e a hierarchical, rather than a paradigmatical
classification. By making subdivisions that are branching out like a tree, different
criteria are used for the separate branches. A taxonomic structure seems to be an
organic solution and most authors have followed this method in one form or
another.
The third solution is to m a k e a paradigmatical classification based on a
limited number of attributes, by selecting criteria that reflect explicit research
questions. Balfet follows this line, without, however, formulating criteria based on
a research objective.
All three solutions can be criticized. The first and the second solution both
result in grouping, rather than classifying. Neither solution results in a
classification with empty classes. Eliminating empty classes is throwing out an
important source of information, since the question w h y certain combinations do
not occur can be as important as the explanation w h y certain combinations do
occur.
The third solution can be criticized, because it gives a very selected and
limited image. Balfet's classification is vulnerable to this criticism, because she
presents her work as 'the' classification of basketry, without making its limitations
explicit. A s long as the classification is linked to an explicit research question,
however, this method is valid. The image m a y be limited, but it is a clearly
selected subject of which the limitations are m a d e known.
In Chapter 9 a classification is presented according to the properties of the
basic structure of basketry techniques." This classification is used as a basis for
understanding the production process of the techniques found in A m a r n a , Qasr
Ibrim, present day Middle Egypt and N e w Nubia. Twenty techniques have been
discerned. The numbering from 1 to 20 has no meaning, other than that it is a
convenient w a y to refer to these techniques throughout the book.
The classification presented in Chapter 9 is a w a y of ordering things, by
simplifying the aspects associated with them. T o o m u c h simplification results in a
misrepresentation, but too detailed a classification is as complex, chaotic and
many-sided as life itself and, therefore, powerless to clarify. It is the task of the
scholar to select the level of simplification necessary for handling existing, and
generating n e w questions. It is important to realise that what is represented is a
thought experiment, a w a y of turning the subject in one's hands and looking at it
from different angles. Ultimately a complete image should be generated, a
narrative that tells us more than the individual pieces can.
The function of classification in the next chapters is to get a grasp on the
subject of basketry technology. It should be realised, however, that it is not more
than a tool, which bluntly cuts off other aspects that are important for the
interpretation of basketry production. These missing facets are taken up by looking
at basket makers as well as at baskets (Chapters 16-19).
The term basic structure will be introduced in Chapters 4 (p. 58) and 5.
CHAPTER FOUR
FIELD WORK METHODS
4.1 Introduction
Fieldwork involves constant decision making. T o keep this procedure of choice
and selection controllable, this chapter accounts for the fieldwork methods used
and the problems that occur. The interpretation of basketry is based on information
gathered during field work and post excavation analysis. Because the latter is
determined by the same research questions as field work, there is a constant
creative interaction between observation, analysis and interpretation. For
archaeologists and anthropologists w h o are writing up their work away from the
field this is noticeable in the often felt need to go back into the field to 'check'
things while formulating the results.
Over the four years that the field work took place there has been a shift in
focus. In 1989 and 1990 the archaeological work concentrated on what could be
indicated as the static aspects of basketry: the primary goal was to understand and
clarify the basketry techniques and the materials used. Attempts were m a d e to
discern the production process, by deducting the working order from a detailed
study of the edges and other technical details.
F r o m 1990 onwards an increasing emphasis has been put on ethno-
archaeological research. At the onset, the main reason for this w a s the apparent
similarity of the ancient and modern basketry. B y using the same method in
recording the archaeological and ethnographic baskets, the similarities are m a d e
explicit, while the dissimilarities are m a d e apparent. Based on this inventory other
aspects, related to the production process and the identity of the basket makers, are
involved. These are, for instance, the raw materials and instruments, the
organization of time and space as well as the professionality, gender and social
status of basket makers.
T h e object of study differs for the archaeological and ethno-archaeological
research. Participating in archaeological field work involves recording the basketry
objects and their archaeological context to contribute to the interpretation of the
archaeological site. The object of the ethno-archaeological research project is the
production process and the context of basketry production. This involves the basket
makers, the place and circumstances of their work, their contacts and economic
organization.
T h e attempt to use questions that are partly designed for ethno-archaeological
research in the study of archaeological basketry opens n e w directions of basketry
analysis. Furthermore, it gives the possibility to use the ethno-archaeological
material not only in analogous reasoning, but also as a confrontation with the
archaeological results.
The fieldwork methods for the research on basketry production involve
recording the archaeological and present day basketry and studying their respective
contexts. Section 4.2 deals with recording archaeological basketry, which involves
a description of the technology, measuring and drawing the shape, identification of
the raw materials, and taking note of the wear, damage and repairs. A detailed
description of the technical analysis of basketry technology has been incorporated
in Chapter 5. M o r e can be found on the subject of recording basketry in W h o is
Afraid of Basketry? (Wendrich 1991c), which deals also with aspects such as
condition, treatment and storage. Although these aspects should be recorded during
field work, they are part of finds processing and fall outside the scope of the
present book. The method used for recording the production process is taken up in
section 4.3. The selection and use of audio-visual means have been accounted for
in section 4.4, while an account of the analysis of video images is elaborated on in
Chapter 6. The problems and pitfalls encountered while doing field work are
related in section 4.5.
4.2 Recording Basketry objects

4.2.1 Technology
The method used to record the basketry techniques featuring in this book, is the
same for archaeological and ethnographical basketry. For each basket or fragment,
five aspects are recorded, as shown in Table 4-1.
1 Basic structure
2 Beginning and end of the basic structure
3 Incorporation of raw material
4 Construction of the shape
5 Decoration
6 Additional features
Table 4-1 Features recorded to describe a basketry technique (cf sections 4 2 1 1 to

4.2.1.6)
The five features mentioned in Table 4-1 have not been listed in the order in which
they appear during the production of a basket, but follow the order in which a
technique is recorded. They will be explained briefly in section 4.2.1.1 to 4.2.1.6.
4.2.1.1 Basic Structure
The basic structure is the basketry fabric of the base or the sides of a basket It is
considered the central feature in basketry recording and analysis, because it is the
part of the basket that is most likely to survive. Even from the smallest fragment
4 Field W o r k M e t h o d s 59
of basketry the basic structure can be recorded. Baskets can be made up out of
several different basic structures, for instance, the base can be made in a different
technique than the sides.
T o compare basketry techniques describing each according to the same
parameters is important. The central concept for this is system, which is a number
of strands which form a fabric by interacting with strands of another basketry
system. Stake-and-strand basketry, for instance, consists of two systems: a skeleton
of stakes, which are woven in with strands.
All basketry techniques can be described as systems that interact in a certain
way. Each system consists of one long strand or several parallel strands, indicated
as elements. The elements of a system are defined as: all strands m a d e of the same
material and with the same orientation and function in effecting coherency in a
basketry technique.
In order to record the basic structure it has to be decided h o w m a n y systems
there are and which strands belong to each system. Secondly, the exact
composition of the systems should be defined and, finally, the interaction between
the systems has to be described. This procedure will be explained and illustrated in
detail in Chapter 5.
4.2.1.2 Beginning and End of the Basic Structure

The basic structure starts at some point, usually the centre of a basket or the
bottom edge of a mat where the strands which form the systems are fastened.
After the main body of the basket or mat has been made, the strands are finished
off. This is usually at the rim of a basket or the top edge of a mat. The centre,
base, sides and rim sometimes show different basic structures, made with the same
continuing strands. These strands m a y run throughout the basket, from centre to
rim, but their function and interaction changes.
Figure 4-2 illustrates the patterns and change in systems in a two dimensional
reproduction of the base, side and rim, as seen from the inside of a sixth century
A D stake-and-strand basket found at Qasr Ibrim. The centre is made of eight
strands, each of which consists of six parallel stems of a species of rushes. The
strands are layed out crosswise and are held in place by two strands which twist
around each other (indicated with number 1 in Figure 4-2).
This technique, k n o w n as twining, is continued in the base. The bundles
which radiate from the centre are split into groups of three parallel strands in order
to fan out and accommodate for the expanding diameter of the base (indicated with
number 2 in Figure 4-2). All strands are fastened off at the end of the base and
n e w elements are inserted to m a k e the side (indicated with number 3 in Figure 4-
2).
Before the sides are plaited, just above the base, a row of waling with three
strands is inserted to put the elements of the side in their proper position (cf.
section 5.1.4.3). The sides are plaited with six parallel rushes in an over 1 I under
1 pattern (indicated with number 5 in Figure 4-2). All strands are fastened off at
the rim, but different methods are followed for the strands which run from right to
left (indicated with number 6) and those running from left to right (finished off at
7 in Figure 4-2).
The rim of a basket, or the edge of a mat, has a double function: it is both
the end of the elements which form the basic structure and the border of the entire
object. The techniques forming the rim, therefore, have to safeguard the coherency
of the object and the strength of the rim itself, since the object and its contents are
often lifted by the rim, or by handles that are anchored in it. Determining the
beginning and end of a basic structure requires insight into the production process.
4.2.1.3 Incorporation of Raw Material

Because baskets are m a d e of materials of limited length, such as strips of palm leaf
strips or grass stems, n e w lengths of material have to be inserted regularly, during
the production of the basic structure. Basket makers also have to compensate for
the irregular shape of their raw materials. Sometimes this involves a separate
production phase to make a more uniform material of unlimited length. Grass is,
for instance, made into string and palm leaves are plaited into long strips.
If such measures are not taken, the limited length of the raw material makes
it necessary that new leaves, stems or twigs are inserted throughout the production
of the basic structure of the basket. A n important feature is the w a y the material
has been inserted and oriented. Willow branches, for instance, have a thick and a
thin end, which have to be distributed evenly over the sides of the basket (see
Figures 3-4 and 3-5 p. 36). The distribution, method of inlay and the degree of
consistency are noted as part of the recording system.
In commonly occurring basketry techniques with little variations, such as
coiled basketry, distinctions are made mainly by the details of adding raw
materials, or the w a y the baskets were started and finished off. Furthermore,
studying present day basket makers reveals that the solutions for technical
problems, such as the insertion of raw materials, are very specific and passed on
from the basket maker to his or her pupil. A detailed study of the insertions and
exhaustions thus gives the best possibility to discern basketry traditions. In these
details w e find a knowledge that is handed d o w n to the next generations, which
forms a clue to tradition.
Figure 4-2 (on page 61) Schematical representation of the inside of a stake-and-strand
basket found at Qasr Ibrim (fourth c. A D ) . Seven parts can be distinguished:
1. The centre, which is also the start of the two systems used in the base.
2. The basic structure of the base, consisting of Z-twining.
3. Transition base-side: the two systems of the base are ended and two
systems for the side are inserted.
4. R o w of waling in order to position the elements for the side.
5. Basic structure of the side (continuous plaiting in \1/1\\1 pattern).
6. Rim: exhaustion of the elements of the first active system of the side.
7. Rim: exhaustion of the elements of the second active system of the side.
4 Field Work Methods
Since the basket makers try to make the transitions of old and n e w strands
invisible, by working them into the fabric as smoothly as possible, this important
feature cannot always be recorded. Unless the baskets are damaged, this clue to
tradition often remains hidden.
4.2.1.4 Construction of the Shape

The shape of baskets is formed gradually while creating the fabric. Since it is an
integral part of the technique, the method to construct and maintain the shape of
the object is recorded in direct connection to the basic structure. In most cases
baskets are shaped by a radical change in angle between the base and the side.
Therefore, special attention should be paid to the transition between the basic
structures of base and side.
The range of variation is considerable and dependent on properties of the
technique and the raw material. In some techniques, such as coiling, and sewn
plaits basketry (cf. Plate 4-3), the transition is done gradually.
Plate 4-3 Sewn plaits basketry: coloured date palm leaf strands are incorporated in a
plaited strip, which is sewn with strips of unspun doam palm leaf to form a
basket (New Nubia, 1992).
In stake-and-strand techniques the base is often finished off and n e w material

inserted to m a k e the sides at the desired angle. This can be seen, for instance, in
Figure 4-2 under 3. The transition of base to side has been strengthened by a row
waling, a technique using three strands, as depicted in Figure 5-12 (f) and (g)
(p. 96). Waling is often applied in shaping stake-and-strand baskets, since the
multi-membered (e.g. three-strand) elements are stronger than single-membered
elements and are used to force the elements of the side in place.
In s o m e objects the side is m a d e in a different technique than the base. This
occurs generally in sieves, where the base is an open trellis work and the sides are
m a d e in the coiling technique in order to provide a strong rim. W h e n the basic
structures of the base and the sides differ, as is the case in a twined sieve with a
coiled rim, the transition between base and side involves the ending of the strands
of the twined trellis and the start of the systems making up the coiled sides.
4.2.1.5 Decoration
Often decoration is applied while making the basic structure, for instance by
inserting regular pattern changes, or by using coloured elements. In present day
Egyptian Nubia palm leaves are dyed and incorporated in plaits which are sewn
into baskets and mats (Plate 4-3). Ancient as well as modern Nubian coiled
basketry is given a decorative pattern by using coloured wrapping strands (Plate 4-
4), a method applied also widely in present day Egyptian Nubia.
Plate 4-4 Decoration of coiled basketry as part of the basic structure: coloured
wrapping strands are used to make a pattern, Qasr Ibrim, fourth c. A D ,
(Courtesy of the Egypt Exploration Society).
The W o r l d According to Basketry
64
Another type of decoration which is made as part of the basic structure is a

monochrome pattern of stitches of different length. The coiled bundle is fastened
provisionally with a first layer of wrapping, which is then covered with a second
layer of stitches wrapping around respectively one, two, three, four and five rows
of coiling (Plate 4-5).
In specific Nubian villages in the vicinity of A s w a n , a decorative technique,
indicated as imbrication is used nowadays: coiled basketry is decorated with
coloured strips of straw, which are fastened on top of the coiled surface. The strips
cover (part of) the surface in an overlapping manner which is reminiscent of
roofing tiles (Plate 4-6, Figure 9-17, see p. 166). '
Plate 4-5 Decoration of coiled basketry, making use of wrapping stitches over several
coils. Qasr Ibrim, fourth c. A D , courtesy of the Egypt Exploration Society.
' Douglas (1935) named this decorative technique after imbrex, the Latin word for tile
(cited by Adovasio 1977, p. 94, figure 113).
4 Field Work Methods 65
Plate 4-6 Decoration of coiled basketry, making use of coloured straw, fastened on top
of the surface as part of the basic structure. The covering stitches have the
appearance of tiles and are hence referred to as imbrication (cf. Figure 9-17,
see p. 166, N e w Nubia 1992).
4.2.1.6. Additional Features

Features are considered additional w h e n they have been applied separately after
the main body of the basket was completed. These include the handles, foot and
separately inserted decoration. Lids are often attached to the basket, for instance
with a number of stitches, but are better considered and recorded as separate-but-
connected baskets, rather than additional features.
Decoration is considered an additional feature only if applied after the
production of the basket, for instance the application of a decoration with paint.
Further examples are decorative stitching (comparable to embroidery) and the
attachment of shells, beads and buttons. Most Egyptian coiled basketry, however,
is decorated with coloured winders or by decorative stitching which is applied
while making the basket and thus part of the basic structure.
Likewise, strengthening can either be applied during the production process
or afterwards. A n example of the first is the insertion of an extra string at the edge
of a mat, which is recorded as part of the basic structure. Inserting a round of
waling is another example of strengthening within the basic structure.
Strengthening is an additional feature only w h e n it concerns a secondary
application, for instance in the form of stitches which anchor the base into the
sides. Sometimes it is difficult to distinguish secondary strengthening from
decoration, or, w h e n applied sloppily, from repairs. Additional features are
inserted at the end of the production process and are clearly part of it. Measures
taken during the baskets' lifetime, to strengthen or repair, are considered in section
4.2.4.
4.2.2 Size and Shape

The size and shape of the baskets are measured and drawn to scale. It is important
not only to measure the object, but also its different components, such as the
diameter and length of the individual strands. A detailed description of the method
used for measuring the objects that feature in this book can be found in Wendrich
1991c (pp. 67, 88-92).
4.2.3 Identification of the Raw Materials

Determining the plant species and the part of the plants from which the baskets
were made is important for several reasons. In respect to basketry technology both
aspects, species and plant part, give information on the original properties of the
material. In archaeological basketry these aspects, such as length, flexibility,
tension strength and colour, have altered over time due to ageing and deterioration.
Knowing the original properties is of importance for understanding the technical
features of the basketry raw materials.
Secondly, the plant materials used provide links to the place of origin of the
baskets and show temporal differences. The 18th dynasty coiled basketry from
A m a r n a , for instance, has been m a d e exclusively of d o a m palm leaf, while the
present day Middle Egyptian coiled baskets are m a d e of date palm leaf.
The identification of the modern materials is done through information by the
basket makers, w h o are usually able to give the vernacular names of plants and
show the places they are collected. The archaeological material can be identified
partly macroscopically, but frequently the analysis of the species has to be done by
microscopic examination of the characteristics of the epidermis.
F r o m the work of Greiss (1957) and Tackholm (Tackholm and Tackholm
1941, Tackholm and Drar 1950, 1954, 1969) it is apparent that the range of plants
used for basketry production in antiquity is limited. Their list of species formed the
basis for the list of criteria used for fibre identification of the A m a r n a and Qasr
Ibrim material.
Because epidermis patterns are used as the main criterion for identification,
only samples in which the epidermis is still present can be identified. Alternative
methods, such as the identification of cross sections, appeared to be unsuitable. In
the course of the research it appeared that w h e n the preservational state w a s such
that the epidermis had worn off completely, making a satisfactory cross section
was usually impossible. Besides macro remains, which still have fragments of
epidermis to be identified, part of the material consists of fibres only. This means
that no epidermis is present and that, therefore, the only method open to identify
them, is to compare the size and the shape of the fibre cells. T h e method of
epidermis pattern identification will be described in detail in Appendix A .
Since most baskets and mats have been m a d e in a two-system technique (one
active and one passive system), each fragment requires minimally two analyses.
The large amount of finds from A m a r n a and Qasr Ibrim, more than 2000
fragments, renders a microscopic analysis of all finds impossible.
4 Field W o r k Methods 67
A deliberate selection was made to economise, abiding the principle that

some materials could be recognised macroscopically, provided they were in a good
condition. Therefore, the group of objects selected for fibre identification is neither
a representative nor a random sample. Especially if the condition of the fibre is
bad, the macroscopical identifications are sometimes doubtful and occasionally
prove to be wrong. Therefore, no precise percentages of materials used are given.
For the basketry from A m a r n a and Qasr Ibrim a diversified sampling
procedure is followed in which three groups of material are discerned. The first
group are plant materials that can be identified macroscopically. Provided they are
in good condition, the two species of palm trees of which the leaves are used in
basket making can be discerned without magnification. Only a number of control
samples were taken to check the macroscopical identification. Therefore, the
number of samples of palm leaf under represents the actual occurrence of palm
leaf.
The second group are the grasses, which can be recognized macroscopically
as different from palm leaf, but discriminating the two tall grass species used in
ancient Egyptian basketry production it is not possible. During thefirstseason at
A m a r n a a random sampling strategy was followed to gain insight into the ratio of
these two grass species. This strategy was abandoned in later seasons, because the
results did not show a specific pattern and from ethno-archaeological work it
appeared that present day basket makers used both grasses without discrimination
(see Chapter 8). This strategy resulted in an over representation of grass samples.
The third group are the less c o m m o n plant species, such as sedges and
rushes. The ratio of these plants in the sample collection is even more askew, since
most objects m a d e of rarely occurring materials were sampled. The sample
collection is thus not representative for the actual occurrence of raw materials.
The plant species from A m a r n a and Qasr Ibrim have been studied and
photographed on site by the author and identified by Otto Brinkkemper and Edwin
van der Heijden. The results of their work are incorporated in Chapter 8 and
presented in Appendix A.
4.2.4 Wear, Damage and Repair

W e a r patterns are important indicators for the use of basketry. Since most basketry
is found as small fragments, the wear pattern shows, for instance, which side was
the baskets' outside. Repairs are also important indicators for the use to which a
basket was put. For carrier baskets repairs are usually found at the base (renewing
the anchoring of the base in the sides) or at the rim and handles. Storage baskets
are usually repaired at the base, while matting repairs are found mostly at the
edge. Sometimes measures of secondary strengthening can be seen. These are
applications of sewing after the basket has been taken into use, but before the
basket was damaged. Since the present book concentrates on production, rather
than use, the aspects of wear, damage and repair will not be dwelled upon in the
following chapters.
4.2.5 Context
The context of the archaeological basketry w a s recorded partly by m e , partly by
others. In A m a r n a the excavations in the workmens' village, the area which had
resulted in most of the basketry finds, had been finished a few years before I
worked on the material. Checking the field notes and talking to the archaeologists
w h o had supervised the excavations w a s the only means to understand the context
of the baskets.
In Qasr Ibrim the situation was very different, because I had the opportunity
to be present most of the season. This gave the possibility to keep in close contact
with the developments on the site. Participation in excavating the areas in which
basketry was found, enabled m e to examine and record the archaeological context
with specific basketry-related questions in mind.
For the understanding of the production process the context did not produce
m a n y n e w insights, compared with the information from the objects themselves.
The basketry at A m a r n a and Qasr Ibrim was deposited mostly in a state of re-use
or discard. Other items, such as tools and raw materials were, however, of great
use for the interpretation. The excavated spaces did not show any particular
features, wich is not surprising in the light of the results of the ethno-
archaeological work. In most cases, basketry production does not require
specialized work shops, or specific installations (see section 4.3).
4.3 Recording Basketry Production

Understanding starts with participation and participation starts w h e n I ask the
basket makers if they can give m e lessons. This is a useful approach for several
reasons. In the first place, a pupil-teacher relation is very different from a relation
between a researcher and an informant. It enhances the self esteem of the basket
makers and in general gives them more credit for their knowledge and experience.
Secondly, trying to make baskets, rather than just looking at them, is invaluable to
get a feeling for the properties of the raw material and the specifics of the
technique.
Apart from participation, the traditional methods used to gain information are
observation and interviews. During field work these three actions preferably melt
to a natural form of behaviour. W h e n I visit basket makers, often and for longer
periods, I try to sit and chat with them. Although I usually ask m a n y questions
during each session, no formal interviews are held. During the chats, I keep in
mind the list of focusing points and the questions that go with these (see Appendix
B ) , but never use interview forms. Observation continues while the work is going
on. Sometimes I stop talking or listening because something interesting is
happening in the process. This is usually picked up and explanations follow. In
1989 and 1990, however, m y limited knowledge of Arabic forced m e to organise a
special session with an interpreter, to ask detailed and abstract questions.
Observation is helped greatly by audiovisual recording: drawings, photographs and
video, wich are introduced in Section 4.4
The interaction with informants brings the human factor back into
archaeology, not in the sense that the present day basket makers are suddenly
thought to be populating the past, but in the mind and daily schedule of the
archaeologist. The exchange of information with basket makers on ancient and
present production brings n e w questions to the forefront. Ancient basketry
production becomes relinked to time, space and the position of people in a family
and a society.
The production method of ancient basketry is inferred from technical details
such as the edges and insertions of n e w strands. Finding similar technical details in
modern basketry, gives a firm base to using present day basketry production for
interpreting the ancient production processes.
T o understand the production process, it helps to split the process in phases
(cf. Section 4.3.1). Involving the aspect of time is a vital step in understanding
production as a process (cf. Section 4.3.2). Studying the context of present day
basketry production gives information that can be linked back to the archaeological
context, but also on aspects that are immaterial, such as the meaning of baskets or
the social status of basket makers (Section 4.3.3). Although linking such
information directly to the archaeological material is difficult and sometimes
impossible, it is exactly this kind of information that makes archaeology alive.
4.3.1 Production Phases

While observing a basket maker, it helps to organise the information by discerning
production phases. Table 4-7 lists seven phases which occur, although not all
baskets need all seven steps, nor are they necessarily executed by the same person.
For most baskets and mats the raw materials need some preparation before the
production is started (a in Table 4-7). Leaves are cut into strips and often the
materials are soaked in water, to regain their flexibility. The phase of preparing
the systems (b in Table 4-7) does only occur with the more complicated
techniques. In the sewn plaits technique, for instance, both the active system
(string) and the passive system (plaits) are prepared separately. Starting the
systems (c in Table 4-7) involves setting up the basket or mat. Usually this is done
at the centre of the basket or the starting edge of a mat. The interaction of the
systems (d in Table 4-7) is the phase in which the basic structure of the object is
made: the base and sides of a basket or the fabric of a mat. This phase include
shaping the object and inserting n e w lengths of raw material. Finishing off the
systems (e in Table 4-7) is the phase in which the rims or edges are made. The
addition of handles, internal rims, feet and added on decorations is done in a
separate phase (f in Table 4-7). The last phase is the finishing touch, in which
irregularities are removed or the basket is pushed and pulled in shape.
a) Preparing the raw materials (soaking, cutting)

b) Preparing the systems (making string, plaiting a long strip)
c) Starting the systems (beginning of basic structure: centre)
d) Interaction of the systems (making the basic structure: base and sides) 1
e) Finishing off the systems (ending the basic structure: rim)
f) Adding features (e.g. handles)
g) Finishing touch (removing irregularities) 1
Table 4-7 Seven phases of the production process.
There are similarities between Table 4-7 and the list of aspects with which to
record a basketry technique (Table 4-1 on page 59). Recording a basket starts with
discerning the basic structure, while studying the production phases shows that the
basic structure is created somewhere in the middle of the production process
(Table 4-7, d).
4.3.2 Timing
If w e want to k n o w h o w m u c h time a production process takes, it seems to be a
logical step to break d o w n the process according to the production phases as
introduced in the previous section. Although such a schematization is very useful
for understanding the process, in practice it is not a good base for timing. A s will
be seen in Chapter 16, the process is more complicated, than s h o w n in Table 4-7.
For timing the process such a simplification is too coarse.
Analysing the actions and movements gives a m o r e precise tool to time the
production process. It also provides the possibility to look at aspects such as
working rhythm and concentration. For this aspect the video has proven to be an
indispensable tool. A brief introduction to the analysis of moving images will be
given in Chapter 6.
Apart from timing the process itself, it is also important to record other time-
related aspects of basketry production, such as the time of day or the time of year
in which production takes place. The former is determined by other daily duties a
(part-time) basket maker might have. T h e latter bears a relation to the harvest of
raw materials or specific seasonal tasks. Unless a complete year can be spent with
the basket makers, this information can only be gained by asking questions.
4.3.3 Context
The material context of basketry production is observable: instruments are
scattered around the work space, the basket makers' actions leave specific traces.
These observations can be related back to the archaeological context, where
similar patterns might be identified. A good example is the mat loom, which is a
simple construction m a d e of four w o o d e n pegs in the ground. These pegs hold two
cross bars in place, on which the warp is stretched. The archaeological remains of
a matting workshop can be no more than four round discolourations of the soil
where the four pegs once were.
Apart from the material context, there are many aspects related to basketry
production which could be dubbed the immaterial context. These are aspects which
are not readily observable at present and certainly do not leave clear archaeological
traces. The meaning of baskets, or the social status of basket makers are
retrievable by interviewing basket makers. In some cases, they also have a
material component. The Nubian baskets given as wedding presents are, for
example, used to decorate the guest room. The location of the baskets and the care
which is taken in hanging them on the wall, is indicative of their special status. For
archaeological finds a special status can be proposed on the basis of the
archaeological context, but the exact nature cannot be retrieved. Ethno-
archaeological examples, however, give possible explanations.
4.4 Audiovisual Means

Photographs and drawings are used widely in archaeology as recording aids. They
represent a given situation in a seemingly objective fashion and thus are thought to
constitute proof of the writers' interpretation. This is not an unproblematic claim,
as will be discussed in section 4.6. In ethno-archaeology, which deals with a world
that moves, the video camera is an obvious tool. It adds not only movement, but
also a sense of time to the representation.
True to the adage that 'one picture says more than a thousand words',
drawings, photographs and video are used to support or replace descriptions. The
importance of using images rather than words, however, is not only that a
description of a basket, a production process or context requires a large number of
words, or that the result is often awkward and unclear, but mainly that pictures tell
a different story than words.
A n important feature of visual means is that it immediately draws the
readers' attention, which is m u c h more difficult to achieve by text alone. The
advantage of text is that certain aspects can be emphasized while a photograph,
although it is a direct call for attention, and determined by the eye of the
photographer, still leaves the viewer to interpret the message. This can be solved
by marking parts of a photograph, or using the caption to give clear directions.
Drawings combine the directness of a visual marker with the possibility to
highlight specific features.
The most important benefit of using drawings, photography and video, is that
they give a possibility to analyse situations or processes in a way which is not
possible by observation alone.
4.4.1 Drawings
Drawing is a slow process which forces the eye and the brain into a thorough study
of the object. M a n y times a basketry technique is not really understood, until a
detailed drawing has been made, which includes the patterns and irregularities.
Drawing often also brings to light where the production is problematic.
Recording basketry and the present day production process requires different
forms of drawing. Drawings to scale are made as part of the record of basketry
objects (cf. Wendrich 1991, 124). Reconstructions, often based on paper models,
are also made to scale. O f some objects it is useful to have axometric drawings,
especially when the shape is irregular or complicated.
Schematical drawings are not made to scale. They represent the w a y in
which the basketry has been structured by showing enlarged or simplified
representations of the technique, sometimes elucidated with colours.
Plans and elevations of the archaeological context at A m a r n a and Qasr Ibrim
were m a d e by different site supervisors. The working areas of the present day
basket makers was planned also.
4.4.2 Photographs1
In the course of the research on basketry production, photography has been a
research tool as well as a recording tool (Collier and Collier 1986). Recording the
general appearance of the basic structure as well as the shape and decoration, a
selection of baskets has been photographed on colour slides and black and white
film. This selection represents both the commonly occurring baskets as the rare
specimens. Close ups are used to record the technical details and aspects of the
raw material.
During archaeological field work photographs are m a d e to record the
context. The excavation files of the Amarna expedition contain photographs of the
excavated areas including some of the baskets and mats in situ. D u e to m y
presence at Qasr Ibrim during the excavations, I had the opportunity to study and
photograph a large number of baskets and mats in situ.
Photography as part of the ethno-archaeological work is used extensively to
record the context and the production process. In relation to this, the use and re-
use to which baskets are put are photographed, the results of which will appear
elsewhere. Recording dynamic processes, rather than static situations is done by
taking series of photographs. In Middle Egypt the actions were observed
"The equipment used for photography are two Olympus O M 1 0 camera bodies, one for
colour slides and one for black and white photographs. In order to avoid distortions I prefer
to use a 50 m m lense. For overviews and inside photography I sometimes made use of a 28
m m wide angle lense. Technical details are photographed with a powerful 105 m m macro
lense. In rare occasions I use a zoom-telephoto lense for outside photography. It gives the
opportunity to photograph people without being obtrusive, but it also has a dishonest side to
capture somebody close up without him or her being aware of it. The film used was mostly
professional 125 A S A black and white or 100 A S A colour slide film It enables photography
in the bright Egyptian sunlight, but also (often with the use of flash) inside houses and work
shops.
repeatedly, and photographs were taken at one specific session. In Egyptian Nubia
observation and photography were done during one visit.
Photographs as a research tool are used to communicate with Egyptian basket
makers. Showing pictures of ancient baskets readily starts a discussion on technical
features, applied materials and use. Photographs of European basket makers
equally provoke and animate the discussions.
The position of the body in different stages of the work and during different
actions is analysed by studying the photographs. Although it is possible to analyse
not only the position, but also the actions from a series of photographs, I chose to
use video.
4.4.3 Video3
Video is not an objective, unbiased medium. It can be maintained that a film is
m a d e during the editing stage. Although it is farfetched to state that the same
footage can be m a d e into a number of entirely different films, it is certainly true
that editing determines m u c h of the atmosphere and the contents. Through editing
a messy production process can be smoothed into an organized one.
T h e objective of the video that is part of this book is to show the production
process, but also to highlight a number of features and offer examples of the way
the actions are analysed (see Chapter 6). The edited video, which lasts one hour, is
a small selection of the total footage which encompasses approximately ten hours.
T h e five production processes which have been represented in one hour, last in
real time between 10 and 500 hours.4 It m a y be clear that filming means selecting.
Not only h o w m u c h is filmed, but also what is filmed and h o w things are
filmed determines the video narrative even before the editing starts (cf. Heider
1976, 46-113). The most important of these choices have been accounted for here.
Selection of Equipment
In order to record moving images, I have chosen to use video rather than film,
even though the latter gives a better quality. There are two main reasons for this.
First of all video is less expensive than film, which gives the possibility to take
long sequences. This gives a better base for analysis of the material and moves the
selection of scenes partly to the editing stage.5
The equipment used in the field was a Super-VHS camera, without external
microphones or light sources. The three hour S-VHS mother tapes were copied onto both
Umatic and Betacam tape. The off-line editing was done on a Umatic editing set by the
anthropologist C van Beek and myself. The on-line editing was performed on a Betacam
editing set.
Sewn plaits baskets take one full day, coiled food covers are made in a period of
three months, in which the basket maker works approximately six hours per day.
Another advantage is that in the editing stage scenes or sound can be used more than
once, since video is edited by copying scenes of the mother tape onto a second tape, while
film is edited by cutting and re-arranging the mother copy.
Secondly, video is much more light sensitive than film. With video one can
film until sunset and inside dim houses without using powerful lamps. It is,
therefore, a less intrusive medium than film. For the same reason, I have not used
external microphones.
Selection of the Subject

Only a small number of basket makers were selected to feature in the film. The
area was limited to Middle Egypt and N e w Nubia, the techniques were those
which were very similar to the ones occurring in antiquity. Preferably those basket
makers were chosen with which I had been in contact for a long period, w h o had
been m y teachers and with w h o m I had a good rapport.
Selection of the Scenes

Although the production sequences of basketry are similar, they are not identical.
At some point one particular sequence is recorded on video. Within this sequence
another selection is made on what and h o w much of the process was filmed.
For short processes, such as turning pots on a wheel, the entire sequence can be
filmed, but when the process takes a day or longer, it is a legitimate choice not to
register the entire procedure, but to take shots at intervals. A m o n g other things,
this means that the complete duration of the process cannot be timed from the
video.
Camera Position, Frame and Angle

The position of the camera and all the actions of the camera reflect implicitly the
research question. Zooming in forces the viewer to concentrate on a specific
subject or detail. Zooming out divides the attention between basket maker and
context. Even an overview of the surrounding area is still a selection by the
cameraman, and excludes more than it includes. The size of the frame (close up,
half total, total or overview) and the position of the camera literally determines the
angle of view and influences the interpretation. More will be said about this aspect
in Section 6.1.
During filming there are constantly moments in which choices have to be

made. Not all the selections mentioned above are well-considered decisions. At the
filming sessions improvisation is unavoidable. It may be clear that, as any method
of recording, filming does not give an objective representation. The film which
accompanies this book is therefore not to be considered as a 'see-for-yourself
proof for the points made in the text. It is a separate account, which partly tells a
different story than the text.
4.5 Problems and Pitfalls

The problems and pitfalls in field work are many and mostly related to
communication. In archaeological work the specialist working on a specific finds
group, needs to communicate with colleagues excavating and those studying other
find groups. Ideally the specialists are present during the excavation season, but
sometimes this is not possible, as was the case in Amarna where the basketry had
been stored since 1986.
Communication is even more important in ethno-archaeological work. The
researcher is removed from the informants in language, culture, social status and
profession. Difference in sexe between researcher and informant sometimes proves
problematic too, although in this respect being an outsider and a w o m a n has the
advantage of giving access to both the men's and the women's world. The
problems and pitfalls are best illustrated by an account of the work of the past few
years.
Introduction to the informants

A good introduction of the researcher is of the utmost importance. In Amarna such
an introduction was given by several people working for years with the
archaeological expedition. During the excavation seasons of 1989, 1990 and 1992,
each of which lasted several months, I was given introductions to visit basket
makers in the three villages in the vicinity of the Amarna dig house.
Such a combination of archaeological field season and ethno-archaeological
work was not possible in Qasr Ibrim. There are no villages in the vicinity, because
the entire area has flooded as a result of the A s w a n High D a m . Around 1960 the
people of Egyptian Nubia were moved to new villages in the south of Egypt around
the city of K o m O m b o . This area is hence known as (Egyptian) N e w Nubia. The
study of Nubian basketry and basket makers was done in two brief concentrated
periods after the excavation seasons of 1990 and 1992. With an introduction by the
Inspectorate of Antiquities I visited several basket makers near the town of Aswan.
In 1990 I made a number of day trips to the N e w Nubian villages as an orientation
on what knowledge and traditions were left after the forced migration of the
Nubian people. In 1992 I stayed for three weeks in the village of Dabud, north of
the city of K o m O m b o . Again the Inspectorate of Antiquities was a great help in
organizing contacts and permission.
This village was the starting point for visiting a large number of Nubian
villages in the area as well as the town of Daraw. The contacts with the basket
makers in these villages were intensive, but less well balanced than the work done
in Middle Egypt. After the first visit, during which the introduction took place and
I observed and participated in the work, a selection was made of basket makers
w h o were considered representative both for the work done in Nubia before the
migration, and for the techniques as they developed afterwards.
Granted that hospitality is a Nubian life rule, the level of rapport was
surprisingly high, considering that the introduction into the villages was minimal
and the period to get acquainted was brief. The only way I can explain this, is that
the informants and the ethno-archaeologist shared an interest in a very specific
subject. Nevertheless, during the work there were regularly communication
problems, which were caused by different factors.
Biases
O n e of the hardest tasks in ethno-archaeological research is the attempt to become
aware of ones o w n biases. In retrospect, the two periods of ethno-archaeological
research can be considered a gradual personal development. In 1989 and 1990 m y
observation of basket makers was steered by questions that had risen from studying
archaeological basketry. M y attention was directed mainly to the production
process and the actions that resulted in particular technical details. It was not until
the second period of ethno-archaeological field work that I gave full attention to
the ethno-archeaological context.
Influence of my presence
M y presence inevitably influenced the situation which I w a s observing or
participating in. Sometimes the entire process was disturbed, because everybody
was paying attention to m y work, rather than their own. At other times w e would
all be working happily along, and apart from the occasional smirk in m y direction
because of m y slow pace, the basket makers would seem to get on with their work,
just as they would without m e being present. It was impossible to rate the amount
to which the 'normal' situation was influenced by m y presence, other than by
noting d o w n the subjective impression that the process and atmosphere seemed to
be disturbed or not.
The video sequence featuring the mat maker Amrit is a prime example of
disturbance of the normal order of things by m y presence (sequence 42:10-47:34).
W h e n the village school was over the pupils streamed into the little square before
the mat makers' house. They were kept at a distance by two helpful neighbours
w h o started beating the boys with sticks. At such moments it is difficult to decide
if the session should be broken off or continued.
Communication problems
A s a result of differences in classification, language and world view, several
communication problems arose. A n example of the first, classificatory differences,
has already been mentioned above (p. 17). W h e n I tried to m a k e inventories of
containers used in and outside the houses I had to name every single type of basket
separately, because the English word basketry has no Arabic or Nubian
counterpart. It was a vivid example of a classificatory problem, the things
belonging together as 'basketry' in m y opinion, being objects of different classes
for m y informants.
Furthermore, it was often difficult to explain the exact purpose of this
excercise to the informants. Although most people were eager to help, the request
to investigate the house, or have all basketry objects being gathered, was
sometimes met with suspicion. Most often, however, looking at old baskets and
pots w a s simply beyond anything considered of possible interest and the request
was not understood.
The interviews presented still other communication problems, because the

questions which I had posed myself, had to be translated into Arabic and had to be
in keeping with the perception and experience of the informants. It was
unconceivable to ask a Middle-Egyptian basket maker h o w he would characterize
his social position. The question that was asked instead, depending on the specific
situation, was something like: "Would you be proud if your son would become a
basket maker too?". The answer was a matter of interpretation, since it not only
reflected a level of 'social status', but also economic motives.
The questions as listed in Appendix B were adapted time and again to the
specific situation of the basket maker. This implies that not all basket makers have
answered the same questions. It was a deliberate choice to give precedence of good
rapport with the basket makers over getting through the questionnaire.
During the first few years, when m y knowledge of Arabic was insufficient,
the problems of using a translator to do the interview sessions became very clear.
In the village nobody spoke English, so the translator was an outsider, w h o was
better educated and usually more wealthy than the basket makers. The major
problem encountered was that sometimes a translator would look d o w n upon the
basket makers, which disturbed the rapport. In several cases an interpretor did not
translate the answer of the informants, but presented his o w n ideas on the subject
instead. Being better educated, he felt there was nothing he could learn from the
basket makers, and that he knew the subject m u c h better than they. In one case the
interpreter, w h o was a Christian, flatly refused to translate a story about a Muslim
Saint w h o was said to have introduced basket making into Egypt.
A s soon as I posed questions on economic aspects, many informants appeared
to become uneasy, especially those w h o m I was not acquainted with all that well.
It was hard to discern to what extent the answers were governed by certain
expectations. Although I presented myself as a scholar w h o was interested in
ancient and modern basketry, questions about the prizes were sometimes
interpreted as a means to investigate the market. Comparing information on prizes
of different basket makers made m e realize that some of them had mentioned
considerably higher prices, perhaps in the hope that I would buy their products at
an expensive rate. In some cases I confronted informants with the accepted market
price. Being caught in an effort to 'cheat' was acknowledged good humouredly.
Usually a number of questions were posed twice in a slightly different
formulation, in order to check answers. Sometimes this was recognized and
informants would remark so or look surprised. At a few occasions the large
quantity of questions provoked irritation and the interview had to be stopped.
Period of research
A number of problems rose because of the limited time that could be spent and
specifically by the time of year in which the research was done. Ideally the
research should take place the year round, to observe what happens in different
seasons. B y the time I realized the importance of long term research, there was no
time left to organize such an enterprise. This was partly solved by asking the
basket makers if their actions were limited to certain periods of the year. From the
answers it appeared that the basket makers of Middle Egypt and Egyptian Nubia
were collecting raw material and making baskets all year round. There was no
such thing as seasonal activities. This is not true for all basket makers in Egypt. A
matmaker w h o m I visited in Dakhla oasis had a specific time of year in which he
collected his raw material. H e used rushes, which had to be collected before a lack
of grass forced the goats to eat rushes.
The second period in which I did m y research in N e w Nubia fell within the
month of Ramadan. During this month Islamic people are fasting during the day
and have two festive meals during the night. In R a m a d a n the entire society is
unsettled, because of lack of food, water and cigarettes during the day and lack of
sleep at night. Especially near the end of the month m a n y people are tired and
irritable. M o r e importantly, during Ramadan Nubian w o m e n are very busy and
have no time, or are to tired to make baskets. It was not until I had arrived in the
village of Dabud that I realized that it was the worst possible time to study basket
makers. Yet, the w o m e n were very helpful and sat d o w n to work in order to show
m e their skills.
Photography and video

In the era of the Adobe photoshop images can be doctored to represent almost
anything, but even photographs which are supposed to give an account which has
not been tampered with, might give a biased representation. The influence of the
photographer is considerable, not only in the selection of the subject, but also in
the manner in which the photo is made. Including the surroundings gives a
different impression than showing a close up of a basket maker.
Officially Islam forbids making images of people. For strict muslims it is
therefore not allowed to be photographed. For most informants this was not a
problem, most of them were very eager. Neither were there objections against
being filmed. In Egypt television is k n o w n in the smallest villages, and video is
increasingly popular to record special occasions such as weddings. However, in a
few cases in Egyptian Nubia, wives whose husbands were not in the village, but
working far away in Cairo, did not want to be photographed or filmed. It was
impossible to decide if this was an Islamic rule, or an attitude ruled by modesty,
but of course their wish was respected.
A s mentioned in section 4.4.3 the use of video does not give an objective
registration of the production process and context, not only since the film maker
has a large influence on the end product, but also because the presence of a camera
influences the situation. Although video was selected because of being less
intrusive, the appearance of a camera usually caused pleasant excitement. In some
cases, however, it disturbed the situation to the extend that the footage became
quite useless. After working with a female basket maker in the vicinity of A m a r n a
for weeks, w e had established a good relation. She was teaching m e and w e spent
several afternoons sitting in the dust of the court yard making baskets. W h e n I
asked her, she declared to have no objections to being filmed. The morning of the
video session, her husband was unexpectedly at home. He had taken the day off to
be present and started to interfere immediately. H e did not want her shown 'on
television' in her old dress. The court yard had to be swept and moments later she
c a m e back in a beautiful dark blue velvet dress. O f course sitting in her usual
shady corner would spoil her garment. Instead she was put in the sunny court yard
on a chair, looking unhappy, uncomfortable and very warm.
Because there was a television in the village, I could show the footage to the
basket makers, and their family and friends. Here m y limited knowledge of Arabic
was a real handycap, because it would have been extremely useful to get some
more feedback.
Consequences
Fieldwork is a step into the unknown and a learning process. There is no
procedure to prepare fully for what will be encountered and improvisation is often
necessary. A multitude of factors are of influence on the result, which are as far
apart as the m o o d in which the participants happen to be in, to the number of
tourists that visit a community. The old fashioned ethnographer describes strange
people and habits, without accounting for the points where he might not have
understood what was happening, but the authoritive ethnography is slowly dying
out and the present generation is aware of the limits of interpretation. Accounting
for the shortcomings of one's field work is not just a matter of being honest, but it
has become part of the research method.
Is it possible to build in a methodical safetey net to accomodate for the
problems occurring during field work? There is no simple answer or fixed
procedure, but the central concept is being aware and explicit about the strength
and weakness of one's research. This is important in all stages of the work. During
field work I attempted to articulate doubts and questions as m u c h as possible in the
interaction between myself and the informants. In writing up the research, I have
indicated where and w h y I a m in doubt.
C H A P T E R FIVE
BASKETRY ANALYSIS
5.1 Analysing the Basic Structure

All baskets are m a d e of strands, which are arranged in a specific manner to create
a fabric, indicated throughout this book as the basic structure. A basket m a y
consist of several kinds of basic structure, the base might be m a d e differently from
the sides, for instance. In order to be able to compare basketry techniques they
have been abstracted into an interaction of systems. Table 5-1 lists the five aspects
necessary to analyse the basic structure: determining the number of systems, the
orientation and activity of each system, the nature of the elements of which the
system consists and the manner in which the systems interact. These five aspects
are the subject of sections 5.1 to 5.5.
A. Number of systems 5.1.1

B. Orientation of the systems 5.1.2
C Activity of the systems 5.1.3
D. Composition of the 1. R a w material (species) 5.1.4.1
systems: 2. Element size, shape and flexibility 5.1.4.2
(width/thickness/diameter)
3. Preparation of elements (half products) 5.1.4.3
4. One element / Set of elements 5.1.4.4
5. Element members (number/orientation) 5.1.4.5
6. Element mobility 5.1.4.6
E. Interaction of the systems (pattern) 5.1.5
Table 5-1 Five aspects involved in analysing the basic structure of a basketry technique
(cf. Sections 5.1.1 to 5.1.5).
5.1.1 Number of Systems

In order to record the number of systems it has to be decided which strands belong
to one system and which to another. Elements of a system are defined as: all
strands m a d e of the same material and with the same orientation and function in
effecting coherency in a basketry technique. T o define the number of systems it is,
therefore, also necessary to look at the orientation (B in Table 5-1) and the activity
of the strands (C in Table 5-1), the raw material (Dl in Table 5-1), the size of the
elements ( D 2 in Table 5-1) and their preparation (D3 in Table 5-1).
1. Are all elements There are at least as many systems as

orientated in the same there are orientations. Continue for
direction? No each orientation separately at question
2, to find out if there are several
systems orientated in the same
direction
Yes
2. D o all elements have the There are at least two systems: one
same function active, one passive. Continue for each
(active/passive)? No system separately at question 3, to
check if there are more than one active
or passive systems.
Yes
3. Is one plant species There are at least as many systems as

used? mere are plant species involved.
No
Continue for each species separately at
question 4.
Yes
4. Are all elements The number of different preparations

prepared in the same way? represent the number of systems.
No
Continue for each method of
preparation separatelywith question 5.
Yes
5. Are the strands all of the If there are separate groups of

same width, thickness, consistently smaller or larger strands,
No
diameter or flexibility? than each of these groups represents a
separate system.
Yes
These strands form one

system.
Figure 5-2 Flow chart for determining the number of systems in the basic structure of
basketry techniques.
5 Basketry Analysis 83
An example of h o w to use these aspects to determine the number and properties of

the systems is shown in the flow chart of Figure 5-2'. By answering five questions
the number of systems can be defined, as well as the criteria by which these systems
are differentiated:
1. Are all elements orientated in one direction? (criterion B in Table 5-1)
If the answer is negative, the number of orientations indicate the m i n i m u m number
of systems. Figure 5-3 shows a mat, m a d e of string and grass leaves, which are
orientated at a straight angle. Since two directions can be discerned, there are at
least two systems. Figure 5-4 shows a technique with three orientations: vertical,
diagonal and horizontal, which implies that there is a m i n i m u m of three systems. It
is necessary to check these three systems separately by answering questions 2, 3, 4
and 5 for each orientation, in order to establish if the minimal number of systems
found by the criterion of orientation is indeed the total number of systems. Figure
5-5 gives an example of a technique in which the orientation of two systems is the
same: the grass rope core and the palm leaf loops are following the same direction:
they are worked from left to right, the linked loops holding the rows of grass rope
in place. If the answer to the first question is affirmative, as in Figure 5-5, then
the second question has to be posed in order to find out if there are more than one
system involved.
Figure 5-3 mat, made of string and grass leaves, which are orientated at a straight angle
(Amarna, ca. 1350 B C )
In Wendrich 1991c a similar flow chart was presented in which the questions were
listed in a different order. The order is of no consequence to the end result, as long as all
aspects are checked.
Figure 5-4 Basketry technique with three different orientations (vertical, diagonal and
horizontal, and four different systems (three passive and one active). T h e
string (system 4) with which the three passive systems are tied together,
follows the same horizontal direction as passive system 3, but differs both in
activity and material (found at the R o m a n fort of A b u Sha'ar, Red Sea coast,
Egypt).
Figure 5-5 Basketry technique with two systems which follow the same orientation
(from left to r.ght). The passive grass ropes form the body of the technique
while the active palm leaf loops cause the coherency of the fabric (Qasr
2. Do all elements have the same function? (criterion C in Table 5-1)

The function of a system can be active or passive. The active system comprises the
strands which are causing the coherency of an object, while the passive system
forms the body. In m a n y cases the passive system is made of less flexible material
than the active system, or made rigid, for instance by fixating the passive strands
in a loom.
In w o v e n objects the difference between passive and active is quite clear: the
warp, fixed in the loom, is passive, while the weft which is woven in, is the active
system. In Figure 5-3 the string is passive, while the grass leaves are active. In
Figure 5-4 the lattice is passive: the elements of three systems are layed out, the
diagonal system 2 being sandwiched between vertical system 1 and horizontal
system 3. A s soon as this contraption is moved, however, it will fall apart. The
coherency is caused by the fourth, active, system, formed by a piece of string
which holds together the entire structure.
In Figure 5-5 the grass ropes are passive, and incorporated in the fabric by
the active palm leaf loops which are linked to a previous row of loops. The two
systems, ropes and loops, are oriented in the same direction, because the basket
maker is following the length of the rope from left to right while making the loops.
Ropes and loops thus have the same orientations, but a different function. The
previous row of loops, to which the next row is linked, is indicated as anchoring
fabric.
Similarly, in the coiling technique shown in Plate 5-6, the palm leaf strand is
stitched into the previous coil. The finished central part of the basket has the
function of anchoring fabric to which the next coil is fastened. The palm leaf
strand and the bundle of shredded fibre follow the same direction. They are both
worked in a coil from the centre of the basket to the rim. The palm leaf wrapping
strand is active, holding the bundle of shredded material in place. The bundle is
the passive system, forming the body of the technique. In Chapter 12 more will be
said on passive and active systems in relation to the production process and the
actions of the basket makers.
If question 2 of the flow chart in Figure 5-2 is answered with "Yes" then w e
can proceed to the next question. If the answer is "No", however, there are at least
two systems: one passive and one active. In that case the third question has to be
answered for each of these separately, in order to check if there are more than one
active or passive systems.
3. Is one plant species used? (criterion D l in Table 5-1)
If the answer is negative, for instance if one set of elements is m a d e from palm
leaf and the other consists of grass, then they are different systems. If the answer
is affirmative, then one can proceed with the next question.
4. Are all elements prepared in the same way? (criterion D 2 in Table 5-1)
Basketry raw materials often retain their original shape and require minimal
preparation. Selection is an important part of the process, because uniformity of
the shape of the raw materials enables the basket maker to make an end product
with a regular appearance.
Plate 5-6 Coiling technique: two systems which follow the same direction, coiling
from the centre to the rim. The passive bundle is held in place by the active
wrapping strand which is fastened with a stitch through the previous bundle
(Middle Egypt, 1992).
For s o m e techniques, however, the preparation goes beyond selection. T o

correct the irregularity of their shape, materials are sometimes cut or split. T o
accommodate for the limited length of the leaves, twigs and stems, strands are
worked up in a separate production phase: grass is m a d e into string, palm leaves
are plaited into long strips (cf. Figure 5-7 and Plate 4-2 on p. 61).
If the answer to the fourth question is "Yes" then one can proceed to question
five. If the answer is negative because, for instance, one set of elements is m a d e of
unworked grass and the other of grass leaves which have been m a d e into a string,
they are different systems and the fifth question has to be answered for both of
them separately.
5. A r e the strands all of the same width, thickness, diameter or flexibility?
(criterion D 3 in Table 4-2). If the answer is negative, for instance w h e n one set of
strands is consistently m a d e of thicker or tougher material than the other, different
systems can be distinguished. This difference occurs, for instance, in stake-and-
strand basketry, where the willow rods which are w o v e n in are usually thinner and
more flexible than the ones which form the passive system. Similarly in the three
system technique
If the last answer is positive, then all strands belong to the same system.
Figure 5-7 Plaiting with 9 strands, as preparation for making a sewn plaits basket. The
strip is plaited with two active systems which are orientated approximately at
a straight angle. The individual plaiting strands turn at the edge of the plait
and are worked back into the plait from the other side.
Figure 5-8 A one-system technique, made of rows of looped palm leaf strands (Amarna.
ca. 1350 B C ) .
If the answer to all five questions was positive, we are dealing with a one
system technique, as is shown, for example in Figure 5-8. In this example there is
one orientation (all strands are looped in subsequent rows). The loops are made by
one ongoing, active strand.2 Only palm leaf is used, which has been prepared by
twisting it to form an ongoing strand. This strand is uniform in diameter and
flexibility over its entire length.
Figure 5-9 gives an example of using the flow chart for the technique
depicted in Figure 5-4. The answer to the first question is that there are three
orientations: vertical, diagonal and horizontal, meaning that w e can at least discern
three different systems. For each of these, question two is answered. All strands
running vertically are passive, the diagonal strands are all passive too, but in the
horizontal direction there are two different systems, one passive and one active.
The latter is the string which ties the whole lattice together. For all four systems
the next questions have to be asked. All elements of the vertical system consist of
rushes, which are unworked and have an average diameter of 2.5 m m . Since no
differences occur within the strands of the vertical system, all vertical strands
belong to system 1. Similarly, the diagonal strands all consist of unworked rushes,
which consistently have a diameter of 2.6 m m and thus form system 2. The
horizontal elements do not have to be split up in more than two systems. The
group of horizontal unworked rushes are all of the same average diameter of 3.7
m m , and thus forms system 3. The active system, which ties the three passive
systems together, consists of cotton which has been made into a string with an
average diameter of 1.5 m m .
5.7.2 Orientation of the Systems

A one-system technique is by definition orientated in one direction, corresponding
with the working direction of the basket maker (Figure 5-9). In a two-system
technique there are m a x i m u m two directions, as is the case with the woven mat in
Figure 5-3, where the grass string warp is orientated at a straight angle to the weft
which is made of individual grass stems. Two-system techniques with both systems
following the same direction occur frequently also. Examples are coiled basketry
(Plate 5-6) and sewn plaits basketry (Plate 4-3, p. 62). The active and passive
system follow the same course, spiralling from the centre to the rim of the basket.
Rectangular sewn plaits mats are not sewn in spiral, but row by row parallel to the
edge of the plaits, just as the matting of loops around ropes (Figure 5-5).
In three system techniques the m a x i m u m number of orientations is three. In
Egypt, however, it is more c o m m o n that two of the three systems follow the same
direction, as is the case in the technique shown in Figure 9-18 (p. 167) and the
decorative coiled inlay technique of imbrication (Plate 4-6 p 65 and Figure 9-17
p. 166).
In a one-system technique the strands are by definition active.

1. Are an No _ dinerem orientations ^-

elements
orientated in •^ -
the same
direction? vertical diagonal horizontal
2. D o all
INO
elements have
the same Yes Yes
function ?
(active/passive)
Passive Passive Passive Active
1 1 1
3. Is one plant 1 1
1
species used?
Yes Yes Yes Yes
Rushes Rushes Rushes Cotton

1 1 1
J
4. Are all 1 1 1
elements
Yes Yes Yes Yes
prepared the
same way?
Unworked Unworked Unworked String
1 1 1
1
5. Are the ^ 1 1 1
strands all of
Yes Yes Yes Yes
the same width,
thickness,
diameter or
average average average average
flexibility?
2.5 m m 2.6 m m 3.7 m m 1.5 m m
System System System System

1 2 3 4
Figure 5-9 Example of using the flow chart (Figure 5-2) to determine the number of
systems and their properties of the technique depicted in Figure 5-4.
The orientation of the systems is not by definition the same as the orientation
of individual strands. The plaited strip used in sewn plaits basketry, for instance, is
m a d e with nine palm leaflets, which are part of two active systems. O n e system is
plaited in from the left, the other from the right. The individual strands turn at the
edge and form part of both systems (Figure 5-7). In terms of the criteria to
determine the number of systems (Figure 5-2) the only thing that discerns the two
systems is the fact that they are follow two different directions, oriented at a
straight angle. There is no other criterion on which to discern these two systems,
since all strands have the same function (active), and are m a d e of the same
material, with the same preparation and the same flexibility. Not only that, the
strands coming alternately in from the left and from the right are physically the
same palm leaflets.
5.1.3 Activity of the Systems

O n e of the criteria to discern systems in a technique is to determine the function of
the strands in making up the structure. All strands that have the same function,
either active or passive, are considered to belong to the same system (cf. Figure
5-2). The passive elements form the body of the technique, while the active
elements cause the coherency.
The passive elements are usually less flexible than the active elements. The
passive elements themselves can be rigid, because a different, for instance woody,
material has been chosen for the passive strands, while the active strands are of a
flexible material. Often, however both systems are m a d e of the same material and
the difference in rigidity is reached by selecting thicker material for the passive
system than is chosen for the active system. Flexible passive strands can also be
m a d e rigid by putting them temporarily under tension. The general term to
describe an external feature to which one or both of the systems are fastened is
frame. A frame can be temporary (a loom) or permanent (for instance a bed or
chair frame on which webbing is made).
Although basketry systems are made up of strands, It is important to regard
individual strands as separate from the system. Individual strands are often part of
different basic structures and their role in the technique can change from active to
passive and back. In decorated coiled basketry, for instance, a passive bundle is
fastened in a coil with differently coloured strips. W h e n the pattern requires that
coiling is done with a red strip, the white strip is layed parallel to the bundle and
thus becomes temporarily passive (cf. Chapter 13). In stake-and-strand basketry
the stakes are passive, but w h e n reaching the top of the sides, they interweave in
order to form the rim. Suddenly the passive stakes have become active.
One-system techniques are by definition active. In two-system techniques in
which both systems are active there is no difference in rigidity of the two systems.
All forms of plaiting, for instance, are made with two active systems (cf. Figure
5-7). Confronted with a structure in which all strands are m a d e of the same
material and have the same flexibility, it is difficult to determine the activity of the
strands. Either both systems are active, and we are dealing with a form of plaiting,
or, if one of these systems was put under tension in a frame, w e are dealing with a
technique with one active and one passive system, such as weaving. Usually this
can only be judged from the edges.
It m a y be clear that the description of the basic structure leans on
presuppositions regarding the production process. In Chapter 13 the definition
given above, which states that the passive system makes up the body, while the
active system causes the coherency, will be replaced by a different one, which
involves the actions of the basket maker.
5.1.4 The Composition of a System

Each system is built of elements, which are the units that interact. A n element can
be a stick or a palm leaf, a plait or a string, or five parallel grass stems. In other
words: elements can be single or composite and have certain characteristics which
are determined by six factors: the material (section 5.1.4.1), the size and
flexibility of the elements (section 5.1.4.2), the preparation (section 5.1.4.3), the
question if one ongoing element is used, or a set of parallel elements (section
5.1.4.4), the number and orientation of the members (section 5.1.4.5) and the
mobility of the elements (section 5.1.4.6). In the following sections, as elsewhere,
the word strand is used as a non-technical term for any length of material used to
make basketry.
5.1.4.1 Raw material

The material of which an element consists is not only distinguished by plant
species, but also by plant part. Leaves, stems, twigs, roots, veins or fibres are all
used and have different properties.
5.1.4.2 Element size, shape and flexibility

The size of the elements, for instance the diameter of willow rods, is related to the
flexibility. Furthermore, elements can consist of single strands, for instance
individual grass stems, or of bundles of material. A n example of the first are the
elements of the active system of the mat of Figure 5-3, where single grass stems
are w o v e n in. A n example of the second are the bundles of fibres used as the
passive system of coiled basketry shown in Plate 5-6, where the active system
consists of individual strips of palm leaf.
5.1.4.3 Preparation of elements

The preparation of the elements varies greatly. Most of the preparations are geared
to making the strands more uniform. Selection of strands of the same size or shape
often takes up a considerable portion of the total production. In stake-and-strand
basketry, for instance, it is extremely important that the willow rods which form
the passive stakes are all the same diameter and are slightly thicker than the
weaving strands. If the selection is not done properly, the result is an unevenly
formed basket, which lacks strength. Cutting or splitting the material is another
way of creating elements which are similar in size and shape.
Sometimes a separate production phase is dedicated to making uniform
elements. For sewn plaits basketry, for instance, both the active and the passive
system are separately prepared: string is m a d e and a long strip is plaited.
Other preparations that occur is the removal of hard or weak parts. Finally,
the material is often soaked to create a suppleness which makes it possible to bend
the elements. After drying the shrinkage of the strands tightens the fabric.
5.1.4.4 One Element I Set of Elements

Plate 5-6 shows an example of two systems, both consisting of one element. The
passive system consists of one bundle, which is coiled, the active system consists
of one strand of palm leaf which winds itself around the passive bundle, and holds
it in place (cf. Figure 9-5, p. 160). In theory, the active strand is one long
element, but in practice the raw materials for basketry do not have a length
sufficient to make an entire basket with one palm leaflet. Since n e w leaflets are not
inserted parallel to, but at the end of an exhausted strand, and because the basket
maker always handles one strip at a time, the sequence of different strips is
considered as one element.
The same is true for the sewn plaits basket shown in Plate 4-3 (p. 62): the
long plait (which is clearly one element) is sewn with m a n y short strips of d o a m
palm leaf. Since they are inserted one by one at the end of the exhausted previous
strip, they are considered one element.
The clearest example of using a set of elements is in the stake-and-strand
technique of French randing (Figure 3-4, see p. 36). The willow rods are worked
in simultaneously and parallel to each other. In the woven grass mat of Figure 5-3,
both systems seem to consist of a set of elements, a number of parallel strands
which have the same function in the process. The elements of a set are used at the
same time, parallel to each other.
In archaeological material it is often difficult to determine whether the basic
structure is made with one element or with a set of elements. It takes a study of the
edges to conclude if parallel single strands were used, or one ongoing strand.
Comparison with the modern production processes is a great help to interpret the
basket rims and matting edges. The fragment of grass matting, drawn in Figure
5-3, for instance, seems to have been made with a set of parallel strings, but from
studying present day mat makers, a likely reconstruction of the production process
is that the warp was made by winding one long string around the cross beams of
the matting loom. The weft consists of a set of elements: m a n y grass stems are
woven in parallel to each other.
Since no edges are present, it is not clear whether the fragment of looped
matting shown in Figure 5-8, is made with one element (similar to knitting with
one ongoing yarn) or with a set of elements parallel to each other. The same
looping technique was used for making a large bagfilledwith doam nuts, which
was found in the tomb of Kha and is n o w on display in the Egyptian M u s e u m in
Turin (Bongioanni 1987, 106; Schiaparelli 1927). From this complete example it is
clear that the bag was made with one ongoing strand.
Edgeless fragments of bed matting, such as found at Qasr Ibrim, appear to
have been made with a set of elements. From present parallels it is clear, however,
that similar bed matting is made with one long rope.
5.1.4.5 Element members

If an element system consists of more than one strand, it has several members. The
element may, for instance, consist of two or three parallel strands which are
woven in together. Whenever an element consists of more than one member, the
orientation of these members is recorded.
Parallel members are indicated by a capital /. W h e n the members are twisted
around each other, their orientation is indicated with a capital S or Z, following the
slant of the middle strokes of these letters (Figure 5-10).
Figure 5-10 Orientation of the members of an element: I, S and Z direction

94
Figure 5-11 (see p. 94)

a. Weaving under 1 / over 1 with an element consisting of one member
b. Weaving under 1 / over 1 with an element consisting of two members
c. Weaving under 1 / over 1 with an element consisting of three members
d. One element consisting of two paired members in S-direction (S-twining)
e. One element consisting of two paired members in Z-direction (Z-twining)
f. One element consisting of three interacting members in S-direction,
pattern: under 1 / over 2 (S-waling)
g. One element consisting of three interacting members in Z-direction,
pattern: under 1 / over 2 (Z-waling)
In Figure 5-11 the variation shown lies in the number of members and the
orientation of the w a y in which they have been twisted around each other. The
active elements consist of one m e m b e r (a), two members (b, d and e), or three
members (c, f and g). The members of a, b and c are woven parallel to each
(I-orientation).
Elements d and e consist of two interacting members, twisting around the
passive system and each other. This technique is often indicated as pairing or
twining and is done either in S-orientation or in Z-orientation. The same principle
but then with three members, can be seen in elements f and g. In basket makers'
terminology such an application of more than two members in a stake-and-strand
element is indicated as waling. Combinations of waling with four or more
m e m b e r s do occur as well. The pattern is adapted to the number of members
involved. For a four-membered element the pattern can be under 2 I over 2 or
under 1 I over 3 or under 3, over 1, as long as the total number adds up to four.
5.1.4.6 Element mobility

The passive elements are usually quite rigid or straight, while the active elements
are mobile, in the sense that they wrap around the passive elements to hold them in
place. The mobility of the active elements can be specified by indicating the
number of planes in which they move.
Mobility in one plane is an undulating movement that goes up and down, but
not left and right. A n example of this is weaving. Mobility in two planes is a
m o v e m e n t that goes up and d o w n as well as left and right. The result is a
corkscrew, such as found in the wrapping strand of coiled basketry. Mobility in
three planes is a movement that goes up and down; left and right; and in addition
turns back onto itself. Strands moving in three planes cross themselves and are a
form of knotting, rather than wrapping.
Perhaps the concept of element mobility becomes clearer, w h e n applied to
actual basketry. Table 5-12 gives an overview of techniques according to the three
kinds of mobility. In addition, the number of members that m a k e up the elements
of the active system are incorporated.
The World according to Basketry
96
X>
E
r i
Table 5-12 (on page 96) Mobility of the active element in one, two or three planes, in
relation to the number of members of which the active element is made up.
Moving in one plane is represented by weaving with single weaving strands

(Table 5-12 a). The active strands are holding the passive elements, here drawn as
sticks to emphasize their rigidity, in place. The weaving strands m o v e up and
d o w n . If two or three strands are incorporated, they are woven in parallel to each
other (Table 5-12, a, b and c).3 The same up-and-down movement is found in
sewing with a basting stitch, such as used in sewn plaits basketry.4
Moving in two planes results in a cork-screw like wrapping. In the table two
examples have been drawn: if the active element wraps around a core which has
the same orientation, then w e see the wrapping as it occurs in the coiling
technique. The second technique in this box involves wrapping a strand around a
number of parallel strands (Table 5-12 d).5 The same movement, with two active
m e m b e r s is k n o w n as twining (Table 5-12e),6 if three active m e m b e r s are involved
the technique is k n o w n as waling (Table 5-12f).7
Mobility in three planes involves that the strands turn back on themselves and
hold the passive elements by binding or knotting them (Table 5-12g).8 If this is
done with two m e m b e r s than w e usually call this tying (Table 5-12h). There is no
term for a technique with three members moving in three planes. Envisaging such
a technique, the closest approximation would be a 'tangle'.
5.7.5 System Interaction

The interaction of the systems defines h o w the fabric holds together. Systems can
be engaged or not. If the active system weaves or wraps around the passive
system, then the systems are not engaged. If, on the other hand, the active system
splits the passive system, then they are engaged. Examples of the former are
weaving, twining, binding and plaiting. Examples of engaged interaction are
pierced basketry (cf. Figure 9-12, p. 163) and sewn plaits basketry. In pierced
basketry, the active system pierces through the passive system. The plaited strips
in sewn plaits basketry are opened up to let the needle with string pass through the
edge.
3Weaving and plaiting: cf. Figures 9-8, 9-9 (pp. 161-162), 9-14, 9-15 (p. 164), 9-20
(p. 168).
Sewing with a basting stitch: sewn plaits basketry (Figure 9-4. pp. 158-159).
Coiling: Figure 9-5 (p. 160); Wrapping: knotless netting (Figure 9-1, p. 155),
grommet (Figure 9-2, p. 156), wrapping (Figure 9-6, p. 160), coiled inlay (Figure 9-17,
p. 166).
Twining: cf. Figure 9-10 (p. 162).
Waling: cf. Figure 9-11 (p. 162). In Europe waling is also done with four or even
five members, but in the Egyptian material these varieties do not occur.
8Binding or knotting: cf. Figures 9-3 (p. 156), 9-7 (p. 161), 9-16 (p. 165), 9-18
(p. 167)
Coiled basketry can be either engaged or non-engaged. The most c o m m o n

form of coiling is engaged: the active strand wraps around the bundle and stitches
through the bundle of the previous row. Another form of coiling wraps alternately
around one and two bundles and is not engaged (Figure 5-13).
Figure 5-13 Engaged and non-engaged coiled basketry.
One system techniques are usually made by a non-engaged connection to the

previous row, thus gradually adding on to the fabric. The fabric to which the new
round of work is linked is indicated as anchoring fabric. Some of the two system
techniques, such as coiling or sewn plaits basketry, are also added on to an
anchoring fabric. The connection of a new row to an anchoring fabric can be
engaged or non-engaged (cf. the two variations of coiling in Figure 5-13).
The techniques that do not attach to an anchoring fabric, usually involve that
the strands of the passive system are layed out and are gradually connected by the
active elements. In order to keep the passive strands from moving about, they are
often fastened in a. frame. The most obvious examples are weaving and twining.
During the interaction of the systems the spacing and patterns of the basic
structure are made. The active, the passive or both systems can be widely spaced.
The spacing often bears relation to the function of objects; to mention two of the
extremes: sieves are widely spaced, while milk baskets are closely spaced. The
interaction can be done in many different patterns, by passing the active elements
around a variable number of passive strands. When patterns are used to indicate
variations within a technique, this implies a hierarchical structure in which pattern
is considered less important than other features of the basic structure.
5.2 A n a l y s i n g t h e Basket as a S t r u c t u r e
After analysing the basic structure, the entire basket can be defined in terms
of one or, often, as a combination of different basic structures. For this, it is
important to make a distinction between strands and system elements. A single
strand might run from centre to rim through the entire basket, but have a function
in different basic structures. It is not u n c o m m o n that the side is made in a different
technique than the base, for instance.
A n example of a basket with has different basic structure in the base and the
wall is depicted schematically in Figure 4-2 (p. 61). The centre is made of eight
strands, each of which consists of six parallel stems of a species of rushes. The
strands are layed crosswise and fastened by means of an active element, which
consists of two Z-twined members (the centre is indicated with number 1 in Figure
4-2).
The twining continues in the base. Bundles of six parallel rushes form the
passive elements, radiating out from the centre. The bundles are split in two
bundles of three rushes to fill up the space between the passive elements, which
grows with the expanding diameter of the base (indicated as number 2 in Figure
4-2). The active element consists of two rushes which are twined in Z-direction.
The passive strands of the base are simply ending, but the active system (the two
twining strands) continue in the side of the basket (not shown in Figure 4-2).
Because the passive strands of the base have ended, n e w passive elements are
inserted for the sides (Figure 4-2 at number 3). They are pushed into the twined
fabric of the base, parallel to the old passive bundles. Then they are bent around
the bundles of the base. This has a double function: it anchors them into the fabric
of the base and at the same time prevents the rows of twining of the base to slip off
the passive bundles.
The newly inserted passive strands, radiating from the base, are put upright
and fastened with the active system. A s said above, the two active twining strands
continue in the side, but they are joined by a third one. The three interacting
strands twist around the passive stems and each other and form a row of waling
(number 4 in Figure 4-2). This row of waling has the function to divide the newly
inserted stems equally over the side and put them in their proper position (Figure
5-14). Waling is often applied in shaping stake-and-strand baskets, since the multi-
m e m b e r e d elements are stronger than single-membered elements and are used to
force the strands of the side in place.
Figure 5-14 Detail of the transition from base to side of the stake-and-strand basket
depicted in Figure 4-2 (p. 61).
After one row of waling, the active system is fastened off. The passive system is
divided into two systems, which are both active (Figure 5-14). T h e rushes are
plaited in an over 1 /under 1 pattern (indicated as number 5 in Figure 4-2).
Both active systems are fastened off at the rim. T h e strands running from
right to left are bent and fixed behind the previous strand (number 6 in Figure 4-
2). T h e strands running from left to right are folded over a stick, plaited back into
the fabric and folded behind the previous strand (number 7 in Figure 4-2).
Discerning different basic structures and the role of the strands in these
(active/passive) forms the link to the production process. They not only represent
different phases of the work (Chapter 15), but are also an indication of the actions
of the basket maker (Chapter 16).
C H A P T E R SIX
AND ACTION...
6.1 Video and Archaeology

W h a t is the use of moving images in archaeology? Generally, this is thought to be
restricted to the documentation of archaeological field work and the
communication of archaeological activities to a broader audience (Struever 1995).
Through an ethno-archaeological approach, however, the scope of film in
archaeology is m u c h wider. The days that 'primitive' society was thought to be an
unchanged continuity of antiquity are long gone, however, so its value lies not in
giving a lively impression of ancient society.
The use of video in archaeology depends strongly on the subject. For the
study of production, with which w e are concerned here, the possibility of a
repeated viewing and analysis of the process, the use of space and time, and the
actions makes video an extremely useful tool for analysis.
Since field work inevitably is a learning process, registration of the
production process on video gives the possibility to correct or supplement earlier
interpretations m a d e with less understanding. This is only true to a certain extent,
however, because the level of understanding of the person w h o holds the camera,
is reflected in the record.
There is a difference between images and words, verbal treatise and visual
narrative (Heider 1976, 4; Collier and Collier 1986, 151; D e France 1989, 37).
The use of the word narrative indicates, however, that in filming and editing there
is an ordering principle at work. T h e film maker uses the subsequent images to
bring across a point. T h e narrative is not only inserted in the editing stage, but
even all unedited footage has some organization to it, determined, often partly
unintentionally, by the person holding the camera.
Still, there is a clear difference between unedited footage, and the narrative
that is a result of editing. T h e video that is part of this book sets out to give an
overview of the production process of several basketry techniques. T h e process
and action analysis, however, have been performed on the long sequences of
unedited footage, while the edited video shows examples. T h e editing and its
narrative, is based on these analyses.1 The timer in the right-hand bottom corner of
the screen enables references from text to image and vice versa. While watching
D e France discerns exposition and exploration, which go hand in hand and are based
on several filmed 'sketches' of the same situation from (literally) different angles (De
France 1989, 317-320).
the video, passages in the book relating to particular sequences can be found
through the video-timer index in Appendix C section 7.
Ethnography, anthropology and sociology have presented us with a wide
range of documentaries and filmed narratives on subjects as far apart as Inuit life,
African lion hunt, preparations of a French fashion model or basketry production.2
Between these films there is not only a difference in region and subject matter, but
also of reflection. There are several methods of using film in analysis of h u m a n
actions and behaviour, which are strongly linked to anthropological theory.
This is partly expressed by the different terms indicating the use of visual
tools in anthropology. O f American origin is the term visual anthropology (Heider
1976, Collier and Collier 1986), while the Dutch tradition (building on the work of
the anthropologist A.A. Gerbrands) prefers ethno-cinematography. Several French
anthropologists avoid the term cinematography, in favour of anthropologic
filmique, because the former term has too m a n y notions of directing and thus
influencing that what is filmed (De France 1989, x). Although this is realized and
accepted (De France, for instance, consequently uses the term mis en scene) it is
also thought that the filmed events have their o w n narrative (its "auto-mis en
scene').
John Collier describes h o w the camera is used to gather raw data, visual
information, which is subsequently analysed. The visual is not an illustration of a
text, but a separate basis of knowledge (Collier and Collier 1986, 13, 169-170,
178). In this positivist approach the camera is thought to record reality, and
enables us to "observe accurately phenomena w e do not understand" (Collier &
Collier 1986, 19). Video is not an objective medium, however, and its use should
not be compared with a laboratory situation where results can be repeated as a
'test'.
Watching the same video sequence twice is a repetition, rather than a reproduction
of results. The value of repeatedly watching the same process, rather than a large
number of similar processes, lies in the option to give a detailed analysis of one
particular sequence. Such an analysis is not possible on even a large number of
observations, because minor changes occur each time an action is executed.
Heider and D e France are concerned with the influence of the filmer at all
different stages of the process. Heider tries to m a k e explicit what makes a film
truly ethnographic (Heider 1976, 113), to be a tool for ethnography. D e France
starts at the other end by wondering h o w the use of film changes the observations
and descriptions in ethnology (De France 1989, 3).
Although all ethnographical film makers are aware of the direct influence of
the framing of the image (cadrage), it is D e France w h o defines exactly h o w this is
done. She distinguishes three types of dominant ethnographic subjects: material,
2Robert Flaherty 1922, Nanook of the North; Jean Rouch 1965, La Chasse au lion a
fare; Annie Comolli 1974, La Toilette; Claudine de France 1965, La Charpaiene see also
Heider 1983. '
6 A n d Action... 103
corporal and ritual techniques. Each of these require a different framing. The
material techniques, such as basket making, are filmed mostly close up, to show
the exact movements of the producers' hands. The corporal techniques, for
instance grooming, dancing or putting on make-up, are filmed in larger totals,
while the ritual techniques are mostly filmed in large overviews. In most sequences
aspects of these three techniques, and thus the framing, are combined.3
Basket making is a material technique and the analysis of the production
focuses, therefore, on the sequence of the process and the precise actions therein.
At the same time, the video that goes with this book, highlights corporal aspects,
such as working position, use of instruments and instrumental use of the body. The
interaction with the surroundings (socializing) of the basket makers could perhaps
be indicated as a ritual aspect, giving the widest frame of reference.
Comparing modern and ancient baskets is the first step in understanding
ancient basketry as a product of h u m a n activity. Studying the present day processes
is the next step and leads to an interpretation of ancient basketry production. Even
more so, because the apparent continuity in Egyptian basketry techniques enables a
comparison to great technical detail. Studying basketry production as a process can
be considered the macro-analysis of basket making (section 6.2).
T h e micro-analysis of basket making involves a detailed study of the actions
of present day basket makers. Such an action analysis gives insight in aspects of
professionality, apprenticeship and skill of the producers. By comparing the
actions with the material result of the end product, this information can be linked
to the ancient baskets and gives us the possibility to pronounce upon the
professionality, apprenticeship and skill of the ancient basket makers (section 6.3).
During filming and editing the film maker can highlight certain aspects,
which automatically results in moving other aspects to the background
(soulignement and estompage D e France 1989, e.g. 27). In the basket makers'
video some aspects have been clearly highlighted, not only by the frame and
editing, but explicitly by using codes, symbols and even freezing the image
(section 6.4). The verbal and the visual are joined in the voice over, which also
directs the attention of the audience. In the basket makers' video the role of the
voice-over has been restricted to a minimum.
6.2 Process Analysis

The analysis of the production process focuses on larger units in the continuous
flow of actions, k n o w n as macro-analysis (De France 1989, 6; Collier and Collier
I a m critical towards the separation of material techniques and non-reflexive corporal

techniques (De France 1989, 40, 122), since both have an actor (agent) and an object
(objet/patient). The corporal aspects of material techniques (the use of body weight, or parts
of the body other than the hands) are of great importance to understanding the physical
requirements of a technique. Nevertheless, to link one type of dominant framing (close-up
or total) to a specific subject matter is helpful both duringfilmingand editing.
1986, 181-182). To distinguish the phases, it helps to play the long sequences of
unedited footage in fast forward. Fluctuations in working rhythm and changes in
working position often indicate transitions from one phase to the next.
During observation of a production process, w e tend to schematize the
sequence, denoting the actions of the basket maker to their 'proper' production
phase. For archaeologists, w h o only have (fragments of) the product itself, this
tendency to simplify the production process is even stronger. Although a simplified
representation of the production has its merits, it helps to understand the process,
organizing and simplifying has serious limits too. By organizing the visual
information in a different ways, putting emphasis on different aspects, room is
given for n e w perspectives. The use of video enables this.
Based on the analysis of the production process through the actions,
recording the working order is possible. Also, the use of work space and the
interaction with the surroundings can be understood. A combination of macro- and
micro-analysis gives an understanding of the movements, the working rhythm, the
working position, and the physical demands of the work. Furthermore, it gives a
means of measuring the time involved.
In Chapter 15 the production process is presented as a schematized phasing,
based on the observation of basket makers at work, while Chapter 16 makes use of
macro-analysis of video images to understand the relation between the actions of
the basket maker and the sequences of the production process. The difference
between production phases and sequences is that thefirstpresent a logical order (a
generalized process), while the latter gives the actual actions of a particular
production process.
6.3 Action Analysis

Just as macro-analysis of video images is helped by watching in fast forward,
micro-analysis benefits greatly from watching the video in slow motion or frame-
by-frame.4 Actions with a steady working rhythm, for instance, can be found by
watching the video in fast forward, but the rhythm can be measured only by
analysing the exact movements.
Both Collier and D e France mention micro-analysis as a method to measure,
track and describe particular aspects (cf. Collier and Collier 1986, 182). The
aspects that I deem relevant for understanding ancient Egyptian basketry are the
actions of basket makers and the results thereof.
Dauer (1980) gives criteria to analyse actions, by focusing on the individual
movements which make up the action. H e uses the term Handlungsaublauf to
indicate the string of activities which leads to a certain result. In those activities
Dauer distinguished a start (Anfang), a course (Verlauf), an end (Ende) and a
transition to the next action (Ubergang). Furthermore, he discerns movements
(Teilbewegung) that are repeated several times and together m a k e up one action
4
A frame is the smallest unit in the video timer (25 frames per second).
6 A n d Action... 105
(Wiederholungeri). The micro-analysis which is performed in Chapter 17 is based

on Dauers' approach, but has been focused on the actions as part of a production
process, the major difference being that all actions are defined in relation to a
material result.
T o emphasize the difference between actions that consist of repeated
movements and actions that do not, two types are discerned: recurring (repetitive)
actions and occurring (non repetitive) actions. T o give an example: to knit a
sweater, m a n y stitches have to be made. This is the recurring action of knitting,
which is a repetition of four subsidiary actions: putting a knitting needle into the
fabric, guiding the yarn around the needles, pulling the yarn through the previous
loop and transferring the loop from one needle to the other. In the Netherlands, this
action is taught to children with the help of a rhythmical rhyme.5
A knitting needle m a y hold 120 stitches, so after 120 repetitions of the
knitting action all stitches are transferred from the left needle to the right needle.
At that point, the right needle, which n o w holds the fabric, has to be taken over
into the left hand and the empty needle is transferred to the right hand, to start the
next row of knitting. This action occurs only w h e n the complete series of knitting
120 stitches has been finished. It is an occurring action.
Determining the actions and the transitions between them is sometimes
difficult. The transition between two actions is often very brief.6 A s a criterion it
has been determined that a m o v e m e n t has to have a m i n i m u m length of one second
to be considered as a separate action, transition or break.
In the video several examples of the action analysis have been incorporated
as small codes in the left bottom corner of the screen. The code letters either refer
to a recurring (r) or occurring action (o), a transition between two actions (t) or a
break in the action (b). The numbers identify the actions, which are listed in
Appendix C section 4. Thus r 12 is the repetitive action of making a coiled basketry
stitch, while ol3 is the occurring action of inserting a n e w length of date palm
leaf.
T h e working rhythm is most strongly present in the recurring actions. Long
periods of repetitive movements are best sustained w h e n the work is done
rhythmically and are sometimes supported by work songs. These long periods of
repetitive movements also constitute the image that onlookers have of the
technique: coiled basket making is perceived as repetitive stitching only. The
occurring actions, such as adding n e w lengths of material, are less obvious and
often appear to be supportive.
"insteken, omslaan, doorhalen, af Iaten gaan"

In m y experience Dauers' distinction between start, course and end of an action is
too difficult to discern and too superficial, at least for m y purpose of analysing production
sequences. Often it is not possible to discern exactly where one action ends and the next
starts. In the analysis of the basket makers' work both are considered as part of a moment of
transition, which is often only one or two seconds long.
Micro-analysis not only gives the possibility to express the working rhythm,
but it also gives a new angle on the production. Rather than considering production
as a linear process, it is presented as a chain of actions (cf. D e France 1989, 169-
176).
D e France gives also attention to 'weak times' where the action is slackening
and 'dead times' where no action is undertaken (De France 1989, 9, 60-61). This
aspect has not been incorporated in m y analysis of basketry production, because
there is no material link to archaeological concerns. All actions that occur during
the production period, but which are not part of the production proper are,
therefore, considered as breaks. It should be noted, however, that although there is
no direct material result of a break, it often has a function in the process. Taking
brief periods of rest, to stretch the legs or relax back and shoulders is important to
be able to work for long periods. Communication with bystanders, an example of a
'weak' period, is of no direct use for the production of a basket, but is part of the
social strategy of the basket maker.
6.4 Highlights in the Video

Five aspects have been highlighted in the video by the appearance of symbols in
the upper right-hand corner of the screen. These indicate sequences in the film
where a particular aspect is clearly visible, but not in all instances where that
aspect occurs. It is a rough guide to basket land, which asks attention for the
properties of the raw material, the working position, the work-eye contact, the
instrumental use of the body and the use of instruments. In Appendix C section 2
the sequences are listed where these symbols occur and specifically what they
indicate.
Properties of the R a w Material
Materials differ in shape, size, tension strength, flexibility,
colour, appearance, and touch7. In the video some scenes give
a good impression of these properties. Several of these scenes
are highlighted with the 'palm tree' symbol.
Working Position
The basket makers are sitting with their legs crossed, or with
one leg stretched in front of them. The bed maker stands bent
over the bed frame. Frequently, a change in position
coincides with a change in action. Apart from the position, it
is relevant to note the orientation of the body towards the
basket, the instruments and the raw materials. The working
position has been indicated with the 'sitting m a n ' symbol
The "touch" of the material is the way in which it feels in the hands of the basket
maker. Grass has for instance sharp edges, palm leaves have large spines.
6 And Action. 107
Work-Eye Contact
Often the basket maker takes the eyes off the work while
talking to persons around him or her. W h e n there is a difficult
part, the eyes return to the work. Little work-eye contact can
be interpreted either as an expression of skill, or as a lack of
concentration, perhaps caused by disturbance due to the
presence of the camera. In the first case, the action is
purposeful and the work rhythm regular. In the second case
the work is hesitant, while the eyes are averted. S o m e of
these transitions between looking up and looking d o w n are
indicated with the 'eye' symbol.
Instrumental Use of the Body

Apart from the hands, often the feet, mouth, teeth, nails or
body weight are used in the production of basketry. This
occurs very generally and easily escapes attention. A number
of instances have been highlighted with the 'foot' symbol.
Use of Tools
The function of tools can be found by asking basket makers
which tools they use. The use of tools can only be recorded
by observation of the action. Repeated viewing and slow-
motion are useful to understand h o w the tools are wielded
exactly. In several cases the use of tools is indicated in the
video with the 'hammer' symbol, especially there where the
same tool is used for different purposes.
T o highlight the concept of active and passive systems, each of the five sequences
has one or more frames that have been frozen. The action is stopped briefly to
show which strand is passive, which active. These are listed in Appendix C
section 3.
CHAPTER SEVEN
SOURCES ALONG THE RIVER NILE
7.1 Selection of Sources

Attention has been focused on basketry from two archaeological sites: Tell el-
A m a r n a and Qasr Ibrim. W h y these two? At the start of this study, not m u c h was
k n o w n about Egyptian basketry. There w a s no basketry dating sequence, no
comprehensive work on the use of basketry, no analysis of basketry technology.
T h e most extensive publication on archaeological basketry in Egypt is Gourlay's
catalogue of the basketry from the workmen's village at Deir el-Medina (Gourlay
1981). His description of the basketry techniques is detailed and comprehensive.
Vital information is lacking, however: there is no analysis of the raw materials, no
references were given to the exact period, nor the context of the objects. Here not
the author of the two volumes is at fault, as m u c h as the excavators, w h o failed to
use field methods and publication standards considered proper in our time.1
In general, basketry features only marginally in excavation reports. There are
brief descriptions at the most, without technical details, drawings or information
on the context. In short: publications of basketry are not to the standard necessary
for the work to be undertaken.
Basketry in m u s e u m collections, often bought from antiquities dealers, has
rarely information on period and provenance, let alone on the exact archaeological
context.
It w a s decided, therefore, to base the study of basketry production o n
archaeological finds from recent excavations, which I could record on site.2 T h e
selection of sites was, therefore, limited to running excavations at which basketry
was found. Because of the generally good preservational circumstances in Egyptian
desert sites, there is a large number of projects that fit these criteria. The quantity
of material at A m a r n a was considerable, at Qasr Ibrim overwhelming. Other sites
were suitable too,3 but the total amount of basketry became so overwhelming that
During the 1934-1935 season Bruyere excavated the workmen's village at Deir el-
Medina and the eastern cemetery. In the course of this work most of the basketry retrieved
from Deir el-Medina was found (Bruyere 1937b, Bruyere 1939).
Since archaeologicalfindsare not allowed to leave Egypt, all staff members work on
site in a field laboratory. It gives the added opportunity to record the finds in situ and to
discuss matters with the excavators and other specialists.
In the same period I have studied basketry of the Dakhla Oasis Project (1990) and the
University of Delaware expedition to Abu Sha'ar (1990-1993).
11 o The World According to Basketry
the number of sites had to be limited. The two selected sites represent different
periods. They are both situated along the Nile Valley, but in different geographical
regions (Middle Egypt and Nubia).
Amarna
Amarna is situated in Middle Egypt, and was occupied in the N e w Kingdom
period, during the end of the 18th dynasty (around 1350 B C ) . All basketry was
found in a workmen's village which existed for a period of about 20 to 30 years
and was abandoned after that. The basketry from the excavations in the workmen's
village, which ran from 1980 to 1986, were stored at the site and studied by m e in
1987, 1989, 1990 and 1994.
Excavations at the beginning of this century in the workmen's village also
yielded basketry find. These have been published as part of an inventory of the
finds from each house and each room. Thus in a number of rooms baskets "of the
usual kind" have been found (e.g. Peet and Woolley 1923 p. 75), without
indication of technique, size and raw material. The finds from these excavations
are partly in the stores of the Supreme Council of Antiquities in Egypt, partly in
M u s e u m collections in Egypt and Great Britain. In many cases it has proven
impossible to find the reference of the baskets from these collections to the
publication. Since the context is not known in detail, the basketry previously
excavated at Amarna is mainly useful as comparative material for the techniques.
For Amarna a number of publications of archaeological material are of
importance.
Important comparative material is also found at another N e w Kingdom
workmen's village. The basketry from Deir el-Medina is ranging from the late
18th dynasty to the Byzantine period. Most basketry probably has to be dated to
the 19th and 20th dynasty. The excavations at Deir el-Medina have been published
by the excavator (Bruyere 1937b, 1939). The basketry has been published in a
separate and much later study (Gourlay 1981). As mentioned above, the drawback
of Bruyere's excavation and Gourlay's publication, is the lack of provenance, and
neither is there information on the approximate date of the objects. Since Gourlay
partly groups his material according to function (cf. Section 3.4.8 p.52-55), N e w
Kingdom and Ptolemaic sandals made in different techniques, appear side by side
without being recognised as belonging to different periods.
Qasr Ibrim
Qasr Ibrim was a fortified town in Egyptian Nubia, the area which is n o w flooded
by Lake Nasser. The remains found at Qasr Ibrim range in date from the Egyptian
N e w Kingdom to the abandonment of the site in the early 19th century (1500 B C -
1811 A D ) . The basketry was found in deposits from the 3rd to the 19th century
A D , with a strong emphasis on the earlier periods (3rd to 6th century A D ) .
Excavations at Qasr Ibrim started in 1963, but it was not until 1986, when
Boyce Driskell and Melinda Blustain started a basketry recording system, that
proper records were made. M y work in Qasr Ibrim started in 1990 at which time
7 Sources Along the River Nile 111
the previously used recording system was not brought under my attention or given
at m y disposal. Apart from the regular preliminary reports on the excavations in
general, some preliminary work on the Qasr Ibrim basketry has been published
(Driskell 1981).
The basketry from the early excavation seasons has largely been discarded.
F r o m 1974 to 1988 finds from Qasr Ibrim were divided between the Egyptian
Antiquities Organization and the Egypt Exploration Society. With the shipping of
boxes of antiquities to Great Britain, a considerable quantity of basketry, mainly
from the Islamic and Christian levels were sent to Cambridge. S o m e of the large
pieces of basketry and matting were used as padding for other materials, such as
pottery, bronzes and stone ware. At the m o m e n t the Oriental Institute in
Cambridge and the British M u s e u m in London both hold part of the collection of
basketry from Qasr Ibrim. This collection is not representative for the finds that
occur at the site, but a study of the collection gives information on techniques and
selection of raw materials. There is only a partial record on the context of these
finds.
Basketry from the present

Between the basketry of Amarna and Qasr Ibrim lies a period of 1600 years and a
distance of 1000 k m . Differences and similarities in the basketry of these two sites
can be attributed to a number of factors, of which the most important ones are
temporal, regional, functional and cultural. By comparing modern basketry from
the regions of Middle Egypt and N e w Nubia an attempt is made to understand the
variation over time, the regional differences as well as the function and meaning of
the ancient and present baskets.
For the A m a r n a region three villages in the vicinity of the archaeological site
were used as a base for ethno-archaeological research. Qasr Ibrim's modern
counterpart was slightly more complicated, since the entire area was flooded after
the building of the A s w a n High D a m in 1961. " N e w Ibrim" is one of the many
Nubian villages which were built near the town of K o m O m b o in the south of
Egypt to house the relocated Nubians. A survey was done not only in N e w Ibrim,
but in a number of other villages as well.
Ethnographic literature, which could help this line of research is scarce. N o
studies have been published on Middle Egypt, but there is a detailed study of the
Upper Egyptian village of Mari Girgis (Henein 1988). Henein has paid ample
attention to material culture, including basketry.
N o publications are available on present-day Nubian basketry, but the
American archaeologist and anthropologist B.N. Driskell, w h o in 1986 has done
field work in the same area as m y survey of 1992, has kindly m a d e hisfieldnotes
available. This information on basketry techniques and the social context of
basketry in N e w Nubia has been incorporated. In order to understand basketry
production, the help of basket makers was sought in the two areas of the ethno-
archaeological study.
7.2 A m a r n a
In this section a brief historical overview is followed by the excavation history of
the A m a r n a workmen's village and a summary of the recent archaeological work
in the workmen's village, which yielded the basketry finds. This is followed by a
descriptions of the village and an attempt to sketch the cultural background of its
inhabitants. The discussion w h o produced the baskets found in the workmen's
village is subject of Section 19.1
7.2.1 Historical Context*

The archaeological site of Amarna is situated in a barren area on the east bank of
the Nile with only a fringe of green land along the river (Figure 7-1).5 The oldest
remains encountered at Amarna are palaeolithic (French 1984). F r o m the Old
Kingdom period onwards the route to the limestone quarries at Hatnub led through
the valley, which was virtually uninhabited at that time. It was not until the N e w
Kingdom period that the area was used intensively. Around 1350 B C , towards the
end of the 18th dynasty, the new capital of pharaoh Amenhotep IV / Akhenaton
was built in the valley, to mark a new political and religious zeal. In approximately
20 years an entire city, including three major temples to the Aton, the single god
which was worshipped, five royal palaces and a large number of private houses
was built, lived in and deserted.
Figure 7-1 (on page 113)

The region of Amarna and the present day villages of el-Till, el-Hagg
Qandil, el-Amariyya and el-Hawata. The limits of the ancient region are
marked by boundary stelae (after Kemp 1985, 309).
high desert
modern settlement
V. edge of cultivation
•X boundary stelae
For a chronology of the Egyptian history see: Baines and Malek 1984, 36-37.
Amarna is also known in the literature as Tell el-'Amarna. This name is probably
derived from a misunderstood rendering of the name of el-Till (also indicated as el-Till el-
Amarna), the village where the Nile is crossed to reach the site. Amarna is not a tell (a hill
formed by archaeological remains deposited over a long period of time). The name Amarna
has possibly been retrieved from the Beni Amran, a nomadic people from the Eastern
Desert, which settled down in the region in the 18th century (Kemp 1985 p 309- Aldred
1988, p. 16).
:••'-. - '«•'.* V f t -,—.-
*' v'**•;'':* , *?*"&'
••- 1 '" ;-:v->:; d ! ^ L i**^ •.•'••*

'ft, -vp" •/^fe&^Jki
V . •£vw^".V' ; i C,.... ,y 4.i-iN \ .•'•
i
z
z
a ^ f ^ = ~ — °l :
,„.#-.jS
z
.,,N ' <\..-'''
»•••"' '_,,,._î"4
..***•* " v CD
"""€'' "'","" ••"'. Q f tf;
...•,--'-.^.->::r' ' .•"•',.-/u •• ••• . ,^*r
•C' ~^^^m fr.\£
.^ •.*<•.' ,._ ...

. __ 1
Excavations have shown that during the relatively brief occupation of the city quite
extensive modifications have been made to both official buildings and private
houses.6 Most of the town was built from m u d brick, as was the majority of the
official buildings. During the short life span of the town, a start was m a d e to
replace the m u d brick with limestone blocks ( K e m p et al. 1989).
In the face of the escarpment to the east of the city, 24 rock tombs were cut
for important officials of the community. The tombs at the south of the valley are
the earliest group. W o r k on the tombs seems to have been stopped suddenly and in
none of them an interment has been found. Neither has a burial ground been found
for the less important people of Akhet-Aton7. Six kilometre east of the town, a
large wadi had been chosen to become the Amarna valley of the kings. A number
of tombs were started, of which only one, the Royal T o m b , has been in use in
antiquity.
In a hilly outcrop near the Royal Wadi two villages were situated, the
workmen's village and the so-called stone village, situated 1 k m . east from it and
never excavated. The workmen's village is thought to have housed the builders
and artists w h o worked on the tombs.
The city Akhet-Aton has been the residence of pharaoh Akhenaton for 12
years of his 17 year reign. Desertion came with the death of the pharaoh and the
subsequent return to the old values and power equilibrium. After desertion of the
town, all stone building elements were demolished, the stone blocks being re-used
by inhabitants of settlements across the river.8
Although indications have been found of h u m a n activity during the 21st, 23rd
and 24th dynasty, it was not until the R o m a n period that the valley was used
Flinders Petrie excavated large parts of the central city in 1891-92, investigating the
Greater Aton Temple, the Smaller Aton Temple, the Great Palace, the King's House, the
Records Office and several private houses (Petrie 1894). In 1896 Barsanti excavated at the
Maru Aten palace (not published) and at the beginning of this century the temples were re-
examined (Bouriant 1903). From 1907 to 1914 the Deutsche Orient Gesellschaft, under
direction of L. Borchardt surveyed the entire region on both the east and the west bank of
the Nile, and excavated the southern and eastern part of the city. This work included the
mansions of high court officials and the house of the sculptor Djehutimes, in which the
famous bust of Akhenaton's spouse, queen Nefertiti, was found (Borchardt 1907-1914,
Borchardt and Ricke 1980). From 1921 to 1936 the Egypt Exploration Society excavated
most of the public buildings on the site (Peet & Woolley 1923 pp. 1-50 and 109-174;
Frankfort & Pendlebury 1933; Pendlebury e.a. 1951). From 1960 onwards the Egyptian
Antiquities Organization has excavated at the southern part of the site, near the modern
village of El Hawata and at K o m el Nana. From 1977 onwards the Egypt Exploration
Society has been involved in a long term excavation project, under directorship of B J
Kemp (Kemp 1980-1987, Kemp et al. 1984-1989).
A n exception to this is the workmen's village, where private tomb chapels and a
cemetery have been found (Peet and Woolley 1923, 92-108; K e m p e.a. 1985).
Many of these blocks have been found re-used in temples at Tuna el-Gebel cf
Figure 7-1 (Cooney 1965, Roeder 1969, Hanke 1978).
extensively again (Kemp 1986, 118; K e m p 1985, 317). In the Christian and
Islamic period the area seems to have been deserted, but no records are available
until thefirstEuropean travellers visited the site in the 18th century.
^ 1 0 ! M .
:: • x.m v.-:-.:- •;•"

•:..-.::;: - * * • ? &••<•}•„„
vV
n.tombI^,'.
•tv.t.
STONE u%,
<*&*£* S}iiV'-'.-.VILLAGE .•<•?•'•*•<••'*-<*?:*X'Z; - W ^
^yjokwEHi; .*•.»•• . ,: • • $>$$ . WyaLÔivib j
' / vil(?age ••:.;. ',,,".i>-- • •' >*'•••'*..?' * . i ^
/ fc. ' 'A?•;•»;•.•!:/•:;••

^ T M A R U ATEW ,&"'w: :-• •
TT7
Figure 7-2 The region of the ancient city of Amarna. fcS| | high desert
The workmen's village is situated in the
foot hills of the high desert modern settlement
(after Kemp 1985, 309).
edge of cultivation
X boundary stele
7.2.2 Excavation History of the Workmen's Village

The basketry found at A m a r n a comes without exception from an isolated
settlement of the N e w Kingdom period, k n o w n as the eastern village (Peet and
Woolley 1923, 51-91), the workmen's village (Pendlebury 1951, vii; K e m p , K e m p
et al.) and the walled village ( K e m p 1987, 24, referring to the village proper).
This village is situated 2 k m East of the main city, in a small hilly outcrop of the
escarpment which forms the boundary of the valley (Figure 7-2).
Despite the small distance, the preservational circumstances of the main city
and the village are very different. Since the precipitation is the same at both sites
and the subsoil water is low both in the main city and the village, the difference
must be attributed mainly to the activity of termites (Isoptera). Indications of
extensive termite activity in the main city were found, for instance, in the
workshop of the sculptor Djhutimose (Aldred 1988, p. 59). T h e only remains of
basketry in the Main City are impressions in m u d of matting used in the
construction of roofs.
In 1920-21 the Egypt Exploration Society excavated 37 houses of the
workmen's village, representing half buildings. The attention was focused on town
planning and architecture. For each house the finds were listed room by room, but
no interpretation was given of their deposition and the descriptions were very
scanty. Directly East of the village 22 private tomb chapels were excavated also
(Peet and Woolley 1923, 92-108).
In 1979 B.J. K e m p resumed excavation of the workmen's village for the
Egypt Exploration Society and he continued working there until 1986 ( K e m p 1980-
1983, 1987; K e m p et al. 1984-1987). H e excavated six house plots inside the
village, three in the centre ('Long wall street' 6, and 'Gate street' no. 8 and 9) and
three in the western section ('West street' no. 1,2 and 3, see Figure 5-3).
The spaces which Newton, w h o planned the village in 1921 as part of Peet
and Woolley's excavation, dubbed 'West Street 1 and 2' were not complete
houses. In 1985 the space of 'West Street 1' appeared to be open, with some small
enclosures built in the southern part against the village enclosure wall. T h e space
assigned to 'West Street 2' was partly an open space too, partly an annexe to the
next house in the row, West Street 3 ( K e m p e.a. 1986, 28-33).
The objective of the excavations inside the walled village w a s to have a
better understanding of the situation in order to interpret the mass of data from the
excavations in the twenties. Most attention, however, was directed to the area
outside the village walls (Figure 5-4). A number of activity areas could be
identified, which raised the possibility of an economic and social interpretation of
the village community. Three more tomb chapels were discovered and excavated,
which led to a re-interpretation of the work done in 1921/22 ( K e m p e.a. 1985). In
1987 the attention was directed towards the connection between the main city and
the workmen's village. The well which supplied the village with water w a s
identified at the edge of the main city in an industrial area with pottery workshops
( K e m p e.a. 1989). The combination of large-scale clearance and selective
7 Sources Along the River Nile .17
question-based excavation has given a good basis for interpretation, although in the
process many new questions have risen (Kemp 1987).
Ii
b J J L L i J "1 \~ h
M •tr u - ^
n
if I—
± _ U i i-
^ *-
I"
t i r~
E L I
U _ t
11 M
W
** i
\-°
r
i—i-
l = U
L l L i
-*,;- L
~\ \
30 N
Figure 5-3 Plan of the Amarna workmen's village (based on Peet and Woolley 1923,
Plate XVI and Kemp e.a. 1986). Numbers indicate the standard numbering
of the houses, as given by Peet and Woolley.
7.2.3 Description of the Workmen's Village

The village measures 69 metre square and has a very regular ground plan. Four
rows of houses form four streets, which Peet and Woolley have named East Street,
Gate Street, Main Street and Long Wall Street. West Street has a row of houses on
both sides. The enclosure wall and the outer walls of the houses are m a d e of Nile-
m u d bricks, similar to those used in the main city. The walls inside the houses are
m a d e of desert marl bricks and show considerable variations. All buildings outside
the walled village are also m a d e of desert marl bricks, which have been quarried
just outside the village. Thus the Nile m u d brick is used for a well-structured,
government provided village structure, which was adapted by the villagers to their
needs and taste with local desert marl brick.
Gate Street 8 probably had a second storey, whereas the excavation of
houses in West Street only yielded traces of light awnings on top of the collapsed
roof. In general the houses measure 5 x 10 m and are divided in a number of
rooms. The interior structure of the houses showed some variation. Usually there
is one room in the front and one large central room which is the main living area.
The kitchen, the bed room and the stairs to the roof are placed in the back of the
house.
There are two exceptions to the uniform size of the houses. In the south-east
corner of the village there is one larger building, probably the house of a village
official. The second exception is the entire west side of the village, which w a s
added on at a later stage and had its o w n entrance gate, being separated from the
rest of the village by the old enclosure wall. In this part of the village (West
Street) the houses are smaller. Since Peet and Woolley did not find m a n y artefacts,
nor a large quantity of roofing material, they suggested that the houses of the west
quarter were stripped of all w o o d and other useful items, because they had been
abandoned before the rest of the village (Peet and Woolley 1923, p.67). During
later excavations this did not appear to be the case. K e m p argued that the western
part was a poorer quarter, which was probably occupied by a police force of about
20 m e n , whose task was to check the desert routes ( K e m p 1987, p.27, 28, 48).
In the basin around the walled village a number of activity areas were
identified. These were found to be m a d e of locally derived materials, their
arrangement was informal and their design was diverse. They can be interpreted as
unofficial private enterprises of the village inhabitants. T o the east of the village,
situated on the side of the hills surrounding the village, at least 23 private chapels
were excavated, established by the local community. Below and in between these
chapels, there were a number of plant beds and animal pens. Although all fresh
water had to be carried into the village on donkey back, the villagers were growing
vegetables and breeding pigs. Seven sets of animal buildings were identified,
which show many differences in lay-out.
A s indicated above, the water w a s supplied by a well in the main city of
Amarna. In an area near the village gate and the agricultural sites, about 50
implacements for large pottery jars were found. It is indicated as the Zir-area,
derived from the arabic word for large water storage jars. Here the water was
distributed, the number of pot implacements suggesting that each household had its
own jar to receive the water supply. A survey of the number of distinctive sherds
found between the village and the main city, identified the route from the zir area
to the well (Kemp et al. 1987, 1-14).
' PERIMETER
ROAD
r ^
LLr
fj-STT \H
rt-3'-~> r-
U
Ftp
CEMETEBr ™
' ANIMAL PENS
=J;ri Si a
•**-*%&
C£%: JANIMAL '••' >,'
PENS
\ QUARRY jT
a ^
->
"T ^ . , ,,,(3
TO MAIN CITY
Figure 7-4 The wider area of the workmen's village, with the road to the main city
(based on K e m p 1983, p.6, fig. 1).
Around and through the valley in which Tell el-Amarna is situated, a

number of desert roads have been identified (Petrie 1894, pp. 4-5, pi. X X X V ;
Davies 1905, pp. 5-6; T i m m e 1917, pp.24-26). O n e of these paths encircles the
workmen's village. O n the spot where this perimeter road met with the main road
from the village to the city, a guard house was found consisting of three well built
rooms and an area in which goats and other animals were kept. This complex was
excavated in 1979, and indicated as site X 1 (see Figure 7-4) ( K e m p 1980).
Other extra-mural activities which could be identified archaeologically were
the quarrying and discarding habits of the villagers. The desert marl for the local
m u d brick was quarried at an area 35 metres south of the village and in three
smaller areas nearer to the enclosure wall. These pits must have been dug at the
time of the construction of the village and have been used as rubbish dumps ever
since. The stratigraphy and nature of the rubbish did yield important information
on the history of the village and the activities of the inhabitants.
7.2.4 Population of the Workmen's Village

The village was designed and built in one go. This ready made village was then
occupied over a very short stretch of time. Perhaps an existing community moved
in, imported as one group from elsewhere. A n alternative interpretation is that
over a short period of time a new community was assembled of w o r k m e n or artists
from all over Egypt. The population of the village consisted in thefirstinstance of
48 households. Nothing is known about the size and nature of these households.
Even the occupations of the inhabitants remain unclear.
Several authors have suggested that Akhnaton had the artists' community of
Deir el-Medina move to this new workmen's village (Peet and Woolley 1923, pp.
51-91, Aldred 1988, pp. 61, 68). This community had been digging and decorating
the tombs in the Valley of the Kings at Thebes from the beginning of the 18th
dynasty onwards. A great deal is known about daily life and work at Deir el-
Medina, since many ostraca were found giving a lively account of the village and
its occupants (Valbelle 1985, 23-25; Bierbrier 1982). The similarity of the village
lay-out and the cult chapels led to the conclusion that A m a r n a and Deir el-Medina
housed the same people.
Apart from these formal features, the Amarna workmen's village did not
yield many comparable remains from the occupants themselves. The A m a r n a
w o r k m e n did not leave any personal messages and letters on potsherds, as the Deir
el-Medina workmen did. However, this seems to be a characteristic of the Deir el-
Medina workers of the 19th dynasty. Secondly, hardly any tools for cutting or
decorating tombs were found at Amarna,9 while at Deir el-Medina chisels, paint
brushes and other tools were found (Bruyere 1939), but these probably stem from
a later date in the history of the village.
9
The only tools from the extensive early excavations are two hammers and a brush for
whitewashing (Peet and Woolley 1923, PI. 19, 3,4).
It proves impossible to decide if the Deir el-Medina village has been

deserted or not at the end of the 18th dynasty, a feature which could be decisive in
the discussion if the community moved for two decades to A m a r n a ( K e m p 1978,
44; Valbelle 1985). Barry K e m p has argued that it is not a given fact that the Deir
el-Medina crew was transferred to Amarna. O n the basis of the activity pattern at
the A m a r n a workmen's village he suggests that the tomb workers might have c o m e
from the region of Memphis, where an equally capable team of workmen,
responsible for building the Saqqara tombs, must have been available.10
The fact that no tools were found at the A m a r n a workmen's village might
be due to the w a y in which the site was abandoned. If the inhabitants would have
m o v e d back to M e m p h i s (or Deir el-Medina), they can be expected to have taken
their tools with them. Alternatively it can be hypothesised that the base c a m p for
the tomb workers was not the walled village, but the stone village, which is
situated one k m to the north east, near the entrance to the Royal Wadi. Since the
stone village has not been excavated this is where the argument has to halt.
The abandonment of the site still is enigmatic (cf. Hari 1984). The
simplistic supposition that A m a r n a and the workmen's village were abandoned at
Akhnaton's death, or a few years after, does not hold. A large number of ring
bezels from the reign of Tutankhamon was found at the workmen's village and a
wall painting discovered during the 1986 excavations at the West street extension
of the workmen's village, depicting a royal figure, was attributed to the reign of
Tutankhamon (Weatherhead 1987). This painting was probably made w h e n the
house w a s first erected, which dates the building of the house to the reign of
Tutankhamon. Thus West street was built at the m o m e n t that the site was thought
to be abandoned.
The stratigraphy of the rubbish d u m p in the Main Quarry suggests that the
village has been abandoned for a while, after which it was re-occupied. Until n o w
it is not clear if the village was lived in by a tomb building community, which
abandoned the site after the death of Akhnaton, and if so, w h o were the n e w
occupants w h o lived in the village. It is not very probable that this was an entirely
different group since the pig breeding went on until the final abandonment. Part of
the village has been occupied by a police force, patrolling the desert roads and
checking the entrance of the village, as has been inferred from the position of the
guard house or check point X I (see Figure 5.2). Such a police force is a likely
K e m p argues that the village at Deir el-Medina has no parallels for the z/V-area.
Another feature that does not occur in Deir el-Medina is the breeding of pigs. As a parallel
to the animal pens at Amarna, he refers to the breeding of pigs in the temple of Ptah at
Memphis in the late 18th dynasty (Kemp 1987, pp. 44-45). Certain characteristics of the
Amarna style are found in the work of memphite artists of the 19th dynasty (Berlandini 97,
250 note 1 and 2).
group of occupants of the West street extension". Guards formed a professional

group which might have been employed to watch the tombs in the turbulent period
after the death of pharaoh Akhnaton. By the time of pharaoh Horemheb the village
was abandoned ( K e m p 1987, 43).
7.2 Middle Egypt

Ethno-archaeological research was done in three of the four villages which are
located near the remains of ancient Amarna. F r o m north to south these are: el-Till,
built partly over the northern suburbs of ancient Amarna; el-Hagg Qandil and el-
Amariyya forming one long string of houses parallel to the river bank near the
southern suburbs of the ancient town; and el-Hawata in the south of the valley (see
Figure 7-1). T o the outsider el-Amariyya and el-Hagg Qandil might appear to be
one village, but the inhabitants experience a clear difference. El-Hawata is the
most isolated village in the valley. Traditionally the inhabitants of the other
villages consider el-Hawata a 'bad' village, of which the population is not to be
trusted. This prejudice is strengthened by the growing and trade of opium which at
present is one of the illegal activities of some of the el-Hawata inhabitants.
Because of this bad reputation m y research was limited to the three other villages.
All three villages have basket makers. O f old, the two villages el-Hagg
Qandil (supported by el-Amariyya) and el-Till are rivals, so w h e n the inhabitants
of el-Hagg Qandil are asked if there are basket makers in el-Till, the answer is
negative. If the el-Till people are asked if baskets are made in el-Hagg Qandil the
answer is negative too. Even though the two villages are only six kilometres apart,
and the inhabitants have relatives in the other village, in both villages the answer
was " W e make baskets, they (from the other village) do not m a k e baskets. But if
you want really good baskets you have to go to Bersheh".12
In Bersheh both m e n and w o m e n make baskets which are sold on the
market in Deir M a w a s or Mallawi. The w o m e n of el-Till, el-Hagg Qandil and el-
Amariyya make the same kind of coiled baskets the w o m e n of el-Bersheh, but they
only produce them for themselves or for friends. Not all the w o m e n k n o w h o w to
make the coiled baskets.
The m e n never make coiled baskets, but they make carrier baskets using the
sewn plaits technique (see Chapter 8). M a n y know h o w to make these baskets, but
only a few m e n in el-Hagg Qandil actually produce them on a regular basis. They
do this part-time besides farming and sell their baskets in the village and
sometimes across the river. In el-Till there seem to be less m e n involved in semi-
professional basket making.
Another indication of the presence of police or guards in the village, is the top of a
wooden military standard, which was found in the Main Chapel in 1986. Such an object
would be more consistent with a policing unit than with tomb builders (Kemp 1987, 46).
"Bersheh is a village just to the North of el-Till, but in the next side valley, separated
by the escarpment of the high desert which forms a natural barrier (see below).
In Amariyya there is a producer of twined mats, used as door mats, donkey

saddles and manure bags. H e does this work full-time, as does the qafas maker,
w h o has a workshop where he and his three sons are making crates out of the
midribs of the date palm leaves. These crates are used to transport the harvest
(especially oranges and tomatoes). For woven matting, which is not made in the
area around Amarna, one has to cross the river. Rope making and knotting large
rope carrier nets is a skill that all farmers seem to have, as is the production of
wind shelters (sabata) from palm leaf and rope.
The range of professions in the three villages is very limited. Most
inhabitants are small farmers. Recently the government has increased the area of
arable land by a large irrigation project. The land is often worked by several
m e m b e r s of the family, producing what is needed for daily sustenance. Crops that
are grown are beans, peas, lentils, salad, potatoes, tomatoes, turnips, carrots,
sesame and lucerne clover. The surplus is sold at the local market or across the
river in the larger population centres at Deir M a w a s or Mallawi.
S o m e of the m e n have a part-time profession apart from farming, being a
village guard, a butcher, shopkeeper or basketmaker. There are very few full time
professionals, even government jobs are performed alongside farming. Most full
time professionals are living and working in the larger communities across the
river where a division of labour is more c o m m o n . In el-Hagg Qandil there is a fish
farm, in which Nile fish is grown and caught with nets.
Apart from growing crops, most families o w n a number of animals:
poultry, ducks, rabbits, goats, sheep and, for wealthy families, cows and water
buffaloes. Donkeys are the most important means of transport. The animals are
often kept in a separate building or enclosure, which also contains the bread oven.
These enclosures are situated in the village near the house.
For a large part the families are self-sufficient, which is the result of the
activities of the w o m e n . All w o m e n k n o w h o w to bake bread and make cheese and
butter. Apart from farming, cooking, cleaning the house and doing the laundry,
m a n y w o m e n know h o w to sew clothes or make baskets.
Until the 1980's most houses in all three villages were built of m u d brick.
With the building of the A s w a n High D a m the yearly floods which brought new
Nile silt have ceased. The government realised that the fertile soil should not be
wasted to make bricks. N o w , under government regulation, there is an increasing
use of fired brick and concrete.
The four villages near A m a r n a were quite isolated. In Middle Egypt the
course of the River Nile runs near the cliffs of the Eastern Desert. The flood plain
of the Nile is 20 k m wide on the west side and only 100 metres wide on the east
side of the river. Only where the cliffs along the east bank recede, forming semi-
circular side valleys, there is space for settlements.
The Nile has to be crossed for each contact with the larger towns on the
west bank, but also tot travel to neighbouring villages on the east bank, because
there is often no space for roads between the cliffs and the river of the west bank
to travel from one semi-circular side valley to the next. The only way to get across
was by sailboat ferry's (felucca's) and rowing boats. Since 1977 the isolation has
diminished gradually with the coming of telephone lines, two car ferries and the
Eastern Desert motor road, which runs a few kilometres to the east on the High
Desert Plain.13
Traditionally there were also contacts with the Eastern Desert population. In
the early eighteenth century a nomadic group from the Eastern Desert, the Beni
Amran, settled in the A m a r n a area. The village across the river from el-Hagg
Qandil is still called Beni Amran. The n a m e of the village of el-Amariyya is also
thought to have been derived from this nomadic group (Aldred 1988 p. 16). The
ancestry of the Beni A m r a n is not important in the identity of the villagers, w h o
never refer to a nomadic background.
The river can be crossed from each village, but extensive contacts exist
especially between east bank el-Till and west bank el-Till. The villages in the
region of el-Till have profited most from the tourism to Amarna. Although the
A m a r n a monuments are not on of the top ten tourist attractions, the northern part
of the town and the North T o m b s were visited regularly, before tourism dropped
severely after the Luxor attack in November 1997. This decline in tourism also
affects the w o m e n of el-Till, w h o had started to sell their baskets to the tourist.
This new market had gradually caused a change in the range of baskets they
were making. Large flat bread baskets, with a diameter of 75 c m do not fit in a
suitcase, so the w o m e n started to make the same model with a smaller diameter to
accommodate the traveller. Also other small baskets, for storing trinkets, or
serving biscuits, were produced because they were attractive for the tourists and
easy to make. The haphazard sales at the tourist ferry, did not lead to a real
professionalization of el-Till basket making.
7.4 Qasr Ibrim

Qasr Ibrim, 237 k m south of A s w a n on the East bank of the Nile (See the m a p
of Egypt opposite page 1), was an eagles' nest on a hill top, overlooking a large
stretch of the Nile river. Steep cliffs ran straight from the fortifications into the
waters of the river, 66 metes below. The construction of the A s w a n High D a m ,
finished in 1960, has caused the River Nile to rise and spread over a large area.
The villages along the river banks and the cemeteries in the valley are n o w 63
metres under water at the bottom of Lake Nasser.
Rescue excavations concentrated mainly on the cemeteries (Adams 1984,
pp. 71-88, Trigger 1965 pp. 36-42). Qasr Ibrim is unique because it is one of the
Since 1977 el-Till has a car ferry, which accommodates the local inhabitant, the
tourists and government trucks transporting limestone blocks quarried near the ancient
quarry of Hatnub. From el-Hagg Qandil there is a regular ferry service to the village of
Beni Amran. Since 1989 el-Hagg Qandil has a car ferry too, which is also used by the
inhabitants of el-Amariyya. The most southern village, el-Hawata, has no regular link to the
west bank.
few settlement sites excavated, it used to have an extraordinary level of

preservation of organic materials and it is the only area in which excavations are
still possible today. Since 1998 this site has come under immediate threat,
however, because the level of Lake Nasser has been allowed to rise to its
maximum.14 Qasr Ibrim, at present no more than three metres above the water
level, n o w is a small island.
In this section a brief historical overview is followed by the excavation
history of Qasr Ibrim. A description of the town, concentrating on the area where
the baskets that feature in the next chapters, were found precedes a sketch of the
cultural background of the inhabitants of Qasr Ibrim in the late Meroitic and early
Ballana period (ca. 100-600 A D ) . The discussion w h o produced the baskets found
in Qasr Ibrim is subject of Section 19.2.
7.4.1 Historical Context

The earliest remains at Qasr Ibrim are from the N e w Kingdom period (1550-1070
B C ) consisting of a stone temple and a number of shrines (Caminos 1968). Other
N e w Kingdom remains are a number of re-used stone blocks and door jambs with
inscriptions, but no evidence of fortifications were found. The habitation centre in
the N e w Kingdom period was across the river, where the town Mi'am (the later
Aniba) was situated. This was the administrative centre of W a w a t (Lower Nubia).
The next evidence for building activity at Qasr Ibrim dates to the Napatan
Period, which runs parallel to the Egyptian Late Dynastic Period (ca. 1000-300
B C ) . A strong dynasty arose in the southern Nubian city of Napata and ruled
Egypt as the 25lh dynasty. During the reign of king Taharqa, at least three temples
seem to have been in function in Qasr Ibrim. Taharqa built a large temple,
opposite wich the denuded remains of a second temple, have been found, perhaps
dating to the same period (Plumley 1974, 228-236). A third temple, dedicated to
several gods, including Isis and Horus of Mi'am, also functioned in the Napatan
period (Driskell 1989).
The first indications for the use of Qasr Ibrim as a settlement and fort are
dated to the Meroitic Period (ca. 300 B C - 350 AD).15 In the early Meroitic period
there were extensive building activities, both monumental and occupational.
Remains have been found of a large Meroitic temple (Plumley 1978, pp. 31-33)
which has been demolished in antiquity. In the temple area and on the desert road
towards Qasr Ibrim a large number of life-size carvings of feet, often adorned with
Meroitic inscriptions, have been found. These indicate that Qasr Ibrim was a
14
The high level is maintained as preparation for the Toshka project, a large scale
government undertaking to create a 'second Nile River' to feed the ever increasing Egyptian
population.
The Napatan court moved from the old capital Napata, which was sacked in 591
B C by the Egyptian/Persian king Psammetik II, to the southern city of Meroe. Trigger
indicated the Napatan and Meroitic period as the Late Nubian tradition and Reisner
designated the term W-group to this culture of strong rulers in Napata and Meroe.
centre of pilgrimage in the Meroitic period. Finds within the Meroitic living areas
at Qasr Ibrim indicate that there were contacts with Ptolemaic Egypt, but there is
no indication of a Ptolemaic occupation of Qasr Ibrim.
The border between Egypt and Nubia at that time was established at
Maharraqa, north of Qasr Ibrim. Soon after the R o m a n occupation of Egypt in 30
B C the Meroites reconquered the north of Lower Nubia and sacked Aswan. This
led the Romans, under the prefect Petronius, to organise a punitive expedition in
23 B C which penetrated deeply into the Meroitic empire. O n the way back, a
R o m a n garrison of 400 m e n with supplies for two years was left at Qasr Ibrim
(Primis) (Adams 1984 p. 340, Kirwan 1959, pp. 24-25). The R o m a n s created a
Nubian-Roman province, the Dodekaschoenus, between A s w a n and Maharraqa.
During late Ptolemaic and R o m a n Periods the Nubian material culture is
named after the type site of Ballana.16 The historical sources of that time sketch a
complicated image. A nomadic people from the Eastern Desert, the Blemmyes,
flocked into Nubia. Because of the threat of the Blemmyan raids the emperor
Diocletian moved the R o m a n border from Maharraqa to A s w a n in 289 A D , and
invited another tribe, the Nobatae or Nobadae, to settle in the abandoned area.
Prim (Qasr Ibrim) was the seat of either a Blemmyan or a Nobatian king. Texts in
Greek found at Qasr Ibrim in 1976 indicate the co-existence of a B l e m m y a n and
Nobatian ruler in Lower Nubia, but it is not clear which of the two ruled from
Qasr Ibrim (Adams 1984, 422-423; Plumley 1977, 74-75; Skeat 1977).
Towards the end of the Ballana period the role of Christianity became
increasingly important. The formal conversion of Lower Nubia (the Nobatian
Kingdom) took place under the emperor Justinian in 543 A D . The churches found
at Qasr Ibrim indicate that Christianity had entered Nubia quite some time before
that date. A Napatan temple which was dedicated to several gods including Horus
of Mi'am, was intentionally destructed in the early Christian period. It has been
maintained that the Cathedral at Qasr Ibrim was built in the sixth century, with
financial and technical support of Justinian, from re-used meroitic stone blocks on
top of an older church (Adams 1984, 474; Plumley 1975, 17).
By the time of the Arab conquest in A D 640, Qasr Ibrim was part of the
powerful Christian kingdom of Makuria, with the city of Dongola in Upper Nubia
as its capital. In Egypt, the Arab conquest was welcomed by the Christian
Monophysites, w h o were suffering from oppression by the orthodox Christians.
Makuria did not welcome the Arabs, however, and the inhabitants of Lower
Nubia, managed to stop the Arab invasion at the Aswan border.
Although an Arab attack on Dongola took place in 652, Nubia did not
become part of the Islamic world. The relations between the Caliphate and the
Nubian Christian kingdoms were arranged in the Baqt treaty, which agreed on a
steady supply of slaves by Nubia in exchange for Egyptian commodities. The
This so-called Ballana-culture is the same as Reisner's X-group and is usually

dated from 350 - 700 A D .
treaty forbade Islamic merchants to settle in the area of the Christian kingdoms
(Adams 1984, 450-458).
The period from A D 850 to 1173, which is indicated as the Classic
Christian period, was again a time of relative quiet and prosperity. Qasr Ibrim held
the seat of one of the three bishoprics of Lower Nubia17. Nubia lived in peace with
Egypt during the Fatimid period and supported the Fatimid rulers against the
Ayyubids. W h e n the Ayyubid ruler Salah ed-Din came to power, this support
resulted in a punitive expedition by Saladins brother Shams ed-Dauwlah. In 1173
Qasr Ibrim was attacked and the cathedral was pillaged. Despite the fact that
Islamic troops were left in Lower Nubia to control the population, the area
remained Christian for at least another century (Adams 1984, 531-536).
F r o m the 12th to the 15th century, Christianity in Lower Nubia came more
and more under pressure by the Bahri Mameluks from Egypt and Arab tribes from
the Eastern Desert. It was an Ottoman army, however, which penetrated into
Lower Nubia to make an end to the Beduin raids, occupied Ibrim and left a
garrison to guard the Ottoman interests (Adams 1984, 609-621).
The garrison consisted mainly of Bosnian soldiers w h o , according to the
European traveller Burckhardt (1819), were payed directly by the Ottoman Sultan.
This rendered the troops independent from the Egyptian Pasha's, which were
ruling Lower Nubia, from the Bosnian conquest in about 1520 onwards, by means
of kashefs (local tax-collectors). The Bosnian garrison remained in the area and
mingled with the local population.
The end of the occupation at Qasr Ibrim came in 1811, when a few
M a m e l u k e nobles, w h o had escaped the attack by M o h a m m e d Ali, fled to the hill
fort. M o h a m m e d Ali's army attacked, conquered and ransacked Qasr Ibrim,
chasing the Mamelukes further south to Dongola. After 1811 Qasr Ibrim was a
ghost town, only visited by occasional western travellers, the people of the area
living in the hamlets along the river bank.
7.4.2 Excavation history of Qasr Ibrim

The archaeological knowledge of Lower, or Egyptian, Nubia is relatively plentiful,
since large area's were excavated in connection with the building of subsequent
dams in the Nile. Three times an increasing area of Nubia was flooded, in order to
ascertain Egypt's water supply and each time a large scale archaeological survey
was organised, in an attempt to record the archaeological remains which were
under threat.
Thefirstd a m was built between 1889 and 1902 and enlarged between 1908
and 1910. In connection to these constructions, the First Archaeological Survey of
Nubia took place between 1907 and 1911 under direction of G.A. Reisner and
The other two traditional cultural centres, Faras and Qurta held the other two
bishoprics, the capital being at Faras.
CM. Firth.18 This survey concentrated on Lower Nubia, the area under immediate
threat. Although at that time the buildings on the mountaintop of Qasr Ibrim were
by no means under threat, the cathedral of Qasr Ibrim was surveyed by the
architect Mileham (published in 1910).
Figure 7-5a (on page 129, top)

Qasr Ibrim area, with the old course of the Nile River (after Mills 1982,
Plate III)
Figure 7-5b (on page 129, bottom)
Plan of Qasr Ibrim (after a plan made by M . Horton in 1982). The dotted
line indicates the water level of 1982.
A Structure X 265, excavated in 1990 and 1992 (cf. Figure 7-6).
B East Gate, destroyed by the water in 1982 (Plumley 1975, 21).
C South-West Gate (Plumley 1970, 17, Plate X X I V 1, 2, 3, Plate X X V 1).
D West stairs.
E Bastion or South Bastion (Anderson, Adams ea. 1987, 33-35).
F Podium (Frend 1984).
G Temple 1, of unknown date (Alexander, Driskell 1985, 23, fig. 1).
H Temple 2, Meroitic, dismantled in antiquity, also indicated as Temple 4
(Anderson, Adams 1978, 32-33; Plumley, Adams, Crowfoot 1977 40-
42).
J Temple 3, New Kingdom temple rebuilt by Taharqa, incorporated in a
Meroitic temple complex and modified into a church, indicated as
Church 3 (Plumley 1974, 228-236; Plumley 1975, 19-20).
K Temple 5, a much denuded rectangular structure, possibly from Napatan
date (Alexander, Driskell 1985, 24-26).
L Temple 6, Napatan temple, modified in the Roman period, destroyed in
the Christian period (Driskell 1989).
M Church 1, the Cathedral or the Great Church, modified to a mosque
(Plumley 1964). Underneath the cathedral the remains of an older church
(Church 4) were found (Gartkiewicz in Alexander, Adams ea 1978 37-
38).
N Church 2, the Small Church or South Church, dismantled by the
excavators (Plumley 1966, 10; Plumley 1974, 214)
On the basis of his excavations of Cemetery 7 in Shellal, Reisner developed a

classification of cultures, which he named A, B, C, D, W , X and Y-group. This cultural
chronology has been used ever since, although some adjustments have been made The term
group illustrates that Reisner interpreted changes in styles of pottery and burial customs as
different peoples which inhabited the area. Later archaeologists have explained changes in
pottery and grave styles as the result of gradual developments rather than disruptive changes
and migration.
ANIBA
REGION
^*i*
tMr.'^.
^fa®K
•"^•ttVSS^v
900
The Second Archaeological Survey of Nubia was directed by W . B . Emery

and L.P. Kirwan from 1929 to 1934. The area under threat stretched from A s w a n
to the Sudanese border at Wadi Haifa. In Qasr Ibrim Emery and Kirwan
performed some small scale excavations in the lower parts of the large cemeteries
from the Ballana and Christian periods in the two valleys North and South of the
town (Figure 7-5).Because the cemeteries were not under immediate threat, the
excavation and the subsequent publication (Emery, Kirwan 1935) did not
concentrate on Qasr Ibrim.
The scale of the third salvage operation was immense, because the building
of the Sad el-'Ali, or Aswan High D a m , which started in 1959, threatened the
entire area of Egyptian Nubia and large parts of Sudanese Nubia as well. With the
help of U N E S C O funds became available not only for spectacular projects, such as
the relocation of the A b u Simbel temple, but also for large scale excavations which
took place from 1960 until 1964.
Year Director Publications
1 1934 Emery and Kirwan (Emery and Kirwan 1935)

2 1961 Emery (Mills 1982; Caminos 1968)
3 1963 Plumley
4 1966 (Spring) Plumley (Plumley 1966)
5 1966 (Dec.) Plumley (Plumley 1967)
6 1969 Plumley (Plumley 1970)
7 1972 Plumley (Plumley 1974, Frend 1974)
8 1974 Plumley (Plumley 1975)
9 1976 Plumley; (Plumley and Adams 1977)
10 1978 Adams and Anderson (Anderson and Adams 1978)
11 1980 Adams
12 1982 Adams (Adams and Alexander 1983)
13 1984 Alexander (Alexander and Driskell 1985)
14 1986 Alexander and Driskell (Driskell 1986, Driskell et alii 1989)
15 1988 Horton (study) (Horton 1988)
16 1990 Horton (Horton 1991)
17 1992 Horton (Horton 1992)
18 1995 Horton (Rose 1995b)
19 1996 Rose (study) (Rose 1996)
20 1997 Rose (study) (Rose 1997)
21 1998 Rose (Rose 1998a,b, Edwardsl998)
Table 7-6 Excavations at Qasr Ibrim

Qasr Ibrim was also expected to disappear under the waters of Lake Nasser.
E m e r y returned to Qasr Ibrim in 1961 to excavate the cemeteries (published by
Mills in 1982). In the same year the epigraphic recording of the N e w Kingdom
shrines in the cliff facade was done by Caminos (published in 1968).
F r o m 1963 onwards Plumley directed excavations within the fortified town
on the top of the hill. In the first Season the attention was focused on the cathedral
and another monumental feature, referred to as the podium (see Figure 7-5 F)
(Plumley 1964). F r o m that year on the Egypt Exploration Society has financed bi-
annual expeditions to the site which are still continued. Until n o w 20 campaigns
have been organised between 1961 and 1998 (cf. Table 7-6).
The long history of the excavations at Qasr Ibrim has k n o w n five different
directors, all of w h o m have implemented their o w n approach. During these years
there was a considerable variation in the composition of the staff. Furthermore, the
approach of archaeology in general has changed considerably between 1961 and
1999.
The changes in approach over time and the large numbers of researchers
and records, which are not all in the same place, make publication an almost
impossible undertaking. Despite this, the utmost effort is made to publish the site
by studying the preliminary reports, archives and filling in the gaps with limited
research on site.
7.4.3 A Meroitic and Ballana Periods House

Qasr Ibrim is unique for two major reasons. First of all it was an important
religious and political centre with an interesting history. Secondly, the
preservational circumstances are extremely good. Apart from botanical and
zoological remains, a large quantity of textiles, leather and basketry has been
preserved. Research at Qasr Ibrim thus gives information on aspects of daily life
which is seldom obtainable to this level of detail from other settlement sites.
The basketry studied for this thesis originates mainly from Meroitic and
Ballana period levels. In 1990 and 1992 a settlement area was excavated near the
south edge of the city (Figure 7-5 A ) . This area had been covered by Islamic
houses, which were removed without proper recording in 1976. Underneath these,
a Ballana structure was found which was designated number X-265.19 Figure 7-7
shows a plan of the house with the room numbers used as means of general
orientation. All features and deposits can be located precisely by their unit
numbers.
19
In 1990 and 1992 this area was indicated as trench 10, excavated by D. Edwards.
Trench 14, situated directly south of trench 10 was excavated by H. Roberton. This was the
southern part of the Ballana house and the area between the house and the girdle wall. A
detailed report of the excavations in 1990 and 1992 is in preparation.
Figure 7-7 Structure X 265, excavated in 1990 and 1992 (after the original plans drawn
by D. Edwards and H. Roberton).
Structure X-265 has known a long history and many modifications. The
earliest remains in this area were dated to the R o m a n or middle Meroitic period.
They consisted of stone walls with cement mortar and pink cement floors, found
underneath rooms 2, 4, 6 and 10. After demolition of this stone building the area
was abandoned until the construction of the Ballana structure. The oldest phase of
this building consisted of the four rooms mentioned above. All walls were made of
m u d brick, the entrance being at the east side of the house. The wall dividing room
2 and 4 is a later addition.
A first extension was built over Meroitic street deposits to the west of the
building (rooms 8 and 26). The walls of these rooms were constructed of rough
stones with heavy plastering. In room 8 three floor levels were found, all dating to
the Ballana period. A square cellar, which has not been excavated, predates the
building of the room. T o the north five more rooms were built at a later date,
extending over street deposits (1, 3, 5, 7, 9). The walls were made of stone
footings with a m u d brick superstructure.
In room 1 a cellar was found with an almost intact roof, consisting of
irregular wooden beams covered with date palm fronts, tied with d o m palm leaf.
The cellar contained organic deposits, many small basketry plates and Ballana
pottery. In room 9 four pottery hearths were found, which identified it as a
kitchen. The hearths were installed late in the Ballana period. R o o m 3 is a deep
cellar, divided by a partition wall. A similar cellar is present under room 9, but as
yet unexcavated.
The full extent of the house is not known, since it runs past the east baulk.
The general impression of this house is that the wealth of the occupants increased
gradually, which can be inferred from the gradual expansion of the house, not only
with rooms, but also with a large number of storage cellars. The finds from the
Ballana deposits give a similar impression. The occupants must have been leading
a comfortable, quite luxurious life.
In the early Christian period the building went out of use, with exception of
the southern part (room 10). At that time the room was divided by a stone wall
laid in herringbone structure. Other early Christian remains were found in area X-
190, which had been excavated partly in 1976.
All other rooms were gradually filled with organic debris interspersed with
lenses of decayed m u d brick and roofing material. The area was levelled in the
Islamic period, creating a platform on which stone houses were built. The only
traces of the Islamic period that were found in 1990, were a large number of
storage pits, all lined with basketry and matting.
7.4.4 Cultural Background of the Population

Since the basketry from Qasr Ibrim is mainly associated with pottery from the late
Meroitic and the early Ballana period (ca. 100 - 600 A D ) , some attention is paid
here to the cultural background of the people w h o inhabited Qasr Ibrim at that
time.
The Ballana culture was originally named X-group by Reisner, who found
unique and novel pottery in his excavations at Shellal, which he interpreted as
belonging to a culture that had swept over Nubia in the 5th century A D (Reisner
1910; A d a m s 1984, 390 -393). Reisner's migration theory was confirmed by the
physical anthropologist Elliot Smith, w h o characterised the X-group people by
their skeletal features as "strongly Negroid aliens" (Reisner 1910, p. 12).
During the second and third archeological survey it became clear that there
was no sudden intrusion of a n e w people.20 The Ballana culture (300 - 700 A D )
was dated immediately after the Meroitic culture (350 B C - 300 A D ) . It became
apparent that the designation of a n e w culture was provoked by a discontinuity of
the wheel thrown pottery, inspired by contacts with Byzantine Egypt21 and
enhanced by the decline of the Meroitic empire ( A d a m s 1984, p. 392)22.
The basketry at Qasr Ibrim shows a clear continuity, however, as does the
hand m a d e pottery, which also seems to be a locally m a d e commodity. The
transition between Meroitic and Ballana period thus seems to be a gradual one.
The discontinuities which occur besides the wheel thrown pottery, such as the
burial customs of the Ballana elite23, do indicate changes but not migrations.
W h o were the people w h o lived in Qasr Ibrim in the Meroitic and Ballana
periods? The Meroites represented a continuation of the Napatan influences in
20
Reisner made his succession of Nubian cultures on the basis of excavations at
Shellal, a site not fully representative for Lower Nubia. The Meroitic phase, which
immediately pre-dates the X-group period at Qasr Ibrim and in most of Lower Nubia, does
not occur as northerly as Shellal. Reisner found the X-group on top of Ptolemaic-Roman
pottery, which made the X-group pottery appear to be a complete innovation. Furthermore,
Reisner did not compare the X-group remains with related Byzantine pottery from Egypt.
" The Byzantine influence on the Ballana pottery is concluded from decoration with
"splash" and "blob" designs, which are considered probably (Adams 1984 p. 402) or
certainly (Trigger 1965, p. 133) a degeneration of the Hellenistic vine motif.
" T h e forces leading to the decline of the Meroitic empire are not clear at present.
Adams supposes that the trade with the Arabian peninsula enlarged the power of the rulers
of A x u m (Abyssinia) and therefore diminished the power of Meroe. The town of Meroe was
not conquered by Axumite kings, however, but probably by another people, the Nuba,
coming from the south (Adams 1984, p. 388).
" The cemeteries of the Ballana elite were found at Qustul and Ballana. From the
latter site. Trigger derived the name Ballana culture for Reisner's X-group. The kings were
buried in large tumuli at Qustul and Ballana, rather than in pyramids, such as found in
Napata and Meroe. In a number of features the Ballana tombs can be compared to the much
older Kerma culture. The tombs are tumuli, covered with white pebbles, the burial took
place on an angareeb (bed with rope matting) and sacrifices of humans and animals were
made. Trigger suggests that the changes in tomb style from pyramids to tumuli is not as
dramatic as it seems. The minor Meroitic burials were never adorned with pyramids, but
continued to be covered with small tumuli (Mills 1982). The Ballana tumuli could be
considered enlarged versions of these. Adams, on the other hand, considers the Ballana
tumuli a reaction against the Meroites and a reference to the Kerma culture (Adams 1984
pp. 413-424).
Lower Nubia. Meroitic seems not to have been the language of a ruling class,
because numerous graffiti in Meroitic are found in the temple areas in Qasr Ibrim
and on the road entering the town from the Eastern Desert. But did Qasr Ibrim
have only Meroitic inhabitants w h o were replaced by the Ballana culture?
A d a m s formulates it as follows: " W e m a y (...) epitomize the riddle of post-
Meroitic Nubia by observing that historians tell us about two peoples, the
B l e m m y e s and the Nobatae, where archaeology discloses only one culture, the
Ballana; moreover, both history and archaeology leave us in ignorance of the fate
of the earlier Meroitic population and culture.".24
Procopius names the two groups, living in the area in the 5"1 century A D .
The Blemmyes are a nomadic people living in the Eastern Desert between the Nile
valley and the Red Sea, the Nobatae are of unknown origin. In the texts there is no
mention of Meroites, which gives the impression that the original population of
Nubia suddenly had vanished with the decline of the Meroitic empire.
Although the cultural continuity in the archaeological material contradicts
this, there is a discontinuity in the written language which needs explanation. The
Old Nubian language, which was introduced at that time and lives on in the
modern Nubian languages, is not related to the Meroitic language. This could be
considered to support the suggestion that the indigenous population disappeared or
was submitted to a group of strangers. Triggers explanation is: " W e must thus
assume that over a period of time Nubian-speaking people entered Lower Nubia in
sufficient numbers to change the language of the region but not in such a w a y as to
break major cultural continuities." (1965, p. 137)
A d a m s connects the gradual introduction of Nubian speaking people with
the introduction of the saqia, the waterwheel which enabled the land to support a
larger population coming in from elsewhere. Both Trigger and A d a m s thus
suppose that a gradual intrusion of Nubian speaking people took place in Meroitic
times. Because the arrival of the Nobatae had no political consequences they were
not mentioned in the official texts, until it was recorded by Procopius that the
emperor Diocletian offered the zone between R o m a n Egypt and the Blemmyes
territory to the Nobatae.
Along the same line of reasoning it could be maintained, however, that the
Nubian speaking people were present in the area from m u c h earlier times onwards.
They would not be visible in written sources because the language that was written
was Meroitic. It was not until m u c h later that the Old Nubian became a written
language. Although there is no clear indication for an early presence of a Nubian
" In the Graeco-Roman harbour of Berenike, recent excavations have brought to light
a distinctive well polished hand made pottery, which seems to occur in the southern Eastern
Desert and the Nile Valley, where it is known from Wadi Qitna and Kalabsha South
(Strouhal 1984, 265). A few sherds have been found at Qasr Ibrim This type of pottery
seems to occur in the area that overlaps remarkably with the Blemmyes territory as
indicated in written sources (Rose 1995).
speaking population, the continuation in the locally produced material culture,

certainly does not contradict this.
The general impression of the houses of the Ballana period and the material
remains found inside them, is that the inhabitants led a comfortable life in
considerable luxury. O n e should be careful, however, with concluding that Lower
Nubian in general was a wealthy area since Qasr Ibrim should not be considered a
typical Lower Nubian settlement, but as a relatively wealthy urban centre within
the Lower Nubian agricultural community.
7.4 New Nubia25

Egyptian Nubia before 1961
After Qasr Ibrim was deserted in 1811, the inhabitants of the area were living at
the river banks in small communities, which together formed the village (balad) of
Ibrim. Those communities, or nugu (singular: nag) consisted of a group of houses
of different size in which members of the extended families were living together.
In general the houses were spacious with a number of rooms opening onto a
central court yard.
Agriculture was the most important means of subsistence. Most crops were
grown for private use, dates being the only crop to be exported in large quantities,
mainly to Egypt. The animals that were kept were cattle, sheep, goats and
donkeys. The cattle was not bred for meat or milk consumption, but employed
mainly to propel the irrigation waterwheel (saqia). Donkeys were used as pack and
riding animals. Fish never was important in the diet, the main source of meat were
goats, sheep and poultry. There were hardly any full time professional craftsmen
in Nubia, the major tasks of the m e n being farming.
Bricklaying and carpentry was done for private needs. Metal, glass and
pottery were imported from urban centres. This is partly due to the fact that
Nubian m e n have a long tradition of working outside Nubia. Nubians have been
employed by the Egyptians as soldiers and servants since the Old Kingdom. This
continued through the middle ages to the present.26 The result of this migration
was that the number of m e n in the Lower Nubian villages was relatively low.
However, the relation between the migrants and their village of origin was
strong and their stay in Egypt was often temporary or intermittent. In the 1960's,
many Nubian m e n working in Cairo, were married to a girl in their village. They
would come h o m e every two years, only to return to town when the financial need
~ This section is mainly based on information from Adams 1984, Curto 1965 Fernea
1973, Fernea and Rouchdy 1987, Geiser 1986, Herzog 1957, Hofmann 1986, Reinisch
1911, Rouchdy 1991 and the notes made kindly available to m e by B.N. Driskell.'
" In his description of his travels to Egypt and Nubia in 1664, J.M Wansleeben
writes: "Die Leute darin (=Ibrim) sind nicht Pech sondern Russ Schwartz, kommen haufig
daraus nach Cairo und dienen sonderlich den Europeern trefflich gem" (incorporated in
appendix of Herzog 1957, p. 189)
7 Sources A l o n g the River Nile 137
was felt (Geiser 1986). Within the towns there were Nubian communities, often
one representing each village.
Perhaps the access to foreign products, both financially and physically,
diminished the need to be self-supporting in other commodities but food. The only
important handicraft in Nubia was that of the w o m e n : basket making. The baskets
were m a d e in their spare time, after work in the fields was finished. The w o m e n
m a d e these baskets for their o w n use, to give away as presents and sometimes also
in order to sell them on the market (see below Section 18.2).
The entire length of the river banks was divided into areas belonging to the
subsequent villages. Urban centres of commerce were situated at both ends of this
string of villages, A s w a n in the north and W a d i Haifa in the south. In the area
between these two towns three languages were spoken: Arabic, Kensi and Fadidja.
The use of these languages was confined to very specific areas. The language
spoken in the area south of A s w a n is called Kensi. A n alternative n a m e the Kunus
use for themselves is Matoki. South of the Kensi-speaking area, around the W a d i
el-Arab, Arabic was the c o m m o n language. In the southern part of Lower Nubia,
the area stretching from Ibrim to W a d i Haifa near the second cataract, a second
Nubian language, called Fadidja, was used. In these three language areas of Lower
Nubia, there is also a distinction between other features such as architecture and
clothing.
The term Kensi for the language and the people (plural Kunus), is connected
to a geographical entity, the W a d i Kunus. The term Fadidja, however, does not
seem to refer to a geographical name.27 Since the inhabitants of N e w Nubia (see
below) identify themselves as either Kensi, Fadidja or Arabic, these three terms
will be used in the following chapters as indication of the three Nubian language
and culture groups. Doing this, it has to be kept in mind, however, that the
cultural identity lies primarily within the balad (village).
Egyptian Nubia after the Resettlement

The building of the subsequent dams in the Nile had far reaching effects on the
people in L o w e r Nubia. O f the small quantity of arable land, an increasing portion
was lost. Not only was most arable land permanently covered by water, but there
was also no n e w supply of fertile soil.
27
" There are indications that the use of this term is quite recent. In the writings of the
early travellers the term Fadidja does not occur, but in the 1960's the people from the
villages north of the 2nd cataract referred to themselves as Fadidja, as a group with a
specific identity. According to Herzog, the people living in the area between Wadi el-Arab
and the second cataract used to refer to themselves by the names of smaller units, such as
the village (Ibrimi). The term Fadidja was probably used by the Kunus to indicate the
language of the entire region with a name derived from one particular area, Fad, situated
opposite Abu Simbel. Herzog considered the term as an innovation introduced by Reinisch
and therefore proposed to use the term Mahasi instead (Herzog 1957, p. 25, 32).
ASWAN
D A B U D M f SHELLAL
Q U R T A W M E L- DAK K A
MAHARRAQAr-
WADI EL- A R A B
ED-DERR
ANIBA
WADI
ABU SIMBE ALLAQI \
FARAS
ABU. H A M E D
0', 150
KM
Figure 7-8 Lower Nubia. Indicated is the old course of the Nile and a number of
settlements, which have been relocated in New Nubia (after Baines and
Malek 1984; Fernea 1973, 2; Herzog 1975, Curto 1965; Rouchdy 1991
74).
Pkalabsha
>EL ALLAQl
• QURTA
EL-DAKKA—•hi'*
DIHIMIT- • WADI EL-ARAB
EL-NASSER
ASWAN
Figure 7-9 N e w Nubia, the relocation area of the Egyptian Nubians (after Rouchdy
1991,75).
Kensi Arab Fadidja I
Shellal (West bank) el-Sebua (West bank) Ginena wa Shebbak (West) 1

Dabud (West bank) el-Malki (West bank) el-Derr (East bank) 1
Umbarkab Wadi el-Arab (East bank) el-Diwan (East bank) 1
Meris (West bank) Shaturma (East bank) Tomas (West bank) I
Dihimit (West bank) el-Sinqari (East bank) Afia (West bank) 1
Qirtassi (West bank) el-Riga (East bank) Qatta
Taffeh (West bank) Korosko (East bank) Ginena (East bank)
Beitel-Wali (West bank) Nag el-Girgawi (East bank) Shebbak (East bank)
Kalabsha (West bank) Shablul (West bank) Khor el-Aqiba (East bank)
Wadi Qitna (West bank) Amada (West bank) Karanog (West bank)
Dendur (West bank) el-Lasiyyah (East bank)
Sabagura (East bank) Aniba (West bank)
Mariya (East bank) Ibrim (East bank)
Gerf Hussein (West bank) Masmas (West bank)
Koshtamna (East bank) Toshka (West bank)
Dakka (West bank) Arminna (West bank)
Quban (East bank) Abu Simbel (West bank)
Qurta (West bank) Abahuda (East bank)
Allaqi (East bank) Gebel Adda (East bank)
Maharraqa (West bank) Gebel Shams (East bank)
Seyala (West bank) Ballana (West bank)
Qustul (East bank)
Adindan (East bank)
Faras (West bank)
Aksha (West bank)
Serra (East bank)
Dibeira (East and West)
Ashkeit (Eest bank)
Argin (West bank)
Wadi Haifa (East bank)
Table 7-10 Egyptian Nubian Villages (Kensi, Fadidja and Arab), approximately from
North to South (sources: Baines and Malek 1984, 179; Fernea 1973, 1)
W h e n in 1959 the building of the A s w a n High D a m was started, it became clear

that the entire region of Lower Nubia, from A s w a n to W a d i Haifa would be lost.
Even large parts of Sudan were affected by this project. The people of Nubia were
involved in two re-settlement projects, one in Egypt and one in the Sudan.
The Egyptian Government reserved an area north of Aswan, near K o m
O m b o , for resettlement. Although an effort was made to compensate the Nubians
by building new villages with good facilities, by providing land and means of
irrigation and by paying each family a considerable amount of money, this could
hardly make up for the loss of their homeland. The Kensi and Fadidja villages
were built in separate areas, divided by the railroad and the main road leading
from A s w a n to Luxor (Figure 7-9, Table 7-10).
Moreover, in building N e w Nubia mistakes were made regarding the
culture and the w a y of life of the Nubians (Fernea and Rouchdy 1987, 373).
Thirty-three villages were built in N e w Nubia, in the same order and with the
same names as the Old Nubian villages. While the villages were rebuilt North of
A s w a n , the distances between the villages were diminished, but the relative
position of the villages was maintained. This meant that villages that had been
situated furthest away from A s w a n in Old Nubia, were rebuilt nearest to the city in
N e w Nubia (Figures 7-8 and 7-9). The new setup was extremely confusing to most
of the Nubian inhabitants: traditionally villages were orientated by their relative
position to Aswan: Ballana and Qustul had been far removed from Aswan, but
n o w were the first villages reached when entering N e w Nubia.
A second mistake was the arrangement of the houses within the village. In
Old Nubia the villages consisted of a string of house-clusters belonging to groups
of relatives. In these hamlets small and large houses were built next to each other.
In N e w Nubia everybody received a house corresponding with the size of their old
house. Since all large houses were built at one end of the village and all small
houses on the other end, the original groups were broken up and scattered over the
village.
Another aspect of the resettlement which had a far reaching influence on
Nubian life was that the distances between the villages were only a fraction of
those in Old Nubia. This had a direct impact on the contacts between the villages
and even influenced rituals such as marriage and funerals.28
The relocation of the inhabitants of Old Nubia did not influence the pattern
of the m e n working in Cairo, A s w a n and other cities. The N e w Nubian villages
still have a relatively large percentage of female inhabitants. Even though N e w
Nubia is in easy reach, many m e n leave their families for months in a row to work
elsewhere. The core of Nubian village life still is formed by the w o m e n .
-)0
Important festivities and rituals used to last several days. With the increase of the
number of people w h o could attend marriages and funerals the costs of those occasions rose
equally. This, and the fact that the villages are all within easy reach by public transport, has
caused the duration of the festivities to diminish (Fernea 1973).
PART T W O
Static Aspects of Basketry

C H A P T E R EIGHT
LEAVES AND BRANCHES
The static aspects of basketry tell us about a basket as a material object. The raw
materials, technique and shape can be recorded and presented as an image of what
an ancient basket probably looked like. O f these three aspects, thefirsttwo can be
recorded from even the smallest fragment of archaeological bakery found. They
are the basis for an interpretation of the dynamic aspects of basketry, which form
the link to h u m a n activity: production, function and meaning of material culture
(part three).
This chapter gives an introduction to the raw materials used in ancient Egypt.
The dynamic equivalent is Chapter 14, which looks into the winning and
processing of these materials. Appendix A , written by O. Brinkkemper, with
contributions by E. van der Heijden, gives the fibre identification of the plant
materials from A m a r n a and Qasr Ibrim. The sampling strategy has been explained
in Section 4.2.3 (p. 66).
In ancient Egypt two palm species and two grasses are used for most of the
basketry. Other materials, which are encountered rarely, are three other grasses,
four species of sedges and two rushes.
8.1 Palm Species (Palmae)

T w o species of palm leaf have been identified at Amarna and Qasr Ibrim. A third,
less c o m m o n , palm species, Medemia argun Wurttemb. ex Mart. (Arabic: argun)
has not been attested.1
It is important do discern the two species, because the palm leaf has different
properties and there appear to be specific preferences. Date palm was not used at
all for basketry in Amarna, the favoured basketry material was d o a m palm leaf. At
Qasr Ibrim both palm species were used, but for one technique d o a m palm was
favoured, while for the other date palm was chosen (see Chapter 11). In the latest
period of occupation, however, d o a m palm leaf was used for the techniques which
hade been made with date palm leaf in previous periods.
'See also Wallert 1962.

Doam palm (also: dom, daum)
Hyphaene thebaica (L.) Mart.

(= Cucifera thebaica Del.)
Arabic: doam
Fadidja: anbi, ambi
Kensi: anbi
drawing after: Germer 1985, 234
The doam palm is a tree with a branched stem and large fan-shaped leaves. Strips
of the leaf, varying in width between 1 and 30 m m have many applications in
basket making. The complete leaves can be used as brooms. At present the
distribution of the doam palm is limited to upper Egypt and the Sudan. In the New
Kingdom period the distribution will have been more general in Egypt and the
doam palm probably was also found in Middle Egypt. In ancient Qasr Ibrim, the
doam palm was probably quite common (Tackholm 1974, 763; Germer 1985, 234)
Date palm
Phoenix dactylifera L.
Arabic: nakhla
Kensi: fend
Fadidja: fend
8 Leaves and Branches 147
The date palm consists of one straight stem branching out in palm fronts. These
are feather shaped leaves, consisting of large midribs of 3-5 m . long, with small
leaflets at two sides. The leaflets are used in basket making. The midribs of the
leaves, which look like slightly curved branches, are used for making furniture,
boxes, roofing and as fuel. The leaf sheaths at the base of the leaf midribs decay
into a fibrous material, used for rope making.
At present, the date palm is c o m m o n in all Egypt, as it probably was in
Egypt of the N e w Kingdom period. Amarna might have be an exception, however:
although many date stones have been found, there is little evidence of other parts
of the date palm, such as stems, leaves, midribs or fibre. In the area of Qasr Ibrim
the date palm was very c o m m o n . (Tackholm 1974, 763; Germer 1985, 232-234)
8.2 Grass Species (Graminae)

T w o tall grasses which were used extensively in both Amarna and Qasr Ibrim, are
k n o w n in Arabic under the same name (halfa). Apart from these there are four
reeds, which are mainly distinguished by size. They are not used for baskets or
mats, but for rigid screens, roofs and furniture. Wheat straw belongs also to the
grass family and has been attested at Qasr Ibrim.
Haifa grass (1)
Desmostachya bipinnata (L.) Stapf.

(= Eragrostis cynosuroides Beauv.)
Arabic: halfa
drawing after: Tackholm 1974, pi. 255
This is a tall, strong grass species of which the leaves and culms are 1 m high.
W h e n flowering, the grass grows spiky dark ears. The grass stands at the edges of
rivers and canals, but also in sandy surroundings, as long as there is some
moisture in the upper layers. Desmostachya was c o m m o n in Middle Egypt in the
N e w Kingdom period, and it still is today. It did grow abundantly in the Qasr
Ibrim area too. Both leaves and culms were used for making basketry and cordage.
Desmostachya occurs more often than Imperata in the basketry and cordage at
Amarna and Qasr Ibrim (Tackholm 1974, 688-690; Germer 1985, 202).
Haifa grass (2)
Imperata cylindrica (L.) Beauv.
Arabic: halfa
drawing after: Tackholm 1974, pi. 279
A tall, strong grass, very similar to Desmostachya bipinnata. When flowering, the
grass grows plumy white ears. Both the leaves and culms were used for making
basketry and rope. At present this grass is very common in the Amarna valley
Imperata outnumbers Desmostachya. In ancient times, both at Amarna and Qasr
Ibrim Imperata did not occur nearly as frequent as Desmostachya. For basket
making, the two grasses are completely interchangeable. Haifa is the Arabic name
for both which is an indication that the two grass species are not distinguished by
the local population. Stipa tenacissima L., a grass species indicated by Germer as
"real halfa-grass" (Germer 1985, 218), does not grow in Egypt (Tackholm 1974,
757; Germer 1985, 224).
Reeds
Phragmites, a tall reed, which can become 5 m high was used widely in ancient
Egypt for making arrows, writing pens, furniture and roofs. Culms have been
found at Amarna in remains of furniture. Impressions in the mud covering of the
roofs, indicate that the culms of Phragmites or Arundo was used as roofing
material in Amarna (Tackholm 1974, 696-697; Germer 1985, 205-206).
Arundo donax is a very tall, bamboo-like reed, which grows 4-6 m high. In
ancient Egypt it was used for making roofs, furniture and flutes. Fragments of the
first two, have been found at Amarna. At present Arundo donax is used in the
Delta for making stake-and-strand baskets.
Common reed
Phragmites australis (Cav.) Trin exSteud

(= P. communis Trin. ssp. altissimus
(Benth.) Clayton
(= P. communis v. isiacus [Del.]
Coss. et DR. = Arundo isiaca Del.)
Arabic: ghaab, boos, local variations
Giant reed
Arundo donax L.
> I*
Arabic: ghaab, boos, local variations 1
drawing after: Germer 1985,284 1
few, 1
The fact that Phragmites australis and Arundo donax have the same n a m e in
Arabic is an indication of h o w difficult it is to tell them apart. Since they are
mainly applied for the same use, exact species identification is not a priority.
(Tackholm 1974, 696; Germer 1985, 203-205)
Other tall grasses

Saccharum spontaneum, wild sugar cane, is a tall grass with dark green very long
stiff leaves, which according to Germer, were used for making mats (Germer
1985, 225-226, 245). This material has not been attested at A m a r n a or Qasr Ibrim,
but the stalks, which are difficult to distinguish from Phragmites or Arundo, m a y
have been used for roofs and shelters. Stalks of Sorghum bicolor, seeds of which
were found in Qasr Ibrim, might have been used for the same purpose.2
Triticum species
Wheat straw occurs seldomly, but a small finger ring has been found at Qasr
Ibrim, which was knotted from a strip of wheat straw. At present, wheat straw is
used for the imbrication technique (see Plate 9-17, p. 166 and p. 245).
8.3 Sedges ( Cypera ceae)

All cyperus species grow in marshy areas, but C. schimperianus is also found on
rocky grounds near the Nile or the Mediterranean coast. The culms of these sedges
are 800 m m to 1 m long, with some leaves around the base.
Papyrus
Cyperus papyrus L.
Arabic: bardi
This marsh plant can grow 5 m high. The soft white internal structure of the culm
is used for the production of papyrus sheets and only the rhind of the culm is used
for rope and basket making. Its use has been attested in Amarna, where it occurs
rarely, but not at Qasr Ibrim.
At present, Egyptian farmers make huts and shelters (sabata) at the edges of their
fields from reeds and stalks of several plant species which had not been introduced in
antiquity, such as sugar cane (Saccharum officinarum), sun flowers (Helianthus annuus L.)
and maize (Zea mays L).
Other sedges
Cyperus rotundus L.
Cyperus alopecuroides Rottb.
Cyperus schimperianus Steud.
Arabic: sa'ad, dees, samaar
left: C. rotundus, right: C. alopecuroides

drawing after: Tackholm 1974, 787
The Arabic terms for these sedges are not specific. The terms differ regionally,
but no distinction is made in Arabic between the three species (Tackholm 1974,
780-790; Germer 1985, 244-250). The long, leafless culms make that sedges are
eminently suitable for mat making, for which they are used nowadays.
The sporadic use of sedges at Amarna and Qasr Ibrim is limited to the
production of coiled basketry and rope. In contrast with C. papyrus, the entire
culm is used. At Amarna C. papyrus and C. rotundus are found. These species do
not occur at Qasr Ibrim, where the sedges used are C. alopecuroides and C.
Schimperianus.
8.4 Rushes (Juncaceae)

Rushes grow at moist fresh water sites and also on salty grounds. The culms are
smooth and grow up to 1 m high. The two species of rushes can be distinguished
by their colour (J. rigidus is a slightly darker green than J. acutus) and from their
flowers and seeds (Tackholm 1974, 662-665; Germer 1985, 200-201).
The properties of the culms of these two species that matter for basket
making (their length,flexibilityand regular shape) are exactly the same. The
Arabic term is used for both types and translates as 'bitter sedge'. At present both
Juncus species are used widely for the production of woven mats.
At Amarna Juncus was used rarely. The few instances that it occurred, the
exact species was not determined. At Qasr Ibrim J. acutus was found in an
intricate basket with a twined base and plaited sides (Figure 4-2, p. 61). J. rigidus
has been attested at a Late R o m a n site for making fish traps in an open, knotted
technique (schematized in Figure 5-4, p. 84) and in long bundles as part of the
roof construction.3
Rushes
Juncus rigidus C.A.Mey.

(= /. arabicus [Asch. et Buch.] Adams.
(= J. maritumus v. arabicus Asch. et
Buch. ex Boiss.)
Juncus acutus L. (drawing)
Arabic: samaar morr
drawing after: Tackholm 1974, 663
11
8.5 Plant Fibres4
Flax (Linum usitatissimum; Arabic: kittaan)

Flax has been grown in Egypt since at least the Old Kingdom, but probably much
earlier. The long fibre cells of the culm were mainly used for the production of
textiles, but also for fine cordage and fish nets. (Tackholm 1974, 314; Germer
1985, 100-101)
Cotton (Gossypium arboreum L.; Arabic: qutun)

Cotton is the fibre protecting the cotton seeds, each fibre consisting of one long
cell. Cotton did not occur in Egypt before the 3rd century B C and was not
cultivated in Egypt until the Islamic period (Watson 1977: 357). The species found
in Egypt in ancient times is Gossypium arboreum. Cotton does not occur at
Amarna, but is used in Qasr Ibrim for textiles and fine string (Tackholm 1974,
356; Germer 1985, 122-123).
Roman fort at Abu Sha'ar (Egyptian Red Sea coast, near Hurghada, 3rd-early 7th
century A D ) . The basketry was studied by me as part of a team of the University of
Delaware, directed by S.E. Sidebotham. The publication is in preparation.
The plants are not depicted, because fibres are worked into yarns which have no
specific shape. This is in contrast with the other plants mentioned here, because for most
basketry raw materials the shape of the plant parts is of importance for the technology.
C H A P T E R NINE
BASKETRY TECHNIQUES, A CLASSIFICATION
The classification of basketry techniques is based on the basic structure of the

baskets. It is the static aspect (result) of production and forms the basis on which
to understand the production process and the producers, although it is just one of
six aspects of basketry technology (Table 4-1, p. 59). A s long as this is kept in
mind, it is useful to limit the classification of basketry techniques to a classification
of the basic structure, the one aspect that can be recorded from all fragments
found. Furthermore, a classification of the basic structure can be done with a
limited number of criteria. Such a deliberate limitation is inevitable if the
theoretically possible techniques and the variation of Egyptian techniques within
that range of technical possibilities are to be shown.
In this Chapter classification criteria are put forward, according to which a
paradigmatical classification is built (cf. Section 2.3.1, p. 13). T h e techniques
occurring in Egypt arefilledinto this classification and an explanation is sought for
the empty classes.
T h e classification is based on the method of analysis of the basic structure as
introduced in section 5.1 (pp. 81-98). Specific terms, such as system, m e m b e r ,
plane, have been introduced in Chapter 5 and are listed in the glossary.
9.1 Classification Criteria

Table 9-0 is a paradigmatical classification of the basic structure of basketry
techniques.1 T w o criteria are involved, which are both dealing with basketry
systems: the activity of the systems and the number of directions in which the
systems are oriented. The relevance of these criteria to basketry production is that
a basket maker follows the same actions in manipulating the strands that belong to
one system. Also, as will be elaborated on in Chapter 13, he handles the active
systems more frequently than the passive systems.
In section 5.1 it w a s determined that basketry strands belong to one system if
the raw material, the orientation and function of these strands are the same (see
flowchart 5-1 on p. 82). T h e basic structure of a technique is described by the
interaction of the systems, in other words, by determining in what w a y the
orientation or the function of groups of strands are different.
The Table number 9-0 is used, because it allows the Figure numbers 9-1 to 9-20 to
correspond with the numbers of the techniques listed in the classification of Table 9-0).
u
« S e
u
en : i
o
rt a "
s £
S om
| flj 00 — On
i M 8 C
£ o — X • O
OJ i 3 1
CT3 E
t/1
a. oa. ? 3
crt ts
: ^\
m = x;_ c
C H .2
>.
_, « D On "O
c : eo oo n « >•
to "a0 c c •e xi -°
'5. -5 ; i
CO o i 3o c
S i^ .^ >%
n3
O. CJ 00 ON uH g
6\ CTn On wi , - c
— o u
DO
X X *v \ ^ aj n u
a.
> c vi .— j= x «
o -2 (fl O3 "> 00 X V \ V ZU
E s i:
•—. 3 O0 ti c i > s
8-* 1C 1E 1^ 1— v
A v
N
> OJ
o 1- (N O D. CL ^
.3
rt m -* .3 «1">1 CNO u LU „
V5 —" —* ~-l *»H
On On On On
E
>, "S S x
w •s -s
<-N| 0) ^ S - a
> Jo oo oo "O oo "a or. -a •° | ™
13 —c .E a.
C Ma C C C C c 00
OJ a| «
1 st
9-10 twii
2st
9-8 wea
sew
9-9 wea
wra
6-11 wal
r « r « •- c
kno
coil
> u
t/5 > i- > u e • — Q. o- i- H
c
CfL
l T3- WN NO t~~ u ^^ aj
0> On On On NO if
DO DO S <
C C £ E -u
nett
nett
9-2 grom met
E
9-1 k notl ess
U o s a
9-3 knott ed
OJn
o >o
>%
to 03 C <« c
r, 1 2
N S ^ U 3 T3
>% c c .. c
'5 oa •c— c(/I c to
> /c o •— o .2 c
o
<•> *.° 5 u o
a// en aj 3 8 5 cj O
2 - C 1- 2 On
o o
I— , .O2 <N
^ o en
4) _U
X
ea
H
9 Basketry Techniques, a Classification 155
In Table 9-0 the aspect number of systems has three attributes, implying that
a basketry technique can be made with one, two or three systems. U p to six-
systems techniques are known to exist, but for the material studied here, the
basketry from Amarna, Qasr Ibrim, Middle Egypt and N e w Nubia, all classes
outside the part of the classification shown in Table 9-0 are empty.
For the one, two and three-system techniques the variation in the activity of
the systems is determined. There is always at least one active system, so a one-
system technique is per definition active. If two systems are involved then there
m a y be two systems active, or only one, while the other is passive. Three system
techniques can have, one, two or three active systems.
The second criterion is the aspect number of directions. This cannot mount
up to more than three, since there are never more directions than systems. By
cross referring the two criteria w e get 18 classes (the 18 squares in table 9.0). In
eight of these, a total of 20 basketry techniques have been listed. The numbers (9-1
to 9-20) refer to schematical drawings of these techniques.
9.1.1 One-system techniques

The first class in Table 9-0 contains basketry techniques made with one active
system, orientated in one direction. These techniques are made with one strand at
the time, or with several parallel strands, which all have the same function in the
technique. The one-system techniques that occur in the basketry corpus under
study are knotless netting (Figure 9-1), grommets (Figure 9-2) and knotted netting
(Figure 9-3).
Figure 9-1 Knotless netting, a fabric made by looping a strand through a previous row
of loops (the anchoring fabric). This is a one-system technique, the active
system moving in two planes.
Figure 9-2 Grommet, a linear band made by wrapping a strand a number of times
around itself. The strand in this one-system technique moves in two planes.
Figure 9-3 Knotted netting, a one system technique. The mesh knots are made with a
yarn, moving in three planes.
One-system techniques are made with one long strand, or with short lengths
of, for instance, palm leaf. A n example of the first is knotted netting, which is
m a d e with yarn, an example of the second is knotless netting,2 where strips of
d o a m palm leaf are added whenever the previous strip has ended. Both knotless
netting and knotted netting are made by adding on a new row to the previous one
(the anchoring fabric).
Excurse: underlying criteria

The fact that three techniques are listed in the same class, means that other
underlying criteria discern these techniques. The difference between a grommet
and the two types of netting are that the first is a linear form, while netting is a
fabric (cf. Table 9-22, p. 171). In this classification the criterion of form (linear or
fabric) is considered less important than the system properties cross referred in
Table 9-0.
A second set of underlying criteria relate to the properties of the active
system only. A s explained in Section 5.1.4.5 and 5.1.4.6 (pp. 93-97) the strands
making up the active system can consist of several members, which move in one,
two or three planes. The classification of active system mobility as shown in Table
5-6 (p. 96), has nine classes. This is another set of criteria that plays a role in the
occurrence of more than one technique per class in Table 9-0.
The underlying criteria have been ordered hierarchically under the
classification according to system activity. They have not been incorporated as
separate dimensions in the classification of Table 9-0, because the exponent of
involving these three additional criteria (form, number of members, number of
planes) results in a classification with 324 classes (Table 9-22, p. 171).
The aspects of orientation and mobility should not be confused. The
orientation of a technique is at system-level, while mobility refers to the
movements made with individual strands. A yarn with which a net is knotted has a
mobility in three planes (see Table 5-12, p. 96), it twists around the previous loop
and around itself to make a knot. The main orientation of the yarn is limited to
going up and d o w n the fabric along one line only (one direction).
9.1.2 Two-Systems Techniques

The two system techniques occur in one or in two orientations. In the sewn plaits
technique, a long plait forms the passive system. The plait is fastened with an
active strand, picking up the edges of two plaits. Both plait and sewing strand
follow the same direction. For a sewn plaits mat the passive system is formed by a
number of plaited strips layed out parallel to each other, for a basket one long plait
is sewn spirally into itself. Figure 9-4a shows a mat, made of wide 17-strand
plaits. The sewing strand is not visible, because the edges of the plaits are drawn
2 Another term used for this particular type of knotless netting is: single interconnected
looping (Emery 1966: 33).
inside each other. The dotted lines display where two plaits overlap. The edge of
the mat is formed by a 9-strand plait sewn on top of the rough ends of the plait.
Figure 9-4b shows the same technique with one long plait which is sewn
spirally. The right side of the start of the plait, in the centre of the spiral, partly
overlaps thefirstround. The left side of the start of the plait is sewn to the right
side of the plait after it has made one turn. Because the two edges are drawn inside
each other, the sewing strand is not visible and the plaited strip seems to form one
ongoing plait. The appearance is thus very similar to the plaiting technique
depicted in Figure 9-14, but the principle is completely different: in continuous
plaiting two systems are involved, which are equally active (see below).
Figure 9-4a A mat made of plaited strips (passive system) which are linked with a sewing
strand (active system). The latter is not visible, because it pulls the edges of
the plaited strips inside one another. The sewing strand moves in two planes.
Figure 9-4b A long plait (passive system) is sewn spirally with a sewing strand (active
system). In this example the start of the plait in the centre of the spiral partly
overlaps thefirstround. The fabric seems continuous, because the edges of
the plait are drawn inside one another.
Coiling (Figure 9-5) and wrapping (Figure 9-6) consist of a passive bundle
and an active winder, both following the same direction. The difference lies in the
form: coiling results in a fabric, a bundle of grass fastened spirally from the centre
to the rim by palm leaf, which is wrapped around the bundle and follows the
direction of the bundle. The wrapping runs parallel to the bundle hand holds the
coil in place. Wrapping in a line or circle is used for instance to m a k e brushes
(wrapping the bundle which forms the handle with palm leaf), or pot stands (a ring
of grass wrapped with string), and also to make three ply string (see page 163). If
the active strand is crossing itself while wrapping around a passive bundle, the
term knotting is used (Figure 9-7).
Figure 9-5 Coiling, a two-system technique with a passive bundle, held in place by an
active winder, or wrapping strand (moving in two planes).
Figure 9-6 Wrapping, a two-system technique with an active strand wrapping (moving
in two planes) around a passive bundle.
nSv-£. |gy Ji'Lr-- IE—'_ '

JlTpI M M j ^ J
M 1~~ i L ' l S i i m ^ K j m M j n T O f t ^ f t r
^"^^W-_S'n--^/V-^^/ ^ ^ J 1 •;' i
C : _st :-ir"Y '• ': [ •
tuf'ft- f l r i £
-£i=i=-3=l S^3ijEjU^^^-iiMi,^3' 't
ll'lc
^ V
Figure 9-7 Knotting with an active strand (moving in three planes) around a passive
bundle. This is an example of linear knotting.
In contrast to the previous two-system techniques, weaving, twining, waling

and piercing are oriented in t w o directions, perpendicular to each other. In
w e a v i n g the passive system, the w a r p , is w o v e n in with the weft. T h e weaving
strand is m o v e d u p and d o w n (one plane). In m a t m a k i n g often single culms of
rushes, sedges or grass are used, which are w o v e n in one b y one. W e a v i n g with
one long ongoing strand (a bundle of rope, for instance), forces the w e a v e r to use
a w e a v i n g shuttle, or to pull the entire strand through the weft with each r o w of
weaving. A solution to this is to leave the total bundle of rope o n one side of the
l o o m a n d w e a v e in loops. This results in weaving with a double strand (Figure 9-
9). T h e basic structure of weaving with t w o strands is very similar to weaving with
one strand. T h e actions of the basket m a k e r , h o w e v e r , are quite different.3
— — - —
-K4L1S /£) /7)

& =
^S5^__
77~Tjji^^ Si ^Se|\^
Z^^7? / ^ 'T^TTfc^ ^7'^'
r\v
Figure 9-8 Weaving with one strand: an active weft is woven at a right angle into a
passive warp. The weft consists of one m e m b e r moving in one plane.
This is an example where the basic structure does not give enough information on the
production, and a study of the edges is necessary to understand the process.
Figure 9-9 Weaving with two parallel strands: the two members moving in one plane.
Jj/A Al)
^_^s£ -*^fc^^ » = a r ^ ^
^s^__xp "J^r-" = 3 ? ^
£%t \ ' ^£&"^r^t
^\'&
Figure 9-10 Twining is a two-systems technique in two directions, in which the active
element consists of two strands (members) moving spirally (in two planes)
around the passive elements.
Figure 9-11 Waling is a two-systems technique in which the active element consists of
three strands (members) moving spirally {in two planes) around the passive
elements at a right angle.
T h e term pierced basketry (Figure 9-12), which occurs in Table 9-0, does not
occur in Table 5-12 (p. 96) which is a classification according to the mobility of
the active system. This is simply, because its active system is not 'mobile', but
either rigid or fixed. Because of this it is often difficult to decide which of the
systems is active, unless the production process is known. In Figure 9-12 the
uprights are the passive system, the horizontal bars are the active system. T h e
active system has been prepared beforehand by punching holes at the proper
intervals. T h e indication 'active' is based on the observation of the production
process, during which the basket maker first positions the uprights and then pushed
on the horizontal bars. The pierced technique of Figure 9-12 is produced in present
day Middle Egypt, but has not been found in A m a r n a or Qasr Ibrim (cf. Table
9-20).
Figure 9-12 Pierced basketry: even if all systems are rigid, at least one system is active.
In this case first the horizontal system (active) is pushed over the two
vertical sticks (passive) to make the corners, and then the remaining vertical
sticks (active) are hammered in through the pre-punched holes in the
horizontal system (passive).
Figure 9-13 In Egypt, two-ply rope is made of two bundles of plant material, which in
the same movement are spun and plied. The bundles are both active and are
twisted around each other.
The W o r l d according to Basketry
164
The next class (two active strands oriented in the same direction) contains
only rope making (Figure 13). The fibre bundles used for rope making are rolled
between the hands and twisted around each other. It is a kind of wrapping
technique in which both strands are active. Three-ply rope, however is made by
fixating a two-ply string and wrapping the third strand around it (cf. Wrapping
p. 160).
Figure 9-14 A continuous plait, made with two active systems, which are orientated
perpendicular to each other. The plait pattern is \2/2\\l (under 2, over 2,
shifting 1) and the strands move in one plane.
Figure 9-15 Plaited strips, made with two active systems, which are oriented in two
directions, perpendicular to each other. At the edge of the plait the strands
change direction. Top:five-strandsplait in a \1/1\\1 pattern; Bottom: nine-
strands plait in a \2/2\\2 pattern. The strands move in one plane.
Plaiting, which together with knotting makes up the next class, is divided in
continuous plaiting and plaited strips. Just like the difference between knotless
netting and grommets, this is a distinction between a fabric (continuous plaiting,
Figure 9-14) and a linear strand (plaited strips, Figure 9-15) which is sewn to form
a fabric (sewn plaits, Figure 9-4a and b). The main difference of plaited strips,
compared with continuous plaiting, is that the strands of the former change
direction at least twice at the edges of the plaited strip. Figure 9-15 shows a plaited
strip m a d e with 5 strands in a \1/1\\1 pattern and a strip m a d e with 9 strands in a
\2/2\\l pattern (cf. Wendrich 1991, 65-66). For the sewn plaits technique the
plaited strips have to be prepared beforehand, in a separate stage of the production
process.
The knotting technique that occurs in the same class, is different from what is
depicted in Figure 9-3, because it is done with two active strands. This technique is
used in Egypt to m a k e a fabric (Figure 9-16). In all knotting techniques the
elements m o v e in three planes, because the strands turn back and cross
themselves.
Figure 9-16 A netting fabric, knotted with overhand knots is made with two active
elements.
166
9.1.3 Three-Systems Techniques

There are four kinds of three-systems techniques in the corpus of basketry
classified here. T h efirstis a decorative coiled technique, termed coiled inlay,
which has two passive systems: the bundle and an extra strand or strands, which
are fastened on top of the bundle with the active winder (Figure 9-17). All systems
follow the same direction.
If m o r e than one strand is used, the appearance is that of a blocked
decoration on top of the coil. This occurs o n s o m e of the coiled basketry rims from
Qasr Ibrim. W h e n the entire coil is covered by a layer of folded wheat straw, the
technique has the appearance of roofing tiles and is k n o w n as imbrication (Figure
9-17).
Figure 9-17 Decorative coiling technique with inlay of a second passive system, which
runs parallel to the bundle and is fastened with the wrapping strand. This
variation, in which a strand of wheat straw covers the coiled bundle, as if it
were covered with roofing tiles is also known as imbrication.
Two variations occur within the class made of two passive systems and one
active system oriented in two directions. T h efirstis looping, of a strand around
two passive systems layed out perpendicular to each other (Figure 9-18). T h e other
is binding,4 which is very similar: the sticks of one passive system are layed out,
Knotting is done with active elements only, while binding involves both active and
passive elements).
after which a second passive system, consisting of a few sticks is positioned

perpendicular across the first passive system. The two passive systems are then
connected with a third, active, system, which follows the orientation of the passive
cross bars (Figure 9-19). The difference between 9-18 and 9-19 is that the active
strand makes loops in thefirstcase and is tied in the second.
Figure 9-18 A passive string is stretched out, on top of which bundles of rigidfibresare
layed out at a right angle. They are connected with an active strand that
loops around both passive systems (moving in two planes).
Figure 9-19 T w o passive systems are layed out perpendicular to each other. They are
connected by an active strand, which binds the passive strand together. The
binding element moves in three planes.
Figure 9-20 Plaiting around a core involves one passive system (the core) and two active
systems (the plaiting strands). It is a linear form of plaiting.
There is one three-systems technique with two active and one passive systems, a
variation of plaiting, around a passive core (Figure 9-20).
9.2 O c c u r r e n c e in Egypt
The techniques listed in Table 9.0 (p. 154) have been found in the excavations at
A m a r n a or Qasr Ibrim. This does not guarantee that they were actually produced
there and in some instances there is reason to believe that some of them were not
(cf. Chapter 10 and 11). For Middle Egypt and N e w Nubia only the basketry
produced at present in the area has been listed in Table 9.0. The frequency of the
different techniques diverges strongly. T o gain insight in h o w c o m m o n or rare a
technique is, the 20 techniques from Table 9.0 have been listed below in Table 9-
21 with an indication of their frequency. For the ethnographic material this has
been based on a general impression, without precise quantification.
\ ^ Provenance archaeological material ethnographic material
Technique ^^~^^ Amarna | Qasr Ibrim Middle Egypt i N e w Nubia

1 knotless netting •
2 grommet •• • •
3 knotted netting •• •
4 sewn plaits ••• ••• •••
5 coiling ••• ••• ••• •••
6 wrapping (brush, ring) ••• •• ••
7 knotting •
8 weaving, one strand ••• •• •
9 weaving, two strands • •• •
10 twining ••• ••• ••
11 waling •
12 pierced •••
13 rope ••• ••• ••• •••
14 continuous plaiting • •
15 plaited strips • • •
16 knotting • •
17 inlayed coiling • ••
18 looping ••
19 binding •• •• ••
20 plaiting around a core •
Table 9-21 Occurrence of the 20 basketry techniques mentioned in Table 9-0.

Indication of frequency: • = seldom; • • = common ; • • • = frequent
T h e horizontal lines in Table 9-21 show the eight classes that contain
basketry techniques from the corpus under study (cf. Table 9-0). Knotless netting
(1) occurs seldom and only in Amarna. G r o m m e t s (2) are found regularly in
A m a r n a , sometimes in Qasr Ibrim. They are still m a d e in Middle Egypt,
especially as pot stands, but in N e w Nubia I have not noticed them. Knotted
netting (3) occurs where fishermen are at work. In present day Egypt and N e w
Nubia nets are not produced, but bought from a factory in Alexandria. The only
work done to them is regular repair.
In the New Kingdom period sewn plaits basketry (4) did not exist in Egypt. It
probably was introduced in the Graeco-Roman period (cf. Wendrich 1997, p. 44).
N o sewn plaits basketry or continuous plaiting was found at A m a r n a (see Chapter
10). In Qasr Ibrim, however, it was found in large quantities and it is perhaps the
most c o m m o n basketry produced in Middle Egypt and in N e w Nubia at present.
Coiling (5) is another widely produced basketry type, which did occur in
Amarna too. It is found in large quantities in Qasr Ibrim and is produced at present
both in Middle Egypt and N e w Nubia.
Wrapped linear objects (6), such as brushes, were found regularly in A m a r n a
and Qasr Ibrim. Very similar brushes are still produced today in Middle Egypt.
Examples of a knotted brush (7), however, have only been found at Amarna.
Mats woven with one strand (8) have been found at A m a r n a and Qasr Ibrim.
At present, they are produced in Middle Egypt. Weaving with two strands (9) has
been found at Amarna and Qasr Ibrim as part of furniture webbing. This kind of
bedding is still produced in the region of N e w Nubia today, but does not occur in
Middle Egypt.
Twining (10) is the technique found most in Amarna. It also occurred
frequently in Qasr Ibrim. It is produced in present day Middle Egypt, but I did not
find any producers of twined matting in N e w Nubia.
Waling (11) was only found in two objects in Qasr Ibrim (Chapter 11). This
type of basketry appeared in Egypt in the R o m a n Period. N o stake-and-strand
baskets are made in present day Middle Egypt or N e w Nubia and no waling
occurs.
Pierced basketry (12) has not been found in Amarna or Qasr Ibrim. It is a
technique which at present is used mainly to produce crates (qafas), cages, tables,
chairs and beds. There are many workshops in Middle Egypt and in the region of
N e w Nubia, although the qafas makers are not Nubian. Rope (13) is used and
produced extensively throughout Egypt, ancient and modern. Often a length of
rope is made on the spot
Continuous plaiting (14) has only been found at Qasr Ibrim, in the N e w
Kingdom Period this technique probably did not occur. It was produced in N e w
Nubia until recently, especially for the production of fans. Likewise, plaited strips
(15) do not occur in Amarna, but they are a half product for sewn plaits basketry
in Qasr Ibrim, present day Middle Egypt and N e w Nubia. Indicated in the table is
the frequency of plaited strips that are not used in sewn plaits basketry. Knotted
fabrics (16) have been found at Amarna and Qasr Ibrim, but are not made at
present.
Inlayed coiling (17) has been found at Qasr Ibrim as a rim decoration, while
a special form of it, imbrication occurs in present day N e w Nubia. Looping (18) is
made in every farmer's household in Middle Egypt, to produce the shinda or
cheese straining mat. Binding (19), which is mainly used for rigid screens, has
been found in several roof constructions at Qasr Ibrim. In present day N e w Nubia
a similar binding technique is still used in roof making. Binding occurs also in
Middle Egypt, for the production of the shelters built by farmers in the fields
Plaiting around a core (20) is very rare, only three small fragments have
been found in Qasr Ibrim (see p. 247). It is not produced in present day Middle
Egypt or N e w Nubia. Similar cords, made in leather, are produced by the Ababda
bedouin in the Eastern Desert of Egypt.
9.3 Empty Classes

Total Number of Classes
In Section 9.1 it w a s said that the underlying criteria, w h e n all given similar
weight in the classification, would give a classification with 324 classes (p. 157).
This has been m a d e visible in Table 9-22, which lists three groups of criteria. T h e
first group cross refers the activity of the systems with the orientation. This
produces the 18 classes (squares) which were shown in Table 9.0.
Subject Criteria
Systems activity of systems 6

(Table 9-0, p. 154) 1 system: act
2 systems: pas/act; act/act
3 systems: pas/pas/act; pas/act/act; act/act/act
orientation of systems (1, 2 or 3 directions) 3 18

Composition of the number of members (1, 2 or 3) 3
active system
(Table 5-12, p. 96) planes in which the members move (1, 2 or 3) 3 9
Form linear or fabric 2 2

Total number of classes (for maximum three systems) 324
Table 9-22 Criteria on which Table 9-0 is based. The third column lists the number of
attributes. The fourth column lists the number of classes in each of the three
sub-classifications. Cross referring these attributes without weighing in one
classification, results in 324 classes (by multiplication of the attributes).
W h e n , however, the criteria of the mobility of the active system are

involved, then w e have to multiply the 18 classes with the nine classes of Table
5-12 (p. 96). This gives 162 classes, which are doubled if the third aspect is
involved, which is necessary to discern linear forms (a wrapped ring) from fabrics
(a coiled basket).
T o keep the classification manageable, it is better to choose the criteria
considered most important for the question at hand, and arrange the other criteria
hierarchically under these. With the use of relational database programmes,
making a n e w classification whenever a n e w question arises or a n e w point of view
is sought, is easy enough, by ordering the data according to different criteria.
Because our main concern is the production process, the attention is focussed
on the activity and orientation of the systems as presented in Table 9-0.5
Why are Classes Empty ?

O f the 18 classes in Table 9-0, eight are filled, ten are empty. Four of these
classes are empty by definition, because the number of systems is the m a x i m u m
n u m b e r of orientations: a one-system technique can only be oriented in one
direction; a two-systems technique either in one or two directions. T h e area of
classes that are empty by definition is shown with hatching.6
Six classes are empty without an obvious reason. Three system techniques,
either with an orientation in three directions, do not occur in our basketry corpus
at all. Furthermore, three-systems techniques with only active systems do not
exist.
S o m e of these combinations are unlikely to occur. It is hard to imagine, for
instance, a three-systems technique with two active systems oriented in one
direction. It would be a coiling technique, for instance, with one passive coil and
two active strands wrapping around the coil. Such a technique is feasible but
complex and no improvement, either structurally or decoratively, to the simple
coiling or the coiled inlay techniques.
Likewise, three-systems techniques with only active elements, are difficult to
realise. This would force the basket maker to handle all strands simultaneously.
Without fixating some of them (which makes them passive, according to our
definition), these strands would soon form a jumble.
O n the other hand, some classes are empty for no obvious or likely reason.
T h e answer to the question w h y these empty classes occur, adds information
beyond what can be recorded from the archaeological record. A n aid in
determining whether or not a class is empty for obvious reasons is to compare the
Egyptian techniques with those from elsewhere (Table 9-23).
Table 9-0 listed Egyptian techniques, in Table 9-23 the same classification is
used for techniques from the Mentawai, a group of islands west of Sumatra in
Indonesia. T h e basketry from the Mentawai islands w a s collected by the
anthropologist R. Schefold from 1967 to 1969 and published by Keppel in 1984.
There are other technical criteria, which are of importance for different research
questions, such as the space between the elements, which is relevant, for instance, when
determining the function of a basket (milk container versus sieve). There is also a difference
between a fabric which is made as a spiral, or up-and-down (cf. Figures 19-4a and 19-4b).
Another set of classes which remains empty by definition, and therefore has not even
been incorporated in Table 9-0, are techniques with only passive elements. In the
classification the classes '1 system, passive', '2 systems passive/passive' and '3 systems,
passive/passive/passive' have been omitted, because they do not refer to a structure, but to a
stack of sticks or strands, passively lying on top of each other. By definition, at least one of
the systems in a basketry technique has to be active.
/
/
1 system 2 systems 3 systems
/
1 / 1
co/ <3
active passive/active active/active pas/act/act act/act/act 1
—
CO
co
CA
CO
3
a.
°
.a.
O
c
linking All
co
'
'
a.
OO
'
'
'o.
if
l~ _
3 -3
weaving B.II.3
c° CO
1) "O
wrapping A 1.2
knotting A.I 3
weaving bii.1.2 wrapping B.I.3 wrapping B.I.2

'
§.i
1 a
B III 3
3 .3
.a -a
c ..
CO CQ CO
binding
twining B.1.5
9 Basketry Techniques, a Classification
(Keppel 1984: 92)

waling B.1.5
: i
orientation in weaving B.II.2 weaving B.II 2

3 directions:
piercing B.I l
^. ^W ^c^ ^c* ^c* ^c*
X>
CD
co
3
'o
•o
c
cn
c-
o.
o.
UCO 0
*-H£
c
CA c
* ^
c u
O" <
1/3 ?
rz u
a B
c
-C -J
</><c
c CO
E CO
1 -un
ON
•oc *—
OJ
o Tuj
173
U eu
0r
l/-. <;.
c j>: ,
/)
The publication is highly suitable for a comparison of techniques, since it consists

of clear descriptions and technical drawings.7
F r o m the classification of the Indonesian material, it is apparent that there
are no technical objections to using three-systems techniques in three directions.
T w o forms of weaving and one form of pierced basketry are m a d e by the
Mentawai basket makers, with materials that have properties comparable to those
available in Egypt (Keppel 1984). O n the other hand, the Mentawai, masters in
three-systems basketry, apparently refrain from making three-systems techniques
with active strands only, nor do they produce three-systems baskets with an
orientation in one direction.
While the Mentawai collection has a strong emphasis on three-systems
techniques, they do occur infrequently in the Egyptian corpus (numbers 17-20 in
Table 9-21). Only looping (no. 18) and binding with three systems (no. 19) occur
regularly. The three-systems coiling technique (no. 17) is mainly used for
decorative purposes. The three-systems plaiting technique (no. 20) gives a
decorative and extremely strong rope with considerable pulling strength, but has
not been used widely (only one fragment was found in Qasr Ibrim).
The difference can partly be explained from the raw materials used. The
most important Mentawai raw material is rattan (Calamus rotang), a slender palm
species, in English usually referred to as 'cane'. F r o m the stems strips of several
metres long are cut for basketry production. In contrast, the length of Egyptian
raw materials (palm leaf, grasses, sedges, rushes) varies between 0.5 and 1.5
metres only. The necessity to add new lengths of material continuously makes the
three-systems techniques less suitable.
Secondly, the functional range of basketry and other available containers is
an important factor. Basketry fabrics vary between fine, dense and coarse, widely
spaced. Until recently, coiled basketry was even used to hold liquids.8 The
Mentawai three-systems basketry is mainly used for the production of coarse
openly spaced carrier baskets and bags. Ancient Egyptian carrier bags with similar
properties are made in knotting or twining techniques. Although function is not the
subject of this study, it will be brought up in Chapters 10, 11 (the ancient and
modern basketry from Middle Egypt and Egyptian Nubia) and Chapter 12. In these
chapters the specific properties of the 20 techniques listed in Table 9-0 will be
surveyed. By looking at the combination of material, technique, context and
function of objects that have been found in A m a r n a and Qasr Ibrim an insight can
be gained in the other important factors that determine the variety of basketry
techniques: cultural preferences and tradition (see Chapter 20).
Since the criteria for the Keppel's classification are slightly different, some of her
classes are divided over several classes, others are grouped in one class.
The Ababda bedouin of the Eastern Desert of Egypt used finely coiled baskets for
milking the camels. In the last ten years these baskets have been replaced by metal and
plastic containers.
C H A P T E R TEN
BASKETRY FROM AMARNA AND MIDDLE EGYPT
In the previous chapter, Table 9-21 (p. 169) lists the 20 techniques occurring in the
basketry corpus from the four selected sites in Egypt. Eleven of these occur in
Amarna, and eleven in present day Middle Egypt. These are not the same
techniques, however. Only six out of these 11 techniques occur both in amarna and
at present in Middle Egypt. This chapter gives a more detailed description of the
baskets found and a better insight in the frequency with which they occur.
T h e frequency with which the different techniques occur, has been measured
not according to the number of fragments that have been found, but according to
their position in the archaeological context. The expedition records the context at
A m a r n a in units, each unit an archaeological feature such as a layer or deposit. All
fragments found in one unit which are showing the same features, have been
counted as one object, unless it was obvious that the fragments represented more
than one object m a d e in the same technique.
There are two remarks that have to be m a d e in relation to the quantification
of objects in Amarna. The count has been based on the material excavated by B.J.
K e m p between 1979 and 1986 and the report of the earlier excavations by Peet and
Woolley (1923). The basketry as published by Peet and Woolley has been
incorporated as far as it was possible to discern the techniques from the
descriptions in the text. The 1920's material, of which only the complete or nearly
complete objects were recorded, comes from a large portion of the walled village.
All basketry fragments were recorded which were found in the recent excavations,
which took place within the walled village and the neighbouring areas (cf. Section
7.2.3, pp. 117-120).
Throughout this section references will be m a d e in foot notes to basketry
registration numbers.1 Technical features such as the insertion of n e w material,
centres and rims are referred to in Chapter 15.
A complete list of basketry finds from Amarna, including the interpretation of the
context, use, function, meaning, re-use and discard, will be published in a separate
monograph.
10.1 One-system Techniques

Knotless Netting (1)
Fabric. 1
One active system (consisting of one I
member moving in two planes). I
Orientated in one direction. I
(Figure 9-1) |
Amarna
This very distinctive technique (Plate 10-1), occurs in six instances in the A m a r n a
material.2 All fragments have been m a d e of strips of d o a m palm leaf (Hyphaene
thebaica), the width of the palm leaf strips show a variation of 6 - 1 0 m m . All leaf
strips have been twisted in an S-direction and in all instances the left side of the
loops cross over the right side (S-over crossing).
I'M i I i Ii
1 2 3 4 6 7 8 9 10 cm
Plate 10-1 Knotless netting re-used as roofing material, embedded in mud used to cover
the roof ( T A W V - 0344). Courtesy of the Egypt Exploration Society.
"Knotless netting: T A W 0303, 0344, 0376, 0581, 0825, 1039.

10 Basketry from Amarna and Middle Egypt 177
Apart from the six knotless netting fragments, a large number of isolated S-
twisted looped fragments of doam palm leaf have been found in 15 different
contexts.3 In total there are 21 units in which fragments of knotless netting have
been found.
In 1921 one fragment of knotless netting was found by Peet and Woolley.
They described this fragment as "chain bag of palm leaf" (Peet & Woolley 1923,
p. 74, pi. XXII,2).
Middle Egypt
The knotless netting technique does not occur in present day Egypt.
Grommet (2)
Linear
One active system (consisting of one
member moving in two planes).
Orientated in one direction.
(Figure 9-2)
Amarna
Depending on the size of the strand and the diameter of the circle, wrapping can
result in a open ring (grommet), or a closed circular object (pad). Further variation
is found by looking at the size and preparation of the strand: the appearance of the
objects is depends on the material used. In the workmens' village 11 objects have
been found, made in this technique. One small grommet has been made by one
culm of a plant (Juncus species) wrapped around itself in three turns.4
Pads are made by wrapping a string or a twisted bundle of fibrous material.
Three such pads were found, all made of tall grass leaves (both Desmostachya
bipinnata and Imperata cylindrica).5 The grass bundle, consisting of complete
grass plants, roots and all, is twisted in S-direction. After thefirstcoil, the grass
bundle is wrapped around itself two or three times in Z-direction. Figure 9-2
3S-twisted doam palm loops: T A W V 0052, 0370, 0423, 0437, 0452, 0458, 0469,
0472, 0476, 0478, 0493, 0541, 0604, 0651, 0659, 0697 (0052 and 0697 are from the same
unit).
"Grommet: T A W V 1067.
5Grass pads: T A W V 0058. 0059 and 1185.
178 The w o r l d According to Basketry
shows the four stages of construction of a grass pad as shown in Figure 10-2.6 The
same wrapping method was used to make a grommet which forms the core of a
pad, covered with string or palm leaf (see wrapping with 2 systems, p. 186).
In Peet and Woolley's excavation report from 1923, only one large grass pad
was reported, described as "twist of 3-ply rope" (p.82, plate X X I , 1).
Figure 10-2 Grass pad, made by wrapping one S-twisted strand around itself in Z-
direction, as shown in Figure 9-2 ( T A W V 0744).
Middle Egypt
Pads and grommets do not occur often in the present day villages around Amarna.
I have not seen any grass pads or grommets. The only ones that are made at
present, are cloth head rings, used by both m e n and w o m e n to balance the loads
they carry on their heads. The wrapping principle is also used to make coils of
rope.
Pad: T A W V 0744.
Knotted Netting (3)
Fabric. •
One active system (consisting of one 1
member moving in three planes). 1
Orientated in one direction. 1
(Figure 9-3) |
Amarna
T w o fragments of knotted nets have been found, both made of z-spun flax.7 The
flax yarns are irregular in diameter, ranging from 0.5 to 0.8 m m . The nets are
made with mesh knots. Peet and Woolley do not mention any netting finds.
Middle Egypt
In present day Egypt netting is made by machine from synthetic fibres.
Nevertheless, the knots used are still mesh knots and repairing the nets is seems to
be done in the same way as in the past.
10.2 Two-Systems Techniques

10.2.1 With one passive and one active system, orientation in one direction
Sewn plaits (4)
Fabric. |
One passive system. 1
One active system (consisting of one I
member moving in one plane) 1
Orientated in one direction. i
(Figure 9-4) 1
Nets: T A W V 0339, 0340.

Amarna
This technique has not been found at A m a r n a during the recent excavations. In the
report of the earlier excavations one photograph of a basket of this type has been
published (Peet and Woolley 1923 pi. XXI,4 a, p. 85). The basket w a s found in
the "bedroom of house Long Wall Street 11" and had been repaired with coarse
textiles. Although plaiting does occur sporadically in the pharaonic period, this is
the only basket in the sewn plaits technique is known to exist in Egypt before the
Graeco-Roman period. The basket from Long Wall Street 11 most likely is a
modern intrusion, considering that the workmen in Peet and Woolley's days, used
similar baskets to m o v e the earth from the excavation to the dumps.
Middle Egypt
Baskets made from plaited strips are at present an important local industry in
Middle Egypt. They are produced in several standard sizes, which are expressed
by the length of the plaited strip in ba'at. O n e ba'a is the length between two
outstretched arms (approximately 1.50 m ) . Thus the sizes are not exactly standard,
but depend on the length of the arms and the width of the basket maker's chest.
W.' wv-v ,j
^ %*>,*'• *£»
Plate 10-3 Modern sewn plaits carrier baskets from the Amarna region: one 'alaaga of
8 ba'at and two bagutat of respectively 3 and 4 ba'at.
10 Basketry from A m a r n a and Middle Egypt 181
The guff a is the largest basket, sewn from a plait with a length of 15 to 18
ba'a. The maqtafi which is, like the guffa used to transport and store goods, for
instance flour, is 11 to 14 ba'a. T w o of these large baskets are used, balanced on
the two ends of a wooden stick, to m o v e goods on donkey back.
The 'aldga (6 to 10 ba'at) is used as a shopping basket, the gauta or baguta
are the smallest baskets (3-5 ba'at) and mainly used for moving garbage, bricks at
building sites and earth from farmland and, for instance, archaeological
excavations (Plate 10-3).
In el-Hagg Qandil only plaits with nine 'uyun (plaiting strands) seem to be
produced. The more professional basket makers of el-Bersheh, a village 20 k m to
the North, make baskets from nine strands plaits and five strands plaits. The
baskets designed for coarse work, such as garbage collecting, are made with five
plaiting strands, with a \1/1\\ 1 plait pattern.
There is quite some variation in the sizes quoted for the different sewn plaits
baskets.8 According to the literature, Middle Egypt has a larger basket, called
zambil (Behnstedt and Woidich 1994, 191), comparable perhaps to the gotwiyya in
upper Egypt. The zambil is used as a carrier basket for dung and earth, or as a
two-part transport bag, but in the region of el-Hagg Qandil, the masbala is used
for this purpose (see below under twined basketry). There are also some
differences with the sewn plaits basketry from Upper Egypt.9
In el-Hagg Qandil, the larger baskets ('aldga and maqtaf) have a slightly
indented base, which makes it easier to carry them on the head This is also made
easier by the use of a head ring, rolled from a cloth or a scarf (p. 178).
A very specialised basket made in el-Hagg Qandil is the tala'a, a basket used
to carry the pollen from the male date palm tree up to the flowers of the female
date palm tree. This small longitudinal flat basket with a long rope handle is made
of a 4 ba 'at plait sewn into a cylindrical tube with a diameter of 50 c m and 30 c m
high. The tube is folded and the bottom is sewn (Plate 18-2, p. 398).
In the last decade the sewn plaits baskets are being replaced by carrier
baskets m a d e of old car tyres. These c o m e in two different sizes and are referred
to also as guffa or zanbtl. The terms are not related to the size of the basket, but
represent regional differences, guffa is used in Upper Egypt, zanbtl in Cairo.
8Badawi and Hinds 1986, 711; Behnstedt and Woidich 1994, 4-386, 389, 321
Q
According to Henein, a seven strands plait occurs in the upper Egyptian village of
Mari Girgis. He also spotted two types of baskets made of nine strands plaits, the 'aldga.
made of a plait with a length of eight or nine ba'at, and the much larger gotwiyya (Henein
1988, 183). These two basket types are also made from seven strands plaits, along with a
maqtaf of 5 ba'at. The five strands plaits are sewn into three basketry types: "couffin"
(maqtaf of 5 ba'at), "panier" (guffa of 10-12 ba'a) and "hotte double" (gotwiyya of 12 ba'a
for each side of this double basket, carried by camels). Thus the terms seem te refer to the
sizes and we find that the maqtaf is the smallest, followed by the 'aldga, guffa and
gotwiyya.
Coiling (5)
Fabric.
One passive system.
(Figure 9-5)
Amarna
Most of the basketry techniques are named after the appearance of actions of the
active elements, but coiling refers to the passive bundle which is held in position in
a coil by a wrapping strand
Fragments of 29 coiled objects have been found in A m a r n a , none of which
complete.10 F r o m the fragments it appears that all coiled objects are m a d e in the
same technique, with a stitch through the previous bundle (Plate 10-4). T h e raw
materials are also the same and the size of the bundle and winders do not show
great variation (Table 10-5).
The passive element is a bundle of grass, either Desmostachya bipinnata or
Imperata cylindrica. T w o exceptions have bundles consisting of d o a m palm leaf,
but these are small fragments from near the centre of the basket, which is often
m a d e from d o a m palm. O n e fragment was exceptional, because the bundle did not
consist of unspun grass, but of grass sZ2 string (Wendrich 1991c, 30-32)."
The active elements of all coiled basketry fragments consisted of strips of
d o a m palm leaf (Hyphaene thebaica). In all fragments the coil of grass was kept in
place by passing the strip of d o a m palm leaf through the previous coil, each time it
wraps around the bundle (Plate 10-4). The variation occurs in the centres of the
coiled basketry.
The size of the systems shows little variation. The diameters of the grass
bundles range from 4.5-16 m m . The smallest bundle is from a fragment of the
centre of a basket. In all baskets the diameter of the bundle is decreasing near the
centre. The largest bundle is from a basket that has almost disintegrated. The
bundles have become loose, which makes it impossible to infer the original
diameter of the bundle.
10,
Coiled fragments: T A W V 0302, 0358, 0359, 0360, 0361, 0362, 0374, 0378, 0404,
0487. 0512, 0552, 0652, 0653, 0657, 0687, 0698, 0734, 0775, 0822, 0965, 1035.' 1113.
1170, 1182. 1184, 3 without a number.
Coiled basket with bundle of grass string: T A W V 0657;
'^
4 ssl
n\t.\g >y\^*"V tt^*- <~s
V ' 3r%3
G
i?*«
Plate 10-4 Detail of the centre part of a coiled fragment from Amarna ( T A W V 0302).
Courtesy of the Egypt Exploration Society.
If w e leave this one out of consideration, all baskets have bundles with a diameter
between 5 and 15 m m , of which 13 m m is the most frequent bundle size. The
wrapping strands range between 4.5 and 12 m m . For 25 baskets w e have
information on the size of both active and passive elements.
^ \ ^ Passive system:
Active ^""^\^ diameter of
system: ^^^bundle 0-5 m m 5-10 m m 10-15 m m 15-20 m m
width of winders ^^
0-5 m m 1 - 3 -
5-10 m m - 8 10 1
10-15 m m - - 2 -
15-20 m m - - - -
Table 10-5 Relation between the size of the active and passive systems of 25 of the
coiled baskets found at Amarna.
Table 10-5 shows the ratio of the size of the bundle and the wrapping strand.
In all fragments the size of the palm leaf strand is the same, or smaller than the
diameter of the bundle. In Table 12-4 ( p. 256) the size of the elements will be
compared with that of the coiled basketry from Qasr Ibrim.
The description of the basketry finds in the publication of Peet and Woolley
(1923) is minimal. For each room that was excavated, a list of finds is published,
in which basketry is described in very general terms, such as: "a basket of the
usual kind". It seems safe to presume that such an entry refers to coiled basketry,
since the 1979-1986 excavations showed that coiling is the only technique used at
A m a r n a to make containers with rigid walls (baskets, rather than bags). Judging
these limited descriptions, the number of coiled baskets recorded by Peet &
Woolley is presumably 19. N o details are given on technique, size and material.
The context is not indicated, except for the street, house and room number.
The total amount of coiled basketry fragments from the excavations at
A m a r n a is 48: 29 from the recent excavations and 19 from earlier excavations.
Considering that the area excavated in the 1920's was about three times as large as
the area excavated between 1979 and 1986,12 there must have been a large number
of coiled fragments. N o estimation of the quantity can be given, since during the
earlier excavations only the complete or almost complete objects were kept, while
the smaller fragments were discarded.
1"The
2„ recent excavations concentrated on the areas outside the walled village, rather
than the houses. Kemp excavated in total 4 houses, while Peet and Woolley have excavated
37 houses.
Middle Egypt
Coiling is a technique that is widespread in Middle Egypt. M a n y w o m e n in the
villages of el-Hagg Qandil, el-Amariyya and el-Till k n o w h o w to m a k e coiled
baskets. T h e basic structure of the coiled baskets is identical to the ancient
basketry, the coiled bundle being fastened with palm leaf wrapping, which is
stitched through the previous coil. Differences occur in the choice of the raw
materials and in the construction of the centre of the basket.
The raw material used at present for the bundle material are fibres of the
zaghauwah, which is the Middle Egyptian term for the curved stem on which the
dates grow. The bundle is fastened with strips of date palm leaf.
^ ^ ^ | f f '
Plate 10-6 A coiled bread basket (tabaq) of three hand spans wide.
The baskets that are made in Middle Egypt are mainly the tabaq, a large flat
bread basket, and the sabat, a deep basket with rigid, straight, slightly flaring
sides, for storing flour. T h e base of the standard tabaq, is three ashbaar wide,
which corresponds with the diameter of the flat bread which is made in the area.
The shibr is the width of a hand span, three ashbaar is approximately 55 c m (Plate
10-6).
For the tourists a smaller version is made of two ashbaar (36 c m ) , or even
one-shibr baskets which can be used for presenting sweets. Another variety on the
tabaq baskets is the seniyya (tray): a round shallow basket with two handles.
Recently some other innovations have appeared, such as small baskets with lids, to
store trinkets and a special sabat-tilifon in which the telephone can be carried from
one side of the room to the other.
All baskets are executed in simple, undyed date palm leaf. Although the
bundles have quite a large diameter (an average of 20 m m ) , the work is not coarse
in appearance, because the stitches are very regular.
Wrapping (6)
Linear.
One passive system.
(Figure 9-6)
Coiling and wrapping require the same movement of the active element, but can be
distinguished by looking at the interaction of the systems. In coiling the active
strand passes through the previous coil to form a fabric. In wrapping a strand is
just wrapped around a passive bundle, such as the core of a ring or the handle of a
brush.
Apart from rings, pads and brushes other wrapped objects are sticks and
reeds, wrapped in palm leaf or sedges. Furthermore, the production of three-ply
rope at Amarna is also a form of wrapping: the passive system is formed by a two-
ply rope and a third strand (the active system) is wrapped around to form a three-
ply rope (cf. Section 15.6).
The dimensions of the wrapped rings differ, as does their function. Small
rings, with a diameter of about 25 m m , are made from string or palm leaf,
wrapped with doam palm leaf. Of these seven were found in A m a r n a recently,13
while none were reported by Peet and Woolley (1923). The large rings are ranging
in diameter between 170 and 550 m m . and consist of a bundle of grass, wrapped
with string or twisted grass. Depending on their sizes, these rings were used as pot
stands, supports for carrying objects on the head and as support for the pointed
bases of amphoras in carrying nets (Figure 9-6). In total 19 large rings, or
3Small palm leaf rings: T A W V 0764, 0768, 0769, 0796, 1070, 1123, 1135.
fragments of rings have been found, seven by Peet and Woolley, 12 during the
recent excavations.14
Seven small pads were found of which the core was not just a bundle of
grass, but a grommet (see p. 177), covered with a wrapping of string and palm
leaf.15 The pads are made in two stages: first the core is made, then the core is
wrapped with palm leaf. In addition there is a group of small pads, with a diameter
of 25 m m , made by wrapping doam palm leaf around a string or palm leaf circle.
Three of these small pads have been found at Amarna."1 Such pads have not been
reported to have been found at the earlier excavations.
Plate 10-7 T w o floor brushes with wrapped handles. At the top a modern brush from
el-Hagg Qandil, made of date palm fibre. At the bottom a well worn brush
with three fingers from the workmen's village, made of doam palm leaf
strips and grass string. (Courtesy of the Egypt Exploration Society).
14Large rings: T A W V 0302A, 0363, 0364, 0541, 0670, 0671, 0688, 0733 and four
without number.
l5Pads with traces of wrapping: T A W V 0296, 0297, 0298, 0299, 0658, 0744, one
without number.
16Small palm leaf pads: T A W V 0766, 0767, 1066.
Brushes consist of a bundle of grass or palm leaves, with a handle made by

wrapping a string around the bundle (Plate 10-7). Most brushes are tied a few
centimeters above the middle of the bundle. The lower side of the bundle is folded
back over the upper half and the string is wound from the folded end upwards to
form the handle (cf. Section 15.6). The same string is sometimes used to form
three to six divisions in the bundle, thefingersof the brush.
In total seven brushes were found between 1979 and 1986, of which six had a
wrapped handle.17 T w o of these brushes are divided in respectively three and four
fingers. The handle of the seventh brush was not wrapped, but knotted (see
below).
Peet and Woolley report to have found seven brushes, two with fingers (three
and six). From the wear pattern and depictions in the A m a r n a tomb paintings, it is
clear that the short handled brushes with fingers were used as floor brushes. Small
undivided brushes bear evidence of having been used as paint brushes (cf. Figure
10-9, for a paint brush with a knotted, rather than wrapped handle).
The last group of wrapped objects are a number of sticks and reeds lined
with doam palm leaf and then wrapped with doam palm leaf or papyrus (Plate
10-8). Not the strong outer layer from the papyrus culm was used for wrapping,
but the soft spungy inside. Although all sticks were found unattached, they were
probably part of furniture. The shrinkage of the spungy papyrus core tightens the
connections between the sticks which form a table, stool, or chest. In total 18 of
such sticks were found between 1979 and 1986.18 None have been reported by Peet
and Woolley.
: :
Plate 10-8 T w o sticks lined and wrapped with doam palm leaf, probably part of
furniture ( T A W V 0453). (Courtesy of the Egypt Exploration Society).
igWrapped brushes: T A W V 0290, 0295, 0300, 0548, 0690, 1015.

Sticks and reeds: T A W V 0266, 0267, 0272, 0273, 0453 (3x), 0454 (2x) 0486 (2x)
0528, 0593, 0709, 0710, 0810, 0811, 1183.
Middle Egypt
Rings, used as pot stands are not found in the modern villages around Amarna.
Likewise, furniture made of wrapped reeds do not occur. Instead, furniture is
m a d e with the pierced technique (p. 201). The only wrapped objects that are found
regularly produced in Middle Egypt are short-handled brushes. They are
commonly used in households for sweeping the floor. Although they are n o w m a d e
from date palm fibre, rather than d o a m palm leaf, they are still m a d e by wrapping
string around a folded bundle (Plate 10-7). They are increasingly replaced by
brightly coloured plastic brooms, however.
Knotting (7)
Linear.
One passive system.
member moving in three planes).
(Figure 9-7)
Amarna
This technique is represented by one brush from Amarna. It is made in a similar
w a y as the wrapped brushes, by tying a bundle of d o a m palm leaf strips just above
the middle, folding back the bundle and fastening the fibres by tying the string
around the bundle, forming a handle with the tying string. In this case, the handle
is not wrapped, but knotted (Figure 10-9). The active system moves in three
planes, that is to say: the strand is pulled through a loop formed by itself.19 The
brush has not been divided in fingers. Remains of lime in between the fibres
indicates that it has been used for whitewashing.
Middle Egypt
This technique has not been attested in the villages around the archaeological site
of Amarna.
19.Knotted brush: T A W V 0291;

Figure 10-9 Construction drawing of a knotted brush ( T A W V 2091). The bottom half is
drawn in cross section to show the knot around the bundle, which is then
folded back and fastened with the same string.
10.2.2 With one passive and one active system, orientation in two directions
Weaving with one strand (8)
Fabric. 1
One passive system. \
member moving in one plane) 1
Orientated in two directions.
(Figure 9-8) 1
W e a v i n g is in the first place a textile technique, but if the warp or the weft are of
limited length (e.g. plant culms or leaves) then it fits the definition of basketry.
Unclear is, however, if weaving with string should be considered basketry. String,
as a half product, has in principle an unlimited length and its shape stays the same
throughout its entire length, even w h e n m a d e of 'basketry' materials such as grass
and palm leaf. Mostly for practicle purposes, the Egyptian objects m a d e with
'basketry' materials but applying 'textile'-techniques are incorporated in this

survey (cf. Wendrich 1991 a and b).20
Amarna
Most w o v e n fragments found at A m a r n a are made with a passive system (warp) of
grass string and an active system (weft) of grass leaves (Plate 10-10). In 17 units a
number of 21 large fragments of this type of matting have been found.21 M a n y of
these fragments have edges, which are useful indicators of the weaving procedures
(see Section 15.8). The distance between the passive strings is consistently around
2 c m , ranging from 18,2-23,5 m m . Apart from the grass matting, a number of
fragments of side edges were found. O n e of woven matting with bundles of straw,
( T A W V 0308), the other with bundles of papyrus culm (Plate 12-3, p. 253).
Plate 10-10 Woven matting made with a warp of grass string and a weft of grass leaves.
To the right: detail ( T A W V 0007/0008). (Courtesy of the Egypt Exploration
Society).
20
A n important reason for not drawing the line too rigid is a practical one: objects
made in a 'textile' technique, but of a 'basketry' material, such as grass or palm leaf, are in
danger of not being studied: a textile archaeologist will not be inclined to regard such
objects as textile on the basis of the raw material.
2'Mat weaving: T A W V 0001, 0002, 0004, 0005, 0006, 0007, 0008, 0009, 0008,
0010, 0012, 0046, 0057, 0353, 0356, 0357, 0575, 0674, 0749, 1144.
The W o r l d According to Basketry
192
F r o m Peet and Woolleys descriptions ("fragments of matting" p.76, 77, 78,

82) it is often not clear to which technique they are referring. O n e fragment,
which is definately woven is mentioned on p. 58 and a photo is available
(pi. XX,4). For this fragment no context is given. M a n y of the mats found by Peet
and Woolley are twined (see p. 194).
Middle Egypt
In Middle Egypt today mats are woven with grass or rushes. The thick grass mats
are used as sleeping mats, and as cushioning for the mastaba, the brick benches in
front of the houses where people sit in the evenings. The rush mats are used as
floor mats. The grass matting is increasingly replaced by rag rugs, while the
plastic mats take over the role of the rush mats.
Weaving with two strands (9)
Fabric.
One passive system.
One active system (consisting of two
members moving in one plane).
(Figure 9-9)
Amarna
Peet and Woolleys photographs show woven chair and bed matting, made of S-
twisted doam palm leaf. Such matting is made by stringing a wooden frame and
weaving in the crossing elements after that. The pattern in which the chair- and
bedmatting is woven is \1/1\\1, with two parallel strands. Peet and Woolley found
three fragments of furniture matting, with parts of the furniture still attached. The
descriptions they gave were: "fragment of basket work" (1923, p. 82, pi. X X I , 1
obj. 22/29), "bed with twisted rush mattress" (1923, p. 63, pi. XVIII,4) and
"fragments of wooden stool with wickerwork seat" (1923, p. 79).
Wickerwork in the original meaning of the word, is basketwork made of
willow rods, a material not used in ancient Egypt. Chairmatting in ancient Egypt is
either made with fine flax yarns, or with twisted doam palm leaf. N o photo of the
Amarna chair matting was published, but the term wickerwork suggests that the
chair had a seat woven of robust material, for which twisted d o a m palm leaves
would qualify. The bedmatting could not be identified from the photograph, but
again the term "twisted rush matting", suggests the use of doam palm leaf. This
material does not resemble rush (Juncus species), but has some resemblence to
bulrush (Scirpus species), a material used for chair matting in England (Peet and
Woolley's country of origin). During recent excavations, no fragments of this
matting type have been found, but several lumps of dried m u d show the impression
(cf. Plate 10-11). This indicates that furniture matting has been re-used as roofing
material.
. * I .> Hz •
Plate 10-11 Impressions in a dried mud roofing fragment of weaving with double strands
Middle Egypt
Today stringing beds and chairs are still practiced, but not in Middle Egypt. In
Chapter 11 more will be said about bed matting, which, in spite of dwindling
demand, is still made in N e w Nubia.
Twining (10)
Fabric. •
One passive system. I
1
members moving in two planes).
1
Orientated in two directions. 1
(Figure 9-10) 1
Amarna
Twining is the most widespread technique used in Amarna. Although the basic
principle of this technique is that of two active strands twisting around the passive
strands and each other, there is considerable variation in the appearance and
functional properties of the twined fabrics. This is determined by several factors,
such as the size of the elements, the space between elements, the preparation of
the elements and the relation of the systems (Table 10-12).
Elements (size and composition)

size passive elements fine (4 m m or smaller)
coarse (larger than 4 m m )
active elements fine (4 m m or smaller)
coarse (larger than 4 m m )
space passive elements close (space smaller than size)
open (space same or larger than size)
active elements close (space smaller than size)
open (space same or larger than size)
preparation passive elements unworked (leaf strip or bundle)
worked (string)
active elements unworked (leaf strip or bundle)
worked (string)
orientation of active elements S-twined
twining Z-twined
combined S and Z twined
Systems (interaction)
number of passive system single (twining around one strand)
rows double (twining around two strands)
active system single (single row of twining)
double (double row of twining)
Table 10-12 Variations in twined fabric can be specified by looking at the size and make
up of the elements and the interaction of the systems.
The size of the elements is divided into attributes fine and coarse, the space
between the elements into open and close. The elements are worked or unworked,
that is to say, they are either used without m u c h preparation ('cut into strips' is
here considered 'unworked'), or they have been prepared for use in a separate
production phase: grass leaves are made into string.
A description of the interaction of the systems gives three possible orientation
for twining: S, Z and a combination of both (see Wendrich 1991c, p. 58, Figure
36 d and e). Furthermore, there are single and multiple passive strands and rows
of twining. Thirteen combinations were found at Amarna (groups A-I and K-N,
see Table 10-13). With the exception of one group (G), all twining is done in S-
direction.
Seven groups of fine twined matting were found at Amarna, which represent
only a small number of fragments. Group A consists of fine S-twisted d o a m palm
leaves (worked), which are held together with widely spaced rows of twining with
fine zS2 string (worked). Only two fragments found were made in this technique.22
Group B is matting, made completely of fine string. The passive strings are
double, held together by single rows of twining.23 Group C is fine matting
consisting of small grass bundles which are fastened with single rows of widely
spaced twining with string.24 Very similar fine twined matting, but with double
rows of twining occurred in one instance (group D).25
Group E is fine twined basketry which is worked open in two directions an
example of which was found by Peet and Woolley. It is a sieve which is at present
on exhibition in the British M u s e u m . Strips of doam palm leaf are evenly spaced
by single rows of twined doam palm leaf.26 This is very similar to a fragment
found in 1985, the difference being that this small fragment has been twined in S
and Z direction (group G ) . The result is a decorative V-shaped pattern.27 Probably
also part of a sieve is a fine open twined fragment with double passive elements.28
Coarse twined matting occurs also in seven variations, but is m u c h more c o m m o n .
The first group of coarse matting (H) is closely twined, both systems made of
string.
2 2 T A W V 0776, 0786 (twining group A).

2 3 T A W V 1150 (twining group B).
2 4 T A W V 1112, 1166, 0718 (plus 0719 from the same archaeological unit), 0726
(twining group C).
T A W V 0717 (twining group D).
2 6 T A W V 0354, Peet and Woolley 1923, PI. XXII.2 (twining group E).
2 7 T A W V 0011 (twining group G).
2 8 T A W V 0583 (twining group F).
size space preparation rows Or.
pas. act. pas. act. pas. act. pas. act.
A fine fine close open worked worked single double S
B fine fine close open worked worked double single s
C fine fine close open unw. worked single single s
D fine fine close open unw. worked single double s
E fine fine open open unw. unw. single single s
F fine fine open open unw. unw. double single s
G fine fine open open unw. unw. single single sz

H coarse coarse close close worked worked single single s
I coarse coarse close close worked worked double single s
J coarse coarse close open worked worked single single s
K coarse coarse close open worked worked double single s
L coarse coarse close open unw. worked single single s
M coarse coarse close open unw. worked single double s
N coarse fine close open unw. unw. single single s
Table 10-13 Variations in twined fabrics found at Amarna. The 14 groups are discerned
by the criteria listed in Table 10-12, cf. Figure 10-14. (Qasr Ibrim: p.232).
Five fragments have been found belonging to group H.29 Very similar in
appearance is one fragment of closely twined matting in which the passive
elements consist of pairs of string (group I).30
Group J has not actually been encountered, but can not be excluded either.
Small fragments, which only consist of one r o w of the twining, could have been
part of matting with a passive system of single strings. Since the passive system is
unknown, these fragments could belong either to group J, K or L.31
29„
T A W V 0288, 0301, 0703 (0713, 0714, 0720 from the same archaeological unit) and
0741 (twining group H ) .
T A W V 0712 (twining group I).
3 1 T A W V 0436, 0928, 0951, 0989, 1165 (twining group J, K or L).
w m w m m m m m m m m m m
Figure 10-14 Variation in twined fabrics found at Amarna, as listed in Table 10-13.
198 THE W O R L D ACCORDING TO BASKETRY
Most probably, however, these are remains of matting group L, since this is the
most c o m m o n group. Group K consists of open twining, m a d e entirely of string,
the passive elements being grouped in pairs.32 A coarse matting m a d e of bundles of
grass or palm leaf, twined with single rows of string L was found in 32 units.33
The matting of group L is very similar to that of group M , which consists of
bundles of grass or palm leaf with double rows of string twining.34
The last group of twining consists of 30 m m wide strips of d o a m palm leaf,
twined with 2 m m wide strips. Only two small fragments of this kind of matting
have been found.35
Reflections on Size
Looking closer at the 14 twining combinations, a few remarks can be made on the
relation between the technique and the properties of the raw materials. First of all
the size of the elements: both fine and coarse elements occur at Amarna, but with
exception of twining group N , the passive and active elements are either both fine
or both coarse. There is a technical reason for this: the coherency of twined
matting would suffer from combinations of fine and coarse.
Coarse passive strands in combination with fine active strands does not work,
because fine strands are not strong enough to hold large bundles of material in
place. Fine passive strands in combination with coarse active strands are no good,
because small bundles of grass would be squeezed and deformed by large twining
rope.
The only instance where coarse and fine are combined is in twining group N ,
of which two small fragments have been found. The passive elements are wide
strips of doam palm leaf, which are fastened with fine strips of d o a m palm leaf.
The solidity of the leaf strips, which are not very flexible, prevents the passive
elements from slipping out of the rows of twining (cf. Figure 10-14).
Reflections on Space
In a few fragments both the passive and the active system are open spaced (twining
groups E, F and G ) , this is always fine twined basketry, with unworked elements.
The fine twined basketry can be divided roughly into two groups: one in which
both systems are made from the same kind of material and the other in which the
active and passive systems consist of different materials. The first group seems to
be confined to sieves, which obviously have to have an open structure. The second
group consists of finely made matting.
3 2 T A W V 0122, 0505, 0702, 0711 (twining group K).

3 3 T A W V 0003, 0045, 0111, 0161, 0216, 0278, 0281, 0286, 0287, 0342B, 0343,
0354, 0365, 0371, 0386, 0405, 0432, 0497, 0614, 0655, 0691, 0692, 0699, 0730, 0742,
0795, 0845, 0846, 0871, 0912, 0947 brush, 1021, 1176, 1191, 1214, 1253, 1254 (5 groups
of fragments from the same unit) (twining group L).
T A W V 0377, 0379, 0656 (twinging group M ) .
3 5 T A W V 0473, 0673, 0677 (twining group N).
In coarse basketry only the active element is sometimes openly spaced: in

coarse twining the passive system forms the body of the matting, which is kept
together by widely spaced rows of twining. The distances between these rows vary
between 20 and 100 m m .
Plate 10-15 Fragment of a finely twined sieve grid, made of strips of doam palm leaf
( T A W V 0011) (Courtesy of the Egypt Exploration Society).
Closely twined basketry in two directions occurs only in coarse basketry,

although Plate 10-15 shows a fragment which seems to be fine basketry which is
closely twined in both directions. This fragment, belonging to twining group G
( T A W V 0011) appears to be closely twined, but the structure is very loose, and
the rows of twining can be moved. It is probably a fragment of an open spaced
sieve grid, with a small mesh.
Orientation (twining direction)

It is a trend in Amarna that twining is done in S-direction. Z-twined basketry does
not occur, which is remarkable, because it would be technologically the most
suitable twining direction. A s with rope making, which is done by spinning and
plying the strands in opposite direction, to ascertain the coherency of the string
(see p. 203, and Section 15.13), twining would ideally be done in the direction
which is opposite of the orientation of the twining string. All twining string is spun
in z-direction and plied in S-direction, which would m a k e Z-twining the logical
orientation. M o r e is said about the orientation of twining in relation to the Qasr
Ibrim twined matting (Section 11.2.2).
Peet and Woolley probably have found five fragments of twined sieves and
18 fragments of twined matting. It is not always clear from their description if the
matting was woven or twined.36
Twining is used for making flexible objects, such as mats or bags. At
A m a r n a one brush has been found which was not m a d e by folding double a bundle
of leaves, but by holding together a number of grass bundles with S-twined zS2
string (Figure 10-16).37
Figure 10-16 Brush made of a rolled twined mat ( T A W V 0947).
Peet and Woolley 1923, pp. 58, 68, 72-79, 82 (plates XX.4 and XXII,2).
T A W V 0947, twined brush.
Middle Egypt
In present day Middle Egypt twined basketry is still used, but there are not m a n y
variations. Closely twined string matting (twining group H ) is produced in the
region of A m a r n a (Sections 15.10 and 16.4). They are made of date palm fibre
string and used as door mats. T w o mats sewn together make a large bag to put on
donkey back for the transportation of earth and dung.
Grass floor mats with widely spaced rows of twining (group L ) are not
produced in present day Middle Egypt. At present floors are covered with (plastic)
w o v e n matting. The finely twined sieves from antiquity have modern equivalents
of metal mesh.
The only other modern application of twining is in the production of rigid
screens, which are part of farmers huts as protection against wind and sun in the
fields (sabata). These are made of the culms of reed or sugar cane, bound or
twined together with S-twined zS2 string.
Pierced (12)
Fabric.
One passive system.
fT—n—ITT—f\ One active system, immobile.
<4Jrr v M&> (Figure 9-12)
H ^
Amarna
This technique has not been found in Amarna.
Middle Egypt
Pierced basketry is made in el-Amariyya. The mid ribs of date palm leaves are
used for making crates, cages, chairs, tables, beds and a bat-shaped appliance for
throwing bread in the oven (matraha, Plate 10-16).
The pierced technique consists of stakes, made of pointed split palm
branches, and strips of the same material, in which holes have been punched with
the help of a mould. The passive elements are always widely spaced. The spacing
of the active elements is open or close, depending on the function of the object.
In the same area, the pierced technique is also with flexible elements: the
string of the active system is 'pierced' through the ply of the string making up the
passive system. All string is made in zS2 orientation out of date palm fibre (Figure
10-17).
Plate 10-17 Workshop for the production of products from date palm mid ribs. A
matraha (bread bat) is made by pushing pierced strips (active system) over
stakes (passive system). To the left, in the corner, a small open cupboard.
made in the same technique.
Figure 10-18 Carrier net. made in a pierced technique: the active string is pushed through
the passive string to form a net.
10.2.3 Two-systems techniques with two active systems in one direction
Rope making (13)
Linear. j
T w o active systems (moving in two
planes).
Orientated in one directions.
(Figure 9-13) 1
Amarna
Here rope making refers specifically to the production of two-ply rope (consisting
of two strands). The grass bundles were spun and plied, in opposite direction, so
that the ply holds the spin and vice versa (cf. Section 15.13).
10.2.4 Two-systems techniques with two active systems in two directions
Continuous plaiting (14)
Fabric.
T w o active systems (moving in one plane)
(Figure 9-14)
Amarna
Peet and Woolley found what seems to be a fragment of continuous plaiting in a
U / 1 W 1 pattern impressed in m u d of(Peet and Woolley 1923, p. 57, pi. XVIII,6).
N o continuous plaiting w a s found during recent excavations and none is k n o w n
from other N e w Kingdom contexts. The possibility must be taken into account that
this fragment is a later intrusion.
Middle Egypt
N o continuous plaiting is produced in Middle Egypt
Plaited strips (15)
Linear.
(Figure 9-15)
Amarna
Plaiting is an old, but not an ancient technique in Egypt. T h e only plaited strips
found during the recent excavations at A m a r n a are plaits with three strands.38 The
date of these plaits is uncertain, since they were found in drift sand, during
cleaning, or in unsealed deposits. Judging from the materials used (papyrus and
grass) they are probably old, but they might not date from the 18th dynasty. A s
mentioned before (p 180), Peet and Woolley reported to have found a basket in the
sewn plaits technique, which is probably not ancient.
While twining and weaving occur from the earliest periods onwards, plaiting
seems to have been introduced into Egypt in the 21st dynasty or later. In m u s e u m
collections there are m a n y plaited sandals, but these are in general badly
provenanced. The earliest dated plaited sandal is on exhibition in the Egyptian
M u s e u m in Cairo and is reported to date from the third intermediate period (1070-
712 B C ) . A m o n g the many sandals found in the 18th dynasty tomb of
Tutankhamon no plaited ones were found, all sandals were made in a variant of the
coiling technique.
Middle Egypt
In the three villages of el-Hagg Qandil, el-Amariyya and el-Till plaited strips are
m a d e for the production of sewn plaits basketry. The strips are m a d e both with
five and nine strands (cf. Figure 9-15, see also p. 181).
Knotting (16)
Fabric.
T w o active systems (moving in three
planes)
(Figure 9-16)
Amarna
In one unit a large number of fragments were found of papyrus string with
overhand knots. These were found in relation to a wrapped ring and a rope handle,
also m a d e of the rhind of Cyperus papyrus. The isolated knots are part of the only
example of knotting with two systems k n o w n from Amarna, a large net for
carrying amphoras. Figure 10-19 shows a reconstruction of the carrying net. The
ring, holding the base of the amphora is shown in Figure 9-6. T h e handles were
made of sZ2[S]3 rope (Figure 10-20).
38T A W V 0584,0805, 0899, 1014. 1097, three-strand plaits.

Figure 10-19 A knotting fabric which forms part of an amphora carrier net (reconstruction
of T A W V 0292).
Figure 10-20 Detail of the handle of the amphora carrier net of figure 10-19.
10.3 Three-Systems Techniques

Looping (18)
Fabric.
T w o passive systems.
One active system (moving in two planes).
(Figure 9-18)
Amarna
Middle Egypt
Special mats are made to drain the young white cheese. These mats are k n o w n as
shinda and are used and made by all Middle Egyptian farmers w h o o w n a c o w or
water buffalo and thus have milk to produce the cheese.
Binding (19)
Fabric.
One active system (moving in three
planes).
(Figure 9-18)
Amarna
Middle Egypt
In order to make the walls of farmers' huts which protect the workers in the fields,
stalks such as reeds or maize are tied to several stems which have been layed
across.
CHAPTER ELEVEN
B A S K E T R Y F R O M Q A S R IBRIM A N D N E W NUBIA
O f the 20 techniques listed in Table 9-21 (p. 169) 16 occur in Qasr Ibrim and eight
in present N e w Nubia. The frequency in which the different techniques occur, has
been measured not according to the number of fragments that have been found, but
according to their position in the archaeological context. At Qasr Ibrim the context
is recorded in units, each unit being an archaeological feature such as a layer or
deposit. All fragments found in one unit and showing the same features, have been
counted as one object, unless the fragments clearly represented more than one
object made in the same technique. Although excavations have taken place from
1964 onwards, the material presented here is from only two areas, excavated in
1990 and 1992.' Throughout this section references will be made in foot notes to
the basketry registration numbers.2
Technical features such as the insertion of new material, centres and rims are
also important indicators of technical traditions, but will be considered in relation
to the production process, subject of Chapter 15.
11.1 One-system Techniques
Grommet (2)
Linear.
(Figure 9-2)
These are areas 10 and 14 excavated by D.N. Edwards and H. Roberton. The
basketry registration numbers consist of an indication of the area, three digits indicating the
unit and three digits indicating the unique basketry number.
"A complete list of basketry finds will be incorporated in a separate publication on the
Qasr Ibrim basketry.
Qasr Ibrim
Three grommets were found, m a d e of three different materials: d o a m palm leaf,
date palm fibre and grass.3 These were all found in Ballana contexts (third to sixth
century A D ) . They were probably used as pot stands.
New Nubia
There is no Information on the use and production of grommets in present N e w
Nubia.
Knotted Netting (3)
Fabric.
member moving in three planes).
(Figure 9-3)
Qasr Ibrim
Despite the wealth of organic materials surviving at Qasr Ibrim, only one small
netting fragment w a s found.4 The net w a s made of cotton yarn and mesh knots.
The context has been dated to the seventh century A D .
New Nubia
As is true in Middle Egypt today,fishermenuse netting made of synthetic fibres in
a factory.
11.2 Two-Systems Techniques

11.2.1 With one passive and one active system, orientation in one direction
Sewn plaits (4)
Fabric.
1
One passive system.
member moving in one plane)
(Figure 9-4) 1
Grommets: 10.103-537 doam palm; 10.133-539 date palm; 10.463-554 grass

Netting: 10.206-544.
11 Basketry from Qasr Ibrim and N e w Nubia 209
This technique was very c o m m o n at Qasr Ibrim. Variations within the class of
sewn plait basketry, which is represented by 181 objects, is found by applying
additional criteria, as was done for the twined matting from Amarna (pp. 194-201)
The variation of sewn plaits basketry is not determined by the same aspects
as used for twined basketry do not grasp the variation in sewn plaits basketry.
There is, for instance, no variation in the space between the elements and the
interaction of the systems: all baskets and mats are closely spaced, since they
consist of plaits sewn into a seemingly ongoing fabric (Table 11-1).
Elements (size and composition)

size passive elements varied depends on plait pattern and size of
plaiting strands
active elements fine 4 m m or smaller: strips of doam palm i
coarse larger than 4 m m : string
space passive and active close edges of plaits pulled inside each other
preparation passive elements worked plaited strips
active elements unworked strand of doam palm leaf, fine
worked string, coarse
plait pattern active elements worked see Table 11-2
Sy items (interaction)
no variation (sewing the edges of the: plaited strips)
Table 11-1 Variations in sewn plaits basketry can be specified by looking at the
preparation of the passive elements (the plait pattern of the plaited strip) and
active sewing strand (strand or string).
Although the size of the elements varies, this is not the core of the matter,
since the size is mainly the result of the kind of plaited strips used. T h e variations
in the plaited strips (the passive system) and the sewing strand (the active system)
are w h a t is of real importance in order to describe the variation within the s e w n
plaits technique at Qasr Ibrim.
Variation of plaited strips

T h e variation lies both in the n u m b e r of plaiting strands and the plait pattern
(Table 11-2, Figure 11-3). A t Qasr Ibrim the most frequently used plait is m a d e
with 9 p a l m leaves at a time. T h e pattern, indicated with the formula \2/2\\l, is
one strand passing under two, over two, with a shift of one (Wendrich 1991c,
65-66).
strands pattern frequency % sewing strand
5 U/1U1 21 ; 1 1 % grass string
7 \2/l\l/2\\l 1 j < 1% doam palm leaf strip
9 \2/2\\l 122 ; 6 7 % string or palm leaf strip
11 \2/l\l/2\\l 1 | <1% doam palm leaf strip
13 \2/2\\l ; 11 j 7% doam palm leaf strip
17 \2/2\\I ; 23 i 13% doam palm leaf strip
21 \2/2\\l \ 2 1 1 % doam palm leaf strip
Table 11-2 Variation in plaited strips. O f the 181 sewn plaits baskets 122 are made with
9-strands plaits (67%).
This pattern occurs 122 times,5 considerably more than the second and third groups
m a d e respectively with 17 palm leaves in a \2/2\\l pattern and with five palm
leaves plaited in a M / 1 W 1 pattern.6
T h e plaits used for sewing into basketry all have in c o m m o n that their edges
are orientated perpendicular to each other (SZ, cf. Figure 11-3). Edges oriented
parallel to each other (SS) cannot be pulled inside each other and are, therefore,
not suitable for making mats or baskets in the s e w n plait technique. This is the
'Qasr Ibrim Plaiting Rule'.
T h e first plait s h o w n in Figure 11-3, is m a d e with fours. It has t w o S-
orientated edges and is, therefore, not suitable for sewing into a fabric. T h e second
Strips plaited with 9 strands V2/2W1: 10.001-397,398; 10.012-391, 393, 395; 10.024-011;
10.026-112: 10.032-008, 401, 402, 403, 404; 10.041-094, 095; 10.045-097 099-' 10 046-037 038'
10.048-026; 10.049-005; 10.054-004, 029a + b, 030; 10.061-206, 411; 10.066-412' 10 069-09L
10.071-006; 10.072-086, 515; 10.073-417; 10.075-172; 10.082-087, 420' 10 095-421' 10 098-105
110, III, 424; 10.099-103; 10.103-082, 083, 106; 10.105-427; 10.121-107- 10 122-198' 10 124-429'
431, 432; 10.133-085, 434; 10.144-096; 10.152-435; 10.162-119; 10.172-514' 10 192-207' 10 197-
208; 10.203-439; 10.303-283. 284, 285, 286, 288, 289, 290, 443; 10.306-291' 10 308-442'- 10 310-
446; 10.324-473; 10.340-495; 10.344-484; 10.349-301; 10.350-491; 10.372-505 506' 10 373-497-
10.376-507, 508; 10.377-510; 10.380-516, 519, 520; 10.401-447; 10.402-448' 10 404-449- 10 405-
323, 452, 455, 456; 10.406-461; 10.410-462; 10.413-326, 465, 467- 10 414-476- 10 419-479- 10 420
481; 10.423-487: 10.428-486; 10.429-488; 10.450-335; 10.480-499- 10 486-503•' 10 487-513- 10 490
356; 14.063-370, 371, 372, 373; 14.157-374, 375, 386; 14.298-272- 14 321-358- 14 327-376 377-
14.393-493, 494; 14.469-366, 367. ' '
Plaits of 17 strands in \2/2\\l pattern (twenty-three fragments of matting)- 10 001-399-
10.012-392, 394; 10.041-093,409; 10.046-036; 10.064-033,034; 10.069-090 410- 10 078-413 415-
10.082-088, 418, 419; 10.098-423; 10.099-102; 10.172-521; 10.404-450; 10 405-453' 454- 10 406-
460; 10.413-464; Plaits of 5 strands in M/1W1 pattern (21 fragments of carrier baskets)-
10.001-396, 10.015-400. 10.041-092. 10.045-100/101, 10.046-039/040, 10 049-005 10 054-003
,0.07.-007, ,0.078-205/414, .0.082-089, ,0.373-498, ,4.063-379/382, 1 4 . 1 3 3 - 3 « i , 3 w S S
plait, m a d e of five strands has edges in S Z orientation, as does the third plait with
seven strands. Plaits with five and seven strands are suitable, but plaits with four
and six strands are not.
Thus it follows from the 'Qasr Ibrim Plaiting Rule' that plaits m a d e in a
U / 1 W 1 pattern are suitable to m a k e into a fabric w h e n plaited with an odd number
of strands, but plaited strips m a d e of an even number of strands are not.
Apart from the M / 1 W 1 pattern, a twill pattern is found: a pattern in which the
shift is smaller than the number of strands that are passed. The twill pattern used
at Qasr Ibrim for making the strips is Y2/2W1 (under two, over two, with a shift of
one). The number of strands used for twill plaits is governed by the same rule: the
edges have to be oriented in an opposite direction (SZ).
Not all strips plaited with an odd number of strands are suitable. Figure 11-3
shows that a plait in this pattern m a d e with seven strands is cannot be sewn into
an ongoing fabric: both edges are oriented in the same direction (SS). A nine-
strands plait, on the other hand, has its edges oriented in different directions (SZ)
and is suitable for using in the sewn plaits technique.
Figure 11-3 Plaited strips of which the edges are orientated perpendicular to each other
(SZ) can be used for making fabrics, while strips with parallel edges (SS)
cannot. From left to right: plait with 4. 5 and 7 strands in a \1/1\\1 pattern,
plait with 7 and 9 strands in a \2/2\\l pattern.
J f e
s S i H l S i
t $ &
I i ! ! !
B f l f l
W § >
W m
l l S l
^^^•4 r^V^vl*/'
*§^j$)jtfj&\ ^^y^^^^c^l\
$ $ ^
^ ^ P I P ^
yrh
(H)
W
Figure 11-4 Left: a seven-strands plait with pattern \2/l\l/2\\l Right: an eleven-strands
plait with pattern \2/2\l/l\2/2\\l. The combination of a \2/2\ plait pattern at
the sides and a \1/1\ pattern in the middle, makes the plait suitable for
sewing into an ongoing fabric (SZ oriented edges).
F r o m the 'Qasr Ibrim Plaiting Rule' it follows that plaited strips m a d e in a \2/2\\l
pattern are m a d e with 5, 9, 13, 17, 21 or 2 5 (etc.) strands.7 M a k i n g a fabric of
twill plaited strips with 7, 11, 15, 19 or 2 3 strands, is not possible. Three
exceptions were found: one twill seven-strands plait and one twill 11-strands plait
T w o twillfive-strandplaits were found at Qasr Ibrim (10.099-538; 10.335-545), but

these are just plaited strips, and are not used in matting. Eleven matting fragments were
found composed of 13-strand plaits: 10.032-405; 10.098-422; 10.103-081; 10.121-108; 10.195-
438; 10.202-214; 10.219-441; 10.380-512, 518; 10.384-522; 10.410-463; Only two fragments of
matting from 21-strand plaits were found: 10.071-006 and 10.098-425. The latter is not
complete and might also be a fragment from a plait made with 25 strands.
with adapted patterns, resulting in an S Z edge (Figure 11-4).8 Also a twill 19-
strands plait was found, which had not been used in antiquity to make a fabric, but
as a stopper blocking the hole of a m u d grain silo.9
In general, baskets were made of five- or nine-strands plaits, mats were
made of the wider strips, plaited with 13 strands or more.
.,*?'.
H
"^^tflcV ft. "* '- •*
Plate 11-5 Sewn plaits basket from nine-strands plaits made of date palm leaf, sewn
with 3 m m wide strips of date palm leaf. The base and side have been cut,
since the basket was reused as pit lining. Ballana Period (third to sixth
century A D ) . (Courtesy of the Egypt Exploration Society).
Periodic variations
In all periods nine-strand plaits were used for making baskets, but for the different
periods found at Qasr Ibrim, differences can be reconstructed for the use of plait
patterns and the material of which the strips were m a d e . T h e sewing strip varied in
composition and material also.
Five-strands plaits did not occur until the Islamic period (twelfth to
nineteenth century A D ) . M o s t of the plaits m a d e of d o a m palm leaf are also found
in Islamic contexts. In the earlier periods date palm w a s the raw material used for
plaited basketry.
T h e variation in the sewing strand is that between string and unworked
strands. In the Meroitic, Ballana and Christian periods the sewing is done with
unspun palm leaf strips, n e w lengths of the strand added on during the sewing
process with overhand knots. It is not until the Islamic period that zS2 string, m a d e
of palm leaf, palm fibre or grass, is used to sew the plaits into a basket.
Plait with seven strands: 14.157-387; plait with eleven strands: 10.191-437.
9Plaited strip made with 19 strands: 10.414-328.
Although this variation seems to be one of preferences over time, the matter
is more complicated. The five-strands baskets did occur in other parts of Egypt
before the twelfth century A D . Indications for this are found in textual and
archaeological material. In a Greek papyrus, arranging the lease of an olive and
date tree orchard, two types of basketry are mentioned: evvedTtAoKov and
7ievcarc[A.]6Kov, which I believe stand for sewn plaits basketry m a d e with nine
and five-strands plaits.10 In a fifth century A D context sewn plaits basketry with
five and nine strands was found in Berenike. The five-strands baskets were made
of date palm leaf and sewn with grass string, while the nine-strands baskets were
plaited from d o a m palm leaf and sewn with unspun strands of the same material
(Wendrich 1995, 74-77).
Most of the sewn plaits fragments are part of carrier baskets. O n the whole,
thefive-strandsbaskets are made with wider plaiting strands than the nine-strands
baskets and have, in general, a coarser appearance (Plate 11-6).
Plate 11-6 Complete carrier basket made of five-strands plaits (doam palm leaf), sewn
with string. Islamic period (after 12th century A D ) . (Courtesy of the EES).
P. Vindob.Boswinkel 8, line 13, cf. Worp 1977, 97 and P. Soterichus 4 line 18

cf. Worp 1988, 280).
They probably were used for the hard work of carrying stones, sand and dung,
while the nine-strands baskets were perhaps more employed for carrying grain,
and other foodstuffs. O n e of the nine-strands baskets has long rope handles and an
oval shape, which is very similar to the pollination baskets made and used at
present in Middle Egypt (cf. p. 181).
The wider plaited strips, which were used for matting show a variation over
time too. Plaits made with 13 strands are found in Meroitic, Ballana and Christian
contexts (third century B C to twelfth century A D ) . Strips plaited with 17 or 21
strands do not occur before the sixth century A D . All matting fragments from the
Islamic period (twelfth to nineteenth century A D ) were made with 17 strands, most
often with d o a m palm leaf.
Although the part of Qasr Ibrim excavated in 1990 and 1992 had not been
occupied during the ninth to twelfth century A D , and only a few storage pits from
that period were found, most variation in plaiting techniques is found in the
Christian deposits.
New Nubia
S e w n plaits baskets in N e w Nubia are lavishly decorated with patterns of three or
four colours worked into the plait. All baskets are made from nine-strands plaits in
a \2/2\\l pattern. The sewing is done with a flat needle and thin strips of doam
palm leaf. W h e n a sewing strand runs out, a new length is knotted to the old
strand. Apart from the small baskets, made of a plait with a length of three ba'at,
larger decorated baskets are made, of six and 10 ba'at (Plate 11-7). These baskets
are indicated with the Arabic word guffa, but they are also called sellan bursh in
Kensi and sellan nibid ('mat-basket') in the Fadidja area. Shibr is a Kensi word for
the smaller sewn plaits baskets of 3 ba'at. The baskets have almost straight sides
and two small rope handles. They are used as shopping baskets and as storage
baskets inside the house, for instance for dates and peanuts.
Plate 11-7 Sewn plaits baskets, made of 9-strands plaits, sewn with strands of doam
palm leaf. The sizes are 5, 6 and 7 ba'at. Aswan 1990.
216
Plate 11-8 Sewn plaits 'bridal' mat, with decorative zigzag patterns. This mat is 14
strips wide (rather than the 'ideal' 16), sewn with strands of doam palm leaf.
N e w Ibrim 1992.
The basket making tradition differs per village. In the Kensi village of
Dabud, for instance, nobody makes sewn plaits baskets, but in the Kensi village of
Dihimit, m a n y w o m e n m a k e baskets, as well as mats. T h e village is well k n o w n
for its basket makers. In the Fadidja village of N e w Ibrim, the w o m e n m a k e sewn
plaits mats, but not sewn plaits baskets. The Fadidja w o m e n state that these are
made by the Kensi w o m e n .
The mats are m a d e by both Kensi and Fadidja w o m e n . Plaited strips are
m a d e of 17-strands of date palm leaf and sewn with knot-connected strands of
d o a m palm leaf. The plaited strips are decorated by incorporating one plaiting
strand that has been dyed either red or green. In a 17-strands plait, this one
coloured leaflet stands out as a zigzag pattern over the length of the strip (Plate
11-8, Figure 9-4a, p. 158). Traditionally, 16 strips are sewn parallel to each other.
The terms used for these mats are in Kensi bursh and in Fadidja Nubian nibid. The
decorated mats are m a d e in preparation for a wedding as present for the bride and
bridegroom and kept in the guest room of the house.
Basketry from Qasr Ibrim and N e w Nubia 217
Apart from the coloured Nubian baskets, the Nubian households also use
large undecorated sewn plaits baskets, which are also called guffa. It is the Upper
Egyptian equivalent of the Middle Egyptian maqtaf, and is made by Egyptian men.
Coiling (5)
Fabric.
One passive system.
(Figure 9-5)
Qasr Ibrim
With a few exceptions, all coiled baskets in Qasr Ibrim are decorated. These
decorations were m a d e during the production of tne basic structure, either with
colours or by using decorative wrapping patterns. S o m e variation occurs in the
Qasr Ibrim coiled basketry in the interaction of the active and passive system.
O f a total of 134 coiled objects and fragments, 99 were m a d e in the 'stitch
through bundle' technique." T h e palm leaf strand is wrapped around a bundle of
grass and stitched through the previous coil (the anchoring fabric). This is the w a y
that all coiled basketry in A m a r n a w a s m a d e .
In 17 objects, the stitch does not pass through the bundle, but only through
the previous r o w of wrapping strands.12 T h e effect is a surface which is m u c h
m o r e smooth than that of thefirstgroup, in which the stitch is passed through both
previous winder and bundle (Plate 4-4 on p. 64). T h e stitch-through-winder
technique is used in combination with coloured winders, forming a decorative
Stitch-through-bundle coiling: 10.001-152; 10.012-183, 184; 10.043-186; 10.045-140;

10 061-046, 047. 048. 158; 10.064-044, 045: 10.066-199. 201. 202; 10.072-061. 160: 10.078-203;
10.082-056. 057. 058. 059, 060; 10.083-146. 147. 148; 10.088-166.167; 10.090-163, 164; 10.095-
150; 10.098-043. 154; 10.110-159; 10.121-064, 065, 066. 143, 144, 145, 157, 161, 162. 165, 181;
10 122-197; 10.123-049; 10.128-155. 156; 10.129-190; 10.133-069. 185; 10.139-142; 10.144-168;
10.154-151. 211; 10 166-440; 10.169-177, 180; 10.180-182; 10.195-174; 10.203-213; 10 216-173,
280, 281, 475; 10.247-062, 063, 217; 10.303-287, 444; 10.317-293; 10.323-472; 10.325-294; 10.335-
296. 483; 10.340-496; 10.341-492; 10.344-185; 10.345-298, 299; 10.347-300; 10.358-304; 10.380-
311. 312, 313, 314; 10.401-321; 10.404-451; 10.405-457,459; 10.412-474; 10.413-469; 10.414-478;
10.419-480; 10.420-482; 10.429-489; 10.485-502; 10 486-354; 14.493-369.
l2Stitch-through-winder coiling: 10.053-051. 052; 10.061-050; 10 067-054; 10.172-278,
279; 10.075-032; 10.083-042; 10.216-282; 10.325-295; 10.352-303; 10.360-305; 10413-470; 10.419-
330; 10.480-344; 10.485-346: 14.493-368.
pattern. Another feature of these fragments is that the bundle is not round in
diameter, but slightly flattened. A n exceptional basket is a widely spaced stitch-
through-winder basket from a late Ballana context (Figure 11-9).13
not to scald
Figure 11-9 'Stitch-through-winder': coiling technique in which the stitch just picks up
the wrapping strand of the previous coil, rather than running through both
the wrapping strand and the bundle. This small basket is a rare example of
widely spaced winders, which leave the bundle visible.
In five fragments stitch-through-bundle coiling is combined with wrapping

around t w o bundles.14 T h e coils near the centre and the rim are fastened with the
stitch-through-bundle, but the fabric of the sides of the baskets is m a d e by
wrapping the p a l m leaf strand successively round o n e and t w o bundles, without the
use of stitches. This results in a closely w r a p p e d , decorative fabric (cf. Figure 5-
13 o n p. 9 8 ) . In the five baskets the a m o u n t of stitch-through-bundle coiling varies
between only one coil just around the centre, to three coils near the centre and nine
coils near the rim. Three of these baskets w e r e found in well sealed early deposits,
w h i c h w e r e dated as early Ballana or Late Meroitic (first to third century A D ) .
T w o baskets w e r e found in disturbed building areas of the Islamic period, in which
remains of demolishing of Ballana buildings w e r e found.
1310.075-032.
Combined stitch-through-bundle and successive wrapping around 1 and 2 coils:
10.061-053; 10.082-055; 10 162-153; 10.172-187; 10.377-310.
In eight fragments the stitch-through-bundle technique is combined with

decorations with wrapped patterns.15 The stitch-through-bundle technique is used
near the centre (cf. Plate 4-4 on p. 65), near the rim and sometimes in between
two bands of the wrapping pattern.
The pattern is m a d e in a second phase, over a fabric which is a fast version
of the stitch-through-bundle technique: the palm leaf strip is w o u n d a number of
times around the bundle, and then fastened to the previous coil, or over two coils,
with one stitch (Figure 11-10). The fabric made in this 'lazy basket makers' stitch'
is not very strong, but is only used to form the basis for the decorative wrapping,
which at the same time strengthens the fabric. The number of coils over which the
wrapping is done varies. The schematic drawing in Figure 11-10 shows a pattern
wrapped over nine coils, the basket of Plate 4-4 (p. 65) has a pattern over five
coils.
With the exception of two small fragments, which were found in a context
of the Christian period, these baskets were all found in Ballana and Meroitic
contexts.
Figure 11-10 The 'lazy basket makers' stitch' covered by decorative wrapping.
Combination of stitch-through-bundle and decorative wrapping: 10.098-216; 10.162-

141; 10.350-302; 10 372-306; 10.405-322, 324; 10.416-509; 10.485-353.
In four baskets a combination of three of the techniques mentioned above was

used: the stitch-through-bundle technique near the centre and rim, the successive
wrapping around one and two bundles for the sides, with one or two bands of lazy
basket makers' stitch, covered with wrapping as decoration.16 All four were found
in early Ballana contexts.
The only coiled object in which there is a variation in the space between the
coils, is a the basket shown in Figure 11-11.17 Comparing this basket with the one
in Figure 11-9 illustrates the different effects of using space in the active or the
passive system.
Figure 11-11 Open spaced coiling (the passive system is widely spaced, while the active
system is closely spaced).
^Combination of three coiling methods: 10.075-041; 10.121-067,068; 10.485-347.

Open-spaced coiling: 10.473-343.
In figure 11-9 the passive system is closely spaced, the active system is widely
spaced. This has no consequences for the coherency, shape or strength of the
basket, but only for the appearance. If the passive system is widely spaced,
however, the entire structure is influenced. In the basket of Figure 11-11 both the
bundle and the winders are made with d o a m palm leaf. The basket was found in a
Ballana period context.
Size of bundle and winders

The diameter of the bundle is in all fragments smaller than the width of the
wrapping strand (Table 11-12). There were 42 fragments (31 % ) of very fine
basketry, with both systems smaller than 5 m m . Most of the baskets were made of
bundles with a diameter between 5 and 10 m m and a winder smaller than 5 m m
(57 % ) , which is stillfine,compared to the coiled objects from other regions (cf.
Table 10-5 on p. 184). This painstaking work of stitching narrow strips of doam
palm leaf around tiny bundles, adds to the delicacy of the lavishly decorated Qasr
Ibrim coiled basketry.
^^v^ Passive system:

Active ^ ^ ^ ^ diameter of
width of winders ^ \ ^
0-5 m m 42 77 5 4
5-10 m m - 5 1 -
10-15 m m - - - -
15-20 m m - - - -
Table 11-5 Ratio between the size of the active and passive systems of the 134 coiled
baskets found at Qasr Ibrim.
Material, shape and function

The bundle material of the baskets from Qasr Ibrim are mainly made of grass
and the hard fibres of the fruit stem of the date palm In some cases straw is used.
For the winders d o a m palm leaf is used mostly, although in some cases wheat
straw was used.
M a n y of the coiled baskets from Qasr Ibrim are complete or nearly
complete. There are not many coiled containers. Most of the coiled basketry is
flat, or slightly curved, which seems to indicate that it was their function to cover,
rather than contain things.
New Nubia
T h e supposition that the coiled baskets from Qasr Ibrim are largely used as covers,
is partly based o n the function of baskets in N e w Nubia. Flat or slightly conically
shaped baskets with a diameter of 75 c m or m o r e , are used to cover aluminium
trays o n which the food is brought in. S o m e of the baskets have a ring handle to
m a k e it easier to lift the basket. In the old days the food w a s not brought in o n an
aluminium tray, but o n a large flat basket W h e n not in use, the baskets are
hanging on the wall as an attractive, colourful decoration.
Smaller, completely flat, baskets with a diameter of 30-40 c m are used to
cover water jars, pots and pans. T h e smallest size are 10 c m in diameter and used
to cover tea glasses. T h e latter are not used often, they are rarely taken of the
wall. Apart from the covers, there are also deep round baskets with a foot, which
is at the same time used as head ring to carry the basket. These are covered with a
flat basket. T h e terms with which these baskets are indicated differ greatly
between villages. A n overview is given in Appendix D .
jr.Jb
Plate 11-12 Recently produced coiled basket and lid, made of doam palm leaf. N e w
Ibrim 1992.
T h e materials of which the coiled baskets were m a d e before the settlement,

are mostly the same as what is k n o w n from Qasr Ibrim. T h e bundle material is the
fibre of the fruit stem of the date palm. T h e wrapping strand is mostly d o a m palm,
although in Kensi villages sometimes wheat straw w a s used. After the
resettlement, d o a m palm leaf is difficult to get and alternative materials are sought
In present day N e w Nubia, the production of coiled basketry is a highly
regarded craft. T h e most important aspect of the baskets is their decoration, which
should be colourful and nicely patterned. M o n o c h r o m e wrapped patterns, such as
found in Qasr Ibrim, do not occur. A specific form of colourful decoration is the
technique of imbrication (see below). In N e w Ibrim I found only one completely

undecorated basket, which was used to take hot bread out of the oven.
The coiled basketry production in Nubia differs from place to place. Four
villages will be considered here: the Fadidja villages of N e w Ibrim and Ballana;
and the Kensi villages of Dabud and Dakka.
Plate 11-13 Small coiled lids, made of coloured wheat straw. They were made in old
Dabud, before the resettlement. N e w Dabud 1992.
In order to make the decorated baskets, the w o m e n of N e w Ibrim dye the

doam palm leaf strips and use the coloured strips in combination with white leaves.
The w o m e n have a preference for d o a m palm leaf, but if they cannot get the
material, they will use date palm leaf, which is thought to be less strong and less
colourful. The w o m e n of Ballana, also a Fadidja village, make baskets mostly
from date palm leaf, although they also would prefer to use d o a m palm leaf.
The lack of d o a m palm leaf is equally felt in the Kensi villages, but the
solution found there is quite different. The Kensi w o m e n do not cling to tradition,
but have found alternative materials. In Dabud the w o m e n make baskets of
brightly coloured cotton yarn. In the village of Dihimit most w o m e n are involved
in the production of sewn plaits baskets, the neighbouring Kensi village of Dakka
has a tradition of coiled basket makers. Since their arrival in N e w Nubia the
w o m e n have experimented with different colourful materials.
Plate 11-14 Innovations in the choice of colourful materials to replace doam palm leaf,
which is in scarce supply. Top: washing line (twenty years old). Bottom:
chocolate wafer wrappers and synthetic ribbon. Dakka 1992.
The results are baskets made of plastic washing line (twenty years old) and baskets
of which the bundle is wrapped with chocolate wafer wrappers, stitched in place
with widely spaced winders of synthetic ribbon (three years old). Details are
shown in Plate 11-14.
Wrapping (6)
Linear.
One passive system.
(Figure 9-6)
Qasr Ibrim
Seventeen brushes were found at Qasr Ibrim which were made in the same manner
as the brushes from Amarna: a bundle of material folded double and wrapped with
string to form a handle (Plate 11-15).18 Of fifteen brushes the brush part was
divided into a number of fingers, ranging from three to seven ."
Other wrapped objects were the handles of baskets, which in three cases
were wrapped with string,20 a stick, wrapped with palm leaf,21 and two jar-
stoppers.22 The jar stoppers were made of a bundle of date palm fibre, wrapped
with date palm fibre string.
Brushes with wrapped handles: 10.001-277; 10.083-149; 10.110-129; 10.121-134

10.122-195; 10.128-215; 10.133-135; 10.195-222; 10.207-133; 10.377-309; 10.469-342
14.065-124 (twined brush with wrapped handle); 14.184-216; 14.207-223; 14.218-125
14.298-138; 14.356-136.
l9Brush with three fingers: 10.083-149, 10.377-309, 14.065-124, 14.218-125; four
fingers: 10.001-277, 10.469-342, 14.356-136; five fingers: 10.122-195, 10.128-215,
10.195-222, 10.207-133. 14.148-126, 14.207-223, 14.298-138; sixfingers:10.133-135;
seven fingers: 10.110-129; brush without fingers: 10.121-134.
^Wrapped handles: 10.099-218; 10.405-458; 10.414-477.
S-wrapped stick from Ballana period context: 10.075-130;
"Jar-stoppers: 10.121-137; 10.128-270.
- i m m
Plate 11-15 Seven brushes found at Qasr Ibrim (top row: no's. 125, 126 and 129; bottom
row: no's. 133, 134, 135 and 136). (Courtesy of the Egypt Exploration
Society).
A wreath, made of a band of d o a m palm leaf, had once been covered with
foliage, similar to vine leaves, and wrapped with small strips of d o a m palm leaf to
keep the leaves in place.23 Wrapped rings and pads, such as were found quite
regularly at Amarna did not occur.
N e w Nubia
In present day N e w Nubia brushes have been replaced largely by plastic brooms.
Brushes that are made in the traditional w a y are not produced by Nubian m e n or
w o m e n , but bought from the Upper Egyptian fellahin, w h o produce them, just like
their colleagues in Middle Egypt.
23Wreath: 10.384-317.
11 Basketry from Qasr Ibrim and New Nubia 227
11.2.2 With one passive and one active system, orientation in two directions
Weaving with one strand (8)
Fabric. |
One passive system.
1
One active system (consisting Of(jne 1
member moving in one plane) I
Orientated in two directions. 1
(F gure 9-8) |
Qasr Ibrim
Thirty w o v e n matting fragments have been found. In all mats the passive and the
active system are of the same material: all mats are closely w o v e n with S-twisted
d o a m palm leaf. It is, therefore, impossible to decide which of the systems is
active, which passive, unless the edges of the mat are present. This does not
present a problem in describing the technique, as long as the weaving patterns are
symmetrical in both directions.
The interaction of the systems, the weaving pattern, is expressed with a
formula similar to that used for plait patterns. T w o weaving patterns were found:
one fragment, from a post Ballana (Islamic?) context, is w o v e n in a \2/2\\l
pattern.24 The other 29 fragments, most of them quite large, represent a kind of
matting which has had a long tradition at Qasr Ibrim, and occurred during all
periods from Meroitic to Islamic.25
The weaving pattern is basically V4/4W1 (Figure 11-16). In most of these 29
matting pieces the orientation of the weaving pattern changes regularly, to decorate
and strengthen the mat. The variation in pattern changes is almost the same as the
number of fragments found. T w o examples are shown in Figures 11-16 and 11-18.
The matting was re-used as pit lining, but the loops which are still attached to
s o m e of the fragments, indicate that its original function w a s that of furniture
matting (Plate 11-17). Apart from these rounded forms, several square palm leaf
loops were found with the matting, indicating that those particular pieces have
been w o v e n around a frame of square beams of 45 x 46 m m .
Woven matting in \2/2\\l pattern: 10.032-010.

" Weaving pattern \4/4\\l: 10.032-407, 408; 10.048-028; 10.103-084; 10.105^426, 428,
10 124-430: 10.133-433; 10.247-104; 10.413-468; 10.427-333; 10.428-334; 10.434-490; 10.450-336.
337, 339: 10.480-500; 10 484-501; 14.298-271 (six large pieces); 14.327-359. 360, 361. 362;
14.469-365;
Figure 11-16 Furniture matting made of S-twisted doam palm leaf in a V4/4U1 pattern.
Note the changes in direction of the weaving (zig-zag pattern).
•j2;£*5*»5 •> '_•> < 11 ~ r • • \'~~- "•'.' -i -r«T*2Lr'i=5iH'

•- us.' -,v-',vM*ui,,"!s^^n'-rii1
*r.-v.'-vi' -.-=L';v,/V/?=i?!LT---.»-'
*^r-;.->r.--r----i it- 'i;il-~ ''--m:!'!T- ' ' V ; ^ ' i --î; ; J';r-_-'^v;/i! -,;;. Hjnr
Plate 11-17 Furniture matting made of S-twisted doam palm leaf. The curled ends
indicate that the matting has been cut loose from a furniture frame.
Figure 11-18 Furniture matting, with a diamond shaped weaving pattern

N e w Nubia
At present no bed matting of this kind is found in N e w Nubia. T h e only w o v e n
matting used in N e w Nubia are floor mats, m a d e of a cotton string warp and a
weft consisting of either rushes or plastic strands. These mats are m a d e by
Egyptians.
Weaving with two strands (9)
Fabric. ~b
One passive system.
members moving in one plane ).
(Figure 9-9) 1
Qasr Ibrim
T h e second weaving technique found at Qasr Ibrim is m a d e with double strands.
Except for a formal difference there is also a difference in production method,
which will be elaborated on in Section 15.9. This weaving technique is used for
bed matting and is m a d e with grass zS2 string. Three fragments have been found:
two of these are w o v e n in a \3/3\\l pattern (Plate 11-19).26 T h e third fragment is
m a d e in a \2/2\\l pattern.27 All three are post Ballana period and probably from
the Islamic period.
^^'•'i(:ffe,-'iv;^%ti!#^
~-J3£T.V \ W > ' ' ;'i: •rl,.7?"(
w-z/ii V H Ul>-' *•" i'F"\\

• " j Z M Z . i i t ;a».'VJ '\ ••' s <•, i . >•••••• ,
Plate 11-19 Bed matting woven with two parallel grass strings. (Courtesy of the EES).
" Grass string matting: 10.032-001; 10.045-098; The \3/3\\l pattern with 2 members
consists of a double strand going under 6 (three double strands), over 6 (three double
strands), with a shift of 2 (one double strand).
" Grass string matting: 10.032-406; the V2/2W1 pattern with 2 members consists of a
double strand going under 4 (two double strands), over 4 (two double strands), with a shift
of 2 (one double strand).
N e w Nubia
Stringing beds with grass rope in a \3/3\\l pattern is still done today. Even the zig-
zag pattern is the same. The work is not done by Nubians, however, but by
Egyptians.
Twining (10)
Fabric. |
One passive system.
(Figure 9-10) |
Qasr Ibrim
At Qasr Ibrim 20 fragments of twined basketry were found. Table 11-20 shows
that there are nine twining variations. The groups H, I and L, are known from
Amarna (see Table 10-13 on p. 196). There are six new groups, defined by the
criteria listed in Table 10-12 on p. 194. The order in which the groups O, P, Q,
R, S and T are presented in the table is determined by their properties.
The fine twined basketry (groups O, P and Q ) are all made of unworked
strips of palm leaf, or plant culms. Group O consists of three objects, with a fine
closely spaced passive system and a closely twined active system in S-direction
(Figure 11-21).28 T w o of them can be dated to the Ballana period, the third is from
a later context (after the third century A D ) . For all three the twining strands are
made of rushes (Juncus acutus). One of the twined pieces is the base of a basket
which is combined with continuously plaited sides.
Group P are objects in which the active strands are twined around double
stakes.29 The two fragments found are made of doam palm leaf and can be dated to
the Ballana period. T w o other pieces, also dated to the Ballana period, were
twined irregularly, sometimes around single, other times around double stakes.30
Group Q is formed by two fragments from the same Ballana period context.31
The passive stakes consist of strips of doam palm leaf, the twining is done in Z-
direction with date palm leaf around double stakes. The fragment is closely
twined, very similar to group O, with exception of the orientation of twining.
O: 10.085-078; 10.122-194; 10.144-071.

P: 10.128-079; 10.463-341.
30O or P (irregular): 10.205-325; 10.427-332.
Q: 10.154-212.
size space preparation rows Or.
pas. act. pas. act. pas. act. pas. act.
0 fine fine close close unw. unw. single single S
p fine fine close close unw. unw. double single S
Q fine fine close close unw. unw. double double Z

H coarse coarse close close worked worked single single S !
R coarse coarse close close worked worked single single SZ

I coarse coarse close close worked worked double single s
S coarse coarse close close unw. unw. single single s
L coarse coarse close open unw. worked single single s
T coarse coarse close open unw. worked double single s
Table 11-20 Nine variations of twined fabrics found at Qasr Ibrim, six of which do not
occur at Amarna (cf Table 10-13 on p. 196). The nine groups are ordered by
the criteria in Table 10-12 (p. 194) and are drawn in Figure 11-21.
The coarse twined basketry and matting at Qasr Ibrim can be divided into six
groups, of which five are represented by only one object.32 The groups H , R and I
consist of closely twined string matting, either orientated in S-direction (Group H
and group I, the latter around double stakes), or in alternately S and Z direction
(Group R ) . T h e effect of the latter is that the fabric seems to be built from V-
shaped stitches (Plate 11-22). The object is a donkey saddle, m a d e of date palm
fibre string. The four rows of V-shapes which stand out clearly in the fabric are
m a d e by two rows of waling, one in S and one in Z direction (see below).
Group S is formed by two fragments of regular closely spaced twining with
unworked strands of doam-palm. Both fragments c o m e from Ballana period
context. Group T consists of a brush, m a d e of bundles of finely shredded d o a m
palm leaf. A mat is formed of loose fibres, fastened at the bottom with one row of
twining. T h e small mat is then rolled up and wrapped with string to form a brush.
This is the same method as k n o w n from A m a r n a (cf. Figure 10-16, p. 200). The
A m a r n a brush was not m a d e of d o a m palm leaf, however, but of grass.
32
H: 10.247-109; R: 10.340-297; I: 10.316-292; S: 10.427-332; T: 10.065-124.
• - j j ^ M r ^ ^ u ^
| '^£'^G . |^^\^1f-^<^x^ ^ ^ <
^j^T' ^ £ {iQl**î<<**%—^ £-a£si? =^^••^7^^ ^•^J^.rj^^fcj-ff^î
^ ^ u ^ j r ^ f ^ ^ ^ X s A ^*\-*,<i^^»'<t^\*^fî g VjC OH' g^*<.'F-U'll
Tv\rE**zf^\^0^^_X~ ' j j ^ ^ * ^
^ r ^ S 2 ^ 2 * ^ c i « X^&k. pITV . jfew~ l^^^f-~T—rT^Y^-_^
îr^c«^?tc^r^^ ^m^^T^ J^l^^^^^^^pj^T^^JL^-_ ^ J^^~*^^~j^&^vQ
f^S-ff^p^^p-f "V^T-^ X S v H i' R "^^TB^>tLT*1
^jc^r^^j&z' ,y -\f,j-
/> a s j g Q ^ ^ "
o * M & &p- jXpwvT] r>»^~^
i\^rU&%&\£ •\^^^sw^^^ I '•^.tJ • • | TIr!X-,L*^
/ 3nw 'j^^\f Jiiff 4 . x* ^*^ j^rjf ^ t ^ / f
"fxpfrvjsi^vj JKrk*^^TL ^ J T p ^ S j r ^ ^ J l ^ ^ i ^ J T^^^-*^^r^^*^^fl8^^^^^^^
J*7^2r^^\5i^^^ ^ ^ ^ 5 -îtf' _ ^ j * -r^Kfcg^^ Jr^ 5r _^f*S^ Z
^ 4 C ^ t jt Jfflmyjr -yÂ |\*. jSfiy ~~^^^. ~~l^>fc.~ -•jfaf^Z'-' - Z L g <
± ^ ^ ^ i ^ y ^ / '^^^r ""* - ^ I f i i ^ l ^ i ^ ^ S ^ 0 ^ j
^^wiZ&dwy ^ & y 'y^'^y ^/tK
t^k^^rl
h ' " M r s ff-MTTtH ' w w u
• , ' i 'I ' —*H' i' ,/l," ' '
[ i n i \ 1 .\\p /
: y^^T^
r L f x j > -^ ^ T "
^ W ^ T y ^ A t r ? ' f*S®V'\^Lrf%rV^
y^^'^^p^0^' >r j* Z jfiy Z jdt. sf-^G%, / ^tT r" «i i;,,',..
"1 i
V.JUI J J j ^ / T o W
f f w l
rl<< \ i M ;
1 •1 1
s f n f r f i l l ' T - ^ l 1 - * ^ i
Figure 11-21 The nine twining variations as listed in Table 11-20 (cf. Figure 10-14 on
p. 197).
The twining technique which occurs most frequently at Qasr Ibrim is the
coarse grass matting, (group L ) already k n o w n from Amarna and there also by far
the most frequently encountered variation.33 Bundles of grass are held in place by
widely spaced twined strings. This type of matting has been found in all periods of
occupation.
33L: 10.032-009, 189; 10.060-209; 10.064-035; 10.401-320; 14.327-388, 389.

i ^ ^ ^ l ^ M m i l
Plate 11-22 Donkey saddle, made of twined date palm leaf string. The V-shapes are the
result of alternate S and Z twining. The striking lines are made by S and Z
waling. (Courtesy of the Egypt Exploration Society).
$!&*% .
v.
l £
Plate 11-23 One of the large grain storage pits, lined with basketry. At the bottom coarse
grass matting, with widely spaced rows of twining, to the right are the
contours of an earlier (and much shallower) pit, lined with cut up fragments
of sewn plaits basketry (Courtesy of the Egypt Exploration Society).
A s almost all matting in Qasr Ibrim, the coarse grass matting was found re-used as
pit lining (Plate 11-23). The area of excavation has been used extensively for grain
storage from approximately the seventh to the nineteenth century A D . For this
purpose large pits were excavated and lined with matting. T h e extremely arid
circumstances in Qasr Ibrim apparently m a d e this possible.
N e w Nubia
In N e w Nubia there does not seem to be any production of twined baskets or mats.
Large closely twined bags (groups H and R ) are used in agriculture, but these are
purchased from Egyptian mat makers.
Waling (11)
Fabric.
One passive system. >
One active system (consisting of three
(Figure 9-11) |
Waling is very similar to twining, but instead of two strands, waling is done with
three strands at a time. Each of these is passed under one passive strand and over
two passive strands, at the same time turning around each other. T h e direction of
this turn can be S or Z. A combination of a r o w of S and Z-waling has the
appearance of long V-shaped stitches. Although waling could be used as the single
technique for making up baskets or mats, it is always used in combination with
other techniques such as weaving and twining.
Qasr Ibrim
In the Qasr Ibrim material waling w a s found twice and both objects have been
dated to the Ballana period.34 In the donkey saddle of Plate 11-22 four decorative
bands of SZ-waling have been inserted. T h e other application of waling is in the
transition between base and sides of the plaited basket depicted in Figure 4-2 on
p. 61. In this basket the sides are plaited with groups offiveparallel rushes. They
have been attached to the basket by pushing them into the base (alongside the
stakes radiating from the centre). A row of S-waling is used to force the rushes in
the proper position to start plaiting the sides.
New Nubia
In present day N e w Nubia no waling is used in the production of mats or baskets.
11.2.3 Two-systems techniques with two active systems in one direction
Rope making (13)
Linear.
T w o active systems (moving in two
planes).
Orientated in one directions.
(Figure 9-13)
3-4SZ-waling: 10.340-297; S-waling: 10.144-071.

Qasr Ibrim
In Qasr Ibrim, as in Amarna, 2-ply rope is m a d e in one process: two active strands
are spun and plied in one action. T h e materials used are grass, date palm leaf,
d o a m palm leaf, but the most frequently used material is the leaf sheath fibre of
the date palm.
New Nubia
A s in Middle Egypt, in N e w Nubia rope production is done casually as part of
other activities. Whenever a length of string is needed, someone makes it on the
spot. Most rope is m a d e of date palm fibre. S o m e date palm leaf is used also,
especially the veins of the leaflets. D o a m palm leaf rope does not occur very often.
Grass rope is used specifically for bed webbing.
11.2.4 Two-systems techniques with two active systems in two directions
Continuous plaiting (14)
Fabric.
(Figure 9-14)
Qasr Ibrim
Continuous plaiting is found 26 times. T w o fragments of a fine continuous plaited
mat m a d e from strands of date palm leaf, were found in a Christian context.35 T h e
plait pattern is \3/3\\l in the middle and \2/2\\l near the edge (Plate 11-24).
Although one of the fragments is m a d e with slightly narrower strands than the
other and plaited in a very regular fashion, the two fragments seem to be part of
the same mat. The basket depicted in Figure 4-2 on p. 61 has continuously plaited
sides, m a d e with five parallel culms of rushes. The plait pattern is \1/1\\1.36
The second group is formed by cylindrical bags, m a d e of continuously
plaited date palm leaflets. There are minor variations in the width of the strands.
The space between the strands varies also, but this is due to distortion of the
object. The plait pattern is always M / 1 W 1 (Figure 11-25). The complete length of
the leaflets is used in the bags and clever use is m a d e of the point of attachment of
the leaflets to the midrib.
Continuous plaited matting: 10.373-307, 308.

10.144-071
>\\
Plate 11-24 Fragment of continuously plaited matting with fine strands of date palm leaf.
Date palm leaflets are the small side-leaves of the feather shaped date palm
leaf. For the bags the long leaflets, which are approximately 500 m m long, are
split along the middle vein into two strands of 20 m m . wide, still connected at the
end. The start of the bag is formed by these connected ends. At the bottom edge
the tips of the leaflets are worked back into the plait. The bag is closed by running
eight n e w leaflets through the holes in the fabric and knotting these at the bottom
of the bag.
In total 24 continuously plaited bags have been found, 23 m a d e of date palm
leaflets and one m a d e from strands of d o a m palm leaf. W h e n the large fan shaped
leaves of the d o a m palm are cut into strips, the strands are not connected at one
end, the w a y date palm leaflets are. Instead, two strands of d o a m palm leaf were
joined with half knots, in imitation of the date palm strands. In total the bag was
m a d e with 13 double strands (26-plait).37
The length of the date palm bags is determined by the length of the palm leaf
strips, the diameter by the number of strands. Table 11-26 shows the variation in
plaiting strands. At the start the leaves are inserted from two sides.
37D o a m palm: 10.026-113.

[ 1 Basketry from Qasr Ibrim and N e w Nubia 239
Figure 11-25 Schematized drawing of a continuously plaited 'disposable' bag, made with
six double date palm leaflets. The hollow plait is fastened off at the bottom.
Eight new strands are inserted with which the bag is closed. This bag is 170
m m long and 60 m m in diameter.
The start of the strands is mostly divided symmetrically, but in seven bags the
number of strands inserted on each side is not the same (e.g. 6+5).38 Seventeen of
the plaited bags were from Ballana period contexts, seven bags, including the one
made of d o a m palm, were from Christian or Islamic contexts.
A s for the function of these baskets: they were all found empty, with
the bags opened at the top, where the plaiting started. The pointed attachment of
the date palm leaf is in all cases ripped open, while the knot which closes the bag
383 + 3: 10.486-355; 4 + 4 : 10.152-115; 10.162-116 5 + 5: 10.048-021; 10.090-123;

10.144-171 6 + 5: 10.384-316 6 + 6: 10.162-118; 14.469-364; 6 + 7: 10.389-319;
10.485-350,352; 6 + 8: 10.162-117; 10.485-349; 7+7:10.048-022; 7 + 8:10.485-351;
8 + 8: 10.485-348: 9 + 9: 10.048-023, 24, 25; fragments (number of strands unknown):
10.098-114; 10.380-511. 517.
after it has been filled, is left intact. This provokes the presumption that these bags
were disposables, for one time use only. Perhaps they were transport bags,
designed to contain a specific quantity of a specific good. O n e of the bags was
found in a cellar with a large quantity of sesame seeds around it. Although the
space between the plaiting strands is not as large as is suggested in the schematized
drawing of Figure 11-25, sesame seeds do not seem to be the most likely
candidate, because they are so small they could perhaps fall from the bag.
double strands
inserted
from the left
9 36 str.
double plait
strands ///
8 28 str. 30 str. 32 str.
double plait plait plait
strands // / /
7 26 str. 28 str.
double plait plait
strands /// /
6 22 str. 24 str.
double plait plait
strands / //
5 20 str.
double plait
strands ///
4 16 str.
double plait
strands //
3 12 str.
double plait double
strands / strands
inserted
3 4 5 6 7 8 9 from the
n = 20 double double double double double double double right
str. str. str. str. str. str. str.
Table 11-26 'Disposable' bags: variation in plaiting strands inserted from the left and the
right. T h e check marks indicate the number of times the combination of
strands inserted from the left and the right occur. T h e total number of
disposable bags found is 20.
New Nubia
Continuous plaiting is used in Nubia to make fans, which are often lined with
textiles. The workmen of the excavation said that in present day N e w Nubia the
disposable bags are still made, but this could not be verified. In Middle Egypt a
basket maker produced one of these bags, a small one made with only 4 palm leaf
strips, as a kind of practical joke: if you put the bag on someone's finger it cannot
be removed again. Attempts to pull it off, make the plaited fabric only tighter.
This kind of plaiting is used in medicine as finger bandages (a 'Chinese finger').
Plaited strips (15)
Linear.
T w o active systems (moving in one
plane)
(Figure 9-15)
Qasr Ibrim
Plaited strips are used mainly as passive element in sewn basketry. The 'Qasr
Ibrim plaiting rule' which explains the absence of plaits with a specific number of
strands in relation to the plait pattern, have been explained on pp. 209-213.
There were some plaited strips that were not made for basketry or mat
production: two strips made with three and four strands in a M/1W1 pattern, two
five-strands plaits and a 19-strands plait made in a \2/2\\l pattern.39
Sandals are made as plaited strips. In total eight sandals have been found,
ranging from a simple sole, fastened with straps, to a shoe construction with an
inner sole, outer sole and plaited sides (Plate 11-27). All sandals are flip-flops,
with a strand passing between thefirstand second toe. From the four sandals with
single soles one was made with six strands of doam palm leaf, two were made of
eight strands (see Figure 11-28) and of one fragment the number of strands could
not be determined, because it was too fragmentary.40 The plait pattern is \1/1\\1.
T w o of the more complex sandals are complete, the inner sole is made in both
cases of 13 strands of doam palm leaf, plaited in a \2/2\\l pattern.
39Three-strands plait: 10.162-436; four-strands plait: 10.078-416;five-strandsplaits:

10.099-538; 10.335-545; 19-strands plait: 10.414-328.
40Six-strands plait: 10.095-132; 8-strands plait: 10.128-080, 10.048-192; number of
strands unknown: 10.065-193.
242
Plate 11-27 Three of the plaited sandals found at Qasr Ibrim (123, 080, 120). The two to
the left are made of doam palm leaf, the one on the right has an inner sole,
outer sole and sides which can be fitted around the foot. (Courtesy of the
Egypt Exploration Society).
Figure 11-28 Schematized drawing of a plaited sandal made as an 8-strands plaited strip.
This sole was attached to the foot with a strand between thefirstand second
toes, leading to the side of the shoe, very much like our present day plastic
flip flops.
[ 1 Basketry from Qasr Ibrim and N e w Nubia 243
In one sandal the outer sole was plaited with five bundles of date palm fibre in a
M / 1 W 1 pattern.41 The outer sole of the other sandal was twined.42 The sides, which
passed over the toe and could be pulled tight around the foot with a piece of string,
were m a d e respectively of plaited strips of nine and 17 date palm leafs in a \2/2\\l
pattern. Remains of two more of such sandals were found. Only the outer sole was
left, with just traces of the inner sole and the sides. In both cases the outer sole
was m a d e with seven bundles of date palm fibre.43 N o specific temporal
differences were spotted in this small fraction of sandals.
New Nubia
In present day N e w Nubia palm leaf sandals have been completely replaced by
plastic flip-flops, which are cheap, and shoes, which give status.
Knotting (16)
Fabric. |
T w o active systems (moving in three I
planes) 1
Orientated in two directions. I
(Figure 9-16) |
Qasr Ibrim
At Qasr Ibrim reef knots and granny knots were used to close the plaited palm leaf
bags. Reef knots were also used to make a palm leaf wreath (Figure 11-29).44 This
curious object was found in a late Christian or early Islamic context. A s with the
plaited palm leaf bags, in this wreath the natural properties of the date palm leaf
were used. The thick midrib of the feather-shaped leaf was split in the middle. The
two halves were formed into two circles, which was made into a wreath by
connecting the leaflets with reef knots.
New Nubia
In present day N e w Nubia no knotted fabrics are made.
Complete sandal with plaited outer sole and 9-strand plaited sides: 10.490-356.
"Complete sandal with twined outer sole, 13-strands plait inner sole (\2/2\\l) and 17-
strands plaited sides (Y2/2U1): 10.316-292.
Outer soles: 10.120-131, 10.457-340.
44Wreath: 10.080-014.
^- * -^ ~ ~-~~^—~—^_
.^-^ ""."-i-—5"\ \V ) ) i^P n>-^—7^3uh)?
./ s- \<<c ^S^v VS^ (/7 2fi!-v\
C/X iffloifcY
X \M\ J/J}
^? m W y/cH //^>x i^C \ t \ / a
M ^ s ^ i S k 5>^—^~^L _- - ~^^zjr/k{ i n
J h^%/£^Lx^:—-^^l ———7^^/^$(mm
/ A.1JI.VYJ5 «vyj
I f f
7\\ #Tl rW/ Mtf W'l KiM <^f'J W /< Z ' w W i
tk*s—~J^) \ n. ^//c\ XJ/Tn^. '^^'V" x y^Z\. <^s*y ^5^ /
Figure 11-29 Schematized drawing of a palm leaf wreath, knotted from a date palm mid
rib which has been split in half and connected by knotting the palm leaflets
of the two halves.
New Nubia
No knotted fabrics were encountered.
11.3 Three-systems Techniques
11.3.1 With two passive and one active systems in one direction
Coiled inlay (17)

Fabric.
One active system (moving in two planes).
(Figure 9-17)
Qasr Ibrim
The rims of small coiled plates, which have an average diameter of 10 cm, are
lined with two coloured strips which run parallel to the bundle and are fastened
with the active wrapping strand (Figure 11-30). In a basket found in the tomb of
Tut ankh A m o n this technique was found as decorative ridge in the side of a finely
coiled basket, but then with three strands on the outside of basket no. 589a
(Figure 11-31).
Figure 11-30 Construction of the coiled inlay rims in several of the small plates found at
Qasr Ibrim (Ballana period).
Figure 11-31 Schematic drawing of coiled inlay in a basket dated to the New Kingdom
period, from a basket in the tomb of Tut ankh Amon.
N e w Nubia
A decorative technique, known as imbrication is used in Toshka el-Ghalalaab and
other villages in the vicinity of Aswan: this is basically a coiling technique of the
'stitch-through-bundle' variety, with a decorative strand fastened on top of the
coiled fabric (cf. Figure 9-17, p. 166 and Plate 4-6, p. 65). The active system
wraps around the bundle and keeps it in place with stitches through the previous
bundle. With the same stitches, the third system is fastened on top of the fabric,
parallel to the bundle. This has mainly decorative value and does not contribute to
the coherency of the fabric. Coiled inlay does not occur in N e w Nubia, but this
technique is used as decoration in some of the baskets made in the Fayoum.
246
11.3.2 With two passive and one active systems in two directions
Binding (19)
Fabric.
One active system (moving in three
planes).
(Figure 9-
Plate 11-32 Roof construction over a cellar Th»

underneath the date pa|m branche w i t h ? "J" T o ™ V'S'ble' they are tied
Tne detail foto is C ^ J ^ J * , * " - N - leaf.
the Egypt Exploration Society). the °Verv,ew (Courtesy of
Qasr Ibrim
A roof, covering a cellar, was made of closely spaced date palm branches, on
which a number of midribs were placed at perpendicularly.45 These were tied with
S-twisted doam palm, following the same direction as the widely spaced passive
system (the cross ribs). The cellar and roof are from the Ballana period.
New Nubia
This roofing technique can still be seen at many places in roofed streets of the old
towns of the Dakhla Oasis. In N e w Nubia the roofs are now made of date palm
midribs tied with metal wire instead of doam palm leaf, but there is a clear opinion
that 'proper' roofs should be made with doam palm binding.
11.3.3 With one passive and two active systems in two directions
Plaiting around a core (20)

Linear.
One passive system.
Two active systems (moving in one plane).
(Figure 9-20)
Qasr Ibrim
Only three small fragments were found of this 3-system technique. One is made
with eight cotton yarns, plaited in a \2/2\\l pattern around a core of flax yarns.46
One consists of a core of date palm fibre around which a plait is made with four
doam palm leaf strands in a M/1W1 pattern.47 The core in the second fragment is
formed by zS2 string around which a plait is made with six strands.48 Both the
string and strands were made of doam palm leaf. The plait pattern is \ 1/3M//1. All
plaits are from an Ballana period context.
Plaiting around a core makes a very strong kind of rope, which can be used
to pull heavy loads.
45Roofing: 10.450-523; remains of roofing: 10.068-072,073,074.

Plait with 8 strands around a core: 10.335-545.
Plait with 4 strands around a core 10.405-551.
Plait with 6 strands around a core 10.458-557.
New Nubia
This technique is not produced in N e w Nubia. It is interesting to note, however,
that in the same area there are villages of settled Ababda bedouin, w h o of old
maintain relations with the Nubian population. Plaiting leather thongs around a
goat hair core is a technique which is known and used extensively by the Ababda
bedouin w h o are living in the southern part of the Egyptian Eastern Desert.
CHAPTER TWELVE
STATUS OF THE STATIC
12.1 Static Statistics

It is time to consider what these hundreds of basket and matting fragments,
abstracted into basic structures, recorded, identified and classified, have to tell us
about basket production in Egypt. A n interpretation of tradition and development
focuses on the properties of the basketry, which are determined by the relation
between technology and raw materials. This is done by comparing quantitatively
and qualitatively basketry from places that are removed in either time, space or
both.
In Table 9-21 on p. 169 the 20 techniques of our basketry corpus are listed
with an indication of which technique is found where. In A m a r n a 11 of the 20
techniques were found, in Qasr Ibrim 16 out of 20, in Middle Egypt 11 and in
N e w Nubia eight. Apart from the presence or absence of certain techniques, there
is also a clear quantitative difference: Table 12-1 and Figure 12-2 show h o w often
these 20 techniques occur at A m a r n a and Qasr Ibrim. In Figure 12-2 the 20
techniques have been plotted on the horizontal axis. The vertical axis indicates the
percentage in which they occur at Qasr Ibrim and Amarna.
In the table and the bar chart rope making (no. 13) is mentioned pro memori.
It is one of the 20 techniques, often used as active or passive system in basket
making, but rope and string is found in such quantities, that incorporating it would
obscure the other techniques in the table and the chart. T w o techniques have not
been found in the archaeological material, but are produced nowadays in Middle
Egypt (pierced basketry, no. 12 and looping, no. 18). All sewn plaits baskets and
mats (technique no. 4) are made of plaited strips (no. 14). Rather than counting
those fragments twice, no. 14 only lists those strips which are not used in sewn
plaits basketry.
Comparing A m a r n a and Qasr Ibrim shows remarkable differences in the
range and ratio of techniques. S e w n plaits (no. 4), by far the most frequently
occurring technique at Qasr Ibrim (40%) does not occur at all at Amarna. A s
indicated on pp.203-204 plaiting techniques were probably not found in Egypt
before the Third Intermediate Period. Since there are few publications about
basketry of the region, it is difficult to determine w h e n and h o w plaiting w a s
introduced into Egypt.
In A m a r n a twining (no. 10) is the most frequently occurring technique
( 3 8 % ) , wrapping (no. 6) the second (22%) and coiling (no. 5) the third (16%). At
Qasr Ibrim the percentage of coiling is higher than at A m a r n a (30%), while

twining and wrapping do not occur that frequent (respectively 4 % and 5 % ) .
This chapter concentrates on mats and containers. The wrapping technique,
which is used to produce a diverse functional array of objects (brushes, pot stands,
head rings, jar stoppers, wreaths and parts of furniture) is considered briefly in the
concluding section, where an interpretation of tradition and development, is
focusing on the continuity and change of Egyptian basketry.
Amarna Qasr Ibrim
1 knotless netting 6 3% 0 0
2 grommet 11 6% 3 <1%
3 knotted netting 2 1% 1 <1%
4 sewn plaits 0 0 179 40%
5 coiling 29 16% 134 30%
6 wrapping 39 22% 24 5%
7 knotting 1 <1% 0 0
8 weaving, one strand 21 12% 30 7%
9 weaving, two strands 1 <1% 3 <1%
10 twining 69 38% 20 4%
11 waling 0 0 2 <1%
12 pierced 0 0 0 0
13 rope pro memori 0 pro memori 0
14 continuous plaiting 0 0 26 6%
15 sewn plaits 0 0 9 2%
16 knotting 1 <1% 1 <1%
17 coiled inlay 0 0 12 2%
18 looping 0 0 0 0
19 binding 0 0 1 <1%
20 plait around a core 0 0 3 <1%
TOTAL 181 100% 448 100%
Table 12-1 Occurrence of the 20 techniques in Amarna and Qasr Ibrim (see also Figure
12-2). The horizontal lines are the borders of the eight classes (cf.
Table 9-0, p. 154). Rope making is included pro memori. Pierced basketry
(12) and looping (18) are produced in the area at present, but not found
among the archaeological material.
12 Status of the Static 251
40
35
30
25
20
15
10
jm. i * h , 1 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Qasr Ibrim Amarna
Figure 12-2 Bar chart of the proportional occurrence of 20 basketry techniques in Qasr
Ibrim and Amarna (for the legend of the 20 techniques and the absolute
quantity of baskets see Table 12-1).
12.2 The Mat, the Basket and the Bag
12.2.1 Sleeping and Sitting
Amarna Middle Egypt
weaving (8) weaving (8)

twining (10) twining (10)
Qasr Ibrim New Nubia
sewn plaits (4) sewn plaits (4)

twining (10)
Amarna
Mats which were probably used to sit and sleep on are woven out of single grass
stems on a warp of grass string (cf. Plate 10-10, p. 191). They have a m a x i m u m
width of 1.20 m , because the length of the individual grass stems is the limit of the
width of the mat. There are two small fragments of woven matting with other
materials: bundles of respectively papyrus and straw on a grass string warp.
Floor mats are thinner and made of grass bundles which are held together
with widely spaced rows of twining. Here the length of the grass determines the
m a x i m u m length of the mat The largest fragment found is 1 m long and 86 c m
wide. In some cases the grass bundles are fastened with string made of d o a m palm
leaf.
Qasr Ibrim
N o woven matting is found in the two excavation areas in Qasr Ibrim. Instead, the
sewn plaits technique is used to make mats to sit on. Most of these matting
fragments, plaited with doam palm leaf and sewn with doam palm string, are
found in the late (Islamic) layers. From earlier (Christian) deposits comes date
palm leaf matting, sewn with unspun strips of doam palm leaf.
Twined matting does occur, however, and it is very similar to the Amarna
mats: bundles of grass with widely spaced rows of twining.
Middle Egypt
At present, no grass matting is being woven in the immediate vicinity of Amarna,
but in the Upper Egyptian village of Nagada I found a mat maker w h o produced
grass matting. The difference with the Amarna matting is that the weaving is done
with small bundles of grass, which is a faster method than weaving individual
grass stems. Individual culms of rushes are also used for mat-weaving, but they
are not produced locally either (I studied a mat maker in the Dakhla Oasis).
Another import, which is rapidly replacing both grass and rush mats, is colourful
matting woven with thin hollow plastic tubes on a cotton warp.
There is no production of twined matting similar to the grass-bundle floor
mats of Amarna and Qasr Ibrim (L in Figure 10-14). A tightly twined mat is used
as doormat (H in Figure 10-14). This is the same fabric as is used for making
dung-bags (see below).
New Nubia
In N e w Ibrim only sewn plaits mats are produced. They are plaited from date palm
leaf and sewn with unspun strips of doam palm leaf. Apart from these, the same
colourful mats which are popular in Middle Egypt are used in N e w Nubia as well.
Similarities and Differences

Although the woven matting of Amarna is made of the same grass as the matting at
present, the difference is that nowadays bundles of grass are used instead of
individual grass stems. This is not a recent innovation, however, since weaving
with grass bundles has a long traditions in Egypt. In thefifthcentury A D deposits

in the Red Sea harbour of Berenike, such mats woven with bundles of grass have
been found in abundance (Wendrich 1998, 253-258). Four small fragments found
at A m a r n a show that weaving with bundles (of papyrus and straw) did occur in
1350 B C also.
Since no grass woven matting was found in Qasr Ibrim (which is
contemporary with Berenike), the question is if weaving with bundles is an Upper
Egyptian tradition. The two matting fragments found at Amarna and made of
twisted bundles of straw or sedges (Plate 12-3) would than have to be interpreted
as imports, while the weaving with individual grass stems might have been
produced at Amarna. A large quantity of woven matting was found in the area of
West Street 1/2 (cf. Section 7.2.3) and could have been produced there.
Ironically, one could argue that the most direct heritage of the Amarna grass
matting are the brightly patterned plastic mats, woven with one straw at a time.
Cm
Plate 12-3 Side edge of a mat woven from bundles of Cyperus papyrus culm. Amarna
1350 BC. (Courtesy of the Egypt Exploration Society).
The coarse twined matting, made of bundles of grass held together with
widely spaced rows of twining, occurs both at Amarna and Qasr Ibrim. At present
this type of basketry is not produced. More will be said on the closely twined
matting in the section on bags.
Sewn plaits matting has been found mainly in the late (Islamic) layers of
Qasr Ibrim. This matting type does not occur in Middle or Upper Egypt. There is
a development in sewn plaits basketry, however, which will be elaborated on in
the section on the bags.
12.2.2 Storing and Covering
Amarna Middle Egypt
coiling (5) coiling (5)
'•'•ititi; Qasr Ibrim N e w Nubia

Iggf coiling (5) coiling (5)
Amarna
Storage at A m a r n a was probably done mainly in coiled baskets, which had the
function of linen chest, cupboard, food container and safekeeping of trinkets. Each
function demands its particular sizes and shapes. If tomb inventories are at all
representative for a household (in this case: palace-) situation, a good impression
can be gained by looking at the basketry from the tomb of Tutankhamon (Reeves
1990, 205-206).
The fragmentary finds from A m a r n a comprise only the bases of coiled
baskets. The preserved diameters range from 100 to 300 m m . The publication of
Gourlay gives a good impression of the variety of basketry shapes from a
workmen's village (and related tombs) during the N e w Kingdom Period (Gourlay
1981).
With one exception, all the coiled basketry from A m a r n a was m a d e from
d o a m palm leaf, wound around a grass bundle. The one exception is a basket with
a winder of papyrus.
Qasr Ibrim
Most of the coiled basketry at Qasr Ibrim is flat and its main function is to cover,
rather than to contain. Small covers, with a diameter of 100 m m are m a d e of
finely coiled d o a m palm leaf winders around a bundle of grass or fibres from the
fruit stem of the date palm. Larger coiled baskets with a diameter of 300 - 400 m m
were probably also meant as food covers.
Only a few coiled containers were found, one of which is depicted on p. 218
(Figure 11-9). These small baskets were also very regularly made and were
probably meant for safekeeping of small and perhaps valuable objects.
Storage in Qasr Ibrim was clearly not done in coiled baskets. Probably, sewn
plaits baskets were sometimes used for this, but more importantly, there were
alternative storage methods: containers of unbaked clay (soma) and simple pits in
the ground, lined with matting (cf. p. 235 Plate 11-23).
Middle Egypt
At present large flat coiled baskets are used for storing bread (Plate 10-6, p. 185),
while high, straight walled coiled baskets are used for storing other things, such as
flour. For storing dates and maize the farmers from Middle Egypt also use soma's,
which are usually built on the roof. In Upper Egypt large soma's are built in front
of the house (Ikram, forthcoming).
New Nubia
Coiled basketry in N e w Nubia is mostly designed to cover things. There are small
plates (100 m m in diameter, cf. Plate 11-13, p. 223), used to cover cups, medium
size covers (300-400 m m ) and large covers, which are convex in shape and have a
diameter from 750 m m up to 1 metre.
There are some shapes, which are clearly meant to contain goods (mostly
food, such as bread), for instance the large basket (diameter of 700 m m ) shown in
Plate 11-12 (p. 222). The Nubian-made sewn plaits basketry is used on a short
term basis to hold goods, but long term storage was until recently done in soma's.
Since the households do no longer need to store large yields, food is put in the
refrigerator and cupboards instead.

In Amarna only doam palm leaf is used for the winders. Date palm leaf does not
occur at all, although the date palm was widespread in Egypt in the N e w Kingdom
period. H o w to explain this complete lack of date palm leaf,fibreor midribs in the
Amarna basketry? There are two possible explanations of the use of d o a m palm for
basketry: doam palm was preferred, or date palm trees were not available. There
are arguments for both interpretations.
The argument that doam palm leaf was considered better than date palm leaf
is supported by comparing Amarna with Qasr Ibrim. The coiled basketry in Qasr
Ibrim was also made of doam palm leaves, even though date palm trees were
available, judging from the large quantities of plaited basketry made with date
palm leaf and the presence of other date palm products on site. In N e w Nubia,
modern basket makers clearly favour doam palm leaf for making coiled basketry.
They maintain that it is stronger.
The argument that date palm was not available at Amarna is based on the
archaeological evidence: apart from many date stones, no other parts of the date
palm tree were found in the workmen's village at Amarna. Not only is d o a m palm
used, rather than date palm leaf, but reeds were used rather than date palm ribs
and grass was used rather than date palm fibre. This is remarkable, because the
stem and the mid ribs of the date palm are useful as roofing material and the fibre
can be used for making rope.
It seems that dates were either transported from a considerable distance, or
that the inhabitants of the workmen's village did not have access to date palm
trees. Probably both arguments are valid: date palm was not available, d o a m palm
leaf was both available and considered a better material.
At present the basket makers from Middle Egypt have no choice, because
d o a m palm trees do not occur north of the Luxor region, while there is an
abundance of date palm trees in the valley.
The bundle material used in A m a r n a is consistently grass. In Qasr Ibrim the
bundle is often made of the fibres from the fruit stem of the date palm tree, which
is also the bundle material used at present both in Middle Egypt and N e w Nubia.
^"^•^^ Passive system:

Active^"^^^ diameter of
width of winders ^~~v-^^
Amarna 4% - 12% -
0-5 m m . 57%
31% 4% 3%
Qasr Ibrim
Amarna - 32% 40% 4%
5-10 m m •"•" ;•,•: 4%
- 1% -
Qasr Ibrim
Amarna - - 8% -
10-15 m m ' ' .
- - - :
Qasr Ibrim
Table 12-4 Relation between size of active and passive element, in percentages
(Amarna n=25; Qasr Ibrim n=134)
In the previous chapters the relation between the size of the bundles and
winders of coiled basketry from Amarna and Qasr Ibrim were listed (Tables 10-5,
p. 184 11-5, p. 221). Table 12-4 shows the percentages of the different sizes of
coiled basketry at both sites. The most frequent combination at A m a r n a is a bundle
size between 10 and 15 m m and a winder of between 5 and 10 m m (40%). In 8 8 %
of the fragments from Qasr Ibrim the diameter of the bundle is smaller than 10
m m and the winder narrower than 5 m m . In A m a r n a only 4 % of the baskets is this
fine. Since the number of coiled fragments represented by this percentage is only
one, a fragment of the centre where the coil and winders start out small, this
comparison is perhaps misleading. Therefore, in addition to the percentages, the
distribution of the relation between the size of the active and passive systems have
been plotted in Figure 12-5.
active system
width of strands
O O
o •
o o o
O 9
o o c
o o
••••• • •
J •• • •
• • • 0 • ••• ••
• ••• ••• •••••
10 m m passive system
diameter of bundle
Amarna: n = 2S O = 1
Qasr Ibrim: n= 134 • =1 •=2 » =3 • >4 =S
Figure 12-5 Distribution of the relation between the size of the active system (the
wrapping strand) and the passive system (the bundle).
In Figure 12-5 the horizontal axis represents the diameter of the bundle,
ranging from 0 - 17 m m , the vertical axis indicates the width of the wrapping
strand. The 25 coiled fragments from Amarna are indicated by circles, the 134
coiled fragments from Qasr Ibrim by black dots.
From Figure 12-5 a several things can be read. N o basket has been made
with a wrapping strand larger in width than the diameter of the bundle. In only
seven objects the active and passive element have the same size. Furthermore, the
Qasr Ibrim basketry is clustered at the left, while the size of the basketry from
A m a r n a shows more variation, but is all situated at the right side. The bundles and
winders of the Qasr Ibrim coiled basketry are considerably smaller than those at
Amarna and more consistently so.
Coiled basketry is produced widely in Middle Egypt and in N e w Nubia. The
difference in size of bundle and winder is detectable nowadays as clearly as it is in
the archaeological material. The basketry from Middle Egypt is m u c h coarser than
the Nubian basketry. Although the finds from A m a r n a are at least 950 years older
than those from Qasr Ibrim, the difference in bundle size and width of winder does
not appear to be periodic but regional.
Another clear difference is that the A m a r n a coiled basketry is plain, while
the Qasr Ibrim coiled basketry is decorated. The difference is even stronger when
comparing the plain Middle Egyptian basketry and the lavishly decorated N e w
Nubian coiled baskets (see below Section 12.4)
12.2.3 Carrying and Transporting
Amarna Middle Egypt
knotless netting (1) sewn plaits (4)

twining (10) twining (10)
knotting (16) pierced (12)
Qasr Ibrim N e w Nubia

* m m
Amarna
In A m a r n a the carrying and transportation is done with flexible, twined bags.
There are several kinds of twined fabrics, making bags with different properties.
The best known twined bag is the open twined 'seed'-bag, which has been found at
Amarna. The bag is made of grass rope, with widely spaced rows of twining (fine
version: A or B in Figure 10-14, p. 197, coarse version: K ) . A coarse dense
fabric is made with closely twined string (H and I in Figure 10-14). This makes a
very strong bag, suitable for heavy loads.
Apart from the twined bags, there are indications that the knotless netting
(technique number 1 in Table 9.0, cf. Figure 9-1, p. 155) is also a transportation
net. In the tomb of K h a such a net has been foundfilledwith d o a m nuts (now on
exhibition in the Turin M u s e u m ) . It seems a valid suggestion that these nets were
made at the time and place of harvest to transport the fruits to their destiny. In
Amarna the net was re-used to strengthen the roof.
The carrier net, knotted from papyrus string is the third technique employed
in A m a r n a for making containers or aids for transportation (Figures 10-19 and 10-
20, p. 205).
Qasr Ibrim
Twined basketry is not used for transportation or carrying in Qasr Ibrim. Instead,
sewn plaits baskets are used. They make strong, flexible baskets which can be
carried on the head, but have handles too. The larger baskets can be hung on
donkey or camel back by balancing two of them on the ends of a stick across the
back of the pack animal.
Middle Egypt
In middle Egypt the sewn plaits baskets are used for carrying on the head (women)
or by the handles (men). They are used for shopping. Heavy loads are not
transported in sewn plaits baskets, however, but in large twined bags. The fabric
of these bags is very similar to that found at Amarna (like H in Figure 10-14, but
twined in S Z direction). For light but bulky loads, such as straw, large nets are
used, which are made in the pierced technique (Figure 10-18, p. 202).
New Nubia
Carrying of shopping in N e w Nubia is done in colourful sewn plaits baskets (Plate
11-7, p. 215). They are also used to display dry commodities, such as dates,
peanuts and beans in shops or on the market. For agricultural tasks basketry is
used which is not produced by the Nubians, but by Egyptian farmers. These are
larger, plain sewn plaits baskets and twined bags and carrier nets.

In Amarna most carrier and transportation baskets and bags were twined. In Qasr
Ibrim the twined basketry has been taken over completely by sewn plaits basketry.
In Middle Egypt there is one twined fabric, which is still produced and used as
heavy duty bag for the transportation of earth and dung. Most of the carrying and
transportation is done in sewn plaits baskets, however. The same is true for N e w
Nubia: the only carrier basket which is produced today is the sewn plaits basket.
This is made mainly for the lighter loads as a shopping basket, while the heavy
duty sewn plaits baskets are bought of the fellahin, the upper Egyptian farmers.
The development of the sewn plaits basket gives an interesting angle to

approach local traditions. Although it is unknown h o w or w h e n sewn plaits
basketry was introduced in Egypt, there seem to be two different traditions, which
can be discerned mainly by the sewing strand. The Egyptian tradition uses a string,
made of date palm fibre or date palm leaf, to sew the baskets. The Nubian
tradition uses unspun strips of palm leaf, which are connected to the next one with
an overhand knot.
In Qasr Ibrim of the Ballana and Christian period sewn plaits basketry is
made mainly of nine-strands plaits made of date palm leaf, sewn with unspun palm
leaf (Table 12-6, cf. Plate 11-5, p. 213). In the Islamic period baskets are made
of nine-, seven- orfive-strandsplaits made of doam palm leaf, sewn with string
(cf. Plate 11-6, p. 214).
In present day N e w Nubia, baskets and mats are made of date palm leaf
plaits and sewn with unspun doam palm leaf, which seems to connect with the pre-
Islamic period finds of Qasr Ibrim.
Amarna Middle Egypt, at present
(no sewn plaits basketry) nine-strands plait, made of date palm leaf
sewn with string
Qasr Ibrim, Ballana and Christian Period: New Nubia, at present
nine-strands plait, made of date palm leaf nine-strands plait, made of date palm leaf
sewn with unspun palm leaf sewn with unspun palm leaf
Qasr Ibrim, Islamic period: Upper Egypt, at present
five-strands plait, nine-, five- or seven-strands plaits,

made of date or doam palm leaf made of date palm leaf
sewn with string sewn with string
Table 12-6 Schematic representation of the variation in the passive system (plaits) and
active system (sewing strand) of sewn plaits basketry.
Looking at the plait patterns does not really help. In present day Upper Egypt
the variety of plaited strips present parallels for the Qasr Ibrim baskets found in
the later (Islamic) layers of the site (Table 12-6). The seven strands plait with a
\2/l\l/2\\l pattern (cf. Figure 11-4, p. 212), found in Qasr Ibrim, is called
saWawi in Upper Egypt (Henein 1988, 182-183), the Arabic term for the nine-
and thefive-strandsplaits are respectively tesc awl ("rimer") and kamsawi..
O n the other hand, the nine- and thefive-strandsplait were known in the
Graeco-Roman period. The Greek terms are evvedTcAoKov, and 7revTa7r[A]6icov
(cf. p. 213). In fifth century A D deposits excavated at the Red Sea harbour town
of Berenike two combinations were found: baskets made offive-strandsplaits (date
palm), sewn with string and nine-strands plaits (doam palm), sewn with unspun
knotted (doam palm) strands.
W h e n comparing the range of techniques and materials it should be taken into
account, that the depositional and post-depositional processes at Qasr Ibrim have
been quite complicated. The areas 10.000 and 14.000 have been occupied during
the Meroitic and Ballana periods. The houses collapsed in the Christian period,
when the area was only partly occupied and mainly used for digging storage pits.
In the early Islamic period there was a continued use of the area for storage and in
the latest phase houses were built on top of the Ballana period remains.
The good preservational circumstances at Qasr Ibrim make it likely that
whenever a pit was dug, a lot of earlier material was unearthed in such a good
condition that it could be re-used. Therefore, the finds from the later periods are
probably mingled with earlier material, the earliest layers being the only ones
which can give a more or less 'unpolluted' image of the basketry production at that
time. Furthermore, the use of basketry from previous periods might have
influenced the production of n e w basketry.
If w e want to define an 'Egyptian' and a 'Nubian' sewn plaits tradition, then
w e should not in thefirstplace look at the plait pattern, nor the species of palm
tree, but mainly to the sewing strip (cf. Wendrich 1995, 82-84).
12.2.4 Sleeping and Sitting 2
Amarna Middle Egypt
wrapping (6)
weaving 2 str. (9) pierced (12)
weaving 1 str. (8) weaving 2 str. (9)

weaving 2 str. (9)
Amarna
Furniture is hardly found at Amarna, but still there are two basketry techniques
which are of importance in the A m a r n a furniture production. The first are light
tables and stools made of reeds wrapped with palm leaf or papyrus (cf. p. 188),
the second is the weaving of the furniture webbing. In A m a r n a only impressions of
furniture webbing with two parallel strands were found. F r o m the impressions it
seems likely that the original matting has been made of S-twisted d o a m palm leaf.
Qasr Ibrim
Furniture webbing in Qasr Ibrim is made by weaving with single strands of S-
twisted d o a m palm leaf, or with double strands of grass string (pp. 227-229). None
of the bed or chair matting has been found in connection to actual furniture. All
matting was re-used as pit-lining.
Middle Egypt
N o bed-matting is made nowadays in the vicinity of el-Hagg Qandil, el-Till or el-
Amariyya, but in the latter village furniture is made of the midribs of the date
palm leaves.
New Nubia
A technique identical to the grass string matting with double strands found at Qasr
Ibrim, is produced in the region of N e w Nubia. The bed matting is not made by
the Nubians, though, but by an Upper Egyptian craftsman.
The 'niche' of the light reed furniture at Amarna has been taken over by the jeriid
(Arabic for date palm branch) furniture in Middle Egypt.
Plate 12-6 Beds with rope webbing (angereeb) for sale on the market for small cattle in
Daraw (New Nubia 1992).
Although the webbing in both Amarna (probably) and Qasr Ibrim (certainly)
is made of S-twisted d o a m palm leaf, the weaving technique is different. F r o m the
impression of the A m a r n a matting it is impossible to say h o w the webbing was
made exactly, since no trace of the edges is available.
For the Qasr Ibrim grass string matting with double strands, w e have a
modern parallel (Plate 12-7). Although the Qasr Ibrim fragments do not have
edges either, it is impossible to tell if the production process was the same as what
is done today, but the appearance of the basic structure is strikingly similar.
12.3 Decoration
It has been mentioned above a number of times: the basketry from Amarna and
Middle Egypt is plain, while the basketry from Qasr Ibrim and N e w Nubia is
decorated.
Qasr Ibrim
Almost every fragment of coiled basketry of Qasr Ibrim has some form of
decoration, either with coloured winders or with wrapped patterns. The Meroitic
and Ballana periods, are distinguished by coiling variations, that are unique for this
period. Wrapping around alternately one and two bundles, the combination of lazy-
basket-makers'-stitch with decorative wrapping and rims with a separate inlayed
coloured strip (cf. pp. 217-221) do not occur anywhere else in Egypt and
disappear in the later periods of occupation at Qasr Ibrim. S o m e of these
techniques, such as the coiled inlay, are known today from East Africa (Eritrea,
Ethiopia).
Small coiled plates, decorated with coloured patterns, on the other hand,
have a long tradition from the Meroitic and Ballana period to the present (cf. Plate
4-4, p. 64 and 11-13 p. 223).
The plaited basketry of Qasr Ibrim does not have this lush colouring,
however. S o m e fragments are decorated with dark brown goat hair string, but only
one mat was made of alternate plaits in two different colours. There was no
decorative pattern within the plaits.
New Nubia
The coiled basketry from N e w Nubia is made of brightly coloured strips of d o a m
palm leaf. The production of undecorated basketry is inconceivable. If coloured
doam palm leaf is not available, alternatives are searched for and found, even if
this means a complete change of the decorative concept: Plate 11-14, p. 224 shows
a coiled basket with coloured winders made of washing line and one made of
candy wrappers. In the first basket the decorative pattern is determined by the
closely spaced winders. In the second, it is the bundle which is coloured, while the
winders are widely spaced in order not to hide the coloured pattern of the bundle.
The plaited baskets in New Nubia are also decorated. The plaited strips are
made with dyed date palm leaflets, plaited into brightly coloured patterns (Plate
11-7, p. 215).
Interpretation of decoration
H o w should these differences in decoration be interpreted? Does decorated
basketry point at a richer society, in which more time can be spent to make the
objects of daily use? Or is it typical for a poorer, or more isolated, society in
which prestige goods had to be made with locally available materials. A third
explanation is a difference in tradition, between Egypt and Nubia.
W h e n asked what the patterns on the coiled baskets mean, the Nubian w o m e n
shrug and without exception say "It is just decoration". Does this mean that the
decoration on the Nubian baskets has no meaning? Perhaps the pattern does not
have a specific meaning, but the fact that the baskets are decorated has. Although
the habit is disappearing at present, 40 years ago in Old Nubia the decorated
basketry plates were made to give as wedding presents and used to decorate the
main room of the house.
But can w e transfer the habits of the present to the Meroitic and Ballana
periods? Does the similarity between the small coiled plates also gives us the
grounds to suggest, for instance, that the Meroitic and Ballana coiled plates were
given as wedding presents? N o , not like that, but what w e can say is that in those
periods basketry was made with care and attention, the decoration pointing at an
other-than-purely-functional use. W e can also conclude that the role of basketry in
ancient and modern Nubia is very different from that in either Amarna or Middle
Egypt. The Egyptian baskets seem to be mostly functional, while the Nubian
basketry has a meaning in the determination of social and regional identity, as well
as social exchange.
12.4 Regionality, Tradition and Change over Time

In the previous section it became clear that there is a regional difference in
basketry. There is not only a strong difference between Egypt and Nubia, but also
between the fellahin (farmers') tradition of Middle Egypt and that of Upper Egypt
(cf. Henein 1988).
The subject of tradition does not really belong in a chapter on the static
aspects of basketry, because tradition itself is not static. There is a constant
development and a reinvention of tradition. Tradition signifies identity, from the
tepid "that's the way w e are doing things here", to the pride of the producer for his
or her product.
Amarna and Middle Egypt

At Amarna, where there were no sewn plaits baskets, basketry for
transportation and agricultural use, such as seed bags, were made by twining grass
or palm leaf intoflexiblebag like containers. The twining technique is n o w used in
Middle Egypt to make large flexible bags in which earth or dung is transported,
but transportation is done in sewn plaits baskets.
Baskets used inside the house were m a d e in the coiled technique and today
still are. The use has changed, however, to a very limited number of functions,
and mainly two shapes: a large flat bread basket and a round basket with high sides
for flour and other goods. This suggests that plaiting might have taken over
functions of twined basketry, while coiled basketry seems to have maintained its
position in the centuries between 1350 and the present.
The only objects that did not occur in the archaeological material are open
worked rigid containers, or crates. They are at present m a d e of date palm midribs,
a material which does not occur at A m a r n a at all.
The variety of basketry has diminished. Table 9-21 on p. 169 shows that
there are eleven techniques both in A m a r n a and Middle Egypt, but the variation
within those techniques (for instance the twining technique) has diminished. At the
same time in present Egypt m a n y basketry objects are being replaced with
alternatives: plastic takes over the role of brushes, cages, baskets and mats.
Qasr Ibrim and New Nubia

The Meroitic and Ballana Period were distinguished by Reisner on the basis of a
variation in pottery. F r o m the basketry point of view there is no difference
between the Meroitic and the Ballana period at Qasr Ibrim. The range of
techniques and the use of materials are the same. Perhaps Reisner's distinction was
based on a local situation, while in reality Meroitic and Ballana was a continuous
occupation of most of the area south of Aswan. Changes occur in the Christian
period, w h e n the variation in coiling gets less and variation in plaited fabrics is the
largest.
The ethno-archaeological survey learns that from the basketry perspective,
there is not one N e w Nubia, but at least three. This does not c o m e as a surprise,
considering the three languages spoken in the area, but it makes us aware that Qasr
Ibrim, the only site in the Old Nubian region that survived and the only site from
which a large corpus of basketry is available, can not be considered representative
for the entire region south of Aswan.
Comparing Qasr Ibrim with N e w Nubia shows a diminishing diversity of
basketry techniques: 16 techniques found at Qasr Ibrim against eight techniques
that are produced in Nubia (Table 9-21, p. 169). This is the situation after the
resettlement. It is clear that in N e w Nubia the coloured and decorated coiled
basketry, as well as the decorated sewn plaits basketry is still m a d e in abundance.
The undecorated basketry for use on the land or as transportation baskets, are
bought from the Upper Egyptian fellahin. This highlights what is apparently
considered important in Nubia: the baskets which are used inside the house (coiled)
or to bring goods to the house (decorated sewn plaits carrier baskets).
Comparing the techniques and materials used at Amarna, Qasr Ibrim and at
present has given some outlines of the possibilities and limitations of an
interpretation of aspects of society, based on the static aspects of basketry.
It m a y be clear that there has not been a linear development. Strong regional
traditions go hand in hand with innovations and (agri)cultural developments.
A m a r n a and Qasr Ibrim have to be considered as two societies with separate
basketry traditions, as is true for Middle Egypt and N e w Nubia.
At the end of this chapter it is clear that an interpretation of the static aspects
of basketry gives information on a number of general technical and social aspects,
but not as m u c h on the 'people behind the baskets'. Being acquainted with the
techniques and the materials, w e can n o w proceed to the dynamic aspects: h o w
were the baskets made (part 4), and w h o made them (part 5).
PART THREE
D y n a m i c Aspects of Basketry
C H A P T E R THIRTEEN
THE INTRODUCTION OF TIME
W h e n time gets involved in the study of basketry production, w e m o v e from the

static to dynamic aspects of basketry. Rather than looking at basketry production
from the structure, w e approach it from the process, as a result of h u m a n actions.
The use of the terms active and passive in the classification of Table 9.0
(p. 154) is a static criterion: the passive system forms the body of the basic
structure, the active system is causing the coherency. Nevertheless, from this
concept w e can make suppositions about the production process: studying the
present day basket makers tells us that the active system is handled more
frequently or longer by the basket maker than the passive system. In mat weaving
the warp is in a fixed position in the loom, and the weft is handled by the mat
weaver. In the production of coiled basketry the bundle is held in a fixed position
with one hand, while the basket maker is concentrating on the wrapping strand.
Describing a technique as an interaction of a passive and an active system is
not a description of the production process. It is at best a description of one phase
of the work. A s is implicitly shown in the drawing of the grommet in the previous
chapter (Figure 9-2 on p. 156), basketry is made in a number of phases, which are
apparent if w e start looking at production time.
The video shows several of these production phases, but not all of the
sequences are shown fully. Thefirstsequence, production of a sewn plaits basket
by M o h a m m e d in el-Hagg Qandil, is the most comprehensive, following the
production from beginning to end (02:53-24:04). The second sequence shows
Rayhayyet in the Middle Egyptian village of el-Amariyya, while she makes the
start of a coiled basket. The sequence shows her also while working on a sabat
basket in a later phase of the work (24:41-32:58). The third sequence features
Khadidja in the Nubian village of N e w Ibrim, making the start of a decorated food
cover (shauwer, 33:34-41:33).
The fourth sequence is again focussing on Middle Egypt, where the twined
mat-maker AmrTt has his work shop in the village of el-Amariyya. The sequence
is picked up half w a y and focusses on the end of the process (42:10-47:34). The
town of Daraw is located in the middle of the Nubian resettlement region. Shahed,
w h o makes grass rope bed webbing, is not Nubian, but of Egyptian parentage.
This sequence also focusses on the middle and end of the process (48:05-58:09).
In the video each sequence is frozen for a brief m o m e n t to show which
strands are active which passive (cf. Appendix C , 3). However, watching the
basket makers at work makes instantly clear that passive and active systems are not
fixed entities.
During the different phases of making a basket, the strands of a system are
sometimes active, sometimes passive. This is clearly visible in the video.
Active and Passive in Coiling

In coiled basketry the bundle is the passive system, the wrapping strand the active
system. Rayhayyet starts making the coiled basket by making a knot in a bundle of
twisted palm leaf that is to form the centre and the start of the bundle (25:08-25:45).
Thus the start of the passive bundle is active, but becomes the passive system when
the wrapping strands are introduced (26:10). W h e n she inserts n e w strands in the
bundle, these are active, until they are part of the bundle and wrapped with the
date palm leaf (28:15-28:25).
The same is true for the coiled basketry from N e w Nubia. Khadidja makes a
small square plait that is to form the centre of the basket (cf. Plate 4-5, p. 65) For
the entire phase in which the plait is made, the four strands are active (34:11-
35:46). A small bundle is placed at one side of the plaited centre (36:06). In this
phase the centre is passive, while the bundle is actively put in position. One of the
plaiting strands is split and half of it becomes the active system (35:58-36:00): it
wraps around the bundle and is stitched through the plaited centre. Both the centre
and the bundle are from that moment on passive, while the wrapping strand is
active. The other three palm leaf strands with which the centre was plaited, are
gradually integrated in the passive bundle.
With the introduction of coloured winders Khadidja is changing wrapping
strands regularly to form a coloured pattern. Only one strand is active at the time,
the other colour is lead along the bundle and made temporarily part of the passive
system (39:34-39:43). If the colour of the passive strand is needed in the pattern, it
becomes active. This strand is n o w used to wrap around the bundle, while the
previous colour is made temporarily part of the passive bundle.
Active and Passive in Weaving and Twining

In the static description of weaving and twining the warp is passive, while the weft
is active. In a dynamic description of these techniques it has to be made clear that
first the warp is made, during which phase the warp strands are active. It is not
until the weft is being woven, that the warp threads become the passive system.
This is particularly clear in the section in which Shahed makes the bed
matting. With the end of the same strand with which the warp is made, he starts
weaving in the weft (50:55-51:31). The actions to make the edges of the warp and
the weft are identical (warp see 50:34-50:42, weft see 54:11-54:19).
The twined matting, made by AmrTt, is also made with a warp and a weft.
First the warp is made, by arranging loops of string around two metal bars (not
shown). After the loops have been equally divided over the width of the bar, the
weft is twined with two active strands (42:24).
13 The Introduction of Time 271
After the twining has been completed, the two metal bars are removed and
the passive system is becoming active once more: the row of loops is fastened by
pulling each loop through the previous one (46:28).
Active and Passive in Rope Making

Making a three-ply rope involves two stages:firsta two-ply rope is made, in
which both strands are active (07:39 and 15:05). Secondly, a third strand is
wrapped around. In this stage the two-ply rope is passive, fixed between the toes
of the basket maker, while the third strand is active (16:14).
When using the terms active and passive we, therefore, have to keep in mind
that they are referring to only one phase of the production process, the phase in
which the basic structure is made into a coherent fabric. The strands of the passive
system might be active in another phase of the production and vice versa.
As a rule the passive elements are lessflexibleduring the production than the
active elements. In coiled basketry the bundle is lessflexiblethan the winding
strands. In stake-and-strand basketry, the passive elements, the stakes, are thick
twigs, while the active elements, the strands wich are woven in, are thinner and
moreflexibletwigs of the same material.
Theflexibilityis not only determined by the kind and size of the raw
material, but also by the production method. Often a frame is used to temporarily
render the passive system lessflexibleby putting the strands under tension in a
frame. The frame can be permanent, as the wooden frame of the bed, or
temporary. The twined matting is made, for instance, on a very simple loom,
made of four pegs in the ground and two metal bars. Often the tight fixing of the
passive system requires that a tool is used to create a space in the passive system
for the active strand to pass through. In the video this is visible in the sequence of
weaving the bed matting: at the last stage of the production a wooden peg is used
as an awl, to force apart the strands of the passive system (57:17).
Coiled basketry is not made on a frame, but is added onto the anchoring
fabric, the previous rows of coiling. In the coiled basketry from Middle Egypt a
small needle is used at the end of the palm leaf, and often a hole is punched in the
previous coil, to pass the needle through the bundle material (29:37). In N e w
Nubia the coiled basketry is made without the help of a needle, but an awl is used
to create the space for the strip of d o m palm leaf to pass (37:33). For sewn-plaits
basketry, a large needle is used to help the sewing string to pass through the edges
of the plait (11:51).
The following sections concentrate on how: how are raw materials gathered
and prepared (Chapter 14) and h o w are the baskets made? The basketry production
is approached from two different angles: the first angle starts at the baskets found
in the excavations and reconstructs the order of actions and time necessary for the
end product (Chapter 15). The second part of the description starts from the
producer: what actions are involved in making the object, in what context does his
work take place (Chapter 16).
The difference between these two angles can be illustrated with an example:
the first approach results in an estimate of the production time, which is the total
time spent on making the object (e.g. eight hours), the second approach is
characterized by an interest in the production period, that is the total period spent
on making the object (e.g. a total of eight hours on four Fridays between the noon
prayer and going back out into the fields).
The importance of the second angle in relation to archaeology is that it
prevents a naive approach of social and economic aspects. A basket maker is not
only a basket maker, but his position in society is also determined by being a
m e m b e r of a family, neighbour, farmer, parent and village watch.
CHAPTER FOURTEEN
CLIMBING THE TREE
14.1 Obtaining and Preparing Raw Materials

Gathering or harvesting the raw materials is mostly part of the basket maker's job.
W e k n o w the range of raw materials used in antiquity, but it is not until looking at
the present day basket makers that w e begin to get an understanding on the time
and costs involved in the process that precedes the actual basket making.
The preparation of the raw materials is divided into two stages. The plants
have to be dried before being used or stored. Cutting, cleaning and sorting is often
also part of the preparatory task. The second stage of the preparation falls usually
just before the materials are to be used. In this stage the strands are cut to size and
soaked. T h efirstpreparatory phase is looked into in this chapter, the second phase
is considered an integral part of the production process and will be dwelt upon in
Chapter 15.
The organization of gathering the raw materials is partly linked to the
seasons. S o m e plants can only be harvested during specific periods, while others
can be gathered year round. Harvesting or gathering at a specific time of year
needs organization and storage space.
With the archaeological remains from A m a r n a and Qasr Ibrim examples of
several of these stages of preparation have been found. Based on these and
interviews with modern basket makers the preparation needed for the different
materials is surveyed.
For each plant species it is surveyed which parts are gathered. Also the
question h o w they are harvested and h o w m u c h time is involved is deemed
important. The ownership of land, plants or trees is touched upon also.
Because this chapter depends heavily on the ethno-archaeological research,
the Arabic terms for the different materials are included. If the Nubian or ancient
Egyptian terms are known, these are mentioned too.1 Since there are m a n y local
differences in the Arabic terms, mainly those used in the vicinity of el-Hagg
Qandil are given. A complicating matter is that most of the basketry terms are
The Fadidja terminology is based on interviews, the Kensi terminology is based on

interviews and literature (Alkwist & Zettersteen 1911, Hofmann 1986), the Arabic
terminology is based on interviews and literature (Badawi and Hinds 1986; Behnstedt and
Woidich 1994; Wehr 1980, Henein 1988), the Egyptian terminology is based on literature
(Helck 1961, Janssen 1975 and Wallert 1962).
The world According to Basketry
274
unknown in modern standard Arabic.2 The Arabic, Nubian and Egyptian terms
have been included in the vernacular glossary of Appendix D.
14.2 Palm Species (Palmae)
Doam palm
Hyphaene thebaica p. 146
p i m used plant parts:
a m i leaves Arabic: waraq ei-doam'

W M L Fadidja: imbi nig
4 S§rv Shv^v 1 Kensi: ambu nig
\ti r > ^
L. Egyptian I 3 w (IV, 109)4
w ^"^vV O
4 p
leaf Arabic: jertt
stem L. Egyptian blj. w
Winning
T h e large fan-shaped leaves of the d o a m palm are cut from the top of the forked
stem, which can grow to a height of 20 metres. T o harvest the leaves the tree has
to be climbed and the leaves are cut at the base. T h e leaf-stems are t w o metres
long and have sharp spikes at the base. T h e organization of the harvest is not
k n o w n . In principle, d o a m palm leaves can be harvested anytime of the year. For
The basket makers which were able to write were asked to spell the names of the
different objects and materials. In many cases their reaction to this request was: "that is a
word that is not written", meaning: it is colloquial Arabic.
Arabic uses: shagar (tree) and waraq (leaf) The doam palm tree is considered a
proper tree, while the date palm is a class of its own. Both species can be considered trees,
in that they are high rising with a stem and leaves, but palm trees are characterised by the
lack of secondary growth in the width (neither species has growth rings). The Arabic
designation of the doam palm as a tree is probably because it has a hard stem, compared to
that of the date palm, which is fibrous.
4 r
L. Egyptian = Late Egyptian. The numbers refer to the volume and page number of
Lesko 1982-1990. Cf. also ps (I 182) for a pack of fan leaves.
14 Climbing the Tree 275
basket making only young leaves are suitable, the old leaves are less flexible and
brittle.
The time involved in gathering enough d o a m palm leaf for one basket
depends on the size of the basket, but the coiled baskets found at A m a r n a and Qasr
Ibrim are quite small and do not take more than part of one of the large fan-shaped
doam palm leaves. It does not take more than 10 minutes to climb the tree and cut
off one leaf. Although there is no specific harvest period for palm leaf, often
several leaves are cut at the same time.
Preparation
Just after cutting, the leaves are dried for at least two weeks before the next stage
of preparation takes place. A half product is prepared consisting of dried and
sometimes dyed strips. The dried fan shaped leaves of the d o a m palm tree are split
with a knife in strips of equal width. The hard veins of the leaf are about 40 m m
apart, which is, therefore, the m a x i m u m width of the strips. In a later stage the
strips are cut into m u c h smaller strands d o w n to 1 m m wide (Chapter 15). What is
stored are not the leaves, but neatly bundled strips. At Qasr Ibrim such bundles
have been found in an Islamic and a Ballana period context (Plate 14-1).5 The
length of the strips in the bundles is 650 and 620 m m , the width is in one bundle
14.2 m m , in the other the width of the strips varies between 2.5 and 14 m m .
tJ8»&-*«-' •'•
Plate 14-1 A bundle of prepared doam palm leaf, tied with a small strip of doam palm
leaf (Qasr Ibrim 10.070-070). The bundle is approximately 13 cm long
At present in New Nubia the dry cut strips are boiled for 30 minutes in
water, to which sometimes alum is added. After drying for a day the strips are
ready to be either used or stored. Part of the d o a m palm leaf is boiled again in a
dye solution. N o w synthetic dyes are used, which are bought on the A s w a n
510.075-031; 10.070-070.
market. If a very deep colour is needed, the dyes are added also the first time the
strands are boiled.
In Old Nubia, which is Nubia before the resettlement following the building
of the Aswan High D a m , natural dies were used, but the m e m o r y of the plants or
minerals used has faded during the more than 35 years that the people have been
living in N e w Nubia.6 The use of natural dies must have been seldom during the
years before the move as well, because even baskets which were brought in the
m o v e from Old to N e w Nubia were made with synthetic dyes. The informants
mentioned Aswan or Wadi Haifa as their source of dyes. The dry strips, whether
plain or dyed, can be stored for a long period.
At Qasr Ibrim another half product was found: bundles of S-twisted doam
palm leaf string, one of which was wound around a stick (Figure 14-2).7 This
suggests that the string for the furniture matting found at the site has been prepared
before the actual weaving started. Before twisting the leaf strips, the material is
soaked. The string has to be wetted again before the weaving can be started.
Figure 14-2 Prepared doam palm leaf, stored for later use, for instance in weaving
furniture matting (Qasr Ibrim 10.416-552). Scale 1:1.
Availability of doam palm leaf

Though doam palm trees were probably more c o m m o n in Egypt in the N e w
Kingdom Period than at present, they may not have been widely available. A n
indication for this is a text stating that packs of doam palm leaves were imported
into Egypt (Gardiner 1937, 11,4,3). In the text there is no indication w h y they
were imported, but it seems likely that the purpose was basket making, since
almost all coiled baskets and most of the twined mats from that period are made of
One informant mentioned three colouring agents: ink, kurkuma (which could be
either the Asian yellow root or a light blue liquid which is put on wounds) and an unknown
substance called sennag addoka (Fadidja Nubian).
710.001-242; 10.416-552.
doam palm leaf. This would imply that the ancient basket makers had to buy at
least part of their raw materials. Not only the leaves of the doam palm are useful,
but the fruits (doam nuts) are edible and the stones are hard and white, a vegetable
ivory (Germer 1985, 234). The stems of the leaves can be used for making roofs
and fences.8 Owning a doam palm, therefore, is an asset.
In Old Nubia, basket makers could obtain doam palm leaf freely from
c o m m o n trees. In N e w Nubia not many doam palm trees are available. S o m e of
these are for c o m m o n use, others are private property, depending on if they are
growing on c o m m o n grounds or privately owned land. In present day Middle
Egypt no doam palm leaf is used.
Date p a l m
Phoenix dactylifera p. 146
used plant jarts:
leaf Arabic: qalb el-nakhla
leaflets Arabic: sa'af en-nakhia.

Fadidja: nintin nig
Kensi: benti nig
leaf mid Arabic: jen't

rib L. Egypt. bLj. w
4 - »
leaf sheath Arabic: i<7
fibre
fruit stem Arabic: zaghauwah

Fadidja: arru
Kensi: jurbah
flower Arabic: sabata, zabata

sheath or za'abata
Winning
All parts of the date palm tree are used, but not all of them for making basketry.
T h e harvesting of parts of the date palm is for s o m e parts done once a year, w h e n
the tree gets its 'maintenance service', other parts are taken of the tree whenever
necessary.
For the stems of the leaves the same Arabic term is used as for the midribs of the
date palm leaves, jerrt. The same is true for the ancient Egyptian term: b'j.w, used to
indicate the midrib of both palm species (Wallert 1962, 53).
Date palm leaves are an example of the latter. The basket makers prefer to
use the young leaves sprouting in the middle of the crown. They are the 'heart' of
the palm tree (qalb el-nakhla, plur. qulub). Not too m u c h of these young leaves
can be taken of a tree, otherwise it will die. The white leaves are usually harvested
from young, not very tall trees, which can be climbed without special tools,
because the old leaves have been cut in such a way that the trunk of the tree is
provided with 'steps' (cf. Plate 14-3).
Plate 14-3 Man collecting young palm leaves and a fruit stem for his wife, who makes
coiled baskets (el-Hagg Qandil 1989) .
The feather-shaped date palm leaves consist of a midrib (jerit) and side leaflets
(sacaf). The expression used in Arabic for the young leaves of the date palm is
khuz abyad (white leaves). The slightly older leaves are indicated as khuz akhdar
(green leaves), which are m u c h tougher than the young leaves and at the same time
quite brittle. Archaeology and textual evidence show that this differentiation
between green and white leaves is very old. In a papyrus from 87 A D part of a
tenancy agreement for land mentions a basket specified as AeuKotpuAAcov, m a d e
of white leaves.9 The archaeological evidence from Qasr Ibrim shows that the
more mature palm leaves were used for coarse baskets, while the average and fine
basketry was made from the young palm leaves.
Once a year, a few weeks after the harvest of the dates, the older leaves at
the base of the crown are cut off as part of the maintenance of the palm tree. The
mid ribs (jerit) of these old leaves are used for different purposes, among which
the production of pierced basketry (Plate 10-17, p. 202), roofs (Plate 11-32,
p. 246) and as fuel. At the same time the remains of the leaf sheaths at the bottom
of the crown are also removed. These brown matted fibres (Iff) are used for rope
making. The leaves have large spines at the base (soka)w which are sometimes
used as awls in making coiled basketry (cf. Plate 14-4). In Qasr Ibrim they were
found stuck in a piece of cloth as a kind of make shift pincushion with pins.
The dates grow on long slightly curved stems, which are cut off at the base
and lowered with the bunch of dates still attached (cf Plate 14-4). These stems
when dry, can be split into rigid fibres, which are used for the bundles of coiled
basketry in Qasr Ibrim (rarely) and modern baskets (commonly).
The term for the fruit stem differs locally. In el-Hagg Qandil two names are
used. The most c o m m o n term is zaghauwa, but it is also indicted as seyf el-zabata
(the 'sword of the fruit sprout'). The bunch of curly twigs that sprout at the end of
the fruit stem and on which the dates grow is called zabata (Behnstedt and Woidich
1994, 181), which is also the term for the female flower bunch before it has grown
into dates.
T w o other terms occur: jurbah and 'arjun. The first is attested in Middle
Egypt and Upper Egypt as a term for the sheaths in which the flowers of the date
palm grow (Henein 1988, 180)." The flower sheaths have a dark brown epiderm,
strips of which are used in el-Hagg Qandil and el-Till to decorate and strengthen
the rims of coiled baskets. In Kensi Nubian the term jurbah el-balah was quoted
as the fruit stem (Hofmann 1986, 87).
The term 'arjun is used in north Middle Egyptian for the fruit stem
(Behnstedt and Woidich 1992, 305) but in Upper Egypt for the female flower
bunch. Henein quotes the use of the term 'arjun for the fruit stem fibre (Henein
9Papyrus Soterichos 4, line 17-18, cf. Worp 1988, p. 280.

This term is found in Nubia. Henein uses the word selld'a (Henein 1988, 403).
The term for this sheath is in Middle Egypt sabatah, zabatah or za'abatah, while
the flowers inside are called za'ab. The sheaths of the male flowers are not used for basket
making.
1988, 189), which makes it the Upper Egyptian synonym for the Middle Egyptian
word zaghauwa. In Fadidja Nubian the fruit stem of the date palm is called arm.
The yearly harvest of parts of the date palm (dates, fruit stems, leaves, fibre)
from old, high standing palm trees is done by specialised m e n , w h o climb the old,
tall trees, which are 20 metres or higher, with the help of a climbing girdle slung
around the tree and the waist. The tools used are very sharp curved knives and a
small hoe.
In Plate 14-4 a m a n is shown climbing a younger palm tree, to cut a white
leaf from the heart of the crown and collect some zaghauwah stems. The fact that
there is still an old fruit stem in the tree shows that it is not maintained regularly.
In el-Hagg Qandil several people o w n date palm trees for their o w n use, rather
than for growing dates on a large scale. There are no specialised professionals w h o
take care of these trees. Rather than harvesting all the dates and trimming back the
tree yearly by removing the old leaves and fibre from the base of the crown,
people walk up to their tree and cut whatever they need at the spur of the moment.
Plate 14-4 With his right hand, the man pulls at an old fruit stem, of which the dates
have fallen. Note the sharp spines with which the base of the date palm
leaves are covered (el-Hagg Qandil 1989).
Preparation
S o m e preparation is needed before the harvested palm leaves and fibre can be used
for making baskets. The young leaves from the centre of the date palm are dried
thoroughly (minimally two weeks) before they can be processed further. Using
fresh palm leaf is not possible because the shrinkage is irregular.
The side leaflets, which w h e n young are still folded lengthwise, are pulled
from the midrib by hand. These leaflets (called sa'af or with a collective n a m e
khuz) are 40 c m long. At some stage (either at the preparatory phase, or just
before the work on the basket starts), the vein is pulled of, leaving two strips of
date palm leaf of approximately 10 m m wide. These strips form a half product that
can be stored until needed. At Qasr Ibrim date palm leaf strips were found, which
were cut into even smaller strands of 4.3 m m wide, with a length of 430 m m long,
tied into a bundle with a piece of grass string Plate 14-5).
Plate 14-5 Bundles of prepared strips of doam and date palm leaf, found at Qasr Ibrim
(10.048-027 and 10.070-070). (Courtesy of the Egypt Exploration Society).
The mid ribs of the leaves (jerit), used for instance for making roofs, but
also for the production of pierced basketry, do not need any preparation, but have
to be used within about five days after they have been cut from the tree, before
they have dried out. They are cut to the required length and split if necessary.
The fruit stems are dried during several weeks and are pounded, to release
the tough fibres. In modern Egypt they are layed out in the streets, to let the
passing cars run over them and do the hard work. After that the fibres can be split
easily with a knife or by hand, using anything pointed. This is usually done while
the basket is m a d e (Chapter 15).
The leaf sheath fibre can be used straight away after a brief period of soaking
in water. For fine string the fibre has to be selected more carefully, using only the
upper parts and soaking for a longer period. Small bundles of fine, clean fibre
were found at Qasr Ibrim.12
12Date palm fibre: e.g. 10.068-243, 10.128-119, 10.236-221.

A vailability
Owning a date palm is important, because it gives access to food (the dates), fuel
(the old leaves and the stem), building material (the stem and the mid ribs of the
leaves) and raw materials for basketry and cordage.
In the workmen's village of A m a r n a only date stones were found, but no
other parts of the date palm were used. All basketry was m a d e of d o a m palm leaf
or grass. It seems that date palm trees were not found in the neighbourhood of the
village, or that inhabitants did not have access to the tree. In Qasr Ibrim most of
the plaited basketry was made of date palm leaf, while the coiled basketry was
made of d o a m palm leaf.
In present day Middle Egypt both coiled and plaited basketry is m a d e of date
palm: the bundle is made of zaghauwah (fruit stem) fibres. In N e w Nubia plaited
basketry is made from date palm leaf, while coiled basketry is mostly made from
d o a m palm leaf or alternative materials. Bundles of jurbah or arru (fruit stem)
fibres are used as the passive system in coiled basketry.
If a basket maker does not o w n a date palm tree, he or she has to buy the
raw material. The price of one young leaf of the date palm varies, but is
approximately 30 piastres, one fruit stem costs about 60 piastres, a large mid rib
one Egyptian Pound (price levels of 1992). N o price indications are known from
the ancient sources.
14.3 G r a s s Species (Gramineae)
Haifa grass
Desmostachya bipinnata p- 147

Imperata cylindrica p- 148
used plant parts:
leaves Arabic: halfa

and Fadidja: 7
culms Kensi: gid
L. Egyptian •nb a 79)
Winning
Since the grasses are perennial, they are always available. The condition of the
grass is best w h e n the young leaves have grown and the plant is not yet flowering,
but in principle grass can be collected as needed all year round.
Preparation
Grass does not need m u c h preparation, just a waiting period of three to five days
to allow the grass to dry. Bundles of grass can be stored dry and have to be wetted
just before they are used. Several of such grass bundles are found at Amarna. At
Qasr Ibrim some complete grass plants are found, which are knotted in circles.
The total length of those grass plants is 900 m m , the diameter of the bundles 20-40
mm.13
Availability
Haifa grass grows abundantly at the edges of rivers and canals, where it can be cut
for free. Although it is feasible that grass was cut whenever necessary, it seems
more likely, considering the m a n y uses it was put to in A m a r n a , that a supply was
made for the months in which the quality of the grass had diminished too much. In
competition with basket making, grass was probably also harvested by farmers
keeping pigs, sheep, goats and donkeys.
Reeds
Phragmites australis p. 149

Arundo donax p. 149
used plant parts:
culm Arabic: ghaab. boos

Fadidja: ?
Kensi: ghaab
L. Egyptian wnr (I, 118)
g3s"
knotted grass bundles: e.g. 10.068-243, 10.070-070, 10.128-119, 10.384-318;

4Germer 1985, 206
Winning
Reeds are cut and used fresh or dried. Probably reeds were never stored, but just
gathered whenever necessary.
Preparation
The only preparation necessary is the removal of leaves and plumes.
Availability
Reeds are available freely, growing along the banks of rivers and canals. They are
best gathered before they are flowering.
Wheat straw
The dried culms of a cultivated grass (Triticum dicoccum) is used in Nubia for
decorating coiled basketry. The Fadidja term for straw is fafa, the Kensi Nubians
use the word brubi, or the Arabic qaSi el-qamh. Immediately after the wheat
harvest the culms are collected and stored until needed.
At present, the glossy material is often dyed in bright colours. The same dyes
are used as for d o a m palm leaf, but the procedure for dyeing straw is simpler: the
culms are split and soaked in the dye for ten minutes without boiling.
Since straw is a by-product of farming, the basket makers w h o do not grow
wheat have to buy the material. Competitively wheat straw is used as inclusion in
clay and m u d brick, as animal fodder and as fuel.
14.4 Sedges (Cyperaceae)
Cyperus papyrus p. 150
used plant parts:
culm Arabic: bardi

L. Egyptian wd (I. 139)
mhj.t, mnh.
twfj'5
Germer 1985, 249

Cyperus rotundus p. 151

Cyperus alopecuroides p. 151
Cyperus schimperianus p. 151
used plant parts:
culm Arabic: samaar, dees,

sa'ad
L. Egyptian wd ? (I, 139)
Winning
Papyrus can be harvested from M a y to August, but the quality is the best in June
and July.'6 For the other sedges, which were gathered in the wild, the best time of
year was probably the same as for Cyperus papyrus.
Preparation
The culms take approximately ten days to dry. They can be stored dry and used
whenever necessary, by soaking them in water for some hours.
Availability
In the Delta sedges were growing abundantly, but at A m a r n a they were probably
not available. This is reflected in the small number of objects m a d e of sedges
found at A m a r n a and Qasr Ibrim.
Information kindly provided by Dr. Ragab's Papyrus Museum, a commercial

organization which has experience with growing C. papyrus for their production of papyrus
sheets.
14.5 Rushes (Juncaceae)
Juncus rigidus p. 152

Juncus acutus p. 152
used plant parts:
culm Arabic: samaar morr

L. Egyptian jsrxl
Winning
Rushes have to be cut in M a y , w h e n they are fully grown, but not hard and dry.
This is also the time that goats and other roaming animals become dependent on
this plant for their diet. The rushes are cut nowadays with a sickle-shaped knife
just above the roots. The rest of the year the mat maker has to work with the
supply of rushes gathered in spring.
Preparation
At present the rushes are sorted by length and then left to dry for several weeks in
a shaded place. There are three different lengths of approximately 1.50, 1.10 and
0.90 metre long. The culms from these bundles are used in weaving three
corresponding widths of mat (1.40, 1.00 and 0.80 metre wide). T h e rushes are
stored dry and w h e n needed bundles of the required length are taken out and
soaked for a few hours in preparation of the work.
Availability
The rushes are available for free. They grow in marshy areas which usually do not
belong to anybody. It seems likely that rushes did not occur in the vicinity of
A m a r n a or Qasr Ibrim. N o indications have been found that they were used for
production on site. The few baskets m a d e of rushes were probably imports, the
F a y o u m being a likely place of origin.
17Germer 1985, 201

14.6 Plant Fibres
Winning
Flax fibres are the fibres of the culm of the Linum usitatissimum (see p. 152),
which means that the complete culm is harvested. The flax plants were not cut, but
pulled out of the ground w h e n ripe. The plants can be harvested w h e n the flax
stems are either green (just before the seeds have ripened), yellow (when ripe) or
brown (old flax stems). The quality of the flax varies with the age in which it is
harvested. The green stems yield a fine fibre (for fine linen textiles), the yellow
stems a mediate, strong kind of flax and the brown stems a coarse fibre, only used
for making string.
In Qasr Ibrim cotton, which was introduced in Egypt in the Graeco-Roman
period (Watson 1977), was widely used. The cotton fibres are the protective cover
of the seeds and are picked of a shrub-like tree (Gossypium arboreum, see p. 152).
Preparation
The cotton fibres have to be separated from the remains of the cotton seeds, but a
m u c h longer process is involved to m a k e fibres from the culms of flax plants. The
plants have to be soaked in water for 10-20 days to ret, a procedure in which the
soft parts of the plant are destroyed by bacterial action, leaving the fibres proper.
The fibres then are combed, to remove the unwanted soft cell structures
(hackling). After drying, the fibres are ready to be spun into a yarn.
Availability
Flax nor cotton was available for free, or could be gathered at low costs. Both the
annual flax plants and the perennial cotton shrub were grown by farmers, flax
since at least the pre-dynastic period, cotton from the Graeco-Roman Period
onwards (Watson 1977).
There is no evidence that flax fibres were produced at the A m a r n a
workmen's village or that either flax or cotton was processed at Qasr Ibrim.
Considering the complex and time consuming preparation of these two fibres, and
the amount of water the preparation of especially flax required, it is likely that not
the plants, but the cleaned fibres, or perhaps even the ready made yarn was sold to
weavers and producers of furniture matting and fish nets.
C H A P T E R FIFTEEN
M A K I N G M A T S , B A G S A N D BASKETS
Although w e have made an inventory of what was produced in Egypt (Chapters 9,

10 and 11), it has not really become clear exactly h o w these objects were made
and only glimpses have been caught in Chapter 12 of why they were made. In this
chapter the production processes will be discussed for each of the twenty
techniques listed in Table 9.0 (p. 154).
The 20 techniques are classified according to their basic structure, which
means that they represent only one phase of the production process. In the next 20
paragraphs those frozen instants in the production process are allowed to unfold.
To express that the descriptions do not represent the production process, 'as it once
was', but an interpretation of h o w it could have been, the present tense is used for
the descriptions. Not only the interpretations of processes that do exist in Egypt
today are based on the observation of modern producers, but also the ones that
only occurred in ancient Egypt.
15.1 knotless netting
Amarna (p. 176) Middle Egypt
prod uction phases doam nuts net time estimate

a. preparation materials soaking d o a m palm leaf 30 minutes
b. preparation systems - -
c. start fixing starting border 10 minutes
d. basic structure twisting and looping 4 hours
e. finish closing the bag with a pull string 5 minutes
f. additions - -
g- finishing touch - -
Knotless netting, of which a few small fragments have been found at A m a r n a has
no modern equivalent. A n ancient parallel has been found in the tomb of K h a (also
from the 18th dynasty). This was a complete net,filledwith d o a m nuts.1
The d o a m palm leaf strips have to be soaked for 30 minutes just before the
production is started (34:09-34:10). A starting strand is fixed between two points,
m a d e of d o a m palm leaf, with a length of approximately 650 m m . Onto this
starting border the first row of loops is linked. The strand is then knotted to close
the circle and the next row of loops can be made, spiralling d o w n until the net has
the required size.
The bag maker twists the palm leaf strands into a loop, which he links with
the previous row of loops.2 N e w lengths of d o a m palm leaf are twisted together
with the strand that runs out. Probably this is done without the help of any
instruments. Since d o a m palm leaf strands are about 650 m m long, a n e w strip has
to be added after every three or four loops. The twisting of the palm leaf strand
and making the loops is done in one action.
The palm leaf strips are twisted in S-orientation and crossed over in S-
orientation. This is not the natural orientation of crossing over w h e n a loop is
made, as is simply tested by strongly twisting a piece of string in S-direction: the
overtwisted strand folds in Z-direction (Figure 15-1). By forcing the S-twisted
strands in an S-oriented crossing, the loops are forced to open up which helps the
bag to retain its shape.
Figure 15- Twisting a palm leaf in S-direction results in a Z-crossing
The bag is started with a small number of loops, fitting the starting strip,
which is increased gradually to enlarge the diameter of the bag. Since only small
fragments have been found at Amarna, deciding the shape of the original bag is not
possible.
.At present this net is on display in the Egyptian Museum in Turin.

'This is a one-system technique, in which the previous row of loops is the steady part
(or anchoring fabric) to which each new row of loops is connected.
15 Making Mats, Bags and Baskets 291
The last phase is finishing off the bag. The small fragments found at Amarna
did not have edges. The edge of the bag in the tomb of K h a has been made by
running a string through the last row of loops. By pulling the string the bag can be
closed. The time involved in making a large bag of knotless netting would be 30
minutes waiting for the leaf strips to soak and approximately four hours for
making the fabric.
Making a knotless netting bag does not require a specific work space, nor
any instruments. Judging from the bagfilledwith doam nuts in the tomb of Kha, it
seems possible that these bags were produced near the place where the d o a m nuts
were harvested. It could well represent an on-the-spot production of transportation
bags, specifically for doam nuts.
15.2 Grommet
•• •
Qasr Ibrim (p. 207) N e w Nubia
prod uction phases pot stand time estimate

a. preparation materials wetting grass or date palm fibre 1 minute
c. start making the first round 1 minute
d. basic structure continuing three rounds 2 minutes
e. finish tucking in the end of the string 1 minute
f. additions - -
Several g r o m m e t s were found in A m a r n a and Qasr Ibrim, m a d e of grass or date

palm fibre, which has to be wetted. T h e moist material is then divided into small
tufts, which are subsequently spun between the hands to form a strand. Details on
this basic rope making technique will be discussed in Section 15-4.
With this strand the first circle is formed around which the next round is
plied. This is done by spinning and plying the fibrous material in the s a m e
m o v e m e n t , in opposite orientation (s-spun and Z-plied or the other w a y around) to
prevent the twisted fibres from unrolling.3 A third and a fourth round are made
after which the end of the strand is tucked into the grommet. Thus the person
making the grommet alternatively picks up a new tuft of material, twists the fibres
into a strand and the strand around the previously formed ring.
There is no passive system: a grommet maker uses both hands in twisting the
fibres and the grommet. The drawing of Figure 9-2 on p. 156 shows the four
rounds as four stages. Looking at the actions, however, it is one ongoing
movement, only interrupted to insert a new tuft of material, during which the
grommet is turned until the four rounds are completed.
Apart from a knife to cut of the ends no instruments are needed. Neither
gathering and preparing the materials, nor making a grommet takes much time.
There are no special requirements for a work space. The production probably was
local and ad hoc, made on the spot whenever a pot stand or head ring was needed.
At present grommets are still made in Middle Egypt as head rings made of rags.
There is no professional grommet production.
15.3 K n o t t e d Netting
••
P>'oduction phases fish net time estimate

a. preparation materials - -
b. preparation systems production of flax string several days
c. start fixing starting border 5 minutes
d. basic structure knotting the net several days
e. finish making a bottom line 30 minutes
f. additions adding floats and weights 60 minutes
g- finishing touch folding the net for use 10 minutes
The first circle forms the anchoring fabric for the second round. These two together
form the anchoring fabric for the third round, and the three of them form the anchoring
fabric onto which the fourth round is plied.
The material used for netting in A m a r n a is flax. Cotton was not k n o w n

before the Ptolemaic Period and was introduced from the South, Nubia being the
first and for a considerable period the only part of Egypt where cotton was grown
(Watson 1977). Although cotton is a suitable material for making nets, at Qasr
Ibrim the one netting fragment found proved to have been made of flax yarn.
Flax needs an extensive preparation (see p. 285). It seems likely that the
preparation of the plants into flax fibres was not done by the same people w h o
were spinning the flax. It is not k n o w n if the processing was done by the farmers
w h o were growing the flax, or by separate producers. It is equally unclear, if
spinning the fibres into yarn was performed by separate spinners or by those w h o
used the yarn (weavers, net makers, furniture makers).
Flax is spun moist, but does not have to be soaked. The most important
instrument in spinning is a spindle whorl. The entire body is involved in spinning.
The flax is regularly wetted with water or saliva, by running the fibres through the
mouth. The spinning whorl is made to turn by rolling the weight with one hand
along the thigh. Spinning a quantity sufficient for a large fish net takes a few days
(depending on the size of the net).
The net maker starts his work by using a string as border, to which the first
row of loops is knotted. This border is fixed, for instance to the wall, or, as
depicted in several tombs in Beni Hassan (Figure 15-2) to a, probably wooden,
support.
Figure 15-2 Depiction of net makers in the Middle Kingdom tombs in Beni Hassan. Left:
tomb 15, right: tomb 17 (after: Newberry 1893).
The process of knotting a net is a continuous repetition of two major actions:

filling a netting needle with yarn and knotting loops until the yarn runs out. The
knots used in making the net are k n o w n as mesh knots. They can be knotted in two
orientations, but in the netting fragments found, and indeed in all present day nets,
only the Z-oriented mesh knot is used. The subsequent rows of knots face the
opposite side of the net: thefirstrow of mesh knots faces the front of the work,
the next row the back. There are two ways in which this can be achieved. Either
the whole net is turned around at the end of each row of knotting, or the net maker
works from left to right with a different stitch than from right to left. The latter
seems the most obvious.
It is not exactly known which movements the net maker m a d e with his
netting needle, but a possible method has been depicted in Figure 15-3. Working
from left to right the netting needle is pulled through a loop of the previous row
from the back of the work to the front (1). Then the yarn is slung in a loop over
the work and the netting needle is passed behind the work through this loop (2).
The netting needle is then pulled d o w n to the right (3) and the knot is fixed (4).
After a row of loops have been knotted, the net maker continues from right to left
by passing the netting needle from the front to the back of the work (5), then
looping the yarn over the loop from the previous row and passing the netting
needle behind the work (6). The netting needle is pulled d o w n to the left (7) to fix
the knot (8).
Figure 15-3 Supposed order of action of making a netting fabric with mesh knots.
15 Making Mats. Bags and Baskets 295
The net is finished by knotting a thicker string at the base. Further additions
such as floaters and weights, which are tied to the net, depend on the use to which
the netting is put. Thefloaterswere not made of cork (Quercus suber), but of the
thick parts of thejenl (midrib of the date palm leaf), some of which were found at
Qasr Ibrim. Net weights are not known from Amarna or Qasr Ibrim. Those found
at the R o m a n fortress of A b u Sha'ar, were made of anything that could way the
net down: stones, pieces of coral, amphora handles and pieces of lead."
The time involved in making a net, depends on the size of the mesh and the
size of the net. A n experienced net maker will need at least a day to knot a square
metre of finely meshed netting (10 x 10 m m ) . This would involve 10.000 knots
and repetitive filling of the netting needle. A square metre of netting with a mesh
of 40 x 40 m m takes approximately half a day.
The most important tool used in making a net is the netting needle. A clear
depiction of one of these is found on the wall painting in the tomb of Kheti in Beni
Hassan. At Qasr Ibrim part of a netting needle, made of jerit was found. Possibly
a gauge stick was used to ensure a constant mesh size.
15.4 s e w n plaits
Amarna Middle Egypt (p. 180)
•••
Qasr Ibrim (p. 208) N e w Nubia (p. 215)
••• •••
production phases sewn plai s basket time estimate

a. preparation materials removing veins, soaking palm leaf 1 hour
b. preparation systems making string and a long plaited 6 hours
strip
c. start sewing the centre 10 minutes
d. basic structure sewing plaits into a basket 4 hours
e. finish fastening the end of the plait 1 minute
f. additions adding edges, rim and handles 1 hour
g- finishing touch removing irregularities 10 minutes
The University of Delaware excavations at A b u Sha'ar (Red Sea Coast), directed by

S.E. Sidebotham. lasted from 1990 to 1994. The publication is in preparation.
a) Preparation of the materials

The dry date palm leaflets are taken up one by one and the hard vein in the middle
is torn of with the help of a knife. The date palm leaf strips are n o w approximately
10 m m wide. The basket maker puts the veins on a separate stack to be used later
for the production of string. In the production sequence of the video a stack of
sorted leaves lies to the right of the basket maker (03:35). These have been soaked
for 30 minutes and are strong and supple. Soaking palm leaves has been well
attested from early Christian sources, since making the sewn plait basketry was a
major occupation of Coptic hermits (Waddell, cf. pp. 78, 86, 121).
The veins, which are also put in water to soak (06.19 and 12:47), are m u c h
harder than the two sides of the leaflet. After a long period of soaking they are still
tough. Just before the basket makers start to use them for making string, he softens
the veins by twisting them vigorously (06:46).
The preparation of the date palm leaf sheath fibre takes m u c h less time: the
fibre is only wetted briefly and then used straight away. A small bunch of the
tough fibre is pulled off and given a little twist (16:01). These little tufts of palm
fibre are used for making the string for the rim and handles.
b) Preparing the systems: plaiting a strip and plying a string

Because the sewn plaits basketry involves the use of both long plaited strips
(Section 15.15) and string (Section 15-13), the production of these are incorporated
in this section.
Plaited strips
A s was shown in section 11.2.1 (pp. 208-217) there are different plait patterns and
there is a relation between the pattern, the number of strands and the use to which
the plait is put.
A twill nine-strands plait is started by laying two double strands across each
other (Figure 15-4 a). Strand A B folds around strand C D (Figure 15-4b). Then
strand C is folded under B and over A (Figure 15-4 c). The upper half of strand C
(CI) is folded under the lower half (C2) and over D. At this m o m e n t there are five
strands (Figure 15-4 d). The upper half of strand B (Bl) is folded under both its
lower half (B2) and the lower half of strand A (A2). Then it is guided over both
the upper half of strand A (Al) and C I (figure 15-4e). This is the start of the
actual plait pattern, which goes under 2 over 2. After dividing strand D in D l and
D 2 there are eight strands. A ninth strand is inserted (C3) and strand C 2 is plaited
under C 3 and D l / over D 2 and Bl (Figure 15-4 f).5 Then alternately the strand on
the right and the strand on the left are plaited towards the middle (B2 runs under
A 2 and A l / over C I and C 2 ; C 3 runs under D l and D 2 / over B l and B 2
etcetera). N e w strand are layed in at the right side (cf. B 2 in Figure 15-4f).
Video sequence 03:47-04:20. The extra (ninth) strand, indicated as C 3 in Figure

15-4f, is
is inserted at 04:16:28.
Figure 15-4 The start of a nine-strands twill plait (pattern \2/2\\l).

The W o r l d according to Basketry
298
Because date palm leaf strips are only about 400 m m long, n e w leaf strips are
inserted behind the outer strands (cf. the new B 2 in Figure 5-4 0-This is repeated
until the plait has the required length.
The time needed to plait a length suitable to make a basket depends on the
size of the basket but for a maqtaf the plaiting takes about 5 hours. If the work has
to be stopped for a while, the plait is fastened temporarily, by locking one of the
outer strands in the plait. A similar w a y of temporarily fastening the end of the
plaiting was found in Qasr Ibrim (Plate 5-5). N o instruments are used and no
equipped work space is needed.
Plate 15-5 Unfinished five-strands plait (\1/1\\1) made of date palm leaf, fastened
temporarily. To the left: small plaited pad of doam palm leaf. (Qasr Ibrim
10.048-027). (Courtesy of the Egypt Exploration Society).
Making string
In present day Nubia baskets are sewn with unspun strips of the d o a m palm leaf,
but in Middle Egypt string is used for sewing. This string is m a d e by the basket
maker out of the veins of the date palm leaflets. All small scale string production,
be it from date palm leaf, veins, grass, date palm fibre, or papyrus epidermis is
done by hand without the help of instruments.
T w o bundles of material are rolled between the hands. The upper hand
(right) moves over the lower hand (left) in a rolling movement away from the
body: the fingertips of the right hand start moving over two bundles of fibre at the
wrist of the left hand and roll the bundles between the hand palms towards the
fingertips of the left hand (Figure 15-6). The two strands are z-spun by this
movement.
Figure 15-6 Making string with the hand rolling method

After each time the rolling movement is finished, the strand nearest to the
fingertips of the left hand is picked up between thumb and forefinger of the right
hand and put d o w n near the wrist, in front of the second strand (07:06-07:58). This
movement creates an S-ply. The result is a zS2 string.
In order to get an sZ2 strand the movement is m a d e in the other direction:
starting with the wrist of the right hand at the finger tips of the left hand and
rolling towards the body, until the finger tips of the right hand have reached the
wrist of the left hand (15:06- 15:46).6 With the right hand the strand nearest to the
wrist of the left hand is lifted over the second strand and put near the fingertips of
the left hand. Both strands are active: both are handled equally by the basket
maker.
The rim of the basket is strengthened by sewing on a separately prepared
three-strands rope in opposite direction (s-spun, Z-plied). T o m a k e an sZ3 strand a
third strand is spun and twisted around the sZ2 strand in a separate phase of action
(15:49-16:27). In this case the sZ2 strand is fixed at two points (passive), while the
third strand is handled by the basket maker and thus considered active. The rope
for the handles is made separately by folding the rope double and twisting it in a
cable twice (19:55-20:13 and 20:13-20:22).
Thus the most important instrument for rope making is the body itself. The
roughness of the hands makes the fibres to twist; the toes of the right foot hold the
end of the string while the basket maker is rolling the bundles (6:49); two feet in
between which the two-strands string is stretched, while the third strand is inserted
(15:49-16:27).
c) Sewing the centre

Before the start is made, the plait has to be soaked (6:20). This gives the palm leaf
its strength back and makes the plait supple and easier to handle. A s long as the
plait is moist, it can be pushed and pulled into the required shape. This is
especially of importance near the centre of the basket where the strip has to be
kneaded into a small circle.
A large needle (mesalla), of 220 m m long is used to which a length of
approximately 1.5 m zS2 string is fastened (Plate 15-7). In N e w Nubia the needle
used to sew the plaits is smaller (105 m m ) and isflat.A needle with a very similar
shape, only slightly longer (113 m m ) was found at Qasr Ibrim (Plate 15-8).
In Middle Egypt the basket maker starts by pushing the needle in the right
edge of the plait over a length of 120 m m . The needle is then strung through the
end of the string, where the ply of the string is twisted open slightly to let the
string pass. By pulling the string, the start of the plait forms a circle.
A n oppositely orientated strand (sZ2) could also be made in two other manners, by:
A: making the same rolling movement, but keeping the string to the side of the thumbs (left)
and the fibre bundles to the right. B: rolling with the left hand from the wrist to the finger
tips of the right hand, while the latter is kept still. However, none of the persons making
rope have been noted to do either of these.
15 M a k i n g M a t s , Bags a n d Baskets 301
Plate 15-7 Mesalla needle used in Middle Egypt to sew plaits. In this particular photo
the basket maker is sewing on the rim of the basket.
Plate 15-8 Sewn plaits needle found at Qasr Ibrim (10.413).
The needle picks up the right side of the plait, after which the second stitch is
made in the left side of the centre coil (9:13-10:07). This action is repeated until the
basket is finished. At regular intervals a n e w length of string has to be added.
At the start of the spiral the plait has to m a k e a small circle with the effect
that the right side of the plait is pushed together, while the left side of the plait is
stretched as far as possible. Both this and alternative starts of plaited baskets were
found at Qasr Ibrim: in s o m e baskets the start of the plait is folded over and sewn
as a flap on top of the left side of the plait (cf. Figure 9-4b, p. 159). The
orientation of the spiral is without exception anti-clockwise (the inside is e-
orientated)?
In N e w Nubia the plaits are not sewn with string, but with strands of d o a m
palm leaf of 3 m m wide and about 650 m m long. Whenever a strand runs out a
knot is m a d e to connect a n e w length of d o a m palm leaf. The knot used for this is
usually an overhand knot, but s o m e basket makers use reef or granny knots. For
The term e-orientated is preferred over anti-clockwise, because it is less ambiguous

(Wendrich 1991: 73). The basket shown in the video is also e-orientated. Indications of the
orientation of the spiral refer in general to the inside of a basket, or the underside of a lid.
This should be specified, because the spiral is reverse on the other side of the object.
mats a long plait is made, which is cut into strips of the required length. The str
are then sewn parallel to each other (Figure 9-4a, p. 158).8
d) Making the basic structure

The plaited strip is passive: it is less mobile than the active sewing strand. The
three components of the basket, the anchoring fabric (which is the part of the
basket which has been sewn), the active system and the passive system are all
handled by the basket maker, but the active system is handled more frequently: the
needle with the string never leaves the right hand of the basket maker (09:20;
17:54). The finished part of the basket and the plait which is fastened are touched
alternately to open the edges of the plaits to the right and the left thus enabling the
needle to enter.
The string pulls the two sides of the plait inside each other. Because of this,
the edges of the plaited strip are invisible and the plait pattern seems to continue,
which gives the fabric the appearance of being plaited continuously over the entire
height of the basket
Apart from his body weight and hands, the basket maker makes extensive
use of his feet. During the entire process of sewing the plait, the basket maker
holds the basket with his left foot (e.g. 16:33).
Timing the production of baskets and mats is difficult, because it is strongly
dependent on the size of the product and the fineness of the work. The size of the
basket is determined by the length of the plait, which is measured in bcfa (p. 180-
181). Sewing a basket of seven bacat takes about 5 hours.
e) Fastening the end of the plait

At a certain point the plait runs out. This is automatically the end of the spiral and
the rim of the basket. The end of the string is used to start the next phase:
fastening the rim decoration.
Mats, consisting of parallel strips, have simple plaited side-edges (the edges
of the two outer strips). The top and bottom edge are formed by the cut off ends of
the plaits.
f) Adding edges, rim and handles

The edges of mats, made of parallel plaited strips which are simply cut off to size,
have to be prevented from fraying. A narrow plaited strip (usually a nine-strands
plait), is sewn at the short side of the mat at a right angle with the sewn strips (cf.
Figure 9-4a, p. 158). Alternatively, ropes are sewn along the edges, as was found
several times in Qasr Ibrim.
The rim of the basket is formed by the edge of the last spiralling round of the
plait, but usually the rim is strengthened by stitches with a leather thong, or palm
Round or oval mats, sewn spirally from the centre, do occur in the Oases in the
Western Desert and in Sudan, but have not been found in either Qasr Ibrim or New Nubia.
fibre string. In Middle Egypt an sZ3 string, made of date palm fibre is sewn onto
the rim with the end of the sewing string (Plate 15-7). The palm fibre string
(passive), fastened with the active sewing string, is there to protect the rim. T w o
handles, made of rope, and prepared separately, are fastened at the edge of the
carrier baskets, the needle piercing the walls (21:04), and are anchored in the
fabric (20:41-23:50). The sewing string is cut off with a knife (22:41).
g) Finishing touch
At the end of the process, the basket maker spends some time removing pieces of
palm leaf which stick out of the fabric.
15.5 Coiling
Amarna (p. 182) Middle Egypt (p. 185)
••• •••
Qasr Ibrim (p. 217) N e w Nubia (p. 222)
••• •••
production phases coiled basket time estimate

a. preparation materials soaking palm leaf 30 minutes
c. start making the centre 10 minutes
d. basic structure stitching wrapping strand around the several days
bundle
e. finish fastening the end of the coil 1 minute
f. additions adding rim decoration 1 hour
a) preparation of the materials

Soaking restores the tension strength andflexibilityof palm leaf. The video gives a
good impression of h o w firm the strands can be pulled without breaking (see 27:27
for date palm leaf strands of 6 m m wide; 36:49 for doam palm leaf strands of
2 m m wide).
For coiling, different sizes of palm leaf strands are used. Coarse baskets are
made with strands of 20 m m wide, while for finely coiled baskets strands of only
2 mm wide are employed. These fine strands, when made of date palm leaf, are
cut from the dry leaves in the preparatory phase of the work. The d o a m palm leaf
strands, however, can be cut into smaller strands after they have been soaked, just
before they are used. They are often torn by hand after starting the tear with
anything sharp and pointed such as a needle, awl or simply one's fingernail (35:58-
36:00). This has two reasons: the boiled and soaked leaves are easier to manipulate
than the tough dry leaves, especially when very fine strips have to be cut.
Secondly, because the doam palm leaves take a lengthy preparation, they are
boiled in bulk. The precise application and thus the precise width of the strip is
decided afterwards.
b) Preparing the systems

Only in one basket a separately prepared bundle was used: from grass a string was
made which was then used as bundle material for a coiled basket, made with a
wrapping system of doam palm leaf.
c) Start of the systems

The start of coiled basketry has to make amends for a bundle which has to be
forced in a small circle at the start of the coil. This is not only a problem in coarse
baskets with thick bundles, but possibly even more so in very fine baskets, where
each irregularity disturbs the smooth appearance. There is quite a variety in the
start of the passive system in coiled basketry. In Figure 15-9 seven different kinds
of a start of coiled basketry are shown.
The first two are distinguished by tapering bundles, the other five have a
separate 'starting system'.
The 'snail' centre starts with a thinned out, pointed bundle, which has been
wrapped in the sewing strand, folded in a small circle and fastened by the first
stitch (Figure 15-9 a, see also Plate 4-4, p. 64). The 'rosette' centre is similar in
design, but the wrapping strand covers a larger part of the centre so that
irregularities are not visible (Figure 15-9 b). Most baskets in Amarna and Qasr
Ibrim are made with either of these starts. The 'snail' centre is also used by the
Kensi w o m e n in N e w Nubia. Twenty years ago they were making snail centred
baskets with wheat straw wrapped around a fruit stem bundle, at present the fruit
stem is wrapped with brightly coloured woollen yarn.
Sometimes the first few coils of a snail centre are covered with a leather
patch, which is fastened by the stitches of the next coil (Figure 15-9 c). This is a
feature that occurs regularly in Qasr Ibrim and still is present in N e w Nubia
(especially the Fadidja area).
Figure 15-9 Variations of coiled centres: a. snail; b. rosette; c. leather-covered; d. extra

coil; e. knotted bundle; f. plaited 2 x 2; g. plaited 4 x 4; h. plaited 6 x 6 .
One basket in Amarna has a separately added centre, consisting of a small

coil of d o a m palm leaf (Figure 15-9 d).9 The stitches of thefirstand second round
are put through this centre coil. The grass bundle is inserted after the second coil.
Although the use of a separate centre coil has been attested only once, several
baskets at A m a r n a probably had a similar centre. This can be concluded from the
baskets with a smooth round hole in the middle, where the separate coil seems to
have become detached and fallen out.
In present day middle Egypt a different solution is found. The basket maker
shreds palm leaf strands into smaller strips and makes a simple overhand knot in
the bundle. The knot is used as the base for the first round of stitches. The ends of
the bundle protruding at two sides from the knot, are folded to one side and form
the start of the bundle (25:08-25:45). After a while the shredded fibres of the fruit
stem of the date palm are inserted in the middle of this bundle of palm leaf (Figure
15-9 e).
The Fadidja w o m e n of N e w Nubia are making a centre that occurs frequently
in Qasr Ibrim also: a separate start is made in the form of a small plait (Figure 15-
9 f; see also Plate 4-5, p. 65; 34:11-35:46). In Qasr Ibrim this centre is used
mainly for baskets decorated with wrapping patterns (see below), the coloured
baskets usually have a snail centre. In N e w Nubia, however, the 2 x 2 centres are
used for the large coloured food covers. Rarely, baskets in Qasr Ibrim are made
with more complicated plaited centres (4 x 4 or 6 x 6, cf. Figure 7-15 g and h).
The strands sticking out of the plait are folded parallel to the sides of the plait and
are incorporated in the bundle. In some instances the centre of the basket has
disappeared, but the quare shape of the hole indicates that the centre has been
plaited. Making the centre does not require any tools.
d) Interaction of the systems

The smooth appearance of m a n y Qasr Ibrim baskets, as well as modern Fadidja
baskets, is made by only picking up the winder of the previous round (Plate 15-
11). A n awl is used, with which a small hole in the previous row of winders is
m a d e (37:33). A narrow strip of d o a m palm leaf is pushed through the hole,
without the use of a needle. The stitch thus does not pick up m u c h of the bundle
material.
In A m a r n a and present day Middle Egypt the basic structure of coiled
basketry does not show any variety. The bundle is fastened in a coil by making
stitches with the wrapping strand through the previous bundle. This is a strong
method of making a coiled basket. The deeper and more irregular the stitch is put
in the previous bundle, the coarser the appearance of the fabric. The winder is
passed through the bundle with the help of a needle (26:58) or both an awl making
a hole through which a needle can pass (29:37).
9
T A 0657, this is the same basket which has the deviating bundle, made of grass
string.
The variety of making the basic structure is much larger at Qasr Ibrim than at any
of the other sites involved. In one basket from Qasr Ibrim, the stitches are very
widely spaced, so that the bundle material is clearly visible (Figure 11-9, p. 218).
The modern parallel is a very recent innovation: basketry made by Kensi w o m e n ,
w h o have started to use alternative materials for d o a m palm leaf. By lining the
bundle with the coloured wrappers of chocolate wafers and fastening these with
widely spaced ribbon, the coloured pattern of the basket is determined by colour of
the passive system, rather than the colour of the active system (winders). This is a
complete break with the manufacturing tradition in order to adhere to the tradition
of lavishly decorating the basketry (Plate 11-14, p. 224).
The use of coloured stitches is very c o m m o n in Qasr Ibrim and N e w Nubia
alike. A depiction of baskets with coloured patterns from the tomb of Rekh-mi-Re
at Thebes shows that in the N e w Kingdom Period even figurative decorations were
made (Davies 1937, Plate 13). D e Garis Davies suggests that these baskets are part
of the tribute from the Oases or Nubia. Judging from the present basketry
traditions the latter seems the more likely.
There are links between the type of centre and the type of wrapping pattern.
The coloured baskets from Qasr Ibrim have either a small rosette or a snail centre,
while the baskets decorated with uncoloured patterns of stitches have a 4x4 or a
6x6 plaited centre. In present day N e w Nubia the small coaster-like plates m a d e by
the Kensi have a snail centre (cf. Plate 11-13, p. 223), while the Fadidja's make
large decorated food covers which are started with a 2x2 plaited centre (34:11-
35:46).
The insertion of new bundle material is done gradually. The diameter of the
bundle is usually slightly smaller near the centre, reaching its standard width after
two or three coils. After every few stitches a small quantity of bundle material is
pushed as deep as possible in the middle of the bundle. In Europe rings of c o w
horn are used to gauge the thickness of the bundle, but the use of such a tool has
not been observed in Egypt, neither is there any archaeological evidence for it.
The thickness of the bundles differs greatly. In Middle Egypt the bundles range
from 10-20 m m , while in N e w Nubia very fine basketry is m a d e with bundles of
only 2 m m in diameter.
The wrapping strand is approximately 650 m m long and, therefore, a n e w
strand has to be fastened after every few stitches. In Middle Egypt and N e w Nubia
this is done by pushing the end of the old leaflet aside in the bundle, while a n e w
strand is layed in underneath the bundle (30:51-31:11, cf. 38:48-38:50). In A m a r n a
no clear examples of insertions have been registered. In Qasr Ibrim the insertions
found were more elaborate: the old and the n e w strands were twisted around each
other, before the old strand was worked into the bundle. The method of changing
colours seems to have been the same in Qasr Ibrim, as it is today in N e w Ibrim:
the wrapping strand with the colour which should not appear in the pattern is
hidden with the bundle inside the wrapping stitches of the current colour (39:34-
39:43).
Giving a time estimate for the production of the base and sides of a coiled
basket is exceedingly difficult, because it all depends on the size, decoration and
finesse of the stitch.
e) finishing off the systems

The end of the systems in coiled basketry is formed by the last winding stitch
which holds the end of the bundle. Often the bundle is thinned out in the last coil.
In badly made baskets the end of the coil is simply cut off, after the last winding
stitch, so that the stumps are visible.
M a n y Qasr Ibrim baskets, have rims which are a continuation of the last coil
of the basket. With this coil often one or two small eyes are formed, which are
meant to hang the basket on the wall (Plate 15-10).
m m m \
Plate 15-10 Small coiled plate from Qasr Ibrim (10.480), with coloured winders, a
coiled inlay rim and an eye for hanging the basket on the wall.
f) adding other features

In one basketry fragment from Qasr Ibrim the coiled centre is covered with an
intrinsic stitched pattern (Figure 15-11). The cover was probably added after the
basic structure was finished and is similar to embroidery. This is in contrast to
leather centre covers, which are fastened with the stitches of the next row of
coiling.
Usually leather-covered centres go hand in hand with rims covered with a
lashing of leather thongs. Rims of coiled basketry are often decorated and
strengthened. W e have no information for the Amarna rim decoration. In present
day Middle Egypt separate stitches are sewn on top of the rim. These rim
decorations are added after the coiling is finished. Other additions to coiled
basketry are an extra coil on the inside of the basket, which form a support for the
rim, handles and feet of the basket. These features are known from N e w Kingdom
basketry from Deir el-Medina, but could not be attested in the fragmentary
remains at Amarna.
Figure 15-11 Plaited cover of the centre of a coiled basket, which has been applied after
the basket was finished.
15.6 W r a p p i n g
« M h Amarna (p. 186) Middle Egypt (p. 189)
••• ••
••
production phases brush with wrapped handle time estimate

b. preparation systems making string 10 minutes
c. start fastening wrapping strand 1 minute
d. basic structure wrapping strand around the bundle 15 minutes
e. finish fastening the end of the strand 1 minute
f. additions -
g- finishing touch - ;
• M w
Figure 15-12 Construction of the brush from Plate 10-7 ( T A W V 0295). A bundle of palm
leaf is tied around a wooden peg, then the lower part is folded back over the
upper half and fastened by wrapping the string around the handle part. The
brush is 237 m m long. Left hand drawing by B. Garfi (Courtesy of the
Egypt Exploration Society).
T h e passive system, that is the sticks or reeds and the bundles that form the
cores of brushes and rings, do not require any preparation, but the active system
is usually either twisted d o a m palm leaf, grass or string, all of which are m a d e in
a separate production phase. Making enough string for wrapping one ring or
brush takes very little time, probably not m o r e than a couple of minutes.
Brushes are wrapped only partially. Usually the core is formed by a bundle
of fibrous material such as grass or shredded palm leaf. This bundle is tied just
under the middle. In the brush depicted in Figure 15-12 a w o o d e n stick is tied in
the middle of the bundle of shredded d o a m palm leaf. The bundle is folded back
and held in place by wrapping the end of the tying string several times around the
core. The brush is divided in three 'fingers'. In this way the round bundle is
shaped into a flat brush. This is only done with large brushes. Usually the
wrapping is ended with a simple knot.
T h e time involved in making rings and brushes was probably not more than
30 minutes per object. N o instruments were used, except perhaps a knife to cut
off the end of the string.
Three-ply rope is another form of wrapping: a two-ply string is kept under
tension, while a third strand is spun and plied around it. This is the way that
three-ply rope is made today (16:14), but in the archaeological material there are
indications that this was the way three-ply rope was made in A m a r n a and Qasr
Ibrim as well. Several three-ply ropes have a little sprig of palm leaf or grass
which has been stuck into the loop of a two-ply string, suggesting that the third
ply was added in a separate work phase.
15.7 KNOTTING
Pioduction phases brush with knotted handle time estimate

b. preparation systems making string 10 minutes
c. start fastening knotting strand 1 minute
d. basic structure knotting strand around the bundle 10 minutes
e. finish fastening the end of the strand 1 minute
f. additions - -
The knotted brush found at A m a r n a is m a d e with S-twisted d o a m palm leaf, which

has to be prepared in advance. T h e brush is started in exactly the same w a y as the
wrapped brushes, but after folding the bundle of d o a m palm leaf back the S-
twisted strand is knotted around the core of the brush fibres, keeping the fibres
together, shaping the brush and forming the handle (Figure 10-9, p. 190). The
brush is not divided in fingers. The end of the string is tucked in the last knot and
cut off. Thus, the only tool used in the entire process is probably a knife.
15.8 Weaving with one strand
Amarna (p. 191) Middle Egypt (p. 192)
••• •
••
production phases grass woven mat time estimate

a. preparation materials wetting grass 5 minutes
b. preparation systems making string 4 hours
c. start preparing loom, setting up the warp 2 hours
d. basic structure weaving in the weft 4 hours
e. finish taking the mat from the loom and 30 minutes
making the edges
f. additions - -
g. finishing touch pulling the warp tight after drying 30 minutes
The grass for woven matting such as found at Amarna needs little preparation, five
minutes soaking is sufficient. Rushes need a longer period of soaking,
approximately a day. For this kind of matting a simple loom is used. Such a loom
is known from a tomb painting of the Middle Kingdom period (Figure 15-13). It
consists of four pegs in the ground and three cross bars. T w o at the end, which
hold the warp threads and one in the middle which the weaver is holding with two
hands.
The Beni Hassan mat weaver weaves a blocked pattern with coloured
strands, probably rushes which are either natural shades (yellow and green) or
dyed. The weaving pattern is \2/2\\2 (a plain weave over two warp threads) which
gives the block pattern. The weaver sits on the finished part of the mat, which is
remarkable, because his weight would push the mat to the ground and disturb the
tension of the warp strands. This would make it difficult to weave in the weft.
JflMjJMui oiihiimi»»HIiiu i y » n
'I L-t-UH j • I '.-.H-i—> I i I I ! i I
InlTWlYTOiffTli)in'nlTl 11) i^JwuflO,
Figure 15-13 Mat weaver from a tomb painting in the Middle Kingdom tomb of Khety at
Beni Hassan (Drawing by A.M. Hense).
Observing present day mat weavers helps us to understand how this could have
worked. At present the mat weavers are also sitting on the finished part of the mat,
but they prevent that their weight pushes d o w n the mat to the ground by putting a
wooden board underneath the finished part of the mat, which is raised on two
stoned. This divides the weight of the weaver and keeps the weft under equal
tension.
Figure 15-14 gives a schematic drawing of a present day mat loom as seen in
the Upper Egyptian village of Nagada. It consists of four pegs in the ground (A/B
and L / M ) to which two cross beams are fastened (E and I). They are tied to the
pegs with four lengths of rope (C/D and J/K), which allows the mat maker to keep
tension on the warp threads. The mat-maker sits on the finished part of the mat,
which is lifted from the ground by a board on top of two stones (F). In front of
him there is a heavy cross beam, through which the warp is running (G). A bundle
of material is put in front of the cross beam, across the warp threads, within easy
reach and ready to be used (H).
The cross beam (G) has a different function than the ones in horizontal looms
for textiles, which are used to make certain that the width of the cloth is kept the
same. Since mats are not woven with an ongoing weft, which, if made with too
much tension will gradually pull the edges inward, there is less danger of uneven
sides.
p f/
[, 1— - L i *
T T
A C [• - — - J L
—
l = ^ _ _
[ ~^r~—-
B D K M
G H
j ^ z r
B E /////////////////////s,
G
S/////S///////S/////
Figure 15-14 Schematic drawing of the mat loom as used in present day Upper Egypt
(Nagada).
The heavy cross beam (G) has a double function: spacing the warp and beating in
the weft. After weaving in two or three strands the mat maker pulls the beam with
force towards him, so that the weight of the beam pushes the grass or rushes into a
compact mat.
The mat maker leans forward to take a strand from bundle H and weaves it
in through the weft just in front of him. H e moves each rush stem or small bundle
of grass up and d o w n through the warp. In one movement, with aflickof the hand
the strand is pushed under a warp strand and pulled up and over it. In this type of
weaving the distinction between passive and active systems is clear: the mat maker
touches mainly the weft strands with which he or she weaves. Plant stems are not
completely regular in shape: the foot is thicker than the top. It is this thickest most
rigid part, which is held by the mat weaver and pushed up and d o w n through the
warp, while the top end follows. The Amarna mats are woven alternately from left
to right and from right to left
After having observed the modern mat loom, it becomes m u c h easier to
understand the details of the mat loom Beni Hassan tomb painting. There, the loom
also exists of four pegs and two cross beams holding the warp. The third cross
beam is depicted as if it is slightly toothed, which might be an indication of the
holes through which the warp threads run. By extensive use, these holes are worn
into slits. It seems that the last cross beam (equivalent of beam I in Figure 15-14)
has the same tooth marks, which might suggest that the end beam also has a row of
holes through which the warp runs.
In the painting one strand has been woven in and the mat weaver has taken
hold of the heavy cross beam to push it into place. O n the other side of the beam
new strands are ready to be woven in (cf. H in Figure 15-14). The side edges have
not been made yet: the ends of the woven strands (probably rushes) are still
sticking out of the weave.
The warp is nowadays made of industrially prepared cotton yarn and has an
odd number of threads. The two outer strands are usually double or thicker than
the rest of the warp, to give more strength to the side edge. The length of the plant
stems determines the m a x i m u m width of the mats. The grass of the side edges of
the A m a r n a mats is folded around the outer warp strand and held in place with
stitches (Figure 15-15a and Plate 10-10, p. 191). The fragments of two mats made
of papyrus and wheat straw have edges which are twisted around the outer warp
string and back into the weave (Figure 15-15b and Plate 12-3, p. 253).
Figure 15-15 Side edges of Amarna woven matting (not to scale). Left: side edge of grass
mats (folded and S-twined, cf. Wendrich 1989, 188). Right: side edges of
mats made of bundles of papyrus or straw (woven back into the mat).
316 The world according to Basketry
Figure 15-16 Top edges of Amarna woven matting. For edge a. the loom beam has been
taken out and the loops are intact. Edges b-f are made of warp loops which
have been cut. Drawings not to scale.
The top and bottom edges of the Amarna grass woven matting show a large
variety of which two are shown in Figure 15-16. When the mat isfinishedit is
taken off the loom. There are two ways of doing this: either the warp loops which
run around the cross beams are cut, or the cross beams are slid out of the loops.
The first method results in loose ends of string, the second in a row of loops.
There are practical reasons to choose for the method of cutting the loops. In some
cases it is simply not possible to slide the cross beams out of the fabric, for
instance if the beams are very heavy, or irregular. But even if the cross beams are
smooth and can be lifted easily, a work space which is not much wider than the
mat itself lacks the space to move the cross beams and take the mat off. In
Amarna both methods occurred. Even taken into account that the top and bottom
edge in a mat can be different, from the variety if these edges it seems likely that
there were mats from several producers at the Amarna workmen's village,
supposing a mat maker will use one particular system of finishing off his mats.
Rush mats, which are woven with quite wet strands, have to dry for a
number of days after the mat is take of the loom. Then the weft, which shrinks
during the drying process, is tightened by pulling the strands with a metal hook.
Furniture webbing is woven directly on the bed or chair frame. In Qasr Ibrim
remains of bed matting were found made of S-twisted strands of doam palm leaf.
The warp was made over the full length of the bed, but because the bed matting
that was found had been re-used as pit lining, it is not exactly clear h o w the weft
was inserted. The edges appear to have been cut, which would indicate that the
matting has been woven alternately from left to right and from right to left,
looping around the side of the bed frame. Because the string had to be pulled
through the warp with each passing, either a weaving shuttle was used, or, more
likely, only short strands were used, adding a piece after each passing of the
warp.
The time involved is one full time day for weaving a mat of medium size
(1.20 m wide and 2 m long). Instruments involved are a loom, a knife and a metal
hook, with which the weft is tightened after the mat is finished. A specific work
space is required. This can be any flat surface, either inside or outside, but
preferably in the shade. This work space is often, but not necessarily, permanent.
The only fixed points are four pegs which are put in the ground and as long as the
cross beams and the warp are not in place the space can be used for other
activities.
15.9 Weaving with Two Strands
Qasr Ibrim N e w Nubia (p. 231)
•• •
production phase grass bed matting time estimate

a. preparation materials wetting grass 5 minutes
c. start setting up the warp in the bed frame 2 hours
d. basic structure weaving in the weft 4 hours
e. finish fastening the end of the weft 5 minutes
f. additions - -
g. finishing touch - -
Grass for the rope only has to be wetted (48:47-48:53). T h e preparation of

the string takes up the largest part of the production process. Weaving with two
strands differs from weaving with one strand, because the work is done from one
side of the frame. Loops of string are woven in, which results in two parallel weft
strands. The side edges are made by pulling the loop through the previous one,
which is similar to crocheting (54:11-54:19).
The tools used are three tapered wooden pegs and a beating stick. The video
sequence on bed making shows the use of these. The pegs are pushed through the
ends of the loops, to prevent them from slipping back. The peg is removed
temporarily when the end of the next loop is pushed through. The peg is replaced
in the n e w loop. For this work the mat maker apparently prefers a smooth peg
over a rough one. In the beginning of the video sequence Shahed stops to change
the pegs (51:39). The weft is regularly tightened by beating with a stick (53:11-
53:16). Towards the end, when the bed is almost filled with string, there is no
space to wield the beating stick and a peg is used to tighten the weft (56:30-57:00).
At the very end of the process the pegs are used as awls to make space in the
fabric to put in the last ends of the weft (57:17-57:23).
The mat is finished off by pulling the end of the weft string through the two
remaining loops of warp and weft (58:00). The time needed to m a k e the string and
to fill one bed frame with matting is approximately 12 hours. The work space
needed is the size of the bed plus a metre on all sides. This space is not
permanently taken up by the work, only w h e n a bed is made.
15.10 Twining
Amarna (p. 194) Middle Egypt (p.201)
••• ••
•••
production phases closely twined mat time estimate

a. preparation materials soaking palm leaf, wetting grass 30 / 5 minutes
c. start setting up the warp 30 minutes
d. basic structure twining in the weft 4 hours
e. finish taking mat off loom, finish the edges 10 minutes
f. additions
g. finishing touch shaking out the dust 5 minutes
The time estimate given above is for a mat, two of which are sewn together to
form a bag. Fragments of similar matting have been found at Amarna. In el-
Amariyya these twined matting bags are m a d e of date palm fibre, which is a
strong, quite course material (42:29). The fabric is made on a horizontal loom,
which is even more simple than the woven matting loom: just four pegs in the
ground with two cross bars (Figure 15-17). The tension of the warp cannot be
adjusted, because the loom bars are simply hooked behind the pegs.
Figure 15-17 Schematic drawing of a loom for twined matting. At a the end of the warp
string continues as woven starting border. The same strand becomes twining
strand 1 (the stippled strand at b). The start of the warp string (which was
temporarily knotted around the bar) is untied and becomes twining strand 2
(the hatched strand at b). The letters c and d indicate respectively the right
and the left side edge. Not to scale.
After the string has been made by the hand rolling method, the warp is set
up. The start of the warp strand is temporarily tied to the metal cross bar (detail
drawing in Figure 15-17). The end of the warp string (the stippled strand at a in
Figure 15-17) is woven back to the beginning of the warp and used as one of the
twining strands. The second twining strand is formed by the start of the warp
string (the hatched strand at b in Figure 15-17). The two strands are then
alternately pulled around the warp strands. W h e n a string is exhausted, a new
length is added by pulling the small tip of the old string through the loop at the end
of a n e w length of string (42:22). The side edges are formed by wrapping both
strings around the outer warp string (c and d in Figure 15-17).
In Middle Egypt the top edge consists of loops which are linked into each
other (Figure 15-18, 46:33-47:20). N o top edges of closely twined matting have
been found at Amarna, to decide if the ancient matting was m a d e on a similar type
of loom.
Figure 15-18 Top and bottom edge of twined matting, made by linking the warp loops in a
chain. The arrow indicates the working direction.
Twined matting is depicted very commonly in tomb paintings (Figure

15-19). The standard mat, a striped rectangle with a number of crossing bands,
seems to be a schematized representation of openly spaced twined matting, which
is a flexible type of twined matting made of bundles of fibres (grass or palm leaf
have been attested in Amarna and Qasr Ibrim), fastened with widely spaced rows
of twining with string.
Figure 15-19 Depiction of matting in tomb paintings. Horizontal stripes represent bundles
of grass, palm leaf or papyrus, vertical stripes the rows of twining.
Figure 15-20 Openly spaced twined matting from Amarna. The bundles are usually grass
or doam palm leaf. The twining is done with zS2 string made of grass or
doam palm leaf. The side edges are characterized by a zigzag line to the next
row of twining.
In Amarna the rows of twining are from 35 to 100 m m apart. The weft is twined
from left to right, zigzag d o w n at the side edge and then is twined from right to
left (Figure 15-20). The widely spaced twined matting probably was not m a d e on a
loom, but fixed on one side only on a set up string.
F r o m the Old Kingdom onwards, the production of matting is depicted
regularly in tomb paintings. Judging from the representation of the mats these
scenes represent the production of openly twined matting such as shown in Figure
15-20. The mat making scenes are mostly related to shepherds w h o work in the
papyrus fields. O n e of the more detailed examples is found in the tomb of Ti in
Saqqara and shows two scenes in two registers (Figure 15-21).
The upper register depicts a m a n laying out strands of papyrus. The lower
register shows to m e n facing each other. They are working away on what seems to
be afinishedmat (the rows of twining are in place), holding dumbbell shaped tools
in both hands. The appearance of this tool suggests it to be a pounder or, more
likely, a small brush. This would indicate a possible end stage of the matting
process: pounding the fibres or cleaning the finished mat from loosefibres.It is an
enigmatic scene, because there are no indications in the extant matting that
pounding of fibres took place. Brushing the mat was perhaps part of the process,
but it would have been the finishing touch, rather than an essential action from a
technological point of view.
Figure 15-21 Mat making scene from the tomb of Ti in Saqqara (after Wild 1953, plate
CXXIV).
Usually the tomb scenes represent the production phase that is the most
'typical' for a production process. This would be the action that busied the mat
makers for the longest period and the kind of action that comes to an outsider's
mind, w h e n asked what a mat maker does. While twining a mat most time is spent
on laying out the fibre bundles for the passive system and connecting these with
the active system (the twining proper). The scene from the tomb of Ti shows the
first phase of laying out the bundles of papyrus in the upper register. If the lower
scene represents part of the actual twining phase, then the dumbbell shaped objects
probably had a function in twining, such as spacing or tightening the rows of
twining. They would then have to be interpreted as wooden combs or stiff brushes
with which the weft was pushed in place.
Apart from twined matting, there are twined sieves and, in Qasr Ibrim, also
baskets. These have in c o m m o n that the passive elements are rigid and can be
made without a loom. Twining is often combined with other techniques. The
sieves, for instance, have a twined grid fastened in a coiled rim.
Only few tools are used in twining: a knife; perhaps a tool to space the rows
of twining and, for closely twined mats, a loom. The work area needed is the size
of the mat and some space around it. A s with the looms for weaving, the fixed
parts of the twining looms are not more than four pegs in the ground. The space
around and in between these pegs, is available for other activities w h e n no mats
are being made.
15.11 W a l i n g (11)
Amarna Middle Egypt
Giving a time estimate is not relevant since waling does not occur in Qasr
Ibrim as a technique in which a complete basket is produced, but only as part of
the production of a basket with a twined base and plaited sides.10 N o tools are
needed for waling.
10,
In Berenike in the early Roman garbage dump a basket is found in which waling
plays an important part. It has four rows of decorative waling. The basket is made of very
fine willow osiers, which is indicative for the probability that this basket was imported from
the north part of the Mediterranean. The publication of this basket is forthcoming in the
report on the Berenike 1998 excavation season.
15.12 Pierced Basketry
Amarna Middle Egypt (p.201)

J s i d t t
••• |
^ytiit* IjjJ> - lyr*" ~m*~^
^ j j - L_ "liil" _hj; ; E' Qasr Ibrim New Nubia .
ST
a US3 § ?
production phases qafas (orange crate) time estimate

a. preparation materials - -
b. preparation systems cutting sticks, making holes 10 minutes
c. start setting up vertical corners and 5 minutes
horizontal cross bars
d. basic structure hammering in the stakes of the sides 10 minutes
and the base
e. finish hammering in the last stake -
f. additions - -
Today the pierced technique is used for all tomatoes and oranges are transported in
qafas crates m a d e in a form of pierced technique, which offers an interesting
contrast at several levels. Both the passive and the active elements are rigid: the
qafas maker takes pointed sticks, m a d e from the split mid rib of the date palm and
h a m m e r s them into a frame of the same material. In this frame holes have been
punched beforehand (Plate 15-22).
The same producers m a k e also a large flat bat-like implement with one
handle, a matraha, used for spreading out dough to form a thin round sheet of
bread which is flung from the tray onto the oven floor to be baked. The matraha is
also m a d e of pre-punched palm midrib, but active and passive are reversed: the
stakes are set upright and the punched parts are h a m m e r e d onto them, (cf Plate
10-17, p. 202)
T h e time involved in making a qafas is about 25 minutes, of which almost
half is spent in the preparatory phase of cutting and punching the sticks. The
workshops have an industrial organization. O n e person cuts and splices the mid
ribs, the next person punches the holes, the third person makes the frames, the
fourth person h a m m e r s in the sticks of the walls. T h e instruments used are large
knifes to cut and splice the date palm midrib, metal punch-awls in several sizes,
hammering sticks and measuring sticks to determine the position of the holes for
each different type of qafas.
Plate 15-22 Qafas maker, hammering pointed sticks into pre-punched holes.
15.13 Rope
••• •••
••• •••
The hand rolling method of rope making has been explained in section 15-4 (cf.
Figure 15-6 on p. 299). Spinning and plying are done in the same movement and
usually in opposite directions: spinning in a z-direction (clockwise), plying in S-
direction. Thus the opposite orientation of spin and ply are locking each other and
provide the coherency of the string. N o instruments, other than the h u m a n body
are needed although there are indications that large scale rope making, for
shipping, for instance, was done with the help of a twisting device with weights at
the end (Teeter 1987).
15.14 Continuous Plaiting
Amarna Middle Egypt
Qasr Ibrim New Nubia (p. 241)
• •
production phases 'disposable bag time estimate

a. preparation materials splitting and soaking palm leaf 35 minutes
c. start setting up the edge 1 minute
d. basic structure plaiting the fabric 5 minutes
e. finish finishing off the edge 2 minutes
f. additions putting in the closing string 2 minutes
T h e time estimate for the coarse disposable bags found in Qasr Ibrim is 10
minutes per bag, not counting the preparatory work, which takes 30 minutes of
soaking and an estimated five minutes to split the date palm leaflets. The end with
which the leaflets once were connected to the mid rib is left intact (Figure 11-25 p.
239). T o plait fans of narrow strips of date palm leaf, in a twill pattern (\2/2\\l),
would take around three hours. T h e fans are approximately 300 x 2 0 0 m m . N o
instruments are needed for making plaited basketry.
15.15 Plaited Strips
Amarna Middle Egypt
• •
The production of plaited strips has been highlighted in section 15.4 (pp. 296-298).
15 Making M a t s , Bags and Baskets 327
15.16 Knotting
•
production phases amphora carrier net time estimate

a. preparation materials soaking papyrus stems 30 minutes
b. preparation systems making string and bottom ring 3 hours
c. start making the handles 10 minutes
d. basic structure knotting the net 3 hours
e. finish fastening net to bottom ring 20 minutes
f. additions putting in the closing string 2 minutes
Papyrus string is m a d e in separate production phase. The amphora net is started by

six strings, three strings cabled together to form a handle (cf. Figures 10-19, 10-20
and the bottom ring Figure 9-6). T w o strings run d o w n from the two handles
which are knotted to form a fabric. Finally the wrapped rope ring is added with
large overhand knots. The time indications above were m a d e on the basis of a copy
of this net."
The shrinkage of the wreath, seems to indicate that it was m a d e of one
young, not entirely dry date palm leaf, which was split along the mid-rib. The
wreath was started by tying the two halves of the split palm leaf mid rib into two
parallel circles, forming the frame. Then the fabric was m a d e by knotting the side
leaflets of the two halves (cf. Figure 11-29, p. 244).
For knotting no instruments were needed. The time depends on the size of
the knotted fabric, but the estimate for the amphora sling proved to be accurate,
compared to the experimental knotting of the net (for which the papyrus string was
imported from Egypt), taking into account that the person w h o knotted the net,
although an knotting expert, had never m a d e an amphora net before.
The copy was made by P. van der Griend for an exhibition on ancient packing
materials in the Allard Pierson Museum in Amsterdam ("De Oudheid Verpakt" 10 October
1997 to 1 February 1998) (cf. Wendrich 1997).
15.17 Coiled Inlay
Amarna Middle Egypt
Qasr Ibrim New Nubia (p. 245)
• ••
A time estimate for coiled inlay, which is mainly a decoration technique for coiled
basketry, has been incorporated in the section on coiling (pp. 303-309).
15.18 Looping
Amarna Middle Egypt
••
production phases sinda cheese mat time estimate

a. preparation materials wetting palm fibre, shredding fruit 30 minutes
stem fibre
b. preparation systems making string 1 hour
c. start fastening the top rim piece 10 minutes
d. basic structure making five tying rows the fibres 3 hours
e. finish fastening the bottom rim piece 5 minutes
f. additions making handles 15 minutes
g- finishing touch smoothing the mat, by taking off 10 minutes
fibres that stick out
The cheese mat is m a d e in farming households, which m a k e their o w n cheese. The

process takes at least half a day. T h e top and bottom of the mat are m a d e of a split
piece of palm mid rib. In between the fibres are fastened with a n u m b e r of widely
spaced rows of looping which create enough space between the fibres to let the
moist of the fresh cheese drip out, but not enough to loose the actual cheese
matter.
m
-Apr "• • -
Plates 15-23 and 15-24 Woman making cheese and cheese drip mat (el-Amariyya 1992).
» f c ^
15.19 Binding
••
•• ••
production phases roof time estimate

a. preparation materials soaking doam palm leaf 30 minutes
b. preparation systems making s-twisted binding string 1 hour
c. start laying out palm ribs 30 minutes
d. basic structure binding ribs to the cross ribs 1 hour
e. finish fastening off the binding string 5 minutes
f. additions - -
g- finishing touch putting on old matting and a layer of 1 hour
mud
T h e time estimate for making a roof or screen of approximately 2 x 3 metres is a

bit over three hours. First the binding string has to be m a d e by twisting d o a m palm
leaf. T h e n a number of three or four palm midribs are layed out, parallel to each
other with a space of half a metre in between. Perpendicular to these closely
spaced palm ribs are placed. Each of these is tied to the cross ribs with d o a m palm
string (Plate 11-32, p. 247). T h e fastening off of this binding string marks the end
of the process. T h e roof then might be covered with old matting and a m u d layer.
15.20 Plaiting around a Core
._ .. _
Amarna Middle Egypt
The fragment found was too small to conclude if new plaiting strips were inserted,
nor could be inferred how the plait was ended.
C H A P T E R SIXTEEN
T H E C H O R E O G R A P H Y OF W O R K
W o r k has its o w n rhythm, its o w n movements, its o w n pace, its o w n use of space.
These are the aspects that this chapter focuses on. The purpose is to organize the
production process in yet another way, putting emphasis on different aspects, to
give room for new perspectives. The use of video enables this.
A n archaeologist in trying to understand ancient production, will tend to
focus on the most obvious parts of the process: the repetitive actions of making the
basic structure. Weaving is thus reduced to putting the warp in the weft, coiling to
stitching the bundle with the wrapping strand. Watching and working with basket
makers teaches that a process is m u c h more than making the basic structure (cf.
the seven phases a-g in Chapter 15).
By analysing several aspects in the video one is forced to go beyond such an
artificial, linear organization of the production process. This chapter will focus on
the working order, the production time, the working rhythm, the use of the body
as instrument, and the movements around the work space.
16.1 The maqtaf
sewn plaits basketry

made by M o h a m m e d in el-Hagg Qandil
video sequence 02:53-24:04
16.1.1 Working Order

Compared with the order presented in Section 15.4, the sequence registered by
video, follows a different, more complex order:
(a) preparing the material (03:32-03:46).
The basket maker sorts strands of palm leaf. Only the very end of this action is
shown, in which the basket maker has a stack of date palm leaf strands lying to his
right.
(b) preparing the passive system: plaiting a strip (03:47- 06:35).
starting the plait (03:47-04:10).
This is shown completely in the video sequence and takes only 24 seconds.
making the plait (04:11-06:35).
A period of approximately four hours is spent on making the plait. In the
video representation this has been shortened to only one minute and 49
seconds. The finishing off of the plait does not occur until much later
(16:57). In preparation of the start of the work, the plait is rolled up and put
in an aluminium pan with water to soak. Meanwhile, the basket maker starts
the preparation of the string.
(a) preparing the material (06:36-06:56).
The veins have been put in water to soak, while the basket maker was making the
plait. H e takes the material out of the water and loosens the fibres by moving them
forcefully around for 10 seconds.
(b) preparing the active system: string (06:57-07:58)
starting the string (06:57-07:24).
Starting the string takes 27 seconds: the basket maker sorts and arranges a
small bundle of veins, folds the bundle double and starts twisting the two
sides of the bundle into two different strands of the string, while plying them
in the same movement. H e then takes off his shoe and holds the start of the
string between his toes.
making the string (07:15-07:58).'
Only one minute and two seconds of this phase are shown, while the actual
time spent is much longer, approximately 30 minutes.
(c) starting the basic structure of the basket (08:05-10:50).
In preparation of the sewing of the basket, the basket maker takes a large needle and
runs the string through the hole, takes up the plait and starts sewing. The entire
sequence is shown in real time.
(d) making the basic structure (09:49-18:20).
The basket maker runs his needle alternately through the edge of the basket and
through the edge of the plait. This is an ongoing movement that lasts several hours
until the basket is finished. In the video this sequence is represented by eight minutes
and 31 seconds. During this part of the process the basket is shaped and regularly new
lengths of string are added to the active system (11:07). Within the sequence, several
things happen, which belong to different phases of the production process:
preparing raw materials
Some date palm leaf veins are left to soak (12:47).
preparing the active system (13:02).
The soaked veins are used on the spot, to lengthen the sewing string after it
has run out. The amount of string prepared earlier by the basket maker
during the time it took the plait to soak was not enough to finish the entire
The frozen image of string making is shown in video sequence 07:39.

16 The Choreography of Work 333
basket, so at a certain point he starts making string, rather than inserting

ready-made lengths of string.
preparing raw materials for the rim and handles (14:25-15:04).
Date palm leaf sheath fibre is soaked and small tufts are made in preparation
for the string production for rim and handles.
production of string for rim and handles (15:05-16:28)
The basket maker proceeds making string from the date palm fibre. This
string is rolled in an opposite direction to the string with which the basket is
sewn. H e makes a two-ply string, which he stretches between his toes. The
fixed, passive, string is then made into a three-ply string by adding a third
strand (the active element) (16:14).
making the basic structure (16:33-16:55)
The basket maker continues sewing the plait.
finishing the passive system (16:49-17:23 and 17:07-17:24)
In the meantime a friend finishes the plait.
(e)finishingthe systems of the basic structure (17:59-18:20).
The basic structure of the basket is finished with sewing the last part of the plait.
Although the passive system (the plait) is now finished, the active system continues.
The basket maker adds a new length of string.
(0 adding other features (18:20-24:04)
With this new length of string the basket maker sews on the palm fibre rim.
preparing the systems (19:47-20:31).
The basket maker prepares the rope for the handles out of palm fibre string.2
T h e handles are s e w n onto the basket.
The video picks up the sequence after the first handle has been finished. The
basket maker folds the basket to measure the position of the second handle,
throws the handle, which is attached to a length of palm fibre string, into the
basket and starts sewing from the inside outwards. The handle is fixed to the
basket by a combination of sewing and wrapping the string around the handle
base. The end of the string is sewn into the fabric of the basket, following
the oblique openings in the plaits of the sides, to provide a strong anchoring
of the handle. The string is cut off and a second, decorative, line of string is
stitched into the plait structure.
(g) finishing touch (23:48 - 24:00)
The end of the sequence is formed by the basket maker cutting off bits of palm leaf
which stick out of the plaited sides (24:04).
16.1.2 Production Time

T h e maqtaf m a k e r has to plan his w o r k in advance, because the time that a palm
leaf needs to dry is t w o weeks. If he has dry palm leaf, the preparation takes only
30 minutes, for the strands to soak.
T h e total production time for the s e w n plaits basket is approximately nine
hours. M a k i n g the plait takes four hours, while making the string takes one and a
The video only shows one of the handles being prepared, while in reality the basket
maker makes two.
half hour. Sewing the basket lasts approximately three hours, while the basket
maker then needs another half hour to m a k e the rim, handles and finish off the
basket. A s the process does not follow a fixed order, the timing is approximate.
Timing the actions shows that the repetitive actions take up most of the time,
while the occurring actions such as adding a n e w length of sewing string, starting
the plait, or making the start of the basket, take relatively little time.3
The occurring actions often result in the features which are characteristic for
a basketry tradition. It is in the method of insertion of n e w lengths of material, in
the start and finishing off of the work, as well as the decoration that idiosyncrasies
or local traditions become apparent. In the entire process these characteristic
features take up a small portion of the time.
While editing the video, a deliberate choice was m a d e to show these
idiosyncrasies. Thus, the edited footage can be misleading: what is shown in real
time are the short procedures, while the long sequences are shortened
dramatically. In the video the process of making a maqtaf takes 21 minutes and 11
seconds, while in reality making the basket takes approximately nine hours, often
divided over two days. Making the plait takes four hours of which 2 minutes and
24 seconds are shown (04:11-06:35), sewing the basket takes three hours, which has
been compacted to just under four minutes (09:49-13:02, 16:33-16:55 and 17:59-
18:20).
Working period
The maqtaf maker works part time, because he also tends hisfieldsand he has to
sell his produce. There is no clear seasonality to the work. The raw materials can
be stored dry and harvested year round. H e makes baskets during all seasons, but
the time he can spend depends on h o w busy he is working the field. Since
agriculture is no longer ruled by the flooding of the Nile, he has no clear busy
period or season w h e n the work is slow.
In general he works in thefieldsin the mornings and makes baskets at home
in the afternoon and early evening. S o m e days he works steadily from early in the
morning to late in the afternoon. H e makes approximately five baskets a week and
on Thursday morning he goes to the market in Mallawi to sell his baskets.
16.1.3 Working Rhythm

The descriptions of lives of the early Christian saints regularly mention that the
hermits were making plaited basketry in their cells. The rhythm of plaiting, string
making and sewing helped them to concentrate and contemplate, while the
products gave them an income (Wipszycka 1986, Waddell 1936, 78, 86, 102,
122).
Starting the plait takes only 24 seconds, starting the basket takes two minutes and 45
seconds.
16 The Choreography of W o r k 335
Most of the time in making a maqtaf is spent on the repetition of the same
movements (recurring actions). It is especially in these periods of recurring actions
that a specific rhythm occurs, wich enables the producer to keep going for a
prolonged period. In sewn plaits basketry the recurring actions are: making the
plait, rolling up the plait, making string, sewing the plait and sewing the rim.
However, w h e n considered closely, although apparently occurring actions are
performed whenever necessary, and are m u c h more diverse than the recurring
actions, they often have a specific rhythm too.
Plaiting is a recurring action, which is interspersed with the occurring action
of picking up n e w plaiting strands and inserting them in the work (04:12, 04:26,
05:33). In Chapter 9 the plait pattern was noted d o w n with the help of a formula,
\2/2\\l. Furthermore, plaiting is described as consisting of active elements only,
which expresses that the basket maker handles all nine strands equally.
These two aspects, all strands are active and the pattern is Y2/2W1, translates
in a rhythm of the work: the basket maker plaits a strand "under 2 / over 2" with
the right hand and "under 2 / over 2" with the left hand. This action is done in a
particular cadence (04:47). Without losing the rhythm, n e w leaflets are added in
the plait, always on the right-hand side. After every plaiting movement with right
and left, the basket maker consistently pulls the strands to tighten the plait. Since
the plaited strips are several metres long, the basket maker continues the same
movements and the same rhythm for at least four hours.
The next recurring action is rolling up the plait. In the video this sequence
has been used as an example of the analysis of actions, by indicating the actions
with a combination of a letter and a number in the left bottom corner. In the
description of the production sequences given above, this sequence has been
described as: "In preparation of the start of the work, the plait is rolled up". In the
28 seconds incorporated in the video (05:47-06:15) the recurring action of rolling up
the plait (rl) is interrupted seven times by an occurring action (o2) of picking off
little ends of palm leaflets which are sticking out of the plait.
The basket maker turns the plait with his right hand, while he is stroking the
plait with his left hand, checking the surface for irregularities and guiding the plait
to the roll (rl: recurring action of rolling until the plait is rolled up). Whenever
bits of palm leaf are sticking out, he takes the plait in his left hand and picks off
the ends with his right hand (o2: occurring action). H e makes seven rolling
movements, then plucks five times, he continues rolling once and plucks twice, he
then rolls one time again and plucks four times. H e repeats rolling once and
plucking four times, he then rolls twice and plucks four times. After this he rolls
once and plucks five times, rolls twice and plucks once. The end of the sequence
shown is that he makes three rolling movements. In this sequence the rhythm is not
found in the number of times that he does the rolling action, but in the regularity
of the movements.
Occurring actions are more diverse and are performed for a shorter period.
Although this would suggest that rhythm is not important, m a n y occurring actions
nevertheless have a specific rhythm. Thus, the occurring actions do not disturb the
rhythm of the process as a whole.
A good example is the removal of excess material while rolling up the plait.
Although this action (o2) is provoked by bits of palm leaf sticking out, it is not
ruled by irregularity. This is particularly clear if all footage (and not only the
edited version) is taken into account (Table 16-1). During the 13 rolling
movements, many palm leaf ends are passing unhindered under the hands of the
basket maker, w h o seems to have a slight preference for pulling off four bits
between each sequence of rolling movements.
rl 7 1 1 1 2 1 2 3 1 1 4 1 1 1 1 13
1
o2 5 2 4 4 4 5 1 7 2 6 5 4 5 1 2 3
J
Table 16-1 Rolling up the plait: number of rolling movements (rl) and plucking
movements (02) in the complete footage of the film. A double line indicates
where the sequence in the edited video (05:47-06:15) ends.
The occurring action of twisting the veins of the date palm leaflets, as a
preparation for string making (06:46-06:56), shows a rhythm too. The basket maker
works on them until he feels the fibres are loose enough to be made into string.
Nevertheless, he does the twisting in a specific rhythm. H e twists the bunch four
times at the thick end, four times in the middle, twice towards the end and another
two times a the tip of the leaflets. The but ends require a longer period of working
than the thin leaf tips, but there is a rhythm of two times four and two times two.
The next sequence with recurring actions is making the string (06:36-07:58).
His right hand moves over the left hand from the wrist to the finger tips (z-
spinning). With each movement the spun bundle is picked up at the finger tips and
put back at the wrist (S-plying, cf. Figure 15-6, p. 299). The rolling movements of
the hands (recurring action) are interrupted by the insertion of new material
(occurring action).
There is some variation in the number of times that the hand rolling
movement is made after each newly inserted bunch of leaf veins. In the video
section the start of the string is made by folding the strands double and making six
rolling movements. The next rolling sequences show five, seven and seven rolling
movements. In a second string making sequence (13:02-13:24) he lengthens the
sewing string on the spot in two series of four rolling movements. H e finishes the
length of sewing string with twenty rolling movements, which are also meant to
tighten the ply. The rhythm of string making is not clear from the number of
rolling movements (respectively 6, 5, 7, 7 and 4, 4, 20), but from the movement
itself, which is at the same time rhythmic and fluent.
After making the string, the start of the basket is made. This sequence, the
analysis of which is shown in the left-hand bottom corner of the video has been
described above, as: "In preparation of the sewing of the basket, the basket maker
takes a large needle and runs the string through the hole, takes up the plait and
starts sewing". This seems to consist of four actions. However, in the two minutes
and 45 seconds shown in the video (08:05-10:50) w e can distinguish nine actions, in
which especially actions olO and rl 1 alternate a number of times.
This sequence leads to the recurring action of sewing the basket. The
preparatory actions are threading the needle (o3), picking up the plait and letting it
drip for a while (o4), arranging the plait on the ground (o5) and picking up the
plait (06). During the entire sequence the basket maker holds his needle in his right
hand, while the string and the plait are handled with his left hand.
Actions o3 to o7 are preparatory. It is the getting ready for the action. The
first part of this sequence which has a direct function in the production process is
the kneading of the plait, to make it supple (o7). This is done between both hands
in vertical and horizontal direction. In total he bends the plait end 14 times. A
large needle (mesalla) is pushed into the side of the plait, picks up four loops and
is pulled through (08). Moving the needle up and d o w n to force it through the
plait, the string is pulled through infivejerky movements.
Figure 16-2 Making the start of a sewn plaits basket: Left: pushing the needle through
four loops (08). Right: pulling the needle through the end of the string (o9).
Drawing not to scale.
The loop at the start of the string sticks out of the end of the plait and by
pulling the string through this loop in three jerks, the end of the plait forms a small
circle (o9, Figure 7-49). This circle has to 'set' and the basket maker therefore
knees the start of the basket, while pulling the string with force (olO).
The rest of the sequence is a combination of the recurring action of stitching
through the edge of the plait or basket (rl 1), and kneading to settle the centre
(olO). T o make up for the small diameter of the centre and to get a good circle the
basket maker picks up two loops of the plait and one loop of the basket.
Start of sewing the maqtaf
no. rll stitching olO kneading
1 2 loops of plait 3 jerks kneading, 1 x tightening

2 1 loop of basket 4 jerks kneading, 3 x tightening; kneading lx tightening
3 2 loops of plait 3 jerks kneading, lx tightening
4 1 loop of basket 3 jerks not much kneading, 3x tightening
5 2 loops of plait 3 jerks kneading
6 1 loop of basket 2 jerks kneading, 2x tightening
7 2 loops of plait 2 jerks kneading, lx tightening
end of sequence
(following is based on complete footage; not incorporated in the edited videotape)
8 1 loop of basket 2 jerks kneading , lx tightening
9 2 loops of plait 3 jerks kneading !
11 2 loops of plait 3 jerks kneading
13 2 loops of plait 2 jerks no kneading
14 1 loop of basket 3 jerks kneading
15 2 loops of plait 2 jerks change of working position
Table 16-3 Number of stitches to make the centre of the sewn plaits basket. Per stitch is
indicated how many plaiting strands are picked up, with how many jerks the
string is pulled through (action rll) and the number of times centre is
tighten and shaped (action olO).
Although the start of the basket seems at first viewing irregular, this scheme
shows quite some regularities. The string is mostly pulled through the edge of the
plait in three jerks. Towards the end, w h e n the string shortens, it is sometimes
pulled through in two jerks. Tightening of the centre is mostly done after a stitch
through the edge of the basket. The stitch through the two loops of the plait edge is
usually just followed by kneading and bending the plait as preparation for the
needle to enter the tightly pulled basket edge at a convenient angle.
The 15th stitch marks the end of making the centre. After that the stitches are
picking up one loop of the plait against one loop of the basket. N o more kneading
is done, the working position changes and the rhythm moves into that of the
recurring action of sewing the basket (rll), which is sustained for approximately
three hours.
The large mesalla needle is pushed through the edge of alternately the basket
and the plait. If the sewing string is very short, the string can be pulled through the
plait with one jerk, but mostly the number of times the string is pulled shows a
clear rhythm, irrespective of the length of the sewing string.
In the beginning of the sequence the rhythm basket 1 /plait 1 (10:58) and
basket 2 /plait 2 (11:03) occur. A new string is inserted at 11:07, but the rhythm
basket 2 /plait 2 continues for three more stitches. Most commonly, in the base of
the basket the rhythm is two jerks to pull the string after stitching in the basket,
one jerk to pull the string through the plait.
In the side of the basket, however, this rhythm changes: the string is pulled
with one jerk through the plait edge and with a two-stepped jerk through the basket
edge, irrespective of the length of the string. Only w h e n the sewing string has just
been inserted and is still very long, the basket maker uses two separate jerks,
which has the same rhythmic effect (e.g. 12:44). The basket maker uses his right
hand, in which he also holds the needle, to open up the plaited edge of the basket.
Even this action does not disturb the overall rhythm.
A n even clearer indication of the working rhythm is a small knock with the
needle on the body of the basket. W h e n making the base of the basket, the basket
maker does this before pushing the needle through the edge of the plait (e.g. 11:42
and 11:50). This changes w h e n he is making the side and rim of the basket,
because then he knocks before stitching into the basket edge (first time at 12:02).
At some point, w h e n he restarts the action after a brief break, he knocks with his
needle on the basket (13:33), refrains from starting, then knocks again w h e n he is
truly restarting the sewing motions (13:43). This rhythmic knock occurs mostly
with the 1 in basket side / 2 in plait jerk rhythm. M u c h later in the process, while
a n e w long string has just been added, the rhythm is still the same: a knock with
the needle on the basket, pulling through the string with one jerk through the
basket side and two jerks through the plait edge (16:35). This same rhythm is found
in sewing string to the rim.
1 '
Sewing the basket: rhythm of pulling the string through basket and plait edge
basket plait remarks video
1 jerk 1 jerk short string 10:58
2 jerks 2 jerks with short string, 11:03

2 jerks 2 jerks with long string, two knocks before stitch in plait 11:46
2 jerks 2 jerks two knocks before stitch in plait 11:53
2 jerks 2 jerks 11:58
1 jerk 1 jerk knock before stitching in basket 12:03
1 jerk 7 - 12:07
2 jerks 1 jerk 12:11

1 jerk 2 jerks knock before stitching in basket 13:36

1 jerk 2 jerks 14:00
2 jerks 2 jerks 16:32
2 jerks 2 jerks knock before stitching in basket 16:35
basket rim string

4 jerks 4 jerks - 19:02
2 jerks 2 jerks - 19:15
1 jerk 2 jerks - 19:22
1 jerk 2 jerks 19:27
2 jerks 2 jerks knock before stitching in rim 19:32
Table 16-4 N u m b e r of small jerks with which the string is pulled through respectively
the edge of the basket and the plait or the basket and the rim string.
While the basket maker is sewing the basket, he pauses a while to make the
string for the handle (15:05-15:34). This is done in a movement opposite to making
the sewing string: the right hand moves from the tip of the fingers to the wrist of
the left hand, picks up the s-spun strand at the wrist and takes it back to the finger
tips (Z-ply). The third strand is inserted with the same movement from fingertip to
wrist.
The material he uses to make the string, date palm fibre, is longer than the
date palm leaf veins which he used in the first rope making sequence. This is
reflected in the larger number of rolling movements (rl) made until a new length
of date palm fibre has to be inserted (o2). A s with the string made of date palm
leaf veins, the rhythm is not found in the number of rolling movements
(respectively 8, 5, 11, 3, 12), but in the regularity and smoothness of the
movement itself.
16.1.4 The Body as Instrument

Looking at the movements in detail, clarifies that the most important instrument of
the basket maker is his body. Studying the physical balance, working position and
the use of hands, feet, eyes, mouth and body weight results in understanding the
instrumental use of the body and the physical demands of the work.
At the beginning of the video the basketmaker squats, his heels on the ground, his
weight in the middle, his elbows resting on his knees. He is sorting leaflets using both
hands, while keeping his eyes on the sorting (03:38).
He keeps this position while he makes the start of the plait.4 His eyes are focused on
the work. His body weight shifts slightly to the left leg whenever he picks up new leaflets
from the ground with his right hand (e.g. at 04:11).
Further in the plaiting process we find him in the same squatted position, but his
weight is gradually more on the left leg, especially when he reaches out to pick up new
leaflets (05:10). His eyes wander from the work, while he talks to his wife, but only
briefly.5 He can afford to look up while putting in a new leaflet, or while tightening the
plait, but when he is plaiting the strands, he looks at his work. He bends the outer plaiting
strand down and pushes it up and over two. The nine plaiting strands are not single leaflets,
but small bunches. He needs to look which leaflet belongs to which plaiting bunch,
something he apparently cannot do by touch, despite his experience.
The close up of his hands (05:25) shows that he holds the plait in his left hand, four
fingers underneath the plait and his thumb on top. With his right hand he puts a new leaflet
on top of the plaiting strand that is most to the right. In the same movement he bends the
plaiting strand down, passing it underneath the second and third strands. Holding the top of
the plait with his right thumb, he flips the plaiting strand up with his fingers between the
third and fourth strands and leaves it lying over the fourth andfifthstrands. He pulls the
leaflets and tightens the plait with his right hand (05:29) and then repeats the process with
his left hand.
A symbol indicating the working position has been inserted at 03:54.

A symbol indicating work-eye contact has been inserted at 05:13.
He only picks up new leaflets with his right hand, sometimes one, sometimes a bunch
which he puts between the little finger and ring finger of his left hand (e.g. 05:34). The new
leaflets are always put in with his right hand on the right side of the plait.
At the end of the plaiting sequence he is still squatting, a position which he has
maintained for four hours and still maintains throughout the rolling up of the plait. The only
variation is that he shifts his weight for a moment: on his left leg while picking up new
leaflets; on his right leg while rolling the plait (06:07). His eyes are focused on his hands
while rolling the plait (rl). With his left hand he picks irregular leaflets from the surface,
his elbows still resting on his knees (o2).
H e maintains the squatting position while soaking the plait. H e pushes the plait under
water with his right hand, while holding the dry end of the plait with his left hand (06:19).
H e lifts a bundle of palm leaf veins, which are soaking in the same pot, and takes it
over with his right hand. With his left hand he pulls out a small quantity of veins, while he
pushes the bundle back with right. Holding the small bunch with his left hand, he loosens
the fibres by making anticlockwise circling movements with his right hand. Eyes focused on
the work, he makes four turns, then releases the left end, picking up the bunch more to the
middle. H e makes another four turns and releases his right hand, grabbing the bunch in the
middle. H e continues with two turns, releases his right hand again and makes another two
turns just over the middle towards the tops of the veins (06:53).
With his right hand he pushes the bunch of veins in the pot under water. H e takes
over the bunch in his right hand and puts it on the rim of the aluminium soaking pan. With
his left hand he picks up a few veins to start making string (07:00).
H e then takes half the bunch in his right hand and turns it over, arranging it in a way
that the two bunches overlap in the middle, with the thick ends towards the outside. H e
reaches out to pick up more veins with his right hand, but drops them again. Holding the
middle of the bunch in his left hand, he picks up the bottom half with his right hand, folds it
double and rolls his right hand over his left hand (07:07). The rolling movement is made by
moving his right hand away from his body, over the two bundles of veins on his left hand,
starting with the fingertips of the right hand at the wrist of the left hand. W h e n he reaches
the point where the fingers of the right hand have passed the left hand, he picks up the first
bundle with the palm of his right hand and deposits it back at the wrist of the left hand,
behind the second bundle (cf. Figure 15-6, p. 299). H e is still squatting with his elbows
resting on his knees, eyes directed towards his hands.
H e picks up a n e w bunch of veins with his right hand and brings it up to the string
which he holds in his left hand. H e reconsiders and moves his body weight to his left foot
while his eyes turn to his right foot (07:18). With his left hand, which is still holding the
string, he puts off his shoe and he sticks the start of the string between hisfirstand second
toes.6
His body weight remains on his left foot, while his right foot rests on its heel, because
the string is still very short. H e continues making the string, picking up n e w bunches of
veins with his right hand. With the lengthening of the string he slowly lowers his right foot,
but his weight is still mainly on the left side.
In the next sequence, in which the basket maker starts the sewing of the basket, he
squats with the needle in his right hand and a piece of string in his left hand. His weight is
in the middle and his elbows are resting on his knees. With his left hand, as well as with his
right hand that comes up to assist, he folds the end of the string double (o3).
6'A symbol indicating instrumental use of the body has been inserted at 07:22.
With the needle in the right hand, he pushes the string with his left hand through the
eye of the needle. H e then uses his left hand to pull the string through the eye and to open
up the string by twisting it slightly. H e twists the short end with his right hand, in which he
also still holds the needle, three times around and through the two strands of the string,
which he has opened with his left hand. H e pulls and stretches the string by holding up the
needle in his right hand, gliding with his left hand along the string to the end. H e changes
the weight of his body from the middle first to his left, then to his right foot en stretches out
his right hand, still holding the needle, to the aluminium pot in front of him. H e
reconsiders, retreats his right hand and takes over the needle with his left hand (t3-4).
H e stretches out his right hand and takes hold of the rolled up plait, which is still
soaking in the aluminium pot. H e lifts the plait and takes hold of the end of the plait, which
is outside the pot, with his left hand. H e sits for 13 seconds holding the plait above the pot
to drip. H e then slowly moves his weight to his left foot (o4).
H e looks to the left, the place where he plans to put d o w n the plait. While he drops
the dry end of the plait from his left hand, he twists his body to the left and puts the plait
down with his right hand (t4-5).
Plate 16-5 M o h a m m e d working in the street in front of his house, surrounded by his
family and neighbours (el-Hagg Qandil, 1989).
With his body slightly twisted to the left, leaning with his upper body on his left knee,
he lets go of the plait for a m o m e n t and then picks it up with his left hand, in which he still
holds the needle. With his right hand he replaces the dry end of the plait. H e then lifts the
rolled up plait with two hands, takes over the plait with his right hand, while his left hand
rearranges the dry end. His body weight is on his left foot. H e puts the plait d o w n with his
right hand and also drops the dry end. H e hits the plait with two hands as if ending the
movement, or establishing its position (o5).
H e focuses on the aluminium pot, turns his body back to the right, picks up the pot
with his right hand and puts it a bit further, out of the way. His right hand moves up, as if
he is chasing aflyor a thought. H e stretches the string and his body, holding the needle in
his right hand and the start loop of the string in his left hand. H e turns left to where the plait
lies, changing his body weight from right to left. H e drops the string from his left hand, his
eyes all the while on what his hands are doing. H e moves with his left hand towards the
rolled up plait. H e rests his left hand on top of the plait, while the right hand moves in
towards the centre of the roll...(t5-6)
... and takes the middle of the plait, pulling it upwards. H e then takes over with his
left hand pulling the plait up a little more, while he lets go with his right hand, so that the
centre of the plait unrolls. H e takes over five times (right behind left behind right behind left
behind right) and then takes the plait with his left hand drawing it twice through the right
hand. All the while his working position is squatted, with his elbows resting on his knees,
his weight in the middle and his eyes focused on what his hands are doing (06).
H e holds the start of the plait with his left hand and bends the start of the plait with
his right hand (t6-7).
H e 'kneads' the plait with two hands, to make the start of the plait a bit more supple,
because it has to form a very small circle (o7).
H e holds the plait with his left hand and probes with the needle, of which he holds the
back end in his right hand, to find the opening in the edge of the plait (t7-8).
With his right hand he moves the needle, with his left hand the plait to guide the
needle through the openings in the right side of the plait. H e forces the needle through four
loops, slightly turning and pushing the needle, holding it between thumb and four fingers of
the right hand.7 W h e n the needle is in place, he pushes it during which action he takes hold
of the needle in three different places, each time a bit further to the back. H e lets go of the
end of the needle and takes hold of the point of the needle. H e moves the needle with his
right hand four times up and d o w n in a jerking movement and pulls through the string.
Because of the length of the string, he has to change his grip on the string three times. H e is
still holding the plait in his left hand, which he n o w uses to stop the end of the string to
being pulled through the plait (08).
With a twist of the fingers of his right hand he opens the little loop of the string. H e
keeps it open with his left hand and pushes the needle through the loop with his right hand,
holding the needle in the middle. H e takes up the point of the needle with his right hand and
pulls through the string in three jerking movements (o9).
With both hands he kneads the start of the plait before pulling the last bit of the string
tight with his right hand. H e kneads again with both hands and then pulls the string very
tight in four small pulling movements, each time letting go and taking hold of the string
with his right hand (olO).
A symbol indicating the use of instruments has been inserted at 09:20.

With his right hand he brings the needle towards the edge of the plait (tlO-11). He
guides the needle into the plait and picks up two loops. With the ball of his thumb the needle
is pushed through and in the same movement the right hand moves forward and takes hold
of the part of the needle that has passed through. In three jerks the string is pulled through
(rll).
Both hands are used to knead the plait, then the string is pulled even tighter with the
right hand and some more bending and kneading is done (olO).
The needle picks up one loop from the edge of the basket centre. The right hand picks
up the middle of the needle and with four jerks the string is pulled (rll). Three times he
kneads the basket centre using two hands. Each time the pushing and bending is followed by
a tightening of the string with his right hand (olO).
With his right hand he pushes the needle through two loops of the plait. The right
hand is shifted to the middle of the needle and with three jerks the string is pulled (r 11). H e
kneads with both hands, his working position still squatting with the weight in the middle
(olO).
With the needle in his right hand the basket maker picks up one loop of the basket
base. H e moves the needle up and d o w n and then pulls the string through the loop in four
jerks (rll). H e moves his body weight to the left and uses his left knee as support for
kneading the basket base (olO).
Bringing his weight back to the middle, he picks up two loops of the plait and pulls
the string in four jerks (rll). H e kneads the basket with both hands (olO).
H e guides the needle through one loop of the basket edge. T w o jerks are sufficient to
pull the gradually shortening string (rll). H e kneads on his left knee (olO) and pushes the
needle through two loops of the plait.
All the while he maintains the same position: squatted, with his elbows resting on his
knees. Only when he uses his left knee as support for tightening the fabric, he twists his
body and moves his weight slightly to the left. His eyes are all the time focused on his
work.
In the next sequence (10:58) his working position has changed: he is sitting with his
legs stretched in front of him and the base of the basket clamped between his knees. The
moment where he changed his position has not been incorporated in the edited version of the
video. It occurs when he finishes the start of the basket, the m o m e n t when he no longer
stitches into two loops of the plait against one loop of the basket. From that moment on he
picks up the easy rhythm of sewing, without the labourious kneading of the plait.
After having finished the base, he changes working position again, sitting cross-
legged with his left foot holding the base of the basket (12:00). With his left hand he holds
the plait, with his right hand he wields the needle. H e looks up, talking to his mother, but
only when he pulls through the string. Whenever he guides the needle into the edge of the
basket or the plait, he looks back to his work.8 At first, when the wall of the basket is still
beginning to form, he presses the basket to the ground with the side of his foot (12:08),
while later his foot rests inside the basket, holding it in place (16:32).9
W h e n his sewing string runs out, he makes a new length, by hand rolling bunches of
palm leaf veins (12:52). H e holds the basket with the weight of his left foot, the string stuck
between the first and second toes. H e is leaning over to the left side, to keep tension on the
A symbol indicating the work-eye contact has been inserted at 12:02

A symbol indicating the instrumental use of the body has been inser
a symbol indicating the working position has been inserted at 12:30.
string. He reaches over with his right hand to pick up a bunch of veins (13:00), whi
first pushes on his right knee to even out the strands and then he puts the d o w n to his right,
picking up part of the bunch to work into the string. While he makes string, he regularly
looks away from his work. String making is apparently done mostly by feel. H e finishes
tightening the ply of the string by making the rolling movement above his head, with a fully
stretched length of string.
H e bends over the end of the string, picks up the needle from his right side and
threads the needle (13:27). H e puts his left foot in the basket and starts sewing again.
W h e n the basket is well underway, he takes a break from sewing by making the string
for the rim and handles. With his right leg bent and his left leg stretched out, he soaks date
palm fibre and rolls small tufts between his hands (14:25-15:04). This working position he
has taken from the moment he started making the string.
H e continues making string by picking up the end he is working on with his right hand
(15:05). H e holds the string with the toes of his right foot, which he places backwards a
little and he leans over to the left to keep the tension. The hand rolling movement is
opposite to the one used for the sewing string: with his left hand he rolls over his right hand
from fingertips to wrist. W h e n adding another tuft, which he picks up with his right hand,
he again moves his right foot slightly backwards. Because he cannot lean over more to the
left and the working end of the string is getting out of reach, he pulls the string backwards
with his right hand and holds it with the toes of his right foot again (15:18) and continues to
make string. H e is talking to the neighbours w h o have c o m e on to the roof and he
occasionally glances at them, but his eyes are mostly on the string. H e moves back the
string a second time and adds two more tufts.
Then he takes up the string with two hands and moves it forward, until the part of the
three-ply string he had been working on before adding a length of two-ply is within reach.
H e anchors the finished string with his right foot, his right leg still being bent. The two-ply
string, to which another strand has to be added, is anchored between the toes of his flexed
left foot, his left leg stretched out (15:47). H e continues adding the third strand. H e again
moves his right foot slightly backwards to arrange the tension.
Maintaining this position, he makes a few n e w tufts of palm fibre, which he adds
straight away to the third strand of the sZ3 string.10
H e continues sewing with his foot in the basket. Meanwhile a friend finishes the plait
(16:56-17:06). This is a rich man, w h o has just returned from Saudi Arabia, where he made
his fortune, but he did not forget h o w to make baskets. H e has the same working position as
M o h a m m e d : squatting with the elbows resting on the knees. The last bit of the plait is
finished off by dividing the plaiting strands in two bundles, which he makes into string. This
he does standing bent over, with his left foot on the plait.
M o h a m m e d finishes the basic structure of the basket, by sewing the string at the end
of the plait to the basket. With his left hand he gives a clockwise twist to open the ply of the
string and with right he pushes the needle through the opening (17:54-18:18).
Just before he starts making the rim he inserts a n e w ready-made sewing string
(18:20). H e drops the needle, bends over slightly to the right to take up the n e w length of
string. With the needle in his left hand he opens the loop of the string and pushes it over the
needle, which he then takes over in his right hand. H e then takes off the needle, which he
puts in his lap, tightening the end of the short string to tighten it. With his teeth he seems to
pull off a loose strand (18:28).
A symbol indicating the properties of the material has been inserted at 16:02.
Action W o r k i n g position Video

A) Preparation of raw materials
sorting palm leaf squatted 03:34-03:47
turning palm veins squatted 06:45-06:53
making tufts of palm fibre sitting, right leg bent, left leg stretched out 14:37-15:04
16:02-16:08
b) Preparation of systems
plaiting squatted 03:47-04:55
16:49-17:05
rolling the plait squatted 05:47-06:15
finishing the plait standing bent over 17:06-17:22
making string squatted 06:57-07:58
sitting, left leg stretched 15:05-16:28
C ) START OF INTERACTION OF SYSTEMS
sewing squatted 09:48-10:51
D) M A K I N G T H E BASIC STRUCTURE
sewing both legs stretched 10:59-11:50
sitting, foot in basket 16:32-16:54
17:27-17:54
E) FINISHING T H E BASIC STRUCTURE
sewing end of plait sitting, foot in basket 17:54-18:18
F) adding rim and handles
sewing rim sitting cross legged 20:41-22:12
twisting handle sitting right leg stretched 19:48-20:32
sewing handle sitting cross legged 20:41-22:12
stitching decoration sitting cross legged 22:15-22:40
G ) FINISHING T O U C H
removing irregularities sitting cross legged 23:48-24:00
Table 16-6 Working position per production phase
H e plies the string back into itself, to fasten the newly attached length. H e then leans over to
pick up the palm fibre string and sits straight, still with his foot in the basket, while cleaning
off adhering fibres. H e takes the sewing string in his right hand, discovers that there is no
needle attached and drops the palm fibre rim string from his left hand, because he has to
take up the sewing string. He picks up the needle with his right hand and wets the st
with his mouth (18:45). Previously he wetted the end of the sewing string in the water of the
aluminium soaking pan (11:26-11:28). He runs the string through the needle and fastens the
string by opening the ply of the string with his left hand and pushing the end through in a
circling movement of his right hand (18:50).
He picks up the palm fibre rope again with left and opens the loop at the start of the
rope. The needle is pushed through this loop and then he goes on sewing, guiding the needle
alternately through the rim of the basket and the string, picking up one of the three strands
(19:25).
After finishing the rim, he makes the handles from the same type of palm fibre string.
His right leg is stretched out in front of him, while he has pulled up his left leg." The sZ3
string is fixated on one end between the toes of his right foot.12 He tightens the ply by hand
rolling the palm fibre string, while he looks up from his work. He picks up the rope at
approximately one third of the distance between his hands and toe, and starts turning it
clockwise back onto itself, in this way doubling the string. He then flips the handle over and
repeats the procedure, adding a third layer (20:12-20:23). The result is a handle which
consists of an sZ3[S]3 cable. He fixates the twist by pushing a loop of the string through the
bend of the cable. He twists the fibres into place, by pulling the handle in front of his chest.
Sitting cross legged, he throws the handle, with a length of string attached, in the
basket. With his right hand, which also holds the needle, he pulls the opposite rim towards
himself, to determine where the second handle should be placed. He holds the basket rim
with his left hand and pierces the wall of the basket from the inside out.13 With his left hand
he places and fixates the handle loop, picking up the loop of the first leg of the handle. With
three stitches (out-in, in-out, out-in) he bridges the width of the handle. With his right hand
he bends the top of the handle down to the rim, while he guides the bend of the rope with
his left hand. He puts the looped end of the handle against the outside of the basket, just
under the rim (21:18).
He holds the handle in place with his left hand and stitches with the needle in the right
hand from inside out. He puts the loop of the rope handle with his left hand around the
needle. He positions the handle with two hands, then makes two stitches (out-in, in-out),
sets the handle and stitches again from the outside to the inside. He turns the sewing string
twice around the second leg of the handle and stitches from the inside to the outside.
Although he keeps his eyes on the work, he stitches twice in the wrong place and has to pull
back the needle. He gives a last tug and positions the handle once more.
With his left hand he rolls the basket on its side. He then pushes the needle through
the fabric with his right hand, to anchor the handle string. He first holds the needle in the
middle, then pushes it through with the ball of his thumb.14
He cuts off the end of the string with a knife, while holding the string with his left hand. He
drops the knife, takes hold of the rim with his left hand and runs the needle along the other
diagonal for decorative purposes (22:49). For the last 10 cm the string is just too short. By
loosening it slightly from the needle, he can just finish the diagonal with this piece of string.
Holding the knife in his right hand, he cuts off the string.
Â symbol indicating the working position has been inserted at 19:48.

Â symbol indicating the instrumental use of the body has been inserted at 20:12.
|4A symbol indicating the use of instruments has been inserted at 21:04.
Finally, he removes irregularities from the outside of the basket, wielding the knife.
He is still sitting cross-legged and still concentrated and looking at the work under his hands
16.1.5 W o r k s p a c e
The basket maker needs approximately four square metres of work space. There
are no special demands, no fixed constructions, so he can sit anywhere and in fact
does so. Sometimes he works, inside the house, sometimes in the street in front of
his house (Plate 16-5). This time he works on the roof, sitting on top of the d o m e
of the room below (Figure 16-7).
H e prepares the space by arranging the things he needs around him, within
easy reach. H e forgot to take the aluminium pan to soak the plait and calls out to
his wife to bring it (06:02). T h e video is made in winter, w h e n the temperature
outside in the sun is pleasant. In s u m m e r he prefers to work in the shadow. This is
not only because of the heat, but also because the palm leaf should not dry out.
The working circumstances areflexible.Although the work is monotonous
and the working position is usually squatted, the basket maker can choose his
surroundings. H e often sits with other people, talking while he works.
Figure 16-7 Sketch plan of the work space that M o h a m m e d uses. H e works on the roof
of his house. The circle indicates the domed roof he sits on (Drawing by
H. Barnard).
16.2 The Sabat
coiled basketry
made by Rauwhayya in el-Amariyya
The working order recorded in el-Amariyya, mostly follows the sequence as

s h o w n in the table of section 15.5 (p. 303):
a) preparing the material (24:41-25:07).

The fruit stems (zaghauwa) have been pounded to loosen the fibres. The date palm
leaflets have been plucked from the palm branch, the vein is removed and the leaf
strips have been soaked in water for 30 minutes (not shown). They are kept moist by
rolling them in a piece of cloth.15 The basket maker sorts a bundle of date palm
leaflets and shreds them, using her teeth.
c) starting the basic structure (25:07-28:28).I6
The basket maker makes a knot in the bundle of shredded palm leaf and folds the ends
of the knot to one side. She then takes a small needle (ibra), runs the point of a date
palm leaf strand through the eye and covers the knot with a row of stitches. She starts
where the ends of the knot stick out to the side and works around in e-direction
(anticlockwise) until she has made one full circle. The centre resembles a snail's
shell.
preparation of raw material 27:59-28:18)
She then shreds some of the zaghauwa.
She pushes the small bundle in the middle of the shredded ends which are sticking out
of the start of the basket.
d) making the basic structure of the basket (28:35-31:37 and 31:42-32:53)
With an awl she pierces the previous bundle and she then pushes the needle through
the hole. She regularly has to put in a new palm leaflet and bundle material.17
preparation of raw material (not s h o w n )
A bunch of palm leaf strands wrapped in a cloth are visible lying in the basket to the
left of the basket maker in video sequence 27:54-27:58
In the sequence the phases b and g do not occur: there is no separate preparation of
the systems and no finishing touch.
The frozen image of coiling is shown in video sequence 29:30.
Every time she inserts some of the zaghauwa, she first makes the fibre into
finer shreds.
e) finishing the basic structure (not shown)
She finishes the basket by letting gradually letting run the bundle thinner and stitching
the last length of the bundle with the wrapping strand.
f) adding features (not shown)
This basket does not have handles. She makes a rim decoration by stitching the top of
the rim with the peel of the seed pod of the date palm.

Making a coiled basket takes three full time days, but the w o m e n use at least a
week to finish a basket because the work is not done full time. Basket making is an
activity the w o m e n do whenever their work in the house and the fields allows them
some time. The tasks of the w o m e n in Middle Egypt differ. The poorer w o m e n go
out and help in the mornings in the field. W h e n the farmers are a bit better off,
they prefer their wives to stay at h o m e . Apart from cleaning, laundry and cooking,
the w o m e n bake their o w n bread and m a k e cheese.
Basket making is not limited to a particular period of the year. A s long as there is
dried palm leaf in store, a w o m a n can start working after a preparatory period of
30 minutes (the time needed for the leaflets to soak).

In the previous section on working rhythm it became clear that the rhythm not only
occurs in recurring actions, but also in occurring actions. At the beginning of the
sequence Rauwhayya shreds the palm leaf strands with her teeth. This is an
occurring action, during which she twice takes leaf shreds away from her mouth
with two pulling movements of her left hand (24:45-24:52 and 24:59-25:05).
The recurring action of wrapping a palm leaf strand around the bundle and
fastening it with a stitch is the main action for making coiled basketry. T h e
sequences shown in the edited video do not have a clear working rhythm and
neither does the complete footage.
The heading of Table 16-8 lists the number of wrapping strands (a to k) and
the number of stitches done with each of these (1 to 8). In the table for each stitch
a count is m a d e of the number of times in which the strand is pulled through.
Indicated is also if this is done with a brief break in between and whether the
basket maker gives an extra sharp pull at the end, to tighten the fabric.
At first a palm leaf strand lasts for seven or eight stitches (Table 16-8,
wrapping strands number a-d). Further on in the base, and also in the wall of the
basket a n e w strand has to be put in after only four or five stitches(Table 16-8
wrapping strands number e-k). This is because the bundle is m u c h thinner near the
centre than in the rest of the base and the walls.
The first stitch (column 1) is done with the strand after it has just been
inserted and has its full length. It gradually shortens with each stitch. T h e
shortening goes relatively slow in the topmost rows (a-d), because the diameter of
the bundle is smaller.
The length of the wrapping strand is not clearly reflected in the number of
times the strand is pulled. The semicolons indicate instances of irregular, disrupted
pulling of the strand. The strand is not tightened after each stitch, and several
times the tightening is done twice or even three times.
The insertion of strand c-f shows regular breaks (in Table 16-8 indicated by a
semicolon) to rearrange the bundle and knead the centre. This is explicable,
because with these strands the stiff bundle material has to be forced into a small
circle. The previous strands were holding the more flexible palm leaf bundle (the
ends of the knot, folded double). These could be considered factors that explain
w h y the working pace is irregular. Looking at the long sequences in fast forward,
however, does not show any development of a steady rhythm, neither does a
pattern emerge in the count of the number of times the strand is pulled tight.18
Wrappin g strands (a-k) and n umber of stitches do ne with ea :h (1-8)
1 2 3 4 5 6 7 8
a 3 (1) 2 (1) 2 3 (1) 3 1 (2) 1 (D 1 (D
b 3 (D 4 3 (D 2 (2) 2 2 (1) 2 (1) 1 (D
c 2 3; 1 (D 1;2 1; 2 2 (1) 1; 1 1
d 1;2 3 i;2 (2) 2 (D 2 l; i (D 2 (2)
e 3 (1) 2; 1 (D l;2 (3) 1 (2) 2 (2)
f 2 (2) 1; 1; 1 2; 1 3 (1) 4
g 3 3 3 (D 3 3
h 3 (1) 4 (D 2 3 2
1 4 (1) 4 4 (1) 3 2
J 2; 1 (1) i;2 (l) 4 3
k 2 3 2 3
Table 16-8 Number of times in which the sewing strand is pulled through the bundle.
Each row (a-k) represents one strand, with which between four and eight
stitches are made. A semi-colon (;) indicates an intermezzo of kneading the
centre or positioning the bundle; Numbers in brackets indicate a tightening
of the strand by pulling it strongly. Above the bold line: base of basket;
under the bold line: wall of basket.
18
The pulling of the strand is counted from the first time she releases the needle and
actually takes hold of the strand itself.
It is not until working on the sides of the basket, that the basket maker finally
develops some kind of rhythm. The interruptions to knead or pull, have been
reduced, occurring only twice (jl and j2). Only strand g shows a truly consistent
pattern of three times pulling the strand, only tightening the strand once in the
third stitch (g3). There is no explanation for the consistent occurrence of three
pulls in the fourth stitch (g4 to k4), other than that it seems to be the length that is
easiest pulled through in three hauls.
The lack of working rhythm seems to be related to a lack of concentration.
Rauwhayya looks up regularly and the work stops. In the course of this chapter I
will also argue that working rhythm is related to skill and professionality.

The basket maker uses her hands and mouth for shredding palm leaf and fruit stem
fibre. She does not really use her feet or her body weight:
Rauwhayya sits cross-legged and is shredding date palm leaf.19 With two hands she
brings a bunch to her mouth, pulling one half of the bunch down with her left hand and
taking the rest out of her mouth with her right hand.20 This she does twice. She puts the
first shredded bundle down in her lap with her right hand (24:54) and picks up the second
bunch, levelling out the ends of the bundle by tapping them against her left hand (24:57).
Then she brings them to her mouth and shreds them in two movements. She picks up the
shreds in her lap with right, takes all the strands in her left hand and arranges them with her
right hand: foot by foot, top by top (25:07-25:11).
While looking up, she starts twisting the bundle clockwise (25:11-25:32).21 She makes
an overhand knot by putting the foot end of the strands over the top part.22 She takes over
the loop in her right hand and takes the top end of the looped palm leaf strands in her left
hand. Then she pushes the top half with the thumb of her right hand through the loop and
pulls the twisted strand from the underside with her left hand (25:22-25:32).
She then tightens the knot by pulling the two ends once. Then she takes hold of the
knot and the right side of the bundle and pulls twice. She changes her grip on the knot and
pulls the knot and the left side of the bundle once (25:32-25:45). Then she changes back and
pulls twice while holding the right side of the bundle and the knot. She looks up while
pulling and looks down while placing her hands.23
She then takes hold of the top half of the knotted strands with her left hand, folding
the bottom ends towards the top and taking the entire knot, which now has been folded
double, in her left hand. She kneads it twice to make the two bundles stay on once side
(25:46-25:49).
She puts the knotted piece in her lap and looks for the needle, her eyes focused on her
hands scanning the basket to her left and the other unfinished centre piece of a coiled basket
19A symbol indicating the working position has been inserted at 28:40.
20A symbol indicating the instrumental use of the body has been inserted at 25:00
~ The result is Z-oriented twisting
" The loop thus created is a Z-crossing, which is visible at 25:26.
23A symbol indicating the eye contact with the work has been inserted at 25:43.
which is lying in her lap. The needle is still attached to this and she pulls it off with her
right hand.24
She puts the pointed tip of a palm leaf strand through the eye of a small needle (ibra),
holding the needle in her left hand and the strand in her right hand. She takes over the
strand in her right hand and picks up the folded knot with left. She is ready to start the first
round of the spiral (25:58-26:02).
She does this by laying the foot end of the palm leaf strand under the folded bundles
that protrude from both sides of the knot. Holding the knot with her left hand, she fixates
the strand with her left thumb. She slides her right hand along the strand until she catches
the needle. Then she pushes the needle through the knot, from the side she is facing to the
back of the work.25 At the same time she chases a fly by shaking her head and making an
involuntary movement with her right hand. She picks up the needle at the back and pulls the
strand through the knot in three jerks (26:09-26:21). After the strand has been pulled
through completely, she gives it one extra tug, to tighten it.26
She then takes hold of the strand at the base and spreads it out by pushing her thumb
in the creased area which protrudes from the hole in the knot (26:24). She again slides her
right hand along the length of the strand, to straighten it out. She puts quite some force
behind this pulling movement. The wet palm leaf can stand a great deal of tension, while it
would break off when this was done to it while dry.27 Then she makes the next stitch. She
kneads the basket to form the centre, pulling the bundle into place (26:46). She looks to her
work while she pushes in the needle and looks up while she is pulling the wrapping strand.
Most of the time she places the wrapping strand with her right hand and holds it in
place with the index finger of her left hand (e.g. 27:00). Sometimes she uses her left thumb
(30:55). At the same time she holds the bundle in place with her left hand (27:36), wielding
the needle with her right hand. She twists the bundle slightly in Z-direction (visible at
27:39).
She leans over to the left, to pick up zaghauwa fibres with her left hand. She holds
them with right and then breaks them at approximately one third of the length. She drops the
short end from her left hand. She looks away while she holds the long ends with her right
hand. With both hands she brings the fibres to her mouth, the right hand being close to her
lips, and while pulling with her right hand, she bites in thefibre,making a small split with
her teeth. She then splits the fibre over the full length with her hands. She again looks away
(28:03-28:06). The basket maker seems distracted by what is happening around her: the dog
barks, the family discusses lunch with her, children are shouting, or come to ask for
attention or to be breastfed, the c o w loudly chews sugar cane.
She breaks and splits another part, lying in her lap. She holds the split fruit stem
fibres in her left hand while she manoeuvres the basket centre with her right hand. With left
she pushes the bundle material deep into the middle of the palm leaf bundle (28:23-28:28) .
Then she takes hold of the basket centre with her left hand, freeing her right hand to
take up the wrapping strand. She pushes a new leaflet around the bundle with her left
24
The extra centre piece is the result of m y interference. Before starting the video
shooting, I asked her if she could start the next day with a new basket. Early in the
morning, when I arrived, it appeared she had already started, so I asked her if she could do
it again, explaining that the start of the basket was perhaps the most important.
^?First wrapping of strand "a" (al) in Table 16-8.
" A symbol indicating the properties of raw material has been inserted at 27:27.
thumb, pulling the bundle around the basket centre with her left hand, while fixating the
wrapping strand in place with her right index finger (29:20). She takes over with her left
index finger (29:22), freeing her right hand to pick up a large metal needle with which she
uses as an awl to pierce the previous bundle.28 The difference with the previous stitches is
that the bundle consists n o w of m u c h harder material. In the sequences following, she
sometimes uses the awl, sometimes not.
She pushes the awl through the bundle with her right hand (29:25). Dropping the awl
in her lap, she picks up the needle, pushes it through the hole with her right hand, lets go of
the needle and takes hold of it again at the back, pulling it through with her right hand.
The next sequence is marked in the left lower corner with the indications rl2, olO and
ol3. The recurring action of holding the basket and strand in place with her left hand, while
piercing the previous bundle with the awl and putting through the needle with right, taking
up the strand at the back with right, pulling up the strand forcefully and arranging it with
her right hand, fixating it with her left hand, while piercing the previous bundle with the
awl..., is indicated as rl2.
Plate 16-9 Rauwhayya making a large flat bread basket with a diameter of three shibr
(el-Amariyya, 1992).
28A symbol indicating the use of instruments has been inserted at 29:37
The rl2 action is shown twice (29:40-30:04, 30:06-30:24), but before the strand is
pulled through completely for the second time, she takes the centre in her right hand and
kneads it with both hands, pulling the bundle strongly back with her left hand (30:04-30:05).
This is indicated as olO.29 The same code was used in the analysis of the previous basket
production sequence to indicate the kneading which was done at the start of the sewn plaits
basket. The movements are similar and the purpose of the action is the same. At the centre
of round baskets, the passive element, which is usually more rigid than the active element,
is pulled in a small circle and has to settle.
She continues by making the next stitch (rl2, 30:20-30:24), without further
interruption, but she kneads the centre as soon as she has pulled through the strand
completely (olO, 30:24-30:27).
She has come to the end of the leaf strand and ol3 indicates the occurring action of
inserting a new strand. She takes of the needle (30:27-30:29). Then she interrupts this action
and proceeds with some more folding and kneading (olO, 30:29-30:32). She reaches out
with her left arm to pick up a new leaflet from the cloth in the basket. All the time she is
still sitting cross-legged. She then looks for the needle in her lap, pulling apart the folds of
her dress with both hands. Shefindsthe needle, picks it up with her left hand and puts the
point of the leaflet through the eye with her right hand. She picks up the basket centre with
her left hand, the leaflet in her right hand, putting the broad end underneath the bundle,
fixating it with her left thumb (ol3, 30:33-30:55).
She then continues the recurring action of piercing with the awl, stitching with the
needle and pulling with her right hand (rl2, 30:57-31:05). She kneads before pulling
through the string completely (olO, 31:05-31:08), and continues pulling and stitching (rl2,
31:08-31:37).
The recurring action rl2 shows many irregularities. Sometimes she holds the basket
base flat on her lap, while piercing it with the awl, sometimes she pierces the bundle on her
left knee, other times she rests the basket base on her right knee, or holds it in the air
without any support. There is unity nor regularity in the movements of the recurring
actions. The occurring actions are by nature not as regular as recurring actions, and
consistent with the lack of working rhythm, show considerable irregularity in the work of
this basket maker.
To compare the physical differences between working on the centre of the basket, and
the basket wall, the basket maker was asked to continue working on the large, almost
finished, basket.
W h e n she starts working on the basket, her working position has changed. Her right
leg is bent, her left leg stretched out in front of her, the body weight slightly to the right.30
This change did occur earlier, however, when she was still working on the centre of the
basket (not shown in the edited version of the video).
She lifts the basket briefly with her right hand, to pick up the awl, which is still lying
in her lap. She throws the awl inside the basket, on top of the cloth with the new lengths of
moist palm leaf (31:42).
With her right hand she lays a new strand in, holding the basket with her left hand.
Briefly she uses two hands to lay the end of the previous wrapping strand parallel to the
bundle (31:52). She places the leaflet with her right hand, takes over with the index finger
29
" The transistion between olO and rl2 marks the end of the kneading and a movement
from her hand to the palm leaf strand to continue the stitch (30:05-30:06).
A symbol indicating the working position has been inserted at 31:47
of her left hand and looks away while picking up the awl with her right hand. She looks
back when she makes the hole. She pushes in the needle, takes hold of the strand at the back
of the work and pulls it through.31
Three stitches are shown in close up. During the third stitch it is clearly visible that
she pushes the needle through the bundle missing the hole she made with the awl (32:39).
The image freezes in the fourth movement (32:54).
76.2.5 Workspace
Rauwhayya works in the street in front of her house. It is a quiet back street,
which leads to one of the main public streets of Amariyya on two sides. The street
effectively functions as a court yard (Figure 16-9). In the same area their water
buffalo and c o w are kept. Behind the doors, in front of which they are standing, is
a roofless enclosure in which goats, sheep and a donkey are housed. T o the right
is a small roofed area used as kitchen. During m y visits R a u w h a y y a sits usually
somewhere in this court yard, often with her back against the wall. Only on
wintery days like today, she likes to sit in the sun. H e r materials are arranged
around her. Making a coiled basket does not require any installations. A bit of
space to lay out materials and tools is enough.
Figure 16-10 Sketch plan of the working area where Rauwhayya makes coiled baskets.
The striped areas are roofed. Drawing by H. Barnard.
3'This is indicated in Table 16-8 as sequence gl.

16.3 The Shauwer
. ~ Iîj rz^\\g4^fmi coiled basketry

m a d e by Khadidja in N e w Ibrim
,7 Sliij

T h e w o r k of the Nubian coiled basket makers follows mostly the order as indicated
in Section 15-5.
a) preparing the material (33:39-34:10).
The preparation of the palm leaf (cutting and dying) has not been shown. At the
beginning of the film Khadidja is sorting the palm leaf (33:39-34:05), which is
soaking in a plastic tub (34:09-34:10).
c) starting the centre (34:10-36:12)."
The centre of a basket is a double square plait. Half of one of the plaiting strands is
then used as the active wrapping strand, the other ends which are sticking out of the
plait are incorporated in the bundle material.
She splits one of the doam palm leaf strands sticking out of the plait, by
cutting the leaf with the nail of her thumb (35:59-36:01). The ends of the
bundle material (shredded fibres of the fruit stem of the date palm, called
imbi in Fadidja Nubian) are wetted in a plastic tub in which also the doam
palm leaf is soaking (35:47-36:12).
d) making the basket (36:13-41:12)
The basket maker pierces the previous bundle with an awl and pushes the pointed tip
of the palm leaf strand through the opening, pulling it tight. She regularly puts in new
palm leaf strands and bundle material. She also changes the coloured strands to make
a pattern.
At several instances she works on the raw material. She wets the palm leaf
strand with which she is working (37:21), or the foot of the bundle material
(38:37-38:41). Most of the times she puts in n e w bundle material, she
divides the fibres into smaller strands (e.g. 40:36-40:54).
e) ending the basket (not shown)
The basket is ended by diminishing the bundle in size and fastening the tip of the
bundle with the last stitch.
f) adding rim decoration (33:34-33:39) and foot / handle (not s h o w n )
A decorative, strengthening rim is added, by sewing a leather, or mock-leather strip
spirally around the last coil. The end of this is seen at the beginning of the sequence.
"What not occurs in this production sequence is the preparation of the systems (phase
b), or the finishing touch (phase g).
On the convex side of the basket, around the centre, a coiled circle is sewn. It has a
diameter of 20 c m and is two rows high. It functions either as foot of the basket , as
support (when the basket is carried on the head) or as handle (when the basket is used
as food cover).

Making a coiled food cover takes approximately 80 hours. T h e w o m e n of Nubia
produce the basketry in their spare time only, after having taken care of cleaning
the house and preparing the food. Still, a considerable part of the day is dedicated
to basket making, so that the hours necessary to m a k e the basket are divided over
approximately two weeks.
The w o m e n usually work in the afternoons only, gathering in the court yard
of one of the houses. There is no specific time of year for basket making. Only in
the month of R a m a d a n no baskets are made, because the w o m e n are too busy with
preparing extra food and cleaning the house meticulously, especially just before the
'Eid al-fitr, the grand feast at the end of Ramadan. T h e eighty hours have been
summarized in the video into a sequence of not more than eight minutes.

In Table 16-11 a micro analysis is m a d e of the w a y that Khadidja performs the
recurring action of making a stitch, by counting the number of times she pulls the
strand after making a stitch.33 This is the same approach as used in section 16.2.3
(coiling in Middle Egypt). Comparing Tables 16-8 and 16-11 shows that Khadidja
is working in a m u c h more steady rhythm than Rauwhayya. Mostly she pulls the
strand through in two times, giving it an extra tug to fasten the stitch. In the table,
this is indicated as 2 (1).
There is some variation in the m o m e n t this small tug occurs. Mostly it
follows the pulling of the strand immediately. Sometimes it is done after the awl
has been put into the fabric. T h e number of times she pulls the strand gradually
lessens from three to one. A second basket maker, working in the same courtyard,
also mostly pulls twice and gives one tug (not shown in the edited video).
The length of the d o a m palm leaf strand exceeds that of the date palm leaf by
approximately 10 c m . The m a x i m u m number of stitches with one date palm strand
is eight (Table 16-8). T h e d o a m palm leaf strand makes easily more .than 13
stitches, not only because the wrapping strand is longer, but the diameter of the
bundle is also less than in Middle Egypt.
The following section is based on the complete footage, since the edited version
shows only brief sequences of the recurring action of putting in the awl, pushing through the
palm leaf strand and pulling the strand through the hole in the previous bundle.
V w-i
t j: SZ
t t
_ l— o ^
CU OO
r> CN t t fN m
^-^
. •ca Tc3 —
^—'
IN -
CU u
_ ^ ^~.
"O ^^ --^
— t - — CN —
^^ ^^, , n
£ w oo -— ^^ "~
© t - - CN t — — t
rO o
2, — CJ — ^, r*. ^2, CU C- •o
sz OS CN t <N CN — fN t CN CJ
o
<u ^ r^^ ^-^ ^-, ^ •a
SZ '— ' ''*—' — '1 ' c
5 90 CN - fN cn CN fN — i
a
,—^ ^-^ .—,
•oo ~ ^, ^, c -—•- C-
'—,
00 r~ CN , r*l cN t CN fN _ CN
x:
o
-^ . , ^_ ^*, ,—. ,—, ,—. .—. .—.
coo
*— *-^ S '•*—' ^^ ^^ -—'
o vo <N f*l m m CN CN CN CN
E ^ -^ ^-v ^> •o ^ m ~
3C C OO CN
1/1 <N m fN cN t CN '5 CN t
•u
e
z* ™ O c C c ~
cu •V CN fN rn CN t CN cN ^ CN
OO
T3
C
c
ua,
OO r«1 CN fN m CN CN CN
OO
(N CN
c
Q.
D. ^^ "*—' •—-' -—' *—' ' '—' -—'
<n rn fN m CN CN CN (N CN
*—' s—' s-' ^^ '-' w" —' *—

— cN r*~i m CN CN co m m
—1
\D ju JLJ ^ •a •a u T3 •a •o w T3
c CL CL — CU IS CU IS 1) CU IS .H CU
1—3 a C-
1—
CJ
IS cq 3
•o C C 3
CL Q. s £ ? s £
HC
O tn ctj _C CJ •a tu 4- CD •— CU OO CU OO — CO —
3 r-
00 H •* 0
-:
3
O S CU oo X- -J
|M C/)
SZ TD
m -a C
-4 * "2
O ;-
T3 a. o £ — s u
\o
(N _* CJ —
t C <U CJ
CU OJ CJ
i; °
^H n cu
~ — °° -^ C >
-
O = , OO S o
FT r. ... H -S
o "^
- = lV.£ [fl u X
c_ -* 73 u 0 T3 :j s. U .
•3 ;: T3 — 3 -a ^
C oo C U u "2 t/3 Cl>
ed _ — m — OJ «J C:Cj 2TJ
o •3 ZZ U Q CJ •"" .^> ^-
- - ir x u-
oo
S CU 73 CJ CU •a CO '— "™ •"'
00 73 c 73 J'. 73 > 73 — iL ifl
- - CL 3 73
73 75 T3 u
r/1 73 b j* C <
t- c 73 CU
- CN C >< c b 73•~ E
* UCU - - t/5 73H
u X U 1 u E 11 >> Si)
u -3 E co u uC
NO -a "3 c> 3
1 — c n 1j,C
VO M . — — C H 3 c t> -X, (
/jt/1 x — w = c7
C P3 - ? tJ
s
3 x; oo 'y.
t 1 % s =1
m O ca oj_' w _ (-
- CN g-2
^ ca
•5 E _- u ^
CU <u cu
Tf m CN
^ o u u ra
m c
CN CN ^ oore."3 •£
- OO CA 3 £
P- * 3 3 "3 X:
o
CN CN w o. -3 a -g •="
•=-
c u
S
i-
•a CN CN ^ XI CU •5
c 5 a
CU CU
3 •5 2 o. ca cu o- IT'S
u OJ
s *
C
C lS S ca i: cu cu o — «
o Q. i/i X: i < S. E
o rn rn
—'
vO £ -a u -3
sz CU CU
_CJ ? S
sz '—' w'
c
i~a "- Zd •- M x>
ca
Strand a (Table 6-11), is replaced by strand b, not because the end is

reached, but because the tip of the leaf is split. Since the basket maker does not
use a needle, she needs to have a pointed, unsplit leaf tip. Khadidja uses a second
strand protruding from the plaited centre (strand b). For the same reason, strand b
is abandoned before the complete length has been used (stitches b 1-8). The nine
stitches of strand c and the 13 stitches of strand d do not represent the full length
of the leaf either, since both sequences are not completely present in the full
footage.
Even while regularly changing the coloured strands (e-k), she keeps the same
rhythm, in which the two pulls and one tug are represented most often. Mostly
three pulls occur only when she does not take a good hold on the strand. Towards
the end she often pulls only once, for instance thefifthto eleventh stitch of strand i
and the eight and ninth stitch of strand k.
Several times w h e n changing colour, she winds the wrapping strand one time
around the bundle without fastening it with a stitch. The end of strand h is
similarly wound around the bundle three times. The ends of the strands are
incorporated in the bundle and fastened with the next stitch of the n e w wrapping
strand.
Another indication of working rhythm is the little knock of the awl against
the finished part of the basket. This does not occur while the centre is very small,
but after three rounds it suddenly starts. It is caused by the movement with which
she swings the awl back in her hand, after holding it between pink and ring finger
(cf. Section 16.3.4) w h e n she prepares to pierce the previous bundle (e.g. 40:25,
40:32, 41:10). This little knock can be compared with the one in thefirstvideo
sequence, where the basket maker slapped his needle on the body of the basket
before making the next stitch.

The first action shown in the video is the sorting of doam palm leaf strips, which have been
died purple. Khadidja picks up leaf strips with her right hand, arranging them in her left
hand with which she holds the bunch. She is sitting on her knees, with her legs folded
underneath her (cf. 35:53).
She then proceeds to make the centre of the basket, holding two purple doam leaf
strands of 15 m m wide (a and b) cross wise on top of each other. She makes two loops with
strand a around strand b. She holds the strands with her left hand and with her right hand
she moves about to arrange them in the proper position (34:11-34:28).
She makes a plait out of the two loops by crossing the two ends of the perpendicular
strand b over and under the loops of strand a (34:28-34:50). With two hands she tightens the
square plait (34:50-35:04).
She turns the plait over and she repeats the entire sequence, but this time she makes
the loops with strand b and fixates them with strand a. Again, she holds the plait with her
left hand, while arranging the loops with her right hand (34:05-35:37).
Then she bends forward, stretches out her right hand, picks up a bunch of fibres of
the date palm fruit stem and dunks them briefly into the water tub (35:50). She takes over
the bundle with her left hand and taps the ends of the bundle against her right hand to level
them (35:52). She moves the bundle from her left to her right hand and back. Then sh
picks up the square plait with her right hand. Holding the bundle between the little finger
and ring finger of her left hand, she takes hold of one of the strands protruding from the
plait, splicing it with both thumb nails. She slides her right hand upwards to split the leaf,
thus creating a strand with the proper width for a wrapping strand (35:56-36:00).
With her right hand she holds the plait and sewing strip, with her left hand she puts
half the purple strand parallel to the bundle, and positions the bundle along the edge of the
square plaited centre (36:07).
Holding the bundle and the centre plait with her left hand, she picks up the awl with
her right hand and sticks it into the plait. She then runs her right hand along the length of
the wrapping strand. Holding the tip of the strand, she makes an anticlockwise stirring
movement with the awl, to enlarge the hole. She pulls out the awl with her right hand and
holds it with her little finger and ring finger, while she pushes the leaf tip through the hole
with her thumb and index finger. She lifts the centre while doing this, probably to see
better. She n o w holds the centre between her left thumb and index finger, while the bundle
is fixated between the little finger and ring finger of the same hand. She takes hold of the
wrapping strand at the other side of the hole and pulls it through in two tugs. She tightens it
with one little pull. Although the strand is narrow (approximately 2 m m wide) it can stand
the tension of pulling, because it is wet.34 She then moves the strand over the bundle back to
the front, also with her right hand, and puts it in place. She holds its position with the index
finger other left hand (36:12-36:28). This action is repeated many times.35
The awl she uses is a metal-pointed tip with a wooden handle. A s an alternative for a
proper awl, any pointed object can be used. The w o m a n w h o is finishing the rim decoration
uses, for instance, a metal part of a sewing machine (33:52). The long, sharp thorns with
which the lower parts of the date palm leaf are flanked (cf. Plate 14-4, p. 280), are
sometimes also used as a basketry tool. In interviews, older w o m e n mentioned that in Old
Nubia no other tools were employed apart from these thorns.36
Khadidja takes the plait in her right hand and pushes the bundle forcefully in an angle
of 90° around the square centre, all the while holding the awl between her ring finger and
little finger (37:04-37:06).
She takes over the centre in her left hand, picks up the wrapping strand with right,
positions it around the bundle, holding it with her left index finger. Swinging the awl in her
right hand from between her little finger and ring finger, she takes it between her right
thumb and index finger and makes stitch a5 (37:09-37:21).
She pushes the awl into the plait to start stitch a 6, then she bends forward to wet the
fingers of her right hand and she runs them along the wrapping leaflet, to keep it moist and
flexible (37:21-37:23). She makes stitch a 6 (37:26-37:31) and stitch a 7(37:34-37:45). She
then takes over the plaited centre in her right hand, while she bends the bundle with her left
hand in a 90° angle, forming the third side of the square (37:45-37:56). She makes stitch
a 8 and puts in the awl to start stitch a 9 (37:56-38:07).
A symbol indicating the properties of the raw material has been inserted at 36:49
35In Table 16-11 this sequence corresponds with wrapping strand stitch a 1. Stitches a
2, a 3 and a 4 are included immediately after at 36:28-37:04.
She is now sitting cross-legged, a working position which she took after finishing the
centre plait.37 In this position she makes three more stitches (38:10-38:34).38 She puts in the
awl, to start the next stitch, but then proceeds bending the bundle along the fourth side of
the plait (38:34-38:42). She stops for a moment and bends her upper body forward to wet
the bundle in the tub (38:38). The concentration is visible on her face. Her eyes are
constantly focused on her work and she talks very little (38:42-38:46).
The edited version of the video picks up the process after a few stitches have been
made with a red strand (e 1-3) and a white strand is about to be inserted (f 1).
Khadidja puts the foot of a white strand under the bundle with her right hand, fixates
it with her left thumb. With her left index finger she pushes the red strand aside, parallel to
the bundle, and fixates the white leaf in place. She moves the awl, pushes the white leaf tip
through the hole and pulls it tight (38:46-39:03). Her working position has changed. She
has pulled up her right knee as a support for the basket.
She makes three more white stitches (f 2-4; 39:03-39:34), then she pulls the white
strip towards her and holds it with her left index finger. With her right hand she takes up
the red strand (e) which was part of the bundle. With her left hand she pushes the white
strip parallel to the bundle and with right she places the red strand on top of the bundle,
holding the strands between her thumb and index finger (39:35-39:44). She takes over with
left and pushes the awl into the previous bundle with her right hand (39:45-40:14).
The next sequence displays the recurring and occurring actions in the left bottom
corner. She makes a stitch (action r 12) with a white strand. Leaving the awl in the fabric,
she takes the basket in her right hand (40:19-40:35). Then she leans to the left, reaching out
for fibre filaments with her left arm (40:35-40:36). She has lowered her right knee, and has
stretched her right leg in front of her (cf. 40:55). This is the transition between the
recurring action of wrapping / stitching and the occurring action of splitting the fibre (action
o 14).
She puts the basket in her lap and takes the fibre bundle in her right hand. She breaks
off the bent part with her left hand. Then she splits the foot of a fibre filament with the nail
of her left thumb (40:42) and pulls the fibre apart with two hands (40:43). The second
filament is also split with the thumb nail, but now she uses the foot end of a yet unsplit
filament to divide the fibre over the entire length (40:49). The third time she uses the awl to
split the entire length of the filament (40:52).
The next occurring action (ol5) is wetting the fibre. She gathers the fibres in her right
hand, then takes the bundle in her left hand and bends forward to wet the foot of the bundle
(40:54-40:57).
Then the action of inserting bundle material (ol6) occurs. She taps the fibres on her
right hand, to even out the foot of the bundle. She then picks up the basket from her lap
with her right hand and pushes the bundle material in the middle of the basket's bundle with
her left hand (40:57-41:04). The transition to the recurring action of wrapping and stitching
(t 16-12) is that she raises her right knee and positions the basket on top (41:04-41:07). The
last sequence shown is the start of another stitch (41:08-41:12).
jgA symbol indicating the working position has been inserted at 38:11
The next stitches have not been related to the sequence in Table 16-11, because they
were taken from different parts in the process, to show different angles and aspects of the
process.
76.3.5 Workspace
The w o m e n gather in the afternoon in one of the houses, sitting in a shaded part of
the court yard. The court yards in Nubian houses are spacious and partly covered
(cf. Figure 7-58). In the past, in old Nubia, the w o m e n went out much more,
herding the animals outside the village. They would sit together under a tree and
make baskets there. N o special provisions are needed, just some space to lay out
the materials and a tub with water to soak them in.
Figure 16-12 Sketch plan of the court yard, most of which is roofed over. Drawing by H.
Barnard.
16.4 TheMasballa
twined matting
made by AmrTt in el-Amariyya
The working order follows exactly the outline of production phases as listed in
Section 15-10, p. 319.
a) preparing the material (not shown)
The preparation of the date palm fibre (wetting and making tufts), has not been
shown.
b) making the string for warp and weft (not s h o w n )
The warp and weft are made of the same string (zS2), which is made with the hand
rolling method as shown in the first sequence (07:15-07:58).
c) setting up the warp (not shown)
With the help of an assistant, AmrTt rolls up the string in wide loops. H e puts two
metal bars in the loop and stretches the loops by hooking the bars behind four wooden
pegs in the ground. Then considerable time is spent on arranging the warp so that the
loops do not overlap and the tension on the warp threads is evenly spread, after which
the twining can be started.
d) making the mat (42:10-46:14)
A process of four hours has been represented here in just over four minutes. The side
edge of the mat is slowly formed during the production phase of making the basic
structure (cf. Table 16-13).
e) taking the mat of the loom and making the edges (46:15-47:19)
Taking the mat off the loom only takes a minute. In the edited video the making of
one edge has been incorporated.
g) tapping the mat (47:19-47:34)
At the end of the process, AmrTt taps the mat with his hand. Perhaps this is meant to
shake out the dust, as a final touch, but more likely it is just an indication that the
work has been truly finished. Immediately after, he offers the mat for sale.

A m r T t can m a k e a masballa in a little over two days. H e spends approximately ten
hours to m a k e one mat. Making a masballa takes twice that time, plus the time to
s e w the two mats together. Making the string takes six hours, setting u p the warp
takes thirty minutes and twining the mat takes at least four hours, after which the
edges are made in less than 10 minutes. He occasionally takes a break to smoke a
cigarette, or drink tea, while sitting on top of the half-finished mat.
AmrTt works full-time on making this type of matting. H e works all year
round. Palm fibre is usually harvested in the autumn, just after the date harvest,
but can be harvested at other times of the year as well. H e has a storage with palm
fibre, because well kept date palm trees are cleaned once a year, when the old
leaves are cut in steps and thefibreis torn off the base of the leaf crown.

AmrTt is a professional mat maker. H e has to make production to sustain his
family. H e works concentrated, although the entire village youth has gathered in
front of his house in the hope to get their faces on 'television'. H e keeps a steady
working rhythm, which varies slightly during the work, depending, for instance,
on the length of the twining strands. Since the recurring action of twining is very
brief (mostly less than two seconds), many repetitions can be seen in the edited
video. Therefore, the timing in Table 16-13 is based upon the sequences in the
edited video tape.
The duration of the actions is calculated in the smallest unit that occurs in the
timer in the bottom right-hand corner of the screen. This unit is not divided in 60
parts, but in 25, indicating the individual video frames. The column timer mm.ssff
(minutes, seconds, frames), refers to the precise location of the 51 twining actions
in the video.39
The sequence starts with a newly inserted strand 1, and a m u c h shorter strand
2. The first strand is inserted at a slower pace than average, but the next five
actions are all done in less than two seconds (50 frames). The seventh twining
action takes just more than two seconds, because the entire length of the string is
pulled through the warp threads. In general, the mat maker performs a number of
twining actions in quick succession, during which he does not pull through the
entire strand, after which he pulls free the twining strands. Actions 7,11 and 17
take longer than the rest, because these are when the entire strand is pulled
through.
The working rhythm is also expressed in a unity of movements. During
action 2 to 10, the mat maker pushes the twining strand into place with the middle
finger of his right hand. During action 15 to 19 he uses either his middle finger,
index finger or thumb. In the unedited footage he clearly mostly uses his middle
finger.40
AmrTt twines from left to right in S-orientation, and from right to left in Z-
orientation. Most of the work is done with the first hand: when he twines from left
39
The option of showing frame by frame is not available on all video players.
This use of the middle finger was not noticed until the analysis of the video. The
possibility should be kept in mind that he had a sore index finger, and was using his middle
finger only temporarily. At the moment of m y visit it did not occur to m e to ask.
to right, he uses his right hand to guide the strand through the warp, working from
right to left his left hand. With the other hand he just pushes the strand in its place
in the mat. Even though different hands are used, there is no difference in working
rhythm between S- and Z-twining. The times of actions 1 to 21 are almost all
between one and two seconds.
Table 16-13: timing of twining actions 1-51
no frames as remarks timer (mm:ss:ff)

working from right to left (Z-twining)
1 64 1 twining with new strand 1 42:23:21-42:26:10
2 48 2 push with right middle finger 42:26:10-42:28:08
5 37 1 pulling strand 1 through the warp 42:31:13-42:33:00
12 31 2 twining around the last warp thread 42:42:24-42:44:05
13 148 1 making the left side edge 42:44:05-42:50:03
working from left to right (S-twining)
14 44 2 push with left thumb 42:50:03-42:51:22
15 36 1 push with left index finger 42:51:22-42:53:08
16 31 2 push with left middle finger 42:53:08-42:54:14
17 29 1 no push (pulling strand with 2 hands) 42:54:14-42:55:18
21 56 1 pulling strand with 2 hands // 42:59:19-43:02:00
22 60 1 pulling through new strand 1 44:04:07-44:06:17
24 124 1 pulling through long strand 1 44:08:13-44:13:12
25 63 2 short strand , hindered by long strand 1 44:13:12-44:16:00
26 167 1 long strand 44:16:00-44:22:17
27 143 2 making the left side edge 44:22:17-44:28:10
Table 16-13 continued:
no frames as remarks timer (mm:ss:ff)

working from left to right (S-twining)
28 90 1 // (frozen image) 44:28:10-44:32:00
29 43 2 short strand 44:39:19-44:41:12
30 89 1 long strand 44:41:12-44:45:01
31 78 2 short strand, but tight warp 44:45:01-44:48:04
32 165 1 long strand, hesitant movements 44:48:04-44:54:19
34 120 1 long strand 44:56:15-45:01:10
35 92 2 new strand 2 (inserted 45:01:10-45:12:10) 45:12:10-45:16:02
37 93 2 long strand 45:19:24-45:23:17
42 79 1 push with left middle plus index finger 45:41:15-45:44:19
43 85 2 push with left thumb, arranging warp 45:44:19-45:48:04
44 56 1 no push 45:48:04-45:50:10
45 70 2 pulling strand through warp 45:50:10-45:53:05
48 202 1 right edge with strand 1 45:58:17-46:06:19
49 80 2 push with right thumb 46:06:19-46:09:24
50 75 1 no push 46:09:24-46:12:24
Table 16-13 Timing of twining actions (numbers 1-51).

as = active strand (indicates which one of the two twining strands is active).
frames expresses the duration of the action (25 frames = 1 second).
timer refers to the video timer in the right bottom corner of the screen and
lists the sequence in minutes, seconds and frames (mm:ss:ff).
// indicates a break in the video registration.
This changes drastically in actions 22 to 51, where the actions take much
longer. These actions take place, approximately three hours after numbers 1 to 21.
Although fatigue m a y play a role, there are reasons inherent to the work stage,
which can account for the slightly slower, but still regular pace. First of all a new
strand 1 has just been inserted. The length of the strand has to be pulled out of the
warp threads and this takes time. Secondly, the mat maker is near to the end of the
mat. The empty part of the warp threads is short, there is not m u c h space to work
in and the tension of the warp is m u c h higher than in the beginning of the work.
Thirdly, the level of the warp strands differs: the odd strands, which run over the
iron bar, are 20 m m higher than the even strands, which run under the iron bar.
At this stage twining strand 1 takes a relatively long time: action no. 24 takes
124 frames, which equals five seconds, action no. 26 lasts 167 frames (almost
seven seconds), no. 30 takes 89 frames (three seconds), action no. 32 needs 165
frames (more than six seconds), and action no. 34 lasts 120 frames (almost five
seconds). Twining with the short strand 2, still takes less than two seconds,
although at one point there is a brief tangle with strand 1 (action no. 25, 63
frames) and the tightness of the warp causes a little struggle in action no. 31 (78
frames, more than three seconds).
W h e n a n e w strand is inserted to replace the short end of strand 2 (no. 35),
the timing evens out and settles between two and three seconds per twining action.
This is almost a second longer than in the beginning of the work, due to the high
tension of the warp threads and lack of space to manoeuver the strands.
Making the side edge always takes longer than a normal twining action.
Three examples occur (no's 13, 27 and 48). The duration of the actions is
respectively 148 frames (six seconds), 143 frames (six seconds) and 202 frames
(eight seconds).

A brief description of setting up the warp is included here, though it does not
occur in the edited video.
The mat maker squats down while he faces his assistant, who is squatting opposite
him. The distance between the two men is approximately one metre, the length of the loom.
The mat maker holds the rope in his right hand and reaches out with his left hand to give a
loop of string to his assistant, who accepts it with right and takes it in his left hand. This is
done 18 times, but the mat maker does not count the loops, he continues until the string runs
out. He has measured the length of warp string while making it.
Both men pick up an iron bar, lying on the floor, near the wall, to AmrTt's left side.
AmrTt, holding his side of the string loops in his right hand, picks up the bar with left. He
and his assistant stretch the warp strands by putting the bars into the loops and hooking the
bars behind the two sets of wooden pegs, which are pushed in the ground at a distance of
approximately 0.95 metre. With a looped half knot (cf. Figure 15-17, p. 320), the mat
maker ties the end of the string around the metal bar, to his left. H e and his assistant work
from left to right, carefully positioning the warp strands, spacing them out evenly and
pulling the string to get the proper tension. This results in excess material, which is wound
around the bars a 19th time. In total there are 38 warp strands. With this, the work of the
assistant is finished.41
With the end of the warp string the mat weaver weaves a starting border from right to
left (Figure 15-17 a, p. 320). Because the warp lies around the bars, there is a difference in
level: the top strands are running 20 m m above the lower strands. By weaving over the top
strand and under the lower strand, he pulls the top strands down and the lower strands up,
to get all strands on the same level. The end of the weaving strand is his first twining
strand, which he lays out over two warp strands. H e needs a second twining strand and for
this he uses the start of the warp, which is still tied around the left peg. H e loosens the knot
and moves twining strand 2 under and over the leftmost warp strand, pushing it under the
second warp strand (Figure 15-17 b). H e brings the strand in position by pulling it. Then he
runs the twining strand 1 under the third warp strand.
The mat maker is squatting, his knees folded in front of his chest, his toes almost
touching the front bar of the simple ground loom. This working position evokes the Middle
Egyptian text Satire of Trades, where it is said of a mat weaver: "with knees against his
chest, he cannot breathe air" (Lichtheim 1973, 188). The depiction of a mat weaver in the
tomb of Khety in Beni Hassan (cf. Figure 15-13, p. 313) also represents a squatting figure.
Plate 16-14 Squatted work position of the mat maker AmrTt (el-Amariyya 1992).
41The distance between the two bars is 910 m m and the width of the mat 760 m m .
Although the technique of woven matting differs from twined matting, the position of the
mat maker in front of the loom and, when the mat gets larger, on top of the finished part of
the mat (from 44:04 onwards), is the same.42
With his left hand, AmrTt picks up a length of string. H e takes the looped start of the
string in his right hand and twists it open with a small anticlockwise turn of his left hand
(42:14-42:20). With his left hand he takes hold of the short strand and with his right hand he
pushes the loop around the short strand. With a small, strong pulling movement, he puts the
loop in place (42:20-42:22).43
With a long strand he resumes twining, working from right to left (twining in Z-
direction). With his left hand he pushes the twining strand under the warp, lets briefly go,
and in the same movement pulls the strand up. With the middle finger of his right hand he
pushes the strand into place (42:23-42:26). This train of actions is repeated hundreds of
times before the mat is finished, only 51 of which have been incorporated in the edited
version of the video. Of these the numbers 1-13, 22-27 and 49-51 are twining actions from
right to left (cf. Table 16-13).
Variation occurs in the direction and orientation of twining. Working from left to
right (no. 14-21 and 28-48 in Table 16-13), the action is repeated in mirror image: the mat
maker pushes the strand through the warp threads with his right hand, pushing them in place
with his left middle finger. The result is a row of S-oriented twining. Apart from using his
middle finger, he sometimes also pushes the strand into place with his thumb or index
finger. These occurrences have been indicated in Table 16-13.
The mat maker constantly manipulates the twining strands, while he touches the warp
strands only briefly with his finger knuckles. The twining strands are, therefore, considered
the active elements, while the tension of the warp stretched between the loom bars, make it
the rigid, passive, system.44
The twining movement around the warp strands is the recurring action of making
twined matting (rl7). This action is interrupted when a new length of twining string is
inserted (occurring action ol8) and when the edge is reached (occurring action ol9).
AmrTt, working from left to right, pulls the twining strand with his right hand. He
pulls strand 1 through the warp, using two hands. Strand 2 is only 30 cm long. The insertion
of a new twining strand (45:01-45:12) starts with AmrTt stretching out his right hand and
leaning over to take up a new length of string. H e holds the string in his right hand and
twists open the loop with his left hand. H e looks up to talk to someone while he puts the
loop around the short end of twining strand 2. He pulls the loop in place with a small jerk.
It is not until he continues twining, that he looks back to the work.45
The side edges are made at the end of each row of twining. At the right edge (Figure
15-17 at c, p. 320: 45:58-46:06) he pulls strand 2 with his right hand around the last warp
strand. He pushes the strand with his left middle finger in place. H e then takes over the
strand in his left hand and pulls it across the warp. With his right thumb and index finger,
he squeezes strand 2 around the warp. H e takes strand 2 in his left hand and picks up strand
1 with right. He pushes strand 1 also underneath and around the last warp strand and pulls
the loop tight, guiding strand 2 with his left hand. Strand 2 is now fixated by strand 1 and he
42J
3A symbol indicating the working position has been inserted at 43:25.
4\
Â symbol indicating the properties of the material has been inserted at 42:29.
44„
The frozen image of making twined basketry is included at 45:31. The warp is the
passive_system, the two twining strands form the two members of the active system.
A symbol indicating the work-eye contact has been inserted at 45:09.
lets go of it to take over strand 1 in his left hand and pulls it across the warp. H e now
squeezes strand one around the warp with his right hand. H e then pulls strand 1 underneath
and around the second warp strand and continues the twining action. The left edge is made
exactly the same, but in mirror image.
The mat is finished when the warp loops have become too tight to twine in more
lengths of string. The mat maker twines the last row from left to right, until the shortest
strand has almost run out. The last 10 c m he twines from right to left. The longer twining
strand protrudes from the mat, without being fastened (not shown).
After finishing the mat, AmrTt stands up and bents over, unhooking the cross bar
from the pegs and lifting it with his right hand while pulling the bar out of the warp loops
with his left hand. Without moving his feet, he stretches out towards the other side of the
mat, and pulls the second bar out in the same fashion. H e then sits down to make the top
edge, which he starts in the middle (46:15-46:26).
Plate 16-15 AmrTt and his family in the narrow hall way of his house, which is at the
same time his work shop (el-Amariyya 1992).
The next sequence shows the finishing of the bottom edge (46:27-47:21). AmrTt is
sitting cross-legged with the mat on his lap and the row of loops in a line perpendicular to
his body. There are eighteen loops and he starts with the loop just left of the middle,
without counting the loops. Working away from his body he holds loop 1 in his right hand
and pushes loop 2 through loop 1 with his left hand. H e then takes over loop 2 with right
and pulls loop 1flatagainst the fabric with his left hand. Then loop 3 is pulled through loop
2 (Figure 15-18, p. 231). His eyes wander away from the work regularly. The chain of
loops ends when he pulls the last loop through the fabric and fastens it with a half knot. The
other half of the row of loops he works towards his body, ending the chain with a second
half knot.
16.4.5 Workspace
AmrTt always works in the small hallway of his house (Plate 16-15). It is the only
place in the house where there are pegs in the floor to hold the loom bars. The
light comes in from the doorway, but he himself is mostly in the shade, especially
in s u m m e r when the sun is high. Working in his doorway, which looks out on the
village square, gives AmrTt regular contact with passers by. H e sits facing the
door opening, with the wall close to his left side and a passage w a y into the house
to his right. Behind him there are a large, dark room and a court yard.
Like his squatted working positing, the work space also reminds a little of
what is said of the mat-weaver in the Middle Kingdom Satire of Trades, where it is
said that he gives food to the doorkeeper to let him see daylight (Lichtheim 1973,
188). M a n y mat makers (both those making twined and woven matting) work
inside a roofed area, although sometimes the large mat makers' looms are placed
in the court yard. Luckily, unlike the mat maker from the Middle Kingdom text,
which is set as an example of the wrong career choice, AmrTt is not being beaten
if he stops working. H e is his o w n boss.
16.5 TheSerirbilHabl
woven bed matting

made by Shahed in Daraw
The working order of making woven matting, with two parallel strands follows
precisely the process phases as indicated in Section 15.9, p. 317.
a) preparing the material (not shown)
The preparation of the grass (wetting), has not been shown.
b) making the string for warp and weft (not shown)
The warp and weft are made of the same string (zS2), which is made with the hand
rolling method as shown in thefirstsequence (07:15-07:58).
c) setting up the warp (49:58-50:58)

The warp is set up between the foot of the bed frame and a yellow nylon rope which
is tied at 20 cm distance from the head of the bed frame. The warp is inserted with
double strands. Shahed pulls a loop of string over the length of the bed. He pulls the
looped end around the bed frame and pushes it through the previous loop. By pushing
a wooden peg in the loop he prevents it from undoing, until it is fixated by the next
loop. Setting up the warp takes a large length of string. Although the bundles with
which Shahed is working contain approximately 25 m string, he regularly has to
connect the end of the old string with the start of a new bundle.
When joining two ropes, Shahed thins out the frayed end of the exhausted
string (48:20-48:53).46 He wets both the thinned-out frayed end of the old
string and the frayed end of the new string (48:47-48:53). Then he hand-rolls
the old string to a thin tapering end and proceeds to join the two by
incorporating the thin tapering end into the new string (49:28-49:13).
d) making the bed (50:58-57:30)
A process of four hours has been represented here in just over seven minutes.
e) ending the weaving (57:31-57:59)
By pulling the weft strand through the last loop of the weft and the starting loop of the
warp, the fabric is ended. The last 50 cm of string are anchored in the weft.

Shahed works only some days per month and only in the afternoons after 14.00
hours w h e n his job with the government ends. H e spent one afternoon of six hours
making the string and one afternoon of six hours for making the bed. If need be,
the grass can be harvested all year round, although the best quality is found before
the summer. H e has a supply of string bundles to be ready w h e n there is a demand
for a bed. Although string beds are for sale on the weekly market in Daraw, he
does not produce these. H e only produce beds w h e n they are ordered.

There is a clear working rhythm in the different stages of weaving the matting of a
bed. At first the warp is inserted, for which string is taken from a bundle that
contains 25 ba'at of string (approximately 37.5 metres).47 The string is pulled from
the bundle over the length of the bed and fastened by linking a loop of the string
through the previous loop. This takes between 10 and 20 seconds (Table 16-16). In
general the direction head to foot takes two to five seconds more than the other
way around, because the bed maker has to pull free the string from the bundle.
The second to fourth strings take relatively long, because the bed maker has
to arrange the first loop of the chain. Around the 22nd and 23rd warp strands the
average of 12 seconds is slowing d o w n to 18 and 19 seconds. There is no obvious
reason for this. All movements are just executed slower than usual, which might
A symbol indicating the properties of the material has been inserted at 44:20
One ba'a is the length of two outstretched arms, so approximately 1.50 m.
be a sign that he is tired. The 25th and 26th strands are a bit slower than average
(15 and 17 seconds) because the string is curling up and has to be unwound. The
bed maker stops working and untangles the bundle of string, laying it out across
the street to prevent it from getting knotted.
Making the W a r p (64 warp threads)

foot to head head to foot
no. seconds no. seconds remarks
1st 13 2nd 18 starting phase of the warp
3rd 19 4th 20 talking, starting phase of the warp
5th 12 6th 12
7th 11 8th //
9th 10 10th 15
11th 10 12th 12
13th 12 14th 12 after the 14th a new string is attached
15th 12 16th 10
17th 13 //
22nd 18 slowing down for unknown reason
23rd 19 24th 13 slowing down for unknown reason
25th 15 26th 17 the string is twisted and has to be unwound
35th 11 36th 11
37th 11 38th 13 //
53rd 13 54th 17 in video (49:58-50:11 and 50:11-50:28)
55th 14 in video (50:28-50:42) //
60th 14
61st 12 62nd 13
63rd 13 64th 14 last warp thread, in video (50:51-50:54)
Table 16-16 The number of seconds used to insert the warp thread, from the foot of the
bed to the head part (odd numbers) and the other way round (even numbers).
In total 64 warp threads are inserted, of which only the ones mentioned here
are visible in the complete footage. // indicates a break in the video
sequence.
The edited video shows only the 53rd, 54th (49:58-50:42) and the end of the
64th warp strands (50:51-50:54). The bed maker does not count the number of warp
strings. Although he will be weaving in and "under 6 / over 6" pattern, it is
enough that he has an even number of warp strings while it is apparently not
important to have a number that is a multiplication of six. In ancient Egypt the

term for making a bed is m h , filling the frame. That is exactly what Shahed does:
he continues setting up the warp until the bed frame has been filled.48
The weft takes up almost the double amount of string: six bundles
(approximately 230 metres). There are 160 weft strings woven in, which means
that recurring action r22 (listed under A in Table 16-17) is repeated 80 times, of
which 34 have been recorded on videotape from beginning to end. In the edited
video only a few of these are incorporated, listed where possible in Table 16-17.
With 80 repetitions it is not possible to relate the 34 weaving actions with
specific weft strings (as was done in Table 16-16 for the warp strings). For the
working rhythm this is not really important, as long as they are presented in
chronological order and breaks in the weaving, such as beating the weft, are
indicated.
Weaving thefirstweft string takes considerably more time than the second to
28th. T o m a k e the first weft strand requires counting the warp strands (50:56-
51:32). Shahed starts with running the loop under six strands, and over six strands.
At one point he loses count. Thefirstweaving movement (1A), therefore, takes 38
seconds.
The next one (2A) takes only 15 seconds, because the bed maker does not
have to count the warp strings. H e just shifts the start of the weft string in relation
to the previous row. The strands in the middle (2A-29A) take an average weaving
time of 15.1 seconds.
Weaving strand 25 takes more time than the average, because the string is
entangled and has to be pulled free. Likewise, the time needed to weave in strands
6, 17 and 29 is above average, because the orientation of the weave has changed:
the string has been pulled through the warp and the pattern shifts. This change in
pattern causes a V-shaped appearance with a large zigzag motive. Shahed pays
more attention and works a bit slower w h e n weaving in the strand after a change
of orientation. The weft strand shown in video sequence 53:50-54:20 consists of
the first part of number 19 and the last part of number 28 (Table 7-65). Here a
jump is m a d e from working on thefirstpart of the bed, to working on the last part
of the bed.
The last six weft strings take m u c h more time, because the warp is getting
gradually more tight and there is not m u c h space to m o v e (30A to 35A). It is even
necessary to employ one of the wooden pegs as awl, to m a k e space for the weft
string to pass.
48In total he uses three complete bundles and part of a fourth bundle of string for the
warp (approximately 120 metres). The inside size of the bed frame is 1.85 x 0.65 m, the
outside size is 2 x 0.8 m. The warp does not span the entire length, but only the 1.65 m
from the tension string to the foot of the bed. With the string length necessary to round the
wooden frame, each warp strand takes approximately 1.80 m of string.
sec A B C D remarks
1 75 38 19 8 10 in video (50:56-52:09)
2 47 15 19 3 10 beating 1 (see Table 7-66) R
i
3 41 15 12 4 10 new length of string //
g
4 34 13 9 4 8 beating 2 // h
5 38 15 10 4 9 change of direction t
6 46 18 10 3 15 beating 3
7 28 13 6 4 5
8 13 9 3 II //
9 32 16 8 3 5 beating 4
10 15 9 3 // //
L
11 33 9 10 6 8 beating 5
e
12 37 15 10 5 7 f
13 39 16 9 7 7 beating 6 t
14 38 15 11 3 9
15 36 14 13 2 7 beating 7 // beating 8; n e w string
16 15 11 change of direction
17 37 17 9 2 9
18 33 14 8 2 9 beating 9 // beating 10
19 40 15 8 3 14 in video (53:50-54:04)
20 38 14 11 3 10 beating 12
21 37 15 9 2 11 unwinding the string R
i
22 40 14 12 4 10 beating 13; laying out string
g
23 30 12 7 3 8 h
24 41 17 11 3 10 beating 14 t
25 20 11 4 II string entangled
26 17 9 2 II
27 41 18 10 5 8 beating 15
28 14 8 (54:04-54:20), change of direction
29 42 19 12 3 8 beating 16 // beating 17
30 73 41 12 13 7 tight warp, beating 18 // beating 19
L
31 104 67 18 9 10 tight warp, beating 20 e
32 123 79 21 9 14 using peg as awl, beating 21 f
33 117 76 18 9 14 beating 22 t
34 123 69 27 13 14 beating 23
35 124 partly in video (57:06-57:36)
Table 16-17 On page 378: Weaving in the Weft.

The column to the far right indicates from which side of the bed the mat
maker is working. The sequences of which footage is available have been
numbered 1 to 35. For these the time in seconds is given of the four parts
which make up the weaving action. This weaving action is interrupted by
periods of beating in the weft. These are listed separately in Table 16-18.
The four columns list the time it takes to perform the following actions:
A = weaving (recurring action r22);
B = making the first edge (occurring action o20);
C = tightening the weft (occurring action o25);
D = making the second edge (also occurring action o20).
// = break in the footage
In column B of the time involved in making the first edge is listed. For
numbers 1-5, w h e n the basket maker is working from the right side of the bed, this
is the left edge. After the pattern orientation change, w h e n the basket maker stands
on the left side of the bed (numbers 6-16), the first edge is the right edge. The
terms left and right are related to the orientation of the bed, viewed from the foot
end. This is opposite from the orientation of the viewer, because the camera is
positioned most of the time at the head of the bed.
Similarly, under D the time involved in making the second edge is listed.49
With the exception of one time (no. 6), the first edge takes (often considerably)
more time than the second edge. This is surprising, because the movements
involved in making the edges are identical. Within the working rhythm it fits,
however. The action of making the edges is part of the larger sequence (A-D) of
weaving, which is repeated 80 times. The first edge (B) is in the middle, the
second edge (D) marks the end of this sequence and is characterized by a
slowdown of the action.
The weaving sequence A - D is interrupted regularly by a beating sequence,
which has the function of tightening the weft. Shahed starts the beating sequence
by running the stick, which is approximately 60 c m long, between the wooden
frame and the warp threads. Then he knocks the weft strand, which are tied
around the bed frame, three times with the back of the stick. Then he runs the
stick again between warp and frame before he starts the beating of the weft proper.
W h e n he stands on the right side of the bed, he starts beating from right to
left. Arriving at the left edge, he repeats the same sequence of running the stick
between warp threads and the bed frame and knocking the string loops with the
back of the stick. H e finishes the sequence by beating from the left side back to the
right edge.
49
Second edge: 1-5 right edge, 6-16 left edge, 17-28 right edge, 29-35 left edge.
Beating the Weft

no no. of beats sec b/s no. of beats sec b/s
from right to left from left to right
1 20 12 1.7 6 3 2 R
2 15 + 2 = 17 12 1.4 10 5 2
from left to right from right to left
3 6 + 6 + 2 = 14 13 1.1 8 + 2 = 10 5 2
2 L
4 15 9 1.7 7 + 3 = 10 5
e
5 2 + 13 + 2 = 17 12 1.4 14 9 1.6
f
6 18 11 1.6 // t
7 11 + 4 = 15 11 1.4 10 6 1.7
8 10 + 11 = 21 9 2.3 5 3 1.7
from right to left from left to right
9 24 11 2.2 8 5 1.6
10 3 + 5 + 10 + 6 = 24 13 1.8 7 4 1.7
11 6 + 2 = 8 5 1.6 R
12 7 + 2 = 9 7 1.3 13 + 2 + 5 = 20 15 1.3 i
13 9 + 7 = 16 10 1.6 8 7 1.1 g
h
14 18 + 3 = 21 11 1.9 20 + 8 + 2 = 30 17 1.8
t
15 11 + 2 + 9 = 22 13 1.7 //
from left to right from right to left
16 26+10 + 9 = 45 24 1.9
17 8 + 2 = 10 7 1.4 12 7 1.7
L
18 29 17 1.7 // e
19 17 10 1.7 f
20 20 12 1.7 t
21 20 15 1.3
22 15 15 1
23 15 13 1.1
Table 16-18 Beating the weft with a wooden stick (occurring action o21).
A number of 23 beating sequences have been counted. The second column
lists the number of strokes (+ is an interruption in the rhythm of the beats).
The third column has the total time in seconds, and the fourth column the
number of beats per second. T w o parallel stripes (//) indicate a break in the
footage.
The second part of the beating sequence is usually shorter and faster. When he
works from the left side of the bed, he beatsfirstfrom left to right.
Table 16-18 lists the number of beats per sequence, which shows a
considerable variation. Within the beating sequences there are sometimes small
pauses. These pauses have been indicated in with a + sign. The regularity of the
beating becomes clear w h e n considering the average speed of the beats, which lies
around 1.7 beats per second. The second part of the beating sequence is usually
slightly faster than thefirstpart. In thefirsttwo sequences, Shahed works from the
right side of the bed and beatsfirstfrom right to left with respectively 1.7 and 1.4
beats per second, while he uses 2 beats per seconds working back from left to
right.
Beating sequences 8 and 9 are both fast in the first sequence, w h e n the beats
are gradually further away from the stand point of the bed maker. H e slows d o w n
when he beats in the opposite direction, towards himself. There is no explanation
for this reversal of the normal beating pattern.
The numbers 16 and 18-21 are only beaten in one direction (from left to
right), because a different system is used: towards the end of the work there is no
space to wield the stick. Shahed, therefore, uses a wooden peg to direct the blows
more precisely. H e beats the peg in a different rhythm, similar to hammering, with
series of double knocks.

Shahed works mostly standing with his back bent. When he makes the warp, he stands in
the bed frame, until this becomes impossible because the frame gets graduallyfilledwith
string. He then stands outside the bed frame working from the right side of the bed.
He stretches his back while he prepares the insertion of a new bundle of string.
Standing on the right side of the bed, he holds the end of a new bundle of string in his left
hand and with his right hand he thins out the frayed end, pulling out the dry grass leaves.50
Het takes the string in his right hand and strokes the frayed end three times with his left
hand (48:20-48:34).
While moving to the foot of the bed, he picks up the short exhausted end, which he
already has thinned out. With his left hand he brings it to the new strand which he holds
with right and lays the two parallel to each other. He bends over and reaches under the bed
to pick up a pot with water, which he holds in his left hand. He wets both frayed ends by
dipping them in the pot and lets them drip for a while, putting the pot back on the ground
(48:34-48:56).
With his left hand he puts down the long end and he runs his hands four times (right
after left after right after left) along the wet tip of the short end. The function of this seems
to be more to mark a phase in the action, than to remove excess water or straightening out
the fibres. He then takes half of the grass bundle in his left hand, the other with right, and
he makes the grass into a thin tapering string by rolling it with his right hand from the wrist
to thefingertipsof his left hand. He repeats the rolling movement nine times, rolls the
finished string once between his two hands and then runs his hands along the finished string
50A symbol indicating the properties of the material has been inserted at 48:20
(left after right after left). Again, this seems to mark the end of a sequence, rather than
having a clear function (48:56-49:13).
Shahed takes the short string in his right hand and picks up the long string with lift.
H e moves back to the right side of the bed, while taking over the short string in his left hand
too. With his right hand he arranges the long strand and puts his right foot on top of it,
locking the string between his foot and the bed frame .51 H e takes the frayed end of the long
strand in his right hand and runs his left hand, with which he also holds the point-tipped
short end of the old warp string, along the wet fibres. H e separates them into two bundles
and, while holding the pointed short end as part of the right bundle, starts the rolling
movement of making string. H e rolls with his right hand from wrist to finger tip over the
left hand, connecting the old and the n e w length of warp string. After two rolling
movements he rearranges the string under his foot, pulling it back to shorten it slightly.
Then he rolls five times, pushes together the frayed end and makes three more rolling
movements. H e then takes the frayed bundle tip between thumb and fore fingers, opens up
the string with a twist of his left hand and pushes the grass tips through the ply of the string
with his right thumb. H e rolls the splice twice between his hands in opposite direction (his
right hand moving from fingertips to wrist of the left hand). Finally, he runs his right hand
twice along the string, marking the end of splicing the string (49:13-49:57).
All the while he has been standing straight with his right foot resting on the bed
frame. N o w he bends over and, while walking to the head of the bed, he pulls the string
over the entire length of the frame, using his right hand. H e pulls the warp string around the
yellow stretching string at the head of the bed frame. H e holds the string in place with his
left hand and pulls the string d o w n with his right hand through the space between the warp
and the bed frame. H e pulls the string up and with his left hand he tips the peg, pushing the
previous loopflatalong the edge. While holding the peg in his left hand, he bends the warp
string into a little loop, which he pushes with his right hand through the previous loop,
pulling it through with his left hand. H e pushes the peg through the n e w loop with his left
hand, while he pulls the long end of the string with right. N o w the loop and the peg are
positioned properly and locked in place (49:58-50:11).
H e then walks back to the foot of the bed, guiding the string with his left hand,
pulling it free from the bundle with his right hand. H e repeats fastening the warp strand at
the foot of the bed (50:11-50:28) and again at the head (50:28-50:42).52
Then the camera moves to one of the carpenters. The sawdust in his hair is an
indication of his work and the fact that he just has taken a break. W h e n Shahed asks him, he
helps out, for instance with pulling the weft string at the m o m e n t the weaving direction is
changed (54:27-55:02). For quite a while he sits on the finished part of the bed (50:45-
50:50), talking to Shahed and others. Shahed, however, is concentrated on the work and
does not really answer. W h e n the bed is almost finished, a small group of people is
discussing different types of medicines, which are layed out on the bed (visible at the top
of the screen 57:06-57:15). Again, Shahed works on without paying attention.
W h e n the last warp strand is fixed (50:51-50:54). Shahed immediately proceeds
making the weft. H e holds the warp string locked in place with his left hand and pushes the
string underneath the first four warp strands with his right hand, pulling them up with his
left hand. H e indicates that he has n o w started making the weft, by moving his left hand five
times across the warp. This is for the benefit of the onlooker (51:00-51:04).
A symbol indicating the instrumental use of the body has been inserted at 49:14
These three sequences correspond with numbers 53-55 in Table 16-16.
Plate 16-19 Shahed 'filling' a bed frame with string grass string matting in a carpenter's
work shop (Daraw 1992).
While he keeps the tension on the weft string with his left hand, he pulls a length of
string through the warp, to create the space to work (51:06). H e then pushes the weft string
under with his right hand, pulling it up six strands further, with left. With his left hand he
also tightens the weft by giving it a little tug (51:11). Because it is thefirstweft strand, he
really has to count the number of warp strands which he crosses over or under (51:11-
51:22). In order to cross the 64 warp strands, he repeats the weaving action six times: the
first time he passes the weft string under four warp strands. Then he passes five times 12
strands running the weft strand over six/under six (50:56-51:32, 1A in Table 16-17).
Shahed then walks to the right top corner of the bed to retrieve a wooden peg, whic
is holding the warp loop at the head. H e replaces it with a coarser, unpolished one (51:40).
With the smooth peg, which he prefers to use while working, he returns to the loop at the
end of the first two weft strings at the left foot of the bed.
Standing with his legs apart, his right leg behind the right side of the bed frame, his
left leg next to the foot end of the bed, he bends over to make the left edge. With his right
hand he pulls the weft strand tight. Holding it in place with left, he pulls up the loop
between the bed frame and the warp. H e briefly looks up (51:48), but continues working.
H e pushes the peg through the loop, thus creating thefirstloop of the left side edge. Before
he lets go, he pulls the weft string with his right hand to keep the loop and the peg in place
(51:43-51:54, IB in Table 16-17).
H e holds the weft string in place with his left hand at the left side of the bed, while
reaching back to the right side of the bed, where the weft string protrudes. H e does not let
go until he has pulled the second weft string parallel to thefirst.With his right hand he
gives an extra tug, while holding leg of the bed with his left hand (51:54-52:00, 1C in Table
16-17).
Then he bends the string around the right side of the bed frame and holds it in place
with his left hand. H e picks up the weft string between the bed frame and the warp with his
right hand. With left he pulls the peg out of the last loop with which the warp string is
fastened. H e makes a small loop of the weft string and he pushes this with his right hand
through the last warp loop, pulling it through with his left hand. H e pushes the peg through
the loop, which n o w forms the first loop of the right side edge. H e turns the peg with his
left hand in the anti-clockwise direction and at the same time pulls the weft string with his
right hand, so the peg is locked in place securely (52:00-52:09, I D in Table 16-17).
The second weft loop is inserted with a shift of two warp strands: starting with pulling
the weft loop under six, instead of under four warp strands, Shahed again crosses the 64
warp strands with six repetitions of the weaving action, which ends with passing under four
strands (52:09-53:24, 2 A in Table 16-17). The third weft loop passes over two and under
six, and ends over six and under two. The fourth sequence has only five repetitions: it
starts over four, ending over six (not in the video).
While making the warp and weft, Shahed constantly has to pull free the string from
the bundle. Whenever bundle threatens to get entangled, he stops the work to lay out the
string. Because the work shop is too small, he lays out the string outside in the street
(52:46-53:08 and 56:21-56:28).
Further on in the process, he is still working from the right side of the bed, standing
bent over, making the right edge (o20 53:09-53:11).The working position of Shahed is very
tough. Standing on one side, he has to span the complete width of the bed (80 c m ) . W h e n
making the opposite edge he leans over and keeps this position with sheer muscle power. H e
cannot support the weight of his upper body with his arms, because he needs both hands for
the work.53
The process is briefly interrupted by a short break, while Shahed is talking to the
carpenter and scratches his head (53:16-53:21). H e then starts beating the weft, standing
upright with his left foot on the finished part of the bed matting.54 H e maintains the tension
of the last weft string by putting his foot on top of the weaving strand. H e beats the weft,
holding the stick in his right hand (o21, beating 10 in Table 16-18, 53:11-53:16 and 53:21-
53:50). W h e n leaning over to beat the far (left) edge of the bed, Shahed supports himself
'A symbol indicating the working position has been inserted at 55:32
A symbol indicating the working position has been inserted at 53:22
with his left hand on the finished matting. On the way back (from left to right) he puts his
foot on the ground and leans over on his left arm while beating.
Between the beating and inserting a new weft loop, Shahed slowly pulls the string
towards him, making a large loop of string (transition between o21 and r22, 53:50-53:56).
Then he leans over again, pushing the weaving loop under with his right hand, pulling
it up with his left hand (r22, 53:57-54:19, composed of no. 19A and 2 8 A in Table 16-17).
H e also uses his right hand to pull up the loop and free enough string to enable him to
weave to the opposite end of the bed frame. H e makes the left edge and stops working (o21,
cf. 28B in Table 16-17). H e talks to the carpenter, before changing the orientation of the
weaving (break 54:19-54:37).
The carpenter helps with pulling the entire length of rope through the warp. This only
has to be done when the orientation of the weave is being changed and is avoided otherwise
by weaving with a double strand. Single strand matting occurs in Egypt only with strands of
limited length, such as individual grass stems, but not with long lengths of string.55
Changing the orientation of the pattern also involves a change in working position: Shahed
now works from the left side of the bed.
Shahed makes the same movements, sometimes identical to the movements performed
from the right side, sometimes he uses the opposite hand (Table 16-20). W h e n making the
edge, he starts the same (Table 16-20 a, b, c). The wooden peg is n o w at his right side, so
in order to tilt it with his right hand (e), he has to take over the loop in his left hand (d). H e
pushes and pulls the loop of warp strand through the previous loop with the opposite hand
(f, g). After having put the peg in the new loop, the fastening of the loop is also done with
the opposite hand (h-1): pulling the string while turning the peg in a clockwise direction.
Working from the right, he turns the peg in an anti-clockwise direction. Weaving is done
always the same (m, n), as is beating (r, s, t). Working from the left side, Shahed puts his
left foot on the edge of the bed frame and not on the finished part of the bed. H e also
supports himself with his left hand, leaning onahe warp threads, rather than the finished
fabric. Tightening thefirstweft strand and pulling the second weft strand parallel to the
first, however, is done with the hand closest to the finished part of the bed (Table 16-20 o,
p, q). Beating the weft with the assistance of a peg is done in opposite orientation (u, v).
F r o m the left side of the bed, Shahed measures the zigzag pattern with the same stick
as with which he beats the weft (o24, 55:05-55:14). H e checks if the change in orientation is
in the right place. During the process he does this three times, after each of the three
changes in orientation. At some point there is a discussion between him and the carpenter
that he has made the change too early (not in the edited video). Nevertheless, Shahed
continues, because he already has pulled through the weft strand.
H e proceeds making the left edge (o20, 55:14-55:15) and weaving in the next loop
(r22, 55:15-55:39).56 Then he makes the right edge (o20, 55:39-55:51), pulls the second
weft string parallel to the first (o25, 55:52-55:56), and makes the left edge (o20, 55:56-
56:02). The sequence is ended with Shahed beating the weft (o21, 56:02-56:21).
In this sequence the only recurring action is weaving. Although the action of making
the edges is repetitive, this is considered an occurring action: whenever the weaving reaches
the end of crossing the warp the loop is pushed through the previous loop.
It is not possible to use a weaving shuttle, because the materials for mat making and
bed weaving are too coarse.
In this sequence tthe frozen m o m e n t indicating active and passive elements is
incorporated from 54:22-55:26
Action standing on right standing on left

side of the bed side of the bed
making the edge (o20) 52:24-52:43 55:56-56:02
a pulling string around bed frame right hand right hand
b holding string in place left hand left hand
c pulling loop between frame and warp right hand right hand
d holding loop right hand left hand

e tilting and removing peg left hand right hand
f pushing weft strand through loop right hand left hand
g pulling weft strand through loop left hand right hand

h pushing peg through new loop left hand right hand
i turning peg left hand right hand
k orientation of turn anti-clockwise clockwise
1 pulling weft string right hand left hand
weaving the weft (r22) 52:10-52:24 55:18-55:39
m pushing loop under right hand right hand
n pulling loop up left hand left hand
0 tightening weft left hand right hand
tightening second weft strand (o25) 52:43-52:46 55:52-55:56
P holding the strand right hand left hand

q pulling the second strand left hand right hand
beating the weft (o21) 53:21-53:50 56:03-56:16
r beating with stick right hand right hand
s foot on bed left foot left foot
t support with hand left hand left hand
beating with stick against peg not in video 56:43-57:00
u holding peg left hand right hand
V beating with stick right hand left hand ;
Table 16-20 Performing movements with left or right hand, working from the left or the
right side of the bed.
street
Figure 16-21 Sketch plan of the carpenter's work shop and the area used for making the
bed matting. Drawing by H. Barnard.
In the next sequence the bed frame is almostfilled.Shahed employs a peg to beat the
weft (56:44-56:59), because the space has become too narrow to wield the big stick. H e
holds the peg in his right hand, the point pushed against the weft, and beats it with the stick
in his left hand. W h e n he is performing the same action from the right side of the bed he
holds peg and stick in the opposite hand. For beating, he uses the coarse, unpolished peg.
During the last phase of making the bed, he does not stand bent over, but he is
kneeling while facing the side of the bed. Inserting the weft is increasingly difficult.
Therefore, he uses a wooden peg as awl: with his left hand he pushes the peg into the warp
to make some space.57 H e pushes the strand from underneath with his right hand, taking
57Three
„ symbols indicating the use of instruments, have been inserted. The tool is the
same, a wooden peg, but the use differs: 56:36 wooden peg, used to secure the edge, 56:47
used to beat the weft, 57:17 used as an awl, to make space in the weft.
over above the fabric with his left hand, after which he tightens the weft strand by pu
with his right hand (57:08-57:26). Sitting, he cannot reach the entire width of the bed. He,
therefore, weaves until halfway the bed, then he stands up and walks around to continue
from the other side (not in video).
Working towards himself while sitting on the right side of the bed, he removes the
two pegs which are holding the loops of the last warp and weft strand with his left hand. He
pushes the end of the weft string through the fabric with his right hand, pulls it up with left,
takes over with his right hand to tighten the weft. Then he guides the strand through the two
loops and down through the fabric with his right hand (57:26-57:56). He anchors the last
length of string by weaving it in (starting 58:03).
76.5.5 Workspace
Shahed works in the workshop of the carpenters w h o made the bed frame (Plate
16-19, Figure 16-21). Around him the work of the carpenters continues as usual.
Shahed only needs a floor space of approximately 6 m 2 and some space to lay out
the entire length of string contained in one bundle. For this he uses the street (cf.
52:46-53:05). N o specific architectural provisions are needed for this work.
CHAPTER SEVENTEEN
THE DANCE OF THE DYNAMICS
The previous chapter concentrated on body movements and rhythm. The

movements were described as a kind of choreography: preplanned and repeated. A
choreography is the outline of singular, and in itself meaningless, movements. It is
not until watching the dance as a whole that the meaning can be sensed.
17.1 Production Process

What is the knowledge gained from the micro-analysis in the previous chapter?
W h y is it important to describe the production process to such detail as to the hand
with which the basketmaker does a certain action? In archaeological basketry this
information cannot be retrieved, so what is the point? Perhaps the most important
reason is that it is a method to force oneself genuinely to look at what is happening
in front of our eyes. While describing the actions in detail it becomes suddenly
apparent that some parts of the process have not been understood completely,
although w e have a superficial grasp of the entire process. Although every
producer acts differently, the forced attention to detail brings to light important
information that would be missed w h e n adhering to a schematized view of the
production process.
The image of a production process is greatly determined by the repetitive
actions. The phase in which the repetitive actions are most obvious is that of
making the basic structure. W h e n making a coiled basket the action of wrapping
and stitching the bundle with palm leaf is repeated over and over again to create
the base and sides of a basket. Likewise, plaiting and weaving involve repetitions
of the same movements until the product has the required size.
Repetitive actions characterize the phase of preparing the systems (for
instance plaiting and rope making). In the processes of making the maqtaf basket,
the masballa bag and the string bed, the recurring actions take up an important
part of the time. The other production phases (preparing the raw materials,
starting the basic structure, finishing off the basic structure, adding features and
giving the finishing touch), are mainly characterised by occurring actions.
Looking at the process in detail shows that although the repetitive actions do
take up the largest portion of the total production time, the occurring actions take
up m u c h more time than expected. The repetitive actions are interrupted constantly
with occurring actions: inlay of n e w leaflets, beating the weft, making the edges.
The working order is not always performed linear from preparation of the
materials to the finishing touch. If, during the work, the raw materials are
finished, new have to be prepared. During the production process periods of

waiting m a y be filled with actions that belong to another phase. A detailed
description as given in the Body as Instrument (e.g. Section 16.1.4) is incorporated
in the text because it guides the attention of the viewer to the working position and
the use of feet, teeth, body weight and the left or the right hand.
Basket makers show a great endurance in making the same movement for a
long period and maintaining a steady working rhythm helps this. The transition to
a new phase is often marked by a change in working position and a brief break.
17.2 Skill
The movements of the repetitive actions show a consistency and regularity which
are prerequisites for developing a working rhythm. These movements are not
necessarily the most practical or leading most directly to the end purpose. While
sewing, M o h a m m e d regularly gives a little knock on the side of the basket (Table
16-4, p. 340). This has no direct function in the production of the basket, but is
important for maintaining the working rhythm.
Scanning the working rhythm (for instance by watching long sequences in
fast forward) shows clear breaks, with a sharp decline in the working speed. These
occur mainly at points where a technical difficulty occurs, for instance the centre
of the basket and the inlay of new material. In the previous chapter it has been
shown, however, that there is not only a rhythm in the repetitive actions, but that
occurring actions are sometimes rhythmic too. Because of the variation in the
movements, defining the latter rhythm is m u c h more difficult. The working
rhythm not only differs with each technique, but also with each basket maker.
The working rhythm goes hand in hand with concentration. Four of the five
basket makers featuring in the video, are extremely concentrated. This can be
inferred from the work-eye contact and the interaction of the basketmaker with his
or her surroundings. M o h a m m e d talks to his wife and to friends w h o c o m e and
visit, but he does not avert his eyes from the work. Khadidja sits in the court yard
with three other w o m e n , but does not take part at all in the general chatting. One
of the other w o m e n , w h o is very talkative, works in an unsteady pace. Her
attention wanders and this is reflected in her basket: the coils and stitches are
uneven. AmrTt has reason to be distracted, the entire school youth of el-Amariyya
has gathered in front of his little work shop, but he still works on steadily. Only
n o w and then is he distracted by the racket going on just in front of his eyes.
Shahed pays but little attention to the m e n of the carpentry workshop, w h o sit
on the half-finished bed, discussing all kinds of interesting topics. H e only talks
during the few brief breaks he takes between the different phases of the work.
The only person w h o is clearly distracted is Rauwhayya. Her working
rhythm is irregular and she looks up from her work repeatedly. This shows in her
work, which is not very regular. In her case, the circumstances are partly to
blame. The filming and a group of children shouting at a distance distract her.
17 The D a n c e of the Dynamics 391
During earlier visits, she worked much more relaxed and steadily (cf. Plate 16-9,
p. 355).
Skill is strongly related to working rhythm and concentration: all skilled
basket makers distinguished themselves by a very steady working rhythm, intense
concentration and a very regular appearance of the product. Inexperienced basket
makers, on the other hand, are not able to maintain a rhythm, they are usually
easily distracted and the result of their work is a basket which is irregular in
appearance.
Skill is more than working rhythm and concentration, however. The
experience of assessing and selecting the raw materials and planning the production
(size, shape, time) are other important aspects of skill.
There is a relation between a steady working rhythm and skill, but there is no
inevitable link between skill and professionality. Khadidja is a prime example of an
extremely skilled nonprofessional basket maker. There is, however a relation
between professionality and speed. The (semi) professional basket makers
M o h a m m e d , AmrTt and Shahed all work in a steady, fast rhythm, while Khadidja
works in a steady, but slow rhythm (cf. section 17.4).
17.3 Tradition, Learning and Innovations

Old basket makers pass tradition by teaching the younger generation. The
transmission of skills, practical knowledge, is done by slowly integrating the pupil
in the work (cf. Bentz Hogseth 1998). It is a visual and tactile process. The skills
and experience to be mastered are the movements, agility, rhythm, but also the
knowledge of materials and the ability to plan the production and reproduce
consistently the same shapes and sizes.
In each basketry technique there are difficult stages. The centres of coiled
basketry, for instance, have the technical problem that a coil has to be started with
a bundle that is too thick and too rigid to form a closed middle circle. The
solutions to such technical problems are manifold. It is especially in these (usually
occurring) actions that w e can discern traditions.
Likewise, shape and decoration convey tradition. Perhaps the most important
and least tangible of teaching basketry production, is to convey what the product,
the basket as a whole, should be like. The appearance of a basket includes material
aspects such as the shape, fabric and decoration, but most importantly it has strong
immaterial aspects of function, meaning and social signalling. These are not taught
as consciously defined properties, but are inherent in the basket. They are the
essence of a basket, the dance rather than the choreography. This essence differs
strongly for Egyptian and Nubian baskets.
In the previous chapters, part of the difference has been defined. At present,
Middle Egyptian baskets are plain and their shapes and functions are limited.
There are two kinds of coiled baskets, for instance, undecorated with very specific
functions: large round flat bread baskets and straight sided round baskets, used for
mainly for storing flour. In N e w Nubia the essentials of basketry seem to be linked
to the function and decoration. There are many differences within Nubia, but the
central issue is that the baskets are mainly used as covers rather than containers,
and have to be decorated.
In the Qasr Ibrim material, two types of decoration occur: monochrome
wrapping patterns, and bicoloured (orange-brown) patterns. In N e w Nubia more
colours are used. The example in the video, a Fadidja basket, has purple, red and
white strands. The Kensi villages employ m u c h more colours. O n e of m y questions
to the Nubian w o m e n was: "Are you a better basket maker than your mother?" The
answer was often "Yes, because I use more colours".
The conscious answers are not about the patterns, but about the colours.
Asking w h y the coloured patterns are made, invariably gets the answer "It is just
decoration". It is not, however. There are clear differences between Fadidja and
Kensi villages in the use of materials, colours and patterns. There are differences
a m o n g Kensi villages too, not as m u c h as the decoration is concerned, but in the
techniques applied (sewn plaits baskets versus coiled baskets).
Innovations occur both in Nubia and Middle Egypt, but they are of a
different character. The Kensi Nubian innovations concern materials, colours,
decorative patterns and even the techniques. T o obey the unexpressed, but
obviously present rule 'the more colours the better', basket makers in Old Nubia
already bought chemical dies for the palm leaf in Wadi Haifa or Aswan. In N e w
Nubia Kensi basket makers are using brightly coloured cotton yarn, washing line
and ribbon. They even adapt the coiling technique, while using shiny candy
wrappers (cf. p. 263). Innovations in Fadidja Nubian villages are limited to the use
of chemical dies, of which only a few colours are employed. Also in the use of
n e w patterns, the Fadidja villages are less innovative, more traditional, than the
Kensi villages.
In Middle Egypt innovations only involve the size and shape of the baskets.
The w o m e n of el-Till started to make a larger variety of shapes w h e n there was a
demand from a new market: the tourists visiting the antiquities sites.
17.4 Feedback to the Archaeology

Tools and Work Space
So h o w are w e to translate the essentials of basketry to the archaeological finds?
Starting with the isolated steps of the choreography again and trying to construct a
dance out of these, slowly builds the interpretation. F r o m the study of the present
it is clear that there are no special requirements for the work space of the basket
makers w h o produce coiled and sewn plaits basketry. The tools needed are very
limited too: a knife, needle and an awl, which can be nothing more than a spine
from the branch of the date palm tree.
Mat makers need a horizontal loom. The archaeological remains of such a
loom will typically be nothing more than four post holes. Probably the loom is
inside or in a shaded area. The substantial wooden parts of the loom are rarely
17 The Dance of the Dynamics 393
found in situ, because they are too expensive to discard or leave. Apart from the
loom, no other tools are needed.
The bed maker needs a space which fits the bed frame and gives him some
walk way around it. N o installations are necessary and the tools used are not very
distinct: three wooden pegs.
Working Rhythm
A n important correction on the current archaeological practice, based on the micro
analysis of actions of modern basket makers, is that more attention should be given
to the material results of the occurring actions. For the study of archaeological
basketry this means that specific attention should be paid to the start of a basket,
the insertion of new lengths of material and the rims and edges of baskets and
mats. Watching the time of the production process reveals that features
characteristic for a basket have taken relatively little time in the entire process.
The start of the plait for the sewn plaits basketry takes, for instance, only 24
seconds. Nevertheless, these features give away much of the skill and training of
the basket makers.
The working rhythm of both occurring and recurring actions is reflected in
the appearance of the basket. A regular appearance of the archaeological find is an
indication for the skill of the ancient basket maker. It is not necessarily a marker
of professionality, however. A comparison of the work of nonprofessional and
professional skilled basket makers shows that the professional product displays
signs of haste. The skill of the professional basket maker can be read from the
degree of regularity of such inaccuracies.
Work-Eye Contact
Apart from a sign of concentration, the work-eye contact is also an important
indicator of what is supposed to be the 'good' side of the work. The modern basket
makers always face the side of the basket that will be visible. While plaiting the
long strip, M o h a m m e d watches the side where the ends of the inserted leaflets are
not visible (they are at the back). W h e n he rolls up the plait he takes off the
irregularities from the outside of the coil (the insertions are on the inside). Later in
the sequence the smooth outside of the rolled up plait becomes the outside of the
basket (cf. 13:54-14:01). The inside not only has the insertions, but the sewing
ridges are also much more pronounced on that side. So in sewn plaits basketry the
outside is more important than the inside, despite the fact that the roughness of the
basket might damage the contents. The sewn plaits baskets clearly have not been
produced to contain vulnerable items, such as soft fruits.
For theflatcoiled baskets, made by Rauwhayya, the 'good' side is the inside
(cf. Plate 16-9, p. 355). This is the side on which the bread lies. The sabat baskets
are also smooth on the inside. For the food covers in Nubia, however, the 'good'
side is the convex (upper) side (33:35-33:55).
The work-eye contact of the ancient basket maker can be recognised in the
archaeological material too. Since the side that the basket maker faces is in general
more regular in appearance than the side that is turned away, the orientation of the
basket towards the producer can be inferred. The two sides of the work can usually
be recognized even in quite small fragments of archaeological basketry. In
addition, identifying the 'good' and the 'wrong' side of a basket helps functional
interpretations.
If we want to understand the production of archaeological basketry as more than a

collection of dispersed descriptions, measurements, impressions and thoughts, then
one important aspect still has to be looked into. W h o made the baskets and for
w h o m did they make them? This will be discussed in part four.
PART FOUR
Basket Makers
C H A P T E R EIGHTEEN
P R O D U C E R S A T PRESENT
18.1 Mohammed and Nabawiyya, Sewn Plaits Basket Makers

M o h a m m e d , living in Middle Egypt, is a semi-professional basket maker. H e also
has some fields, which he works in the mornings, while he uses the afternoons to
make baskets. H e learned the trade from his father, but he does not want his sons
to become basket makers. H e wants his sons to go to school to be educated and
become civil servants to prevent them from being as poor as he is.
M o h a m m e d taught his wife h o w to make the plaits, and his mother also
knows h o w to do it. They sometimes help him with making the plaits. H e earns a
very small income of his baskets. Because he does not o w n enough palm trees, he
has to buy palm leaf and fibre. T o buy enough of the qalb el-nakhla, the young
white leaves in the centre of the palm tree, to make one aldga basket costs him
approximately 30 piastres, while the price he gets for the basket is L E 1.50 if he
travels to Deir M a w a s , the nearest small town, to sell the baskets himself. If he
uses a middle m a n , then he gets only L E 1.40 for the aldga basket. For two
afternoons work, he thus receives maximally L E 1.20= (approximately $ 0.35). In
comparison, the day rate for workers in Middle Egypt is L E 8,= (price level of
1992).
basket name size of plait price of material price of basket §
maqtaf 10 ba'a L E 0.50 L E 2.00 1
aldga 7 ba'at LE0.30 LE1.50 1
gauta, baguta 3 or 4 ba 'at L E 0.20 L E 1.00 J
Table 18-1 The prices of raw materials and end products of sewn plaits baskets in
Middle Egypt (price level of 1992). cf. Plate 10-3 p. 180.
M o h a m m e d states that he is a poor man, and other people in the village

characterize him as poor also. Nevertheless, he owns two m u d buildings, a house
to live in, and a stable for his animals. H e has a c o w and some goats. The
sustenance of the family is based on what he grows on his land. The w o m e n of the
family m a k e bread and cheese. Meat is rarely on the table. The income from the
baskets is mainly to buy the things that the family cannot produce.
The clothes he and his family are wearing are in good condition, but the
house is furnished very sparingly: the living room has one table with three legs,
leaning against the wall, a small black and white television on top of it. The family
sits on the floor and if there are guests a plastic mat is rolled out. The bed room
boasts a double bed. There are no cupboards for clothes, which are all hanging
over a piece of string. The kitchen area in the court yard contains a m u d oven on
which all the cooking is done. The family does not o w n a refrigerator, that first
sign of wealth.
M o h a m m e d makes his baskets for farmers (the maqtaf), workers (the small
baguta basket), and households (the aldga, which is mainly used as a carrier
basket for shopping). In spring, M o h a m m e d makes 'the pollinator' a special basket
which is used to carry the pollen from the male date palm tree, up to the flowers
of the female date palm tree (Plate 18-2).
Plate 18-2 The talla'a basket, used to pollinate the date palm trees.
18 Producers at Present 399
Nabawiyya
In the N e w Nubian village of DihimTt sewn plaits baskets are m a d e by w o m e n on
a semi-professional basis (Plate 4-3, p. 62). They are lavishly decorated with
patterns of three or four colours (Plate 11-7, p. 215).
basket type size of plait palm leaf dye price of basket
large guffa 10 ba'at L E 1.00 L E 1.50 LE5.00
middle size guffa 6 ba 'at LE0.50 L E 1.00 L E 3.00
small guffa 3 ba'at L E 0.25 LE0.50 L E 1.50

(shibrl sellan nibid)
Table 18-3 The prices of raw materials and end products of sewn plaits baskets in N e w
Nubia (price level of 1992).
The quantity of palm leaf needed for a medium basket are two young date palm
leaves, at 25 piastres per leaf. Apart from the 50 piastres for the palm leaf, the
basket maker has to invest L E 1 to buy the dyes. The basket costs L E 3 so the
profit for a full day of work is L E 1.50 ($ 0.45). Nabawiyya does not work full
time, however. She makes sewn plaits baskets and mats whenever she has time.
Depending on the season, this is usually not more than two hours a day.
A s long as there is palm leaf in the house, she can start the work after
waiting 30 minutes for the palm leaf to soak. The plaits can be made anywhere and
this activity does not require a large space. Doing this work whenever there is
some spare time is easy and she can stop the work at any m o m e n t when other tasks
(cleaning, baking, cooking, shopping) ask her attention.
She makes the baskets mainly for her o w n use, although she does sell on the
market sometimes. It gives her a little extra income, on top of the salary that her
husband brings in. Like most of the m e n from DihimTt, her husband is working in
Cairo and only comes h o m e once a year. Basket making as a nonprofessional
occupation for w o m e n has a long tradition and the craft is considered an important
asset. Nabawiyya considers selling the baskets a useful contribution to supporting
the household.
Like most Nubian houses, Nabawiyya's house is spacious and clean. She
owns a colour television, a refrigerator and a fan. The guestroom and bedrooms
are nicely furnished with colourful flowery patterns and frilled bedcovers.
Nabawiyya wears colourful dresses with a black overdress w h e n she leaves the
house. In general the standard of living seems to be m u c h better than that of the
Middle Egyptian basket maker.
Nabawiyya learned h o w to make the sewn plaits basketry from her mother.
She started w h e n she was still very small, approximately seven years old. In Old
Nubia the w o m e n would cook breakfast early in the morning and then go out with
the animals. While watching the animals, they would sit under a tree and make
baskets. Now, in New Nubia, they preferably make baskets together still. They get
together in the court yard of one of the houses.
The colourful plaited baskets are mostly used as shopping baskets, or for
displaying goods (dates, peanuts, spices) in the shops of the souk. For agriculture,
undecorated sewn plaits baskets, made by Egyptian men, are used. Nabawiyya
commented on this saying that the Nubian men work the earth, the Nubian women
the palm leaf (Jehus), while the Egyptian men work with both earth and khus. She
implied that the Egyptian men somehow had their materials mixed up (making
baskets is women's work).
18.2 Rauwhayya, Umm Ali and Khadidja, coiled basket makers

Rauwhayya makes two kinds of baskets: the tabaq, the large flat bread basket and
the sabat, a deep basket with rigid, straight, slightly flaring sides. She makes these
baskets for her o w n use and perhaps for friends or neighbours. She never sells
them.
Her husband o w n s ten date palm trees and in general the family is quite well
off. They have a water buffalo and a c o w , w h o just has given birth to a calve; four
goats, one goose and ducks. They also o w n a donkey, but this does not belong in
the category of 'household animals' and thus is not mentioned w h e n asked h o w
m a n y animals they o w n . The house is not large, but boasts a colour television and
a fridge. Rauwhayya's husband is a khafffr (antiquities guard), and farmer.
Umm Ali
U m m Ali lives in the village of el-Till, 5 k m to the North and has baskets for sale.
Since this village is on a tourist route, she sends her neighbours' daughter out to
sell baskets to the tourists. She works precisely and makes regular stitches which
cover the bundle material completely.
basket type days work price of material price of basket
sabat 3 full days LE 1,50-LE 2,= LE 6, = ;

large tabaq 3 full days LE 1,50-LE 2,= LE6,=
small tabaq 1.5 days LE 1,= LE3,= !
large seniyya 3 full days LE 1,50 LE6,=
small seniyya 1.5 days LE1,= LE3,=
Table 18-4 The prices of raw materials and end products of coiled baskets in Middle
Egypt.
Her husband owns three palm trees, but this is not enough for the amount of
baskets she makes. She has to buy part of the raw materials. She complains that
nobody wants to give her zaghauwa (fruit stem) and that she is forced to use halfa
(grass) instead. But only, she assures m e , for the tourist baskets. Egyptians would
never accept this inferior material.1 For the large tabaq she needs five or six of the
young date palm leaves, which cost 20 piastres each and five or six fruit stems at
10 piastres each (cf. Table 18-4).
She is a semi-professional basket maker, with a production of two large, or
four small tabaq baskets per week. With her work she makes approximately
L E 60,= a month, which is half her husbands' income. This money is used for
extra things, "such as cigarettes and tea". The family has a new, government-built,
house. In the court yard U m m Ali keeps chickens, ducks, rabbits and a turkey.
Her husband is a retired government employee.
She was taught h o w to make baskets by a w o m a n of the village. She has
taught her nine-year-old daughter and a neighbour girl of fourteen years old.
Several w o m e n in el-Till know h o w to make baskets, but according to her
husband, U m m Ali makes the best baskets in town. H e recognizes her baskets
anywhere. U m m Ali says that all w o m e n in el-Till know h o w to make baskets,
while the w o m e n from the nearby villages of el-Hagg Qandil and el-Amariyya do
not. The people from these villages come to buy baskets in el-Till while the people
of el-Till buy their sewn plaits baskets in el-Hagg Qandil, where the m e n are
involved in basket making. The suggestion that m e n could also make coiled baskets
is met with horror: "That would be shameful. W h y ? It just would be shameful".
The claim of U m m Ali that coiled basketry is only made in el-Till and sewn
plaits basketry only in el-Hagg Qandil is decidedly not true. In all villages w o m e n
are involved in making baskets. The number of m e n in el-Till making sewn plaits
basketry is low, however. The division between the sexes, coiled basketry being
made by w o m e n , sewn plaits baskets by men, is very strong, although it is
acceptable that w o m e n assist in making plaited baskets (cf. M o h a m m e d ' s mother
and wife, section 18-1).
Several people mentioned the village of el-Bersheh, which is located
approximately 10 k m to the north, as a centre of basketry production. M a n y
persons, both male and female, produce for the markets in Mallawi and el-Minya.
The division between men-made sewn plaits, and women-made coiled basketry is
strongly present here as well.
Although town people might look d o w n on it, in the village it is an asset for a
w o m a n to have the skill of making baskets. Nobody is ashamed of being a basket
maker and m a n y young girls are learning h o w to make them.
' Once upon a time grass was the favourite material for the bundles of coiled basketry
in this region, cf. p. 182.
Khadidja
The basket makers in N e w Ibrim are making a large variety of coiled baskets (see
Appendix D ) , mainly used for covering food. W h e n the villagers partake in a
funeral or celebration food is brought in on trays, and uncovered the m o m e n t the
meal starts. In the old days the food was not brought in on an aluminium tray, but
on a large flat basket, which was also called shauwer. W h e n not in use, this basket
is hanging on the wall as an attractive, colourful decoration. The kontee is a deep
basket with a head ring underneath. The 'good' side of the basket is the outside, in
this case not because the kontee is used as cover, but because it is carried on the
head, covered with aflatbasket, the waltl (Plate 18-5).
Plate 18-5 A deep basket (kontee) covered with a flat lid (walit). The basket is made of
cotton yarn. This family lives in an Arab-Nubian village. (New el-Malki,
1996).
basket type days work price of material price of basket
shauwer 4 L E 13.= L E 20, =
karadj 2 LE6,= L E 10,=
kontee 4 LE 13,= L E 20, =
walTI 2 LE6,= L E 10,=
Table 18-6 The prices of raw materials and end products of some of the Fadidja coiled
baskets in N e w Nubia (price level of 1992).
The large decorated Nubian food covers take approximately 14 days of part
time work, which translates into four full-time days. A smaller food cover, with a
diameter of 300 m m takes two full time days, provided "one has good eyes and is
not lazy".
Khadidja makes the baskets mostly for her o w n use, or to give to friends and
daughters of friends. At a marriage of a village girl, her friends and relatives make
baskets for her, to decorate the house. A bride receives about 30 to 50 baskets.
The mother of the bride makes a sewn plaits marriage mat, which after the
marriage is rolled up and hung from the wall, together with two coiled food
covers. The tradition of giving baskets at weddings has not survived in the Kensi
villages, as it does in the Fadidja villages.
In Old Nubia the markets were far away. S o m e w o m e n told that they used to
make baskets to sell in the North (Aswan) or the south (Wadi Haifa). At present,
with many husbands working in Cairo or abroad, and the local markets being
easily accessible, selling basketry is becoming increasingly c o m m o n . The prices
quoted (Table 18-6) show that the money made in a full day's work is L E 2, = .
Khadidja's house and court yard, like most houses in N e w Nubia, is spacious
and meticulously clean. The level of wealth seems to be higher than in Middle
Egypt. Her family owns a fan, fridge and colour television. The guest room and
bed rooms are well furnished. N o animals are kept in the court yard or around the
house. Goats, sheep and chickens have their o w n space, often a separate, old,
house or barn.
Khadidja learned to make the coiled baskets from her mother and passed her
knowledge on to her daughter. Coiled baskets can be made whenever there is a
short span of time available, and it can be easily put d o w n to return to other tasks.
The only preparation necessary is soaking the palm leaf. Once moist, it can be
kept for a while in a tub of water, or in a wet cloth.
In Kensi villages, w o m e n are organizing workshops where girls are taught
h o w to crochet, sew clothes, make bead work and baskets. The baskets made, are
those sewn with coloured cotton yarn (cf. Plate 18-5). These are sold on the
weekly markets. It is an outlet created by the workshops, to sell their produce.2
Khadidja complains that the young Fadidja w o m e n do not want to learn the
craft. They do not have the patience required, because they want to watch
television. In Kensi Nubia the workshops teach the young w o m e n to make baskets,
but these are not made with the traditional materials.
18.3 AmrTt, Twined Mat Maker

AmrTt is not a rich man. His galabiyya is frayed at the sleeves and his head cloth
is threadbare. Of course, this is his working outfit. W h e n he finishes work, he
changes into a better galabiyya (men's dress) to go to the coffee shop. His sons go
to school and, like M o h a m m e d the maqtaf maker, he hopes that they will have a
better job.
Although he is a professional mat maker, he does not have a separate
workshop, but he always works at the same spot in his house. The pegs of his
loom are fastened in the ground of the narrow hallway, facing the door. Only
when he is preparing the string, he sometimes sits outside. H e lives in a large
almost windowless m u d brick house: two rooms with high ceilings, which give the
impression almost of a cave dwelling. There is hardly any furniture in the house,
just some mats to sit on. The family life concentrates itself in the small, half
roofed, courtyard, where the bread oven and cooking utensils are.
AmrTt does not make a large amount of money. O n each masbala he makes a
profit of L E 2, = . He can make only two masbala bags a week, which includes
making string, twining four mats, and sewing them on three sides into a bag, with
one long side open. Put on the back of a donkey, the points of the bag hang down
along the donkey's belly and the bag opens automatically. Empty, the masbala is
flat as a mat and can be used as saddle. Apart from masbala bags AmrTt also
makes door mats (rigleen) and donkey saddles (barda'a).
18.4 Shahed and Karima, Mat Weavers

The workshop of carpenter Attia, produces timber, wooden donkey saddles (called
barda'a, just like the ones made from twined matting), and beds. Most beds are
finished with a jerid bottom, the frame is filled with the date palm midribs.
Sometimes, a client wishes to have a serir bil habl, a bed with a woven webbing.
Then Shahed is asked to come.
Shahed used to be a professional bed maker, but since demand for this
traditional type of bed is dwindling, he had to take on another job. H e works for
Apart from newly made baskets some women sell at least 35 years old baskets on
these local markets. These date from before the migration to New Nubia. One woman on
the market explained that the colours were still well preserved, because she had been
keeping them all that time in a dark chest. She decided to sell them, because she needed
some cash money.
the government, which means he has an office job that lasts until two in the
afternoon. A s almost everybody, he does different work in the afternoons, for
instance bed making. H e makes approximately one or two beds a month, for which
he orders the frame from the woodworkers for L E 15, = . The rope he makes
himself from halfa grass, which costs him only the time to cut and dry the grass.
The total production time for the bed is twelve hours, six to make the rope and six
to make the bed. For a complete bed Shahed asks L E 35, = . If he would be a full-
time bed maker, he could make five beds a week, the rest of the time he needs for
collecting the raw materials. With full demand he could be making L E 100 a
week, which is a reasonable income, compared with government jobs that pay
L E 120 to L E 200 a month. Cheap metal beds are slowly replacing the wooden
beds with jerfd or string.
The serir bil habl is also called angereeb, a term used in upper Egypt, Nubia
and the Sudan. Shahed is an Egyptian w h o sells his beds to both Nubians and
Egyptians. His father taught him h o w to make the beds. His elder brother also
knows h o w to do this work, but his younger brother works as a farmer and has
never learned the trade. Apart from beds, Shahed also knows h o w to make twined
matting: masbala and a barda'a.
The video shows weaving with a double strand: Shahed pulls loops of a long
string through the warp and hooks them into the previous loop. It is a much faster
method than weaving with one strand a the time, because the string is double and
does not have to be pulled entirely through at each passing of the warp.
Karima
Floor mats and sleeping mats are woven with single strands, on a loom. In middle
and upper Egypt, mats woven of bunches of halfa (grass) are c o m m o n . In the
oases of the Western Desert, such mats are made of single stems of rushes
(samaar). The mat weavers are in general full-time professionals, w h o often have
part of their house or a separate workshop set aside for the trade.
The mat weavers are usually men, but in the upper Egyptian village of
Nagada there is a female mat maker. She is a widow w h o took over the trade of
her late husband out of sheer necessity. She works full time. She is not a wealthy
person and has the responsibility to feed her family.
The raw materials for these mats are "belonging to everybody", growing on
uncultivated land, often at canal banks and near small lakes or pools. The floor
mats are rapidly replaced by woven mats made of brightly coloured plastic straws.
The costs of the plastic mats are higher than those of the grass mats, but they are
less heavy, come in larger sizes and are thought modern.
18.5 Professionality
In the sections above the terms full-time professional, part-time professional and
nonprofessional were used. This distinction is made not as m u c h on the time
involved in working as a basket maker, but on the dependence of the income from
this craft. A full-time professional is completely or largely dependant on an income

out of basket making. The part-time professionals have other occupations, farming
for instance, or are responsible for only part of a family's income.
Nonprofessional basket makers only work for themselves, or for friends and
family. They do not sell their produce on the market.
According to this definition Rauwhayya and Khadidja, w h o make coiled
basketry in el-Amariyya (Middle Egypt) and Ibrim ( N e w Nubia) respectively, are
nonprofessional basket makers. U m m Ali, w h o sells coiled baskets to tourists who
visit the antiquities at Amarna, could be considered a part-time professional,
although she does not really produce for the market regularly. The money she
brings in is nevertheless a considerable contribution to the household income.
M o h a m m e d w h o makes the sewn plaits baskets in the afternoons, but works
on his land in the mornings, is clearly a part-time professional. H e spends one day
a week to sell his produce on the market across the river.
AmrTt is a full-time professional mat maker, as is Karima, w h o weaves
mats. The bed maker Shahed used to be a full-time professional, but had to give up
because there was not enough demand, although mats and beds are commodities,
that are still bought and sold on the market.
The basket makers that are full-time professionals work in trades that require
planning and a special work space. Making flexible twined matting requires a loom
and thus a fixed work space. The same is true for woven matting, where the loom
is m u c h bigger and takes up even more space.
Both twining and weaving are time-consuming processes, which in principle
can be interrupted at any m o m e n t (except perhaps w h e n the warp is set up). They
have specific requirements for space: a large shaded even floor space is needed.
These requirements seem to have a relation to the professionality of the producers.
All the other basketry techniques can be picked up and taken to any place to work
on, the looms are in a fixed place and prevent the space to be used for something
else. This is true as long as the mat is on the loom, but w h e n the mat is finished
and the loom is empty the only things that remain are four pegs in the ground. At
that moment the space can be used for other activities.3
18.6 Gender
Several basket makers had very firm opinions on w h o should be making the
baskets. U m m Ali thought it would be shameful if a m a n would make coiled
baskets. Nabawiyya thought it was a bit ridiculous that Egyptian m e n were making
sewn plaits baskets.
A similar use of an industrial installation has been noted by Nicholson in his work on
the potters of Deir-Mawas. Hand made pots are produced by female potters. With an anvil,
the clay is pounded into shape in a clay-lined hollow in the ground. There are three
differently sized hollows for the different types of pot. The part of the day the hollows are
not in use, they fill up with rubbish, which is lying around in abundance, and are walked
over (Nicholson 1987).
^^^^ Production Middle Egypt
Used by M a d e by m e n M a d e by w o m e n
men in the house / court yard
to the house (shopping) sewn plaits baskets
outside the house sewn plaits baskets

(fields, souk, transportation) twined matting
woven matting
transportation net
pierced qafas crates
women in the house / court yard pierced matraha coiled sabat basket
pierced qafas cage coiled bread basket
woven matting sinda cheese mat
to the house (shopping) sewn plaits baskets coiled sabat basket
outside the house

(fields, souk, transportation)
~^^^ Production N e w Nubia
Used by * - ^ b y _ (Egyptian) m e n (Nubian) w o m e n
men in the house / court yard sewn plaits mats

coiled food covers
to the house (shopping) sewn plaits baskets sewn plaits baskets
outside the house sewn plaits baskets sewn plaits baskets

(fields, souk, transportation) twined matting
transportation net
pierced qafas crates
women in the house / court yard coiled food covers

sewn plaits mats
sewn plaits baskets
to the house (shopping)
outside the house | sewn plaits baskets

(fields, souk, transportation)
Table 18-7 The production and use of basketry by m e n and w o m e n , for Middle Egypt
and N e w Nubia.
On the other hand, it was completely accepted that Karima had taken over her
husbands' profession as mat weaver. Her circumstances are exceptional and
perhaps pitiful, but not frowned upon. Thus there are situations where w o m e n take
over the work of m e n , while I have found no cases in which the m e n take over
what are considered women's basketry techniques.4
W h a t is considered a men's or a women's technique is different in Middle
Egypt and N e w Nubia. Sewn plaits baskets, for instance, are made exclusively by
w o m e n in N e w Nubia. At the same time the w o m e n are aware that the coarse
baskets their husbands use in the fields, are made by (Egyptian) men.
Table 18-7 shows the relation between the production and use of basketry
techniques in Middle Egypt and N e w Nubia. In Middle Egypt there is a much
clearer difference between baskets produced and used by w o m e n than in Nubia.
Mostly, the Middle Egyptian w o m e n produce the basketry they need themselves.
They make the large bread baskets (tabaq) and the storage baskets (sabat), as well
as the cheese mat (shinda). The m e n do not use these baskets. Sometimes the
w o m e n use the sabat to get shopping.
Egyptian m e n use men-made baskets in thefield,but they do not necessarily
make these themselves. Professional basket makers produce most of these baskets.
S o m e w o m e n are involved in working the land (although this is not general and
certainly not considered a good situation). In that case they use the same 'outside'
baskets as the men.
The Egyptian w o m e n use professionally made basketry too: inside they use
the matraha, the bat like board, used in bread baking. Cages made of palm mid
ribs (qafas) are bought to keep chicks. W o v e n matting used to be part of the house
inventory, but they have been replaced mainly by plastic woven mats. The grass
mats are n o w mainly used outside by m e n while doing their jobs (especially
guarding things involves prolonged sitting on grass mats). Outside the w o m e n use
the sewn plaits carrier baskets for doing shopping.
In Nubia the w o m e n make all the basketry. These items are used mainly in
the house and for shopping. The coarse baskets used in agriculture are bought from
outside. It is made by professional Egyptian men. There is a m u c h less clear
division between men's baskets and women's baskets as the use is concerned.
Nubian m e n use the same baskets as Nubian w o m e n . There are some subtle
differences in the use of the coiled food covers, however. The w o m e n use them to
cover the food, the m e n to uncover the food.
The colourful sewn plaits baskets are used in shops to display goods such as
peanuts, dates, beans and spices. These baskets are mostly produced in the village
of DihimTt where w o m e n work as part-time professionals.
4
There is one exception. A man in the Kensi village of el-Dakka wanted to show that
he too knew how to make baskets. When he stitched the basket together, he just pierced the
needle through the fabric, rather than stitching and pulling the edges together. Upon a closer
look, it appeared he had made a plait with six rather than five strands, which makes it
impossible to sew the edges of the plait (cf. pp. 209-213).
C H A P T E R NINETEEN
PRODUCERS IN THE PAST
19.1 Amarna Basket Makers

The basket makers from Amarna, w h o were they, what was their position in the
village. W e r e they professional basket makers or nonprofessionals?
If w e would plant the present situation onto the N e w Kingdom workmen's
village then the picture would be as follows: coiled baskets were made by w o m e n ,
from grass, which they could collect at 5 k m distance near the Nile and from doam
palm leaf, which perhaps would have to be bought. They were nonprofessional
basket makers, producing for their o w n needs only.
Twined and woven matting would be made by part-time or full-time
professional m e n in small workshops which would have enough space to place a
horizontal loom. The grass would be gathered for free, the doam palm leaf
perhaps had to be bought.
Other items, such as brushes and pot stands would be made by the
households themselves (men or w o m e n ) for their o w n use. The amphora net was
probably imported, because papyrus probably did not grow in abundance in the
vicinity. The knotless net, made of d o a m palm leaf was probably imported too. It
came perhaps with the d o a m nuts from wherever these were imported.
Is this a feasible construction? The workmen's village is thought to have
housed the families of the m e n w h o worked in the tombs. Most of the work
consisted of low-skilled labour, excavating the rock to make the rough outlines of
the tombs. S o m e of the half finished tombs at Amarna show that rough stone
hacking was going on at the same time as fine relief carving. The baskets with
rubble were probably carried past the people finishing the reliefs in the narrow
entrance way. Even if the relief carvers were also living in the village, the
community seems to consist more of workmen, rather than highly skilled and
valued artists, with as special position, as was the case in the 19th and 20th
Dynasty community in Deir el-Medina.
Did such a community have its o w n basket makers? The answer is probably
'yes'. In the twenty years of its existence, the village developed a lot of economic
activities, from pig breeding to growing garden plants. In the front rooms of
several houses limestone beam sockets were found, which have been interpreted as
emplacement blocks for (vertical) textile looms. They usually c o m e in pairs and
two were found also in the south annexe of West street 2/3, excavated in 1986
(Kemp et alii 1987, 4-6). Spindle whorls have been found in large quantities which
are a further indication of textile production in the village.
No bundles of prepared basketry materials were found, but it seems very

likely that coiled basketry was made in the village and probably by w o m e n . The
tomb paintings, which show scenes of daily life, do not show basket makers. This
can be taken as an argument that coiled basketry was not made at a professional
level, but as a household activity by nonprofessional w o m e n .
W a s there a mat maker in the village? The most likely area for a mat
weaving work shop would be West street 2/3. This is a house with an annexe,
excavated in 1986 in the vicinity of which most of the woven matting fragments
were found ( K e m p et alii 1986, 30; 1987, 4-11). West street 2/3 did not have any
features, however, which point at the one time presence of a horizontal loom (e.g.
in the form of four post holes).
The woven grass matting was found in the open area to the south of West
street 2/3 (indicated as West street 1 cf. K e m p et alii 1986, 28-33), directly on the
surface of the organicfill,as if blown there. They are taken as an indication that
the annexe of West street 2/3, just to the north of this area, was covered with a
light roof of mats and reeds ( K e m p et alii 1987, 4). Within the annexe of West
street 2/3 more grass woven matting was, lying directly on a m u d floor. Similar
roofs have been found by Peet and Woolley (Peet and Woolley 1923, 67-68). In
their publication probably only a small part of the basketry found was mentioned
and the descriptions give not nearly enough detail to be of any use for a study of
basketry production and work shops in the workmen's village.
O n e of the arguments against this matting being made in a workmen's village
atelier, is that the woven mats found have so many different top and bottom edges.
Unless w e propose that a mat maker makes different edges all the time, this would
indicate at least two or three producers. If w e accept that, then it seems more
likely that the mats come in from (several) work shops in the main city.
The twined mats lead to a different conclusion. Although not many edges
have been preserved, which might tell us more about the producers, the large
quantity of this kind of matting and the lesser demands to the size of the work
space, make it probable that at least some forms of twined matting were produced
in the workmen's village. There are no indications whether or not furniture
matting was made in the village.
19.2 Qasr Ibrim Basket Makers

The situation at Qasr Ibrim is slightly more complicated than at Amarna, because
the periodicity is much more widespread. Fluctuations over time should be
incorporated. A second problem is that the deposits of the different periods are
very different in character. From the late-Meroitic and the Ballana periods w e
have house hold deposits. Especially the excavation of a number of cellars yielded
a large number of coiled baskets. Evidence from the Christian and Islamic periods
comes mainly from pits dug into the earlier layers. The pit lining, consisting of re-
used baskets and mats, presents a very specific selection of materials (only flat
objects which can cover a large surface, such as mats and cut up carrier baskets).
Prod ucers in the Past 411
Then there is the question of the relation between the habitation of the hill top
and the villages d o w n at the river side. W a s there a difference in habitation or
social grouping between the inhabitants of Qasr Ibrim and those of the villages?
W a s Qasr Ibrim an independent town, with its o w n craftsmen, or was there a
regular interaction with the lower villages. W e r e the villagers retreating to Qasr
Ibrim in difficult, unsafe periods? W e will probably never k n o w the exact relation,
because the settlements on the valley floor have not been excavated, and are n o w
under 70 metres of water.
A n attempt of picturing the basket makers from Qasr Ibrim, based on the
present would be as follows: In the late-Meroitic and the Ballana periods (first to
sixth century A D ) coiled baskets are made by nonprofessional w o m e n . The Qasr
Ibrim coiled baskets are decorated with wrapping patterns or colours. The
decorative techniques have no parallels in the Egyptian basketry, but seem to be
linked to East African traditions.
S e w n plaits basketry from the same period is probably also made by w o m e n .
This suggestion is based on three arguments. Small bundles of prepared raw
materials were found in house refuse dated to the Ballana period. A flat needle,
probably used for sewing the plaited strips, was found in late-Meroitic domestic
refuse. In the third place, the sewn plaits matting fragments found were m a d e of
twill date palm leaf plaits, sewn with unspun strips of palm leaf (either date or
doam). These thin strips were connected with an overhand knot.
In present day DihimTt, w o m e n use a very similarflatneedle to sew the twill
plaits with unspun strips of d o a m palm leaf, which are connected with the same
knot.
W a s it considered a valuable thing that w o m e n could m a k e baskets? In
present day N e w Nubia it is. The basketry covers are part of a social exchange
system. Judging from the large number of small coiled plates, decorated with
coloured winders, which were found in a very small area at Qasr Ibrim, it seems
very likely that the basketry then was considered of importance too. Whether they
had a similar role in society as at present is impossible to say with certainty, but it
is not unlikely.
The excavated area in Qasr Ibrim, saw an expansion of the built up area on
top of the collapsed Ballana period houses. The expansion took place in the Islamic
period, after the advent of a military contingent of the Turkish army. The, mostly
Bosnian, mercenaries never left and gradually mixed with the local population.
The basketry of this period, found in grain storage pits, features a m o n g other
things, coarse sewn plaits baskets, which are m a d e in afive-strandstabby plait of
doam palm leaf. They are sewn with string, rather than unspun strips of knotted
palm leaf.
Could w e interpret this basketry as an influence from Egypt, where at present
sewn plaits baskets are sewn with string? W e r e the soldiers involved in basket
making? It seems quite likely that the army had their o w n craftsmen, for
maintenance or armour, and perhaps also for the production of coarse utilitarian
basketry. It seems not improbable that (part time) professional m e n were involved
in the production of sewn plaits basketry sewn with string and in making twined
and woven matting. Whether the matting and furniture workshops were in Qasr
Ibrim itself, or in the villages down at the river, remains a matter of speculation.
19.3 Some New Kingdom Basketry Terms

Although I have stated the difficulties in identifying present day terminology, let
alone ancient terminology (p. 16-17), it is, nevertheless, tempting to try and see if
the information on ancient basketry production can tell us anything new in relation
to the ancient Egyptian terms. It should be stressed however, that this is nothing
but a first survey based on the standard work of Janssen on the Deir el-Medina
ostraca (Janssen 1975).'
The basket makers from Deir el-Medina lived between 60 and 250 years after
those from Amarna and it is often difficult to decide to which period the texts in
which the basketry terms are mentioned exactly belong (Janssen 1975, 15-17).
Although the origins of the Amarna workmen are not at all clear, they might have
come from Memphis, rather than Upper Egypt, which means that there may be
regional differences as well. If the present situation can be taken as comparison,
then this would not have influenced the appearance of the baskets, but it would
have made a difference in the terminology. Janssen's remark that the words used
for c o m m o n objects in the Deir el-Medina workmen's village differ "from that
used for it by the rest of the Egyptians" (Janssen 1975, 133), is stressing the
uniqueness of Deir el-Medina on the basis of the wrong aspect: the basketry
terminology probably differed greatly between all regions of Egypt.
It should be kept in mind, while using the terms and qualifications of the
Deir el-Medina ostraca as a lead to identify specific basketry techniques, that the
writers of the ostraca not necessarily used the basketry terminology consistently or
rightly. The many misnomers in English with terms such as 'reed' baskets (while
none of the English baskets are made of reed, but of willow rods which is not even
vaguely similar) may be a vivid warning.
In thirteen sections, Janssen lists different terms and the prices quoted for
these. H e also tries to identify the type of basketry technique, the shape, size and
material of the baskets.
The kbs-basket and the Techniques ofnbd and hnd

(Janssen 1975, 133-139)
Janssen proposes that the kbs is the grain basket shown in many agricultural
scenes. The kbs -basket is in many instances specified as beingribd,which seems to
For this survey I do not base myself on the original texts, which is something that
should be done in future research. Rather, Janssen's conclusions are held to the light of the
new information built on a synthesis of well dated archaeological material and ethno-
archaeological work.
Producers in the Past 413
indicate the technique.2 Janssen proposes that nbd means coiling on the basis of
several arguments. T o explain the different contexts and meanings of the word
nbd, he takes as basic meaning 'to twist'. In coiled basketry the twisting of the
strands indicates the wrapping of the winder.
term specifications suggestion Table 10-13

kbs nbd (twined) twined grain basket fine: A or B
coarse: J or K
dnit nbd (twined) container for fruit, incense,
s m ' (fine) etc.
ht.ti (coarse) coiled or twined?
krht nbd (twined) container for fruit
ht. ti (coarse) coiled or twined?
mtrht nbd (twined) twined sieve E, F or G

m n d m + nkr nbd (twined) (coiled?) basket with twined E, F or G
sieve as lid (?)
irgs, irks nbd (twined) sack made out of a twined mat Horl
kskst nbd (twined) large twined bag
'rk, sack (also of leather)?
h'w cheap basket, no specifications
htp twined matting, container? L
tm3 sm 'n swy fine twined matting with grass C
sm 'n n w h fine twined matting with string A, B
rdmt matting of sedges
sdr + tm3 hnd (woven?) woven grass sleeping mat with
twined under mat?
skr, skr' nbd (twined) ?
find (woven?)
Table 19-1 Baskets mentioned in Deir el-Medina texts, as given by Janssen (1975). The
third column lists suggestions for a re-interpretation, based on the Amarna
basketry and the ethno-archaeological survey. The letters in the fourth
column refer to the twining techniques listed in Table 10-13 and depicted in
Figure 10-14 (pp. 196-197).
Usually nbd is translated as 'to plait', but Janssen remarks justly that (pharaonic)
Egyptian baskets are never plaited (Janssen 1975, 136).
This basic meaning of 'to twist' would be equally suitable to indicate twining,
but Janssen prefers the identification of nbd with coiling, because Egyptian baskets
are usually coiled. H e states that "hnd cannot m e a n anything other than 'twined'.
Its rare occurrence corresponds too well with the scarcity of twined basketry"
(Janssen 1975, 139).
In the bar chart of Figure 12-2 (p. 251), it appears that twined basketry is the
technique that occurs by far the most in A m a r n a ( 3 8 % of all finds is made in a
form of twining). This reverses Janssen's argument completely. There is another
good reason to take nbd as meaning 'twined'.
If, as w e tentatively concluded above, coiled baskets were made mainly by
nonprofessional w o m e n , then they would presumably not appear that often as
exchange goods. The coiled technique can be expected to feature less in the texts
than the twined technique.
These arguments lead up to the most convincing one: the grain bags which
are depicted frequently in tomb paintings and of which fragments were found at
Amarna, were made in a twined technique, usually of grass or doam palm leaf
string (twining techniques A , B, J and K, cf. Table 10-13 and Figure 10-14, pp.
196-197).
Janssen argues convincingly that the term nbd probably developed into the
general meaning of 'making baskets'. In the Graeco-Roman period this term was
probably transferred to the sewn plaits technique, which mostly took over the role
of the twined basketry. Although remarkable, I do not have the expertise to judge
if the present Kensi and Fadidja term nibid for sewn plaits baskets and mats can be
taken to stem from the same root.
The dnit-basket
(Janssen 1975, 140-143)
This basket is known from a large number of price quotations too. The dnit is
several times characterised as being nbd. This would refer to a twined basket, but
the function, it is known as container for fruit and incense, points m u c h more to a
coiled basket. There are several arguments for and against this.
If w e take nbd to mean in all instances 'twined' than w e should interpret the
term dnit at least in those cases where the word has been specified with nbd, as a
twined, rather than a coiled basket. If w e interpret nbd as a more general term of
'basket making' than dnit could be a coiled basket.
Following the argument that the coiled baskets were produced as a house
hold activity then coiled baskets, even though they are very c o m m o n , would not
be sold as often as twined basketry. In two ostraca the dnit is said to belong to a
w o m a n . This could be used to argue for or against the idea that the dnit is a coiled
basket made and used by w o m e n . Mentioning the normal situation can be
considered superfluous, which would be an arguement that the dnit basket is
usually not, or not specifically, in the possession of w o m e n .
Janssen connects dnit with other meanings of this root, such as 'dam',
'dyke', and the areas marked out by such boundaries (e.g. 'share' and 'registry of
real properties'). Since coiled basketry is the only technique in ancient Egypt
which results in baskets with rigid walls, the term dnit might be related to the
aspect of a 'dam', as a surrounding rigid wall.
The dnit basket is sometimes specified as sm' fine), or ht.ti (coarse). Both
fine and coarse coiled baskets occur, but this is not a decisive argument because
fine and coarse twined basketry exists too. In short, it is not clear what kind of
basket the dnit is.
77rc? krht-basket
(Janssen 1975, 143-145)
This basket, which is not mentioned often, is once said to contain fruit. In two
texts it has been specified with the qualification nbd, which strictly speaking,
would m a k e it a twined basket. Janssen expresses the hope that with the study of
the basketry of Deir el-Medina the shape of this basket, which he considers to be
coiled, can be defined, because the term does not occur before the Twentieth
Dynasty. The publication of the Deir el-Medina basketry, however, does not give
any firm dating of the material and w e have to give up this hope.
77z<? mtrht-sieve and the mndm with nkr

(Janssen 1975, 145-149)
With some reservations the term mtrht is translated as 'strainer'. Sieves and
strainers in ancient Egypt consist of twined grids, held in a coiled rim. The small
sieves have a rim of only one coil, larger sieves and strainers have coiled walls
and a twined base.
The nkr is also thought te be a sieve, but cheaper and therefore perhaps
smaller and less deep than the mtrht. The combination of m n d m with nkr occurs
regularly, leading to speculations that the nkr is a smallflatsieve which functions
as a lid for the m n d m basket. They seem to form a pair, m u c h as a dustpan-and-
brush do. In an Eighteenth Dynasty context the m n d m occurs as a basket for fruit.
Both the m n d m and the nkr are combined with the specification nbd.
The irgs or irks basket

(Janssen 1975, pp. 149-150)
This basket is also characterised as being nbd (twined). In one ostracon it is said
that one fine mat makes one irs (abbreviation of irks). This is reminiscent of the
masballa bag in Middle Egypt, which is sewn of two small twined mats. T w o
fabrics similar to that of the masballa bag of today were found at A m a r n a (twining
variations H and I in Table 10-13).
Figure 19-2 Decorated coloured baskets, tribute from Nubia, as depicted in the tomb of
Rekh-mi-Re (drawing after Davies 1937, Plate XIII).
The 'nbr basket.

(Janssen 1975, pp. 150-151)
Not m u c h is k n o w n about this term. It is quite an expensive basket, which features
twice in a text with a list of funerary equipment. N o further specifications are
given in the Egyptian texts. Purely speculating, w e could propose that, since well
made, decorative coiled basketry w a s part of the funerary equipment, this basket
might represent such special coiled basketry, perhaps imported from Nubia, like
the baskets depicted in the tomb of Rekh-mi-Re probably were (Figure 19-2). This
would explain their relatively high price.
The kskst-basket, the 'rk-sack (?) and the h 'w basket

(Janssen 1975, 151-154)
These three items do not occur often and very little can be said about them. The
kskst is specified to be nbd. It appears to be a large object and could, for instance
be a large twined bag. T h e 'rk is in one case said to be of leather, and Janssen,
therefore defines it as a kind of sack. The h'w baskets are k n o w n to be cheap, but
not m u c h more can be said about them.
Matting: hip, tm3 and sdr

(Janssen 1975, 154-161)
The matting found at A m a r n a is either woven out of halfa grass, straw or sedges,
or twined out of halfa grass or d o a m palm leaf. In tomb paintings the htp mat is
always depicted as a twined mat, probably with widely spaced rows of twining (cf.
Figure 15-19 p. 320, variation L in Table 10-13). Janssen thinks, however, that in
the price ostraca the word htp refers to a container. No specification of the
technique is given, so there is no basis for further speculation.
The t m 3 mat also seems to be a twined mat. Specifications in the ostraca
include the combinations s m 'n swy (fine, made of hay), s m 'n n w h (fine, made of
string) and rdmt (a sedge type). The first two specifications seem to point to
twined matting, which can be made of bundles of dried grass (perhaps indicated as
'hay') or of string (variations C, A and B in Table 10-13).
Sedges are made into twined matting too, if w e m a y belief the tomb paintings
where the herdsmen are making twined matting from papyrus culms. In Amarna
the side edge of a mat of bundles of sedges (papyrus) was found (Plate 12-3, p.
253), but this was the edge of a woven, rather than a twined mat. For the moment,
considering the large quantities of twined matting found at Amarna, w e presume
that the tm3 was twined (floor) matting.
Janssen translates the combination of sdr and tm3 (also occurring as 'the sdr
and its tm3') as a sleeping mat (or pallet) and under-mat. sdr is derived from the
verb 'to sleep, to lie down' and the mat is specified as being find, which Janssen
translates as 'twined' or 'twisted'. In two texts the mat is said to be made of rwy
('straw').
The finds at A m a r n a which are the most likely candidates to be interpreted as
sleeping mats, are the woven grass mats (Plate 10-10, p. 191). They have the right
length (1.65 m ) and width (0.90 m ) to have this function and they are well made
and thick enough to make a comfortable sleeping mat. At A m a r n a two small
fragments of such a woven mat were found to have been made of bundles of straw
(p. 191). Most of the woven matting, however, was made of single grass stems.
The last basketry term listed by Janssen is the skr or skr'. This basket is
sometimes said to be nbd (twined?) at other times hnd (woven?). If this refers to a
container, than the translation 'woven' seems to be out of the question, however,
since no woven containers are known to exist.
The inevitable conclusion from this brief attempt is that a more systematic
survey of basketry terminology is needed, based on a thorough investigation of the
original texts. Even then the expectation that it will be possible to make one to one
identifications of terms with archaeological finds or depictions in wall paintings is
too optimistic. The present Arabic terminology lacks consistency and differs
sometimes between one village and the next. The situation will not have been
much different in ancient Egypt, where there was not even the unifying force of
Television Channels 1 and 2. Furthermore, it is not unthinkable that the ancient
writers were not familiar with the local or specialised terminology and made up a
terminology of their own.
CHAPTER T W E N T Y
SUMMARY AND CONCLUSION
20.1 The Textual, the Audio-visual and the Tactile

The report on the work on ancient Egyptian baskets and present day basket makers
is given in words. Language requires a mental ordering of perceptions and
observations. In the previous chapters the angle of view has shifted from detailed
(basic structure) to the broader spectre (the world of the basket makers). However,
"There is no simple correspondence between word and world" (Shanks 1992, 38).
The world according to basketry is m a d e up of textual, audio-visual and tactile
aspects. The first two have been represented in the book and the video
respectively. They have been cross-linked in different manners.
The tactile aspects are also part of the research, but the m e d i u m to transfer
this kind of knowledge is through shared and gained experience (apprenticeship).
Sitting and working with the basket makers is what I did, but what is impossible to
duplicate. "I will show you" in this context is often not visual at all, but means in
effect "I will make you feel". Getting a 'feel' for something is at the same time
very concrete (handling materials) and abstract (when trying to put that experience
into words).
Numerical aspects have not been mentioned in the title of this section,
because they can be considered a tool to translate the visual or tactile into
language. O n e of the functions of measuring (counting, weighing) is to create a
type of knowledge that can be reproduced in a book. The working rhythm (audio-
visual) is expressed in units of time or movement (Chapter 16). S o m e aspects of
the tactile could be expressed in measured units too (for instance the tension
strength of rope). Used as such, it is in effect a deficient translation of adequate,
but nonliteral information transfers.
Inadequate as the translation m a y be, a m o n g academics it is valued higher
than practical knowledge (cf. Benz Hogseth 1998). If there was ever a monolithic
nature of academics, apart from the discussions arguing against it, it is rapidly
dissolving, however (e.g. Shanks 1992, 35; Hodder 1999, 29). This gives the
opportunity to discuss different kinds of knowledge.
The ancient baskets have been expressed in language with the help of
classification. The work of the present day basket makers has been linked to this
organised scheme (the twenty techniques in Chapter 15), in the form of a construct
description of the production process. The video does not present an unordered
world, either. The sounds and images have been selected during filming and edited
into an ordered narrative.
20.2 Classification a n d T e r m i n o l o g y
T o write a text in which archaeological basketry is caught in words, a description
has to be made. Each basket has a myriad of properties, and some of those are
selected to describe the techniques. These are ordered in a classification, which
concentrates on the basic structure of the baskets: those aspects that can be
recorded, even if only a small fragment of a basket is found.
Following these criteria 20 techniques are listed, which were either found in
the excavations (at A m a r n a or Qasr Ibrim) or are presently produced in Middle
Egypt or N e w Nubia. This classification is just one possible w a y of ordering the
basketry. It can be augmented with more aspects or a completely different angle
can be chosen.
The classification is set up in combinations of binary oppositions (passive /
active), tripartite divisions (elements moving in one, two or three planes) and in
counts of numbers (number of systems, number of orientations, number of
elements). Present users of Egyptian basketry have a very different classification,
which centres around the function and size of the baskets. The criteria behind such
an emic classification can be found partly by looking at the terminology used.
Terminology is a set of words to discern different objects. A n y terminology
is based on a set of criteria, and on an -often implicit- classification. The
classification as presented in Chapter 9 is a West-European scholarly endeavour,
which has no connection with the w a y these baskets were regarded in the past.
Is it possible to retrace the ancient classificatory organization of material
culture? Attempts have been made, for instance by Arnold w h o states that "(...)
cognitive systems should warrant some attention as a significant factor in the
production of artefacts" (Arnold 1971, 22). Shanks and Tilley see the reproduction
of an ancient taxonomic system merely as a description of cognitive systems, while
what is needed in their opinion is an explanation related to social practices and
patterning. "Material culture should be regarded as not merely a reflection of
cognitive systems and social practices, but actually involved in the formation and
structuring of those practices" (Shanks and Tilley 1987, 85).
In m y opinion even this is too limited an approach. If at all possible, the
identification of an ancient cognitive system not only relates to social practices, but
in a m u c h broader sense to the ordering of the world, which is what classification
and terminology ultimately are about. It should be noted, however, that within a
society there is not just one ordering principle, but a complexity of m a n y (and
often contradictory) senses of order.
20.3 Systems, Activity and Actions

T o put in words those aspects that are visible in basketry fragments as well as in
the actions of basket makers w e resort to the terms "systems" and "activity" to
describe basketry techniques. In Chapter 5 m a n y pages have been filled with an
explanation of what the term system refers to. By discerning group of basketry
20 S u m m a r y and Conclusion 421
strands that have the same role in the technique and describing the interaction of
different systems the basketry fragments are recorded according to uniform
criteria.
M o r e important, by looking at the activity of the systems, (active or passive),
a link is m a d e with the actions of the basket makers. T w o definitions have been
given in the cause of this book, which express the change in emphasis from
structure to action. First it was said that the (rigid) passive system makes up the
body of the technique, while the (flexible) active system causes the coherency
(p. 85). Later the definition changed and stated that the strands of the active
system are handled more frequently or longer by the basket maker than those of
the passive system (p. 269). Both definitions are useful, because they highlight
different aspects: one looks at the structure as a result of a process, the other at the
actions of the basket maker. It is a step towards understanding the production
process of the ancient basketry, to get from the static to the dynamic.
The sequences shown in the video are also described in the text (Chapter 16).
Such a detailed description helps in watching the process and the movements of the
basket maker. Although every producer acts differently, the forced attention to
detail brings to light important information that would be missed w h e n adhering to
a schematized or overall view of the production process.
The actions of basket makers have been divided into repetitive and occurring
actions. The repetitive actions determine the image of a production process. They
are the thousands of wrapped stitches that m a k e up the fabric of a coiled basket, or
the hundreds of weft yarns, w o v e n into the warp.
The occurring actions are m u c h less obvious. They are 'supportive' to the
ongoing repetitive actions. Adding n e w lengths of material, turning at the edge of a
row of weaving, smoothing the fabric from irregularities, are examples of
occurring actions. The results of the occurring actions are often not obvious either.
The insertions are hidden in the fabric. The edges are looking smooth, but unless
studied in detail, their structure is not clear. In the production process the
occurring actions are difficult to register. W h e n making the edge, it is often not
exactly clear what the basket maker does, because the m o m e n t passes quickly.
Because of the fleeting character of the occurring actions, they cannot easily be
copied or imitated. Unless a basket maker slows d o w n on purpose to show and
explain what he does exactly, it is difficult to understand the movements. It is,
therefore, specifically in the occurring actions that the passing on of knowledge
and tradition can be spotted. The basic technique m a y be wide spread and show
little variation, but there are different solutions to technical details of insertions
and making the centre, rim or edges.
There is a link between the production phase and the dominating actions. The
start of the work is usually characterized by occurring actions. The basic structure
is dominantly repetitive. The end of the basic structure is characterized again by
occurring actions. Those actions that result in characteristic features usually take a
relatively short period in the total production process. The start of a plait, for
instance, takes 24 seconds.
The movements of the repetitive actions show a consistency and regularity

that are a prerequisite for developing a working rhythm. Nevertheless, the
occurring actions are often done in a certain rhythm also, although this rhythm is
more difficult to distinguish. The working rhythm differs with each technique, but
also with each basket maker.
Skilled basket makers work in a steady regular rhythm. They usually work
with intense concentration. Their movements are not necessarily the most
economical. Several basket makers, for instance, were found to slam their needle
against the wall of the basket before making a stitch. This does not have a function
in sewing the basket, but helps maintaining the working rhythm. Changes in the
working rhythm indicate changes in production (transitions). Studying the work-
eye contact indicates which moments in the process are difficult. During transitions
from one phase to the next, the eyes often wander. During occurring actions and
mostly also in the repetitive phases of the work, the eyes are focussed on the work.
Ancient Egyptian tomb paintings represent production processes by showing
a selection of the most 'typical' scenes. For many activities, such as harvesting,
weaving, beer brewing and leather working, the scenes considered representative
by the ancient artists reflect the production phases that are characterized by
repetitive actions. For mat making scenes the situation seems to be different.
There is one scene with a mat weaver, which shows the occurring action of
reaching out for the cross beam of the loom, to tighten the weft (cf. p. 313). The
twined mat making scenes are as yet unexplained. They seem to represent the
beginning of the work (laying out the materials, sometimes making the string) and
the finishing touch (brushing the mat?). Basket making does not occur in the tomb
paintings.
The working order is not always performed linear from preparation of the
materials to the finishing touch. If, during the work, the raw materials are
finished, new have to be prepared. During the production process periods of
waiting are filled with actions which 'logically' belong to another phase. The
production phase of the finishing touch is almost a ritual moment, which marks the
end of the process. The actions in this phase could be characterized as
'symbolically functional'.
In Egypt it appears that regional traditions can be discerned by the basic
structure, the use of materials, the size of the systems and the decoration, while
local traditions are characterized by small differences in start, finish, insertion of
lengths of materials and also the decoration.
20.4 Context and Meaning

Context and meaning can be considered tactile aspects in the sense that they are
dealing with things that either can be touched (the material context), or have to be
grasped (meaning). The meaning of basketry can be considered an immaterial
context, which reflects more than the social status of the basket maker. Production
and use of basketry are linked to both group identity and personal expression.
How are we to grasp the meaning of basketry? Hodder states that answers to
questions such as 'what does that mean' are "often constructed, forced and
unhappy"(Hodder 1992, 204). Meaning is usually not a conscious construction, and
it cannot be separated from the objects, nor can it be separated from other
dimensions of the production (economical, social, ritual, magical, political, cf.
Tilley 1998, 59).
Trying to formulate meaning into words is truncating it. Basketry items such
as the bilum (knotless netting bag) in N e w Guinea and the Yekuana basketry for
the A m a z o n region of lowland South America are a complex of associations,
which are not at all vague, but have a very concrete and powerful meaning (Tilley
1998, 62-72).
The Nubian basket makers maintain that their colourful stitches are purely
decorative and nothing more. In the interviews that both Boyce Driskell and I did,
the basket makers state that they consider themselves better basket makers than the
persons w h o taught them, because they use more colours and m a k e n e w patterns.
O n a conscious level, basketry is purely functional, one of these functions being
their use as wall decoration. Despite innovations, however, the villages show a
definite character in the use of the same materials, colours and patterns, which is
distinguishable from other villages. There is a clear difference between the Kensi
and Fadidja decorated coiled basketry.
There is a gender division between the persons w h o use and m a k e the
baskets. This is strongly present in Middle Egypt, where the w o m e n m a k e the
baskets that they use (mostly inside the house), while the m e n make baskets used
in the streets and in the fields. In N e w Nubia only the w o m e n make baskets, while
coarse basketry for work in the fields is purchased from an outside group (pp. 406-
408).
In Middle Egypt there are professional, nonprofessional and part-time
professional basket makers. Studying the present day basket makers gives the
opportunity to look at the way they work, the place they work in, the tools they
need and the results of their work. It appears that professional basket makers are
skilled, they need more equipment than nonprofessional basket makers and have a
specific work space. They work concentrated with a very steady rhythm. The
result of their work is uniform: one basket is the same size and has the same
appearance as the next.
If professional basket makers are skilled, w e cannot switch around the
argument, stating that skilled basket makers are by definition professionals. M a n y
nonprofessional basket makers are very skilled and work concentrated for hours in
a steady working rhythm. Baskets from an archaeological context, which are m a d e
with extremely regular stitches, tell us that their producers were skilled basket
makers, but it does not indicate the level of professionality (cf. pp. 390-391).
Professionality is linked to the physical demands of the work: material
resources and bodily strength. If tools and a specific work space are needed, the
producer is more likely to be a professional. O n the other hand, endurance and
hardship are not necessarily linked to professionality. All basketry techniques
require long periods of repetitive actions. All skilled basket makers work in a
steady rhythm, the only difference noticed being that nonprofessionals work slower
and more precise, than professionals. Consistent irregularities in the ancient
basketry, suggesting a hasty, rapid working rhythm, are a more telling indication
of professionality than a regular meticulous appearance (working from the
presupposition that in the past professional basket makers had to work fast to make
production).
At present the only full-time professional basket makers are the mat weavers,
the twined mat maker and the qafas maker. All have work shops, or a specific
designated area to work in. The mat makers (both weaving and twining) use simple
looms. The qafas maker has a hole range of tools, from large knives and cleavers
to measuring sticks and hollow awls.
Basketry production depicted in ancient Egyptian tomb paintings is rare.
There is one depiction of a mat weaver (Middle Kingdom period). In many tombs
scenes of twined basket making occur. This is part of the standard repertoire of the
tomb decoration and occurs from the Old Kingdom onwards. It does not seem a
coincidence that the basketry techniques depicted are those, which, based on
physical requirements, are thought to have been a full-time profession in antiquity.
Net knotters, rope makers and textile weavers are depicted regularly too. In
the Middle Kingdom tomb of Khety at Beni Hassan two groups of flax spinners
have been depicted. Spinning for cloth is a female activity, spinning for nets is a
male activity.
It is remarkable that neither the production of twined, nor of coiled baskets
are represented in tomb paintings, while there are thousands of depictions of the
baskets themselves in agricultural scenes and in offering processions (Wendrich
1999). This can be used as argument that twined and coiled basketry were not
m a d e by full-time professionals, but by nonprofessionals and most likely w o m e n .
The present day Middle Egyptian labour division of agricultural baskets being
made by m e n , while coiled baskets for uses inside are m a d e by w o m e n is,
perhaps, an argument against the suggestion that in antiquity twined grain baskets
were produced by w o m e n .
At present, the social position of basket makers is different in Middle Egypt
and N e w Nubia. In Middle Egypt town people look d o w n on basket making as a
village activity. The w o m e n of the village, on the other hand, are proud of their
work. Even the w o m e n of the richest family in the village of el-Hagg Qandil are
engaged in basket making (just as they are still baking their o w n bread and making
their o w n cheese). The professional m e n , on the other hand, seek to improve their
position. They want their children to be educated and find better work, preferably
with the government.
The Nubian w o m e n , on the other hand, are unambiguously proud of what
they do. The elder w o m e n think it a shame that the young girls are not interested
any more in making baskets. They blame the television.
There is one ancient Egyptian text that gives an indication that in ancient
Egypt mat weavers were in a poor condition, both socially and economically.
Since this is probably a satirical text, it cannot be taken at face value (p. 371,
Lichtheim 1973, 188).
20.5 Continuity and Change

What would you like to emphasize? There are many indications of continuity and
many changes over time, which have been surveyed in Chapters 10 and 11. Here it
may suffice to indicate the broad outlines.
The continuity is strongly regional. The basketry excavated at Qasr Ibrim has
many similarities with the modern Nubian basketry, while the basketry from the
workmen's village at Amarna corresponds with the modern Middle Egyptian
basketry. There are more differences between the present day Middle Egyptian and
Nubian basketry than between the ancient and modern basketry from the same
region.
Coiled basketry in both Amarna and present day Middle Egypt is undecorated
and quite coarse. The Qasr Ibrim coiled basketry is decorated with patterns made
with coloured strands or wrapped winders, while the present day Nubian basketry
is decorated with lavishly coloured patterns. The coiled basketry in both Qasr
Ibrim and N e w Nubia is, contrary to the coiled basketry from Amarna and Middle
Egypt, not used as containers, but as lids or covers.
Twined basketry is by far the most frequently occurring technique in
Amarna. Since the introduction of the sewn plaits basketry, probably in the
Graeco-Roman period (third century B C ) , this technique has gradually taken over
most of the applications of twined basketry. It is used for carrier bags (in Qasr
Ibrim, N e w Nubia and Middle Egypt) and matting (in Qasr Ibrim and N e w
Nubia). In present day Middle Egypt only closely twined sacking is still produced.
S o m e of the sewn plaits basketry is sewn with string, which seems to be an
Egyptian tradition, introduced into Qasr Ibrim in the Islamic period with the influx
of troops from the North. Most of the sewn plaits basketry found at Qasr Ibrim is
sewn with unspun strips of palm leaf, knotted together. In present day N e w Nubia
the plaits are still sewn that way, with a distinctiveflatmetal needle. One such a
needle has been found at Qasr Ibrim in a late Meroitic context (third century A D ) .
In Qasr Ibrim there are some indications for imported baskets. A small
number of fragments were found made of rushes. Parallels for these baskets were
found in excavations at Karanis in the Fayoum, 900 k m to the North. The rushes
were made in a stake-and-strand technique (mostly twining). At present stake and
strand basketry is not produced in N e w Nubia or Middle Egypt, but these baskets,
which differ from the locally produced basketry in that they have rigid walls, are
imported from the Fayoum and the Nile Delta. They are made of the split culms of
the giant reed or the branches of the henna shrubs.
During the research there appeared to be continuity on some levels,
differences in the use of techniques or materials on other levels. This continuity is
not essential for the application of ethno-archaeological research. The purpose of
ethno-archaeology is not to state that 'since w e have concluded that feature A is
similar, we may conclude that feature B, on which we have no information, is

probably similar too'. It m a y be clear that, in the same vein, the video is not used
to give an impression of past lives. Instead, ethno-archaeology is used to
understand the production process and the demands of the work. It helps to
interpret the material remains and to be aware of the immaterial features.
Baskets can be described as a combination of properties, a combination of
materials, techniques and function. Limiting the research to this would be missing
the most important feature, however. Apart from technical and functional aspects,
there are specific preferences and traditions of (sub)cultures. The meaning of
undecorated coiled basketry in Kensi Nubia and Middle Egypt, for instance, differs
widely. In Middle Egypt the undecorated coiled baskets are present in every
household in a purely functional role. Their 'value' is neutral. They are neither
special nor shameful. In Kensi Nubia, however, undecorated palm leaf baskets are
considered as old and backward. They are used inconspicuously, for instance to
cover the water jar in the kitchen. W h e n I wanted to record these baskets, the
w o m e n thought it was quite shameful, while they proudly showed m e their
decorated baskets, which have a prominent place in the guest rooms of the Nubian
houses. The present coloured decorations m a k e the decorated basketry from Qasr
Ibrim look bleak. Traditions are not static, but they are subject to constant re-
inventions, changes and shifts in meaning.
The narratives that are based on the study of ancient and modern basketry
m a y choose to emphasize the continuity or the changes. The arrival of sewn plaits
basketry, probably in the early Ptolemaic period, can be taken as the cultural
influence of a foreign people, causing a break with the old traditions. The
continuity of the plain coiled basketry in Middle Egypt can be used as argument
that the Egyptian fellahin are the true heirs of the pharaohs. Both interpretations
lack balance, because society is not monolithic. The introduction of n e w ideas is
mixed with inventions and adapated to the local situation.
In the world according to basketry w e catch glimpses of a segment of ancient
society which is otherwise not accessible. Rural w o m e n and illiterate peasants
rarely have a voice in officially documented history. Basketry gives thus an
entrance, but at the same time a limited view. Nevertheless, it presents us with a
narrative that w e can expand through cooperation with other archaeological,
egyptological and historical disciplines.
APPENDICES
APPENDIX A
FIBRE IDENTIFICATION CRITERIA
O. Brinkkemper and E. van der Heijden
A.l Introduction
This appendix aims to provide the identification criteria of the botanical macro
remains used for the artefacts described in this thesis. For basketry, both parts of
leaves and of stems m a y potentially have been used in the past. The cell pattern of
the epidermis, the outermost cell layer of these parts, offers the criteria used in
identification. These criteria include the form of the cell walls, the presence of
appendages (hairs, prickles) and the presence (in leaves) or absence (in most
stems) of stomata. These stomata are pores in the epidermis to allow for exchange
of oxygen and carbondioxyde. Besides, epidermes of grasses (Gramineae) often
have silica cells, with distinct shapes and taxonomic significance. Figure A-l
shows typical epidermes of a Gramineae leaf and stem.
subsidiary cell guard cell cork cell silica cell

A
Figure A-l Epidermis of a grass (sugar cane) in surface views. A. lower epidermis of leaf
with stomata. B. epidermis of stem with cork cells and silica cells (after Esau
1960, 66, fig.7.3).
A.2 Methods
Part of the remains was identified by V a n der Heijden (1988). His study yielded 58
identifications belonging to the following families and species:
Gramineae Desmostachya bipinnata leaves and leaf sheaths

Imperata cylindrica leaf sheaths and possibly flower remains
Arundo donax culms.
Cyperaceae: Cyperus papyrus culms and Cyperus rotundus culms.
Palmae: Hyphaene thebaica leaves.
Linaceae: Linum usitatissimum fibres.
His study made use of transmitted light microscopy ( L M ) as well as scanning

electron microscopy ( S E M ) . Small parts of the ancient material were soaked for a
couple of hours in a 1 0 % KOH-solution, with 2 5 % a m m o n i u m or bleach (Eau de
Javelle). The separated epidermes were stained with Astra blue, dehydrated in an
alcohol series and embedded in D a m m a r resin. For S E M . studies, the material was
cleaned by ultrasonic treatment, dried and sputter coated with gold/palladium for
ca. two minutes.
The remaining material was studied by Brinkkemper. This study yielded 110
identifications of material from Qasr Ibrim, 92 identifications from Amarna.' The
following families and species were found:
Gramineae: Desmostachya bipinnata leaves and leaf sheaths

Imperata cylindrica leaf sheaths
cf Phragmites australis culms
culms of an unknown grass species (Gramineae type A )
Cyperaceae: Cyperus papyrus culms

Cyperus rotundus culms
Cyperus cf alopecuroides culm
cf Cyperus schimperianus culm
Juncaceae: Juncus arabicus culms

Juncus acutus culms
Palmae: Hyphaene thebaica leaves

Phoenix dactylifera leaves and fibrous leaf sheaths.
W e would like to thank the Egypt Exploration Society for their kind permission to
study basketry from Qasr Ibrim and Amarna in store in the Oriental Institute in Cambridge.
This material was brought to Great Britain after official divisions of material between the
Egypt Exploration Society and the Egyptian Antiquities Organization.
Appendix A, Fibre Identification Criteria 431
Because of the abundance of material to be identified, a less elaborate

preparation method was sought. The epidermes could be studied by simply cutting
thin slices of the other part of the stems or leaves by means of a razor blade. The
epidermes could then be observed in water with a light emitted microscope and,
where appropriate, mounted in g u m m y syrup (apathy). The advantage of this
method is its speed, but the resulting slides of course cannot compete with those
obtained by V a n der Heijden. Especially the amount of air bubbles present in the
slides often prevents clear photographs. However, this method can be
recommended for a fast identification, after which interesting species might be
treated in S E M .
The identifications were made by means of several earlier studies on ancient
Egyptian material and on recent plant anatomy. The main publications used were
those by Greiss (1957), w h o described the anatomy of 17 species commonly used
in ancient Egyptian material, and by Metcalfe (1960, 1971), w h o described the
anatomy of recent monocotyledon plants. Greiss' study was also used as a basis to
make a reference collection from material present in the States Herbarium of the
Netherlands in Leiden.
A.3 Description of the epidermes
A.3.1 Grasses (Gramineae)

Epidermes of grass leaves show a longitudinal division into costal and intercostal
zones. The intercostal zones overlay the assimilating tissue and thus contain
stomata, the costal zones overlay the fibres and are devoid of stomata. Epidermes
of grass culms, in contrast, are more homogeneous without separate zones. They
may or may not contain stomata. Cells of grass epidermes can be divided in 'short'
and 'long' cells, which is typical for this family. The short cells often contain
highly resistant silica bodies, with a variable shape. A short cell containing a silica
body is often accompanied by a short cell without a silica body. Guard cells of the
stomata are characteristically dumb-bell shaped, which among others also occur in
the Cyperaceae.
Desmostachya bipinnata
The leaf shows two types of epidermes. The abaxial side (Plates A-2 and A-3)
contains costal and intercostal zones. The costal zone is composed of rows
containing numerous silica cells and few short cells, alternating with rows
containing few silica cells and numerous long cells. The long cells (ca. 15-105 p.m
x 6-10 ^ m ) are slightly undulating or almost straight-walled, short cells and silica
cells are 7-10 p.m in diameter and virtually isodiametric. These typical round silica
cells are highly characteristic for this species. The intercostal zone of the abaxial
side also shows the round silica cells, but the rows containing stomata lack these
short cells. The stomata occur in one or two rows in each intercostal zone.
JL'... ;:.'".• ^.\P&'~iP*i*&&^^:.VC-

ivaiiStfrfijJ fbM-4': <JjiiW3fcafe.V^-.r^TaL'&«.-«
'X
P f e t . - f e C « H « " S
Plate A-2 Abaxial side of the leaf Plate A-3 Abaxial side of the leaf
epidermis of Desmostachya epidermis of Desmostachya
bipinnata (LM, 200x) bipinnata (SEM, 500x)
T h e adaxial side of the leaf (Plate A-4) has costal and intercostal bands,
resembling those of the abaxial side. Besides, a third band is composed of
isodiametric or rectangular, m u c h thinner walled cells. T h e costal zone is often
m u c h wider than on the abaxial side.
' • • ' . • s ' V . i V ' i
Plate A-4 Adaxial side of the leaf

epidermis of Desmostachya
bipinnata (LM, 260x)
Appendix A. Fibre Identification Criteria 433
The leaf sheath of Desmostachya bipinnata shows on the abaxial side a great
abundance of the typical round short cells, a differentiation between costal and
intercostal zones can hardly be observed, only the stomata indicate the intercostal
zones (Plate A-5). The adaxial side shows no silica cells or short cells. The cells
are relatively thin-walled, those in the costal zone are longer and less wide than
those in the intercostal zone. The cells enclosing the stomata are conspicuously
pitted (Plate A-6).
Plate A-5 Abaxial side of the leaf Plate A-6 Adaxial side of the leaf
sheath epidermis of sheath epidermis of
Desmostachya bipinnata Desmostachya bipinnata
(SEM, 225x) (SEM, 375x)
Imperata cylindrica (cogon grass, halfa grass)

O f this species, only leaf sheaths were found. The abaxial side shows wide costal
zones shows long cells with slightly undulating walls and characteristically dump-
bell shaped silica cells (Plates A-7 and A-8). The adaxial side lacks these typical
silica cells. The costal and intercostal zones are more or less similarly composed
of broad and elongated cells (100-150/xm x 25-35 /mi), with straight or slightly
undulating cells. The less characteristic adaxial side was not found by
Brinkkemper. The preparation method used will mainly produce abaxial material,
as these convex sides are easier to cut than the concave adaxial part.
Plate A-7 Abaxial side of the leaf Plate A-8 Abaxial side of the leaf
sheath epidermis of Imperata sheath epidermis of Imperata
cylindrica (LM, 350x) cylindrica (SEM, 400x)
Arundo donax (giant reed)

A culm of this species was found by Van der Heijden. The epidermis is not
differentiated in costal and intercostal zones (Plate A-9). The long cells are
comparatively short (22-40 /mr x 17-23 /xm), with conspicuous zig-zag cell walls,
which are strongly thickened and densely pitted. Short cells are more or less
crescent shaped and thin walled, silica cells are roundish (5 p m in diameter) and
accompanied by a short cell. The relatively long stomata (ca. 40 p m ) show
typically triangular subsidiary cells (Plate A-10).
»§ ''*22>« ^ " ] V ' « N t . • >-"**•
Plate A-9 Culm epidermis of Arundo Plate A-10 Stomata in culm epidermis
donax (LM, 480x) Arundo donax (LM, 720x)
Appendix a. Fibre Identification Criteria 435
cf Phragmites australis (reed)

The culm epidermis concerned shows an alternation of one long cell with one short
cell and/or one silica cell. The cell walls are undulating, but somewhat thinner
walled than the material from the reference collection. Greiss (1957: 65) observed
that the epidermal cells of side branches have thinner walls, which m a y explain the
thinner walls in the ancient material. However, it cannot be excluded that the
material belongs to a grass species not present in our reference material, hence the
tentative identification (Plate A-ll).
i*'"*»»ftllJ{*.ViL
Plate A-l 1 Culm epidermis of

Phragmites australis
(LM, 500x)
Gramineae-type A
The culm has straight-walled epidermal cells, long cells are extremely elongated,
up to 400 urn in length and 15-20 p m wide. Short cells are more or less square,
not wider than long cells (Plate A-12).
-» •"'y'" ,T —
•^
.
'* t
\:- i - -
..^C
•:.*""*~~
Plate A-12 Culm epidermis of .... 1
Gramineae-type A
(LM. 350x) ^^siag^^^--i.v&«r> ~13=3— ""
A.3.2 Sedges (Cyperaceae)

Culms of Cyperaceae, at least of Cyperus, the only genus found, are divided into
narrow costal and intercostal zones. This subdivision occurs in leaves of
Gramineae, but not in culms. The epidermis cells do not show the presence of
short cells as the Gramineae do. The stomata have dumb-bell shaped guard cells as
in the Gramineae.
Cyperus papyrus (papyrus/paper reed)

The cell walls, both in the costal as well as in the intercostal zone, are strongly
thickened and densely pitted. Their length (25-45 /mi x 7-10 pxri) is relatively
small in comparison with Cyperus rotundus and C. schimperianus. The stomata of
C. papyrus are arranged in single rows, the subsidiary cells show strongly
thickened outer walls, in contrast to C. rotundus. The stomata of C. papyrus
measure ca. 40/mi x 12-15 p m and are ca. three times as long as wide, which
gives a slender appearance (Plates A-13 and A-14).
3^s5^^^^Si^^©sto
Plate A-13 Culm epidermis of Cyperus Plate A-14 Stomata in culm epidermis
papyrus (LM, 250x) of Cyperus papyrus
(LM 550x)
Cyperus rotundus (nut grass)

The cell walls in the costal zone are elongated (90-175 p m x 10-12 /mi). Their
walls are thinner than in C papyrus (Plates A-15 and A-16). Cyperus
schimperianus has still longer and smaller cells. The stomata of Cyperus rotundus
culms measure ca. 40 x 20 /mi, they are plumper than in C. papyrus and shorter
than in C. alopecuroides (Plate A-17) and C. schimperianus.
Plate A-15 Culm epidermis of Cyperus Plate A-16 Culm epidermis of Cyperus
rotundus (LM, 250x) rotundus (SEM 400x)
Vi -fc-
Plate A-17 Culm epidermis of Cypertti- & * f f'"'

alopecuroides (LM, 500x) .:-»2«LF..»-.« T E ^ — » . . . a ; . • '—
Cyperus schimperianus.
The cell walls in the costal zone are thinner-walled than in C. papyrus and more
elongated than in C. rotundus. They measure ca. 45-100 /xm x 10-15 /mi. The
stomata are different from those in C. rotundus by the narrower subsidiary cells.
Differences with C. alopecuroides are discussed below.
Cyperus alopecuroides.
The epidermal cells measure ca. 60-90 /mi x 15-20 /xm. Their large width gives
the cells a brick-like appearance. The intercostal bands can only be recognized by
the presence of stomata, their epidermal cells are similar to those in the costal
bands (Plate A-17). The difference with C. schimperianus is slight, only the width
of the epidermal cells is somewhat larger.
A.3.3 Rushes (Juncaceae)

Culms of Juncaceae have cells which are more or less isodiametric and square to
hexangular. This type of epidermis cells does not occur in Gramineae or
Cyperaceae. The stomata have dump-bell shaped guard cells in contrast to the
Palmae discussed below.
Juncus acutus
The culm of this species cannot be divided clearly into costal and intercostal zones,
the epidermis cells in these zones show a comparable thickness. The cross section
of the culm is undulating owing to sunken intercostal zones (Plate A-18).
Plate A-18 Culm epidermis of Juncus Plate A-19 Culm epidermis of Juncus
acutus (LM, 250x) arabicus (LM, 250x)
Juncus arabicus.
The intercostal zone of the culm shows strongly sclerotized, thick-walled cells,
which is a conspicuous difference with the previous species (Plate A-19). The
costal zones has thinner walled cells, which resemble those of J. acutus.
A.3.4 Palm Trees (Palmae)

Leaves of the palm species found have stomata with bean-shaped guard cells,
which differ clearly from the dump-bell shaped guard cells in Gramineae,
Cyperaceae and Juncaceae. The abaxial and adaxial leaf epidermes are similar in
the species found.
Apart from leaves, leaf fibres of Phoenix dactylifera have been found. The
epidermis is no longer present, but the combination of spiral thickenings in the cell
walls and the presence of star-shaped crystals is very characteristic. The
macroscopic appearance of these fibres resembles the fibres on coconuts.
Hyphaene thebaica (Egyptian doam palm)

The best characteristic for the leaves of d o m palm are the large peltate hairs
present in the costal zones. Unfortunately, these hairs, appearing as yellow dots in
ancient material, are easily lost. Their absence is therefore not conclusive. The
stomata of Hyphaene thebaica are also larger (ca. 40 /im) than those of Phoenix
dactylifera, so this feature can be used to distinguish these species if peltate hairs
are absent (Plates A-20 and A-21). Furthermore, the epidermis cells are often
wider than long, especially in the intercostal zones. The cells in the costal zones in
Phoenix are considerably longer than wide.
^ f ^ ^ g p P j
Plate A-20 Leaf epidermis of Plate A-21 Leaf epidermis of

Hyphaene thebaica Hyphaene thebaica
(LM, 300x) (SEM, 130x)
Phoenix dactylifera (date palm)

The absence of peltate hairs and the relatively small stomata (ca. 25 /xm)
characterize the leaves of the date palm. T h e cells in the costal zone are longer
than wide. T h e intercostal zone in general shows two parallel rows of stomata,
which is often three or four in Hyphaene thebaica. T h e fibrous leaf sheaths are
characterized by rows of spiny crystals (Plate A-22)
<® * & + # & $ ,
Plate A-20 Fibrous leaf sheath with

crystals of Phoenix
dactylifera (LM, 600x)
APPENDIX B
QUESTIONNAIRE E T H N O - A R C H A E O L O G I C A L S U R V E Y
The focusing points listed in the tables below reflect the subjects that were covered
during the ethno-archaeological survey. These points of interest were partly
covered by observation, partly by interviews. T h e questionnaire used for these
interviews is given below, divided per subject ( A to M).1
A. Personal information
Focusing name, age and personal circumstances

points
Questions What is your name?

H o w old are you?
Are you married?
H o w old is your wife / husband?
D o you have children, how old are they?
What is the profession of your husband/wife?
What is the profession of your father?
What is the profession of your wife's / husband's father?
B. R a w material
Focusing 1. Vernacular name

points 2. Latin name
3. Origin and growth circumstances
4. Ownership
5. Management
6. Method of winning or harvesting
7. Period of winning or harvesting
8. Quantity of winning or harvesting
9. Method of processing and storage
10. Costs per year and per basket
The interview schedule has benefited greatly from the work done by dr. B. Driskell,
who made the notes of his ethno-archaeological survey available to me.
Questions What raw material do you use?

Can you write the name down in Arabic?
D o you buy or collect your raw materials?
Where do you buy your materials?
In what quantities do you buy your material?
H o w much does it cost you per basket?
When do you collect your raw materials?
H o w much time does that take?
D o you have to pay for collecting the material?
Do you have your own land? What do you grow there?

At what time of year do you harvest the raw material?
What do you do with the material after collecting?

After how much time can you use the material?
H o w long can you keep the material before using it?
Where and how do you store the material?
Do you sell the material to others?

D o they use it for basket making too?
At what price do you sell?
What is the best material for your baskets?

W h y is it the best?
Could you use other material if you would not have enough?
Could you use ...? (alternative or anciently used material)
W h y is it better than ...? (alternatively or anciently used material)
Is this best material available around here?
Is there other material which you sometimes use?
Specifically about palm trees:

What kind of material do you prefer, date or doam?
Could you use ...? (the opposite of the previous answer).
D o you own date palm trees?
D o you own doam palm trees?
Are there special trees for harvesting the raw material?
Are all trees suitable for harvesting raw material (male/female trees)?
Are all leaves suitable for basket making?
What do you make rope out off?
What colours do you use?

What are the names of the dyes?
Where do you buy them?
H o w much do you need for one basket?
What are the costs?
Appendix B, Questionnaire 443
C. Tools
Focusing 1. N a m e
points 2. Description
3. Number of each in use
4. Sizes
5. Material
6. Function (it has been designed to do what?)
7. Use (how is it used?)
8. Source or producer
9. Mobility (where does it come from)
10. Costs
Questions What are the names of the tools?

W h o made them?
What are they made of?
H o w much did they cost
What are they used for?

H o w many do you need?
D o you need different sizes?
How are they to be used?

Is this all you need to make baskets?
Did your father / mother use these (same) tools?
Did your father / mother use similar tools?
D. Work space
Focusing 1. Permanency of arrangement

points 2. Lay out and surroundings
3. Organization
4. Costs
Questions D o you always work here? / Where do you usually work?

Are other people working there too? What are they doing?
What other activities do you do here?
D o you always work in the same spot?
Would it be possible to work somewhere else?
D o you work in the sun or the shade? W h y ?
D o you have to pay rent for this space / D o you own this place?
E. organization of time
Focusing 1. During the day

points 2. During a larger time span (week, month)
3. During a year
Questions D o you make baskets every day?

D o you work continuously, the whole day long?
H o w long does it take you to make a basket?
H o w long does it take you to collect the raw material?

H o w long does it take you to prepare the raw material?
H o w long does it take before you can use the raw material?
H o w many baskets do you make each week?

D o you make a basket only when people ask you for one?
What time of year do you collect the raw material?

What time of year do you make baskets?
F. Production process
Focusing 1. Preparation of work space

points 2. Preparation of raw materials
3. Preparation of elements (string / plait)
4. Start of the work
5. Base
6. Transition base - side
7. Side
8. Rim, edge
9. Additional features (handles, lids, added decoration)
10. Insertion of new materials
11. Decoration
12. Determination of the size
13. Irregularities and mistakes
14. Timing of the entire process and the different parts
15. Analysis of actions and movements
16. Position during work
17. Working rhythm
18. Concentration on the work
19. Interaction with the surroundings
20. Final result
21. Repairs
22. Terminology
Appendix B, Questionnaire 445
Questions Can you teach m e how to make a basket?

Can you teach m e how to dye the raw material?
Do you need to prepare your material just before you start working?
How do you dye your material (exact procedure and times)

Did your father/mother do it exactly the same?
Are all the baskets (of this type) the same size?
D o you have a name to indicate the size?
H o w do you determine the size?
Why are you doing it (= specific technical detail) in this way?

Did your father / mother do it exactly the same?
Did you chose to make this decoration? Why?

Does this pattern have a name?
Does it have a specific meaning?
Did your father/mother use the same method of decoration?
Did your father/mother use the same pattern?
D o you decorate every basket in a different way?
H o w do you decide what you are going to make?
D o other basket makers use the same design?
D o you sometimes copy designs of other basket makers?
D o you use designs from other objects, such as pots?
Are there designs that are only used by old women? W h y (not)?
D o young people make these designs? W h y (not)?
Will you make this design when you are older?
or: Did you make this design when you were younger?
Do you repair baskets as part of your work?
What do you call:

the centre, base, side, rim, handles, action of making, etc.
G. Gender aspects
Focusing 1. Gender of basket maker

points 2. Gender of those who use the products
3. Gender of producers and users of other commodities
Questions D o m e n make (these) baskets? D o w o m e n make (these) baskets?

Does your wife/husband help you sometimes?
D o other people help you making your baskets?
For w h o m are these baskets made?

.
H. Economic aspects
Focusing 1. Rate of professionality

point Full-time professional: dependent on income
Part-time professional: partly dependent on income
Semi-professional: occasional additional income
Non-professional: no income, for o w n use
2. Possession of raw materials
3. Costs and yield
4. Distribution (house, village, region, country, export)
5. Use of non-local baskets
6. Replacement of basketry containers by other materials
7. Replacement of mats by other materials
Questions What does it cost you to buy raw materials?

What does it cost you to buy tools?
D o you sell the baskets?

H o w much do you get for a basket? Is that enough?
Are you doing other work besides basket making?
Would it be possible financially if you would stop making baskets?
For w h o m are you making your baskets?

D o you give baskets away? For free, in exchange or as present?
Are you working alone? Are other people helping you?
H o w many rooms has your house?

D o you o w n any animals? D o you o w n land?
D o you have a refrigerator, television, fan?.
•
I. Function
Focusing 1. N a m e of basket
points 2. Description
3. Function, use and re-use
4. Comparison with other containers, floor covers or screens
6. Function within groups of society
Questions What is the name of this basket?

What is this basket used for?
D o you ever use it for something else?
D o other people use it for something else?
D o you use the baskets (also) to measure quantities?

If not: what do you use for measuring
If yes: specify measures
Appendix B. Questionnaire 447
J. Meaning of baskets in social contacts
Focusing 1. Baskets for special occasions

point 2. Baskets for special groups
3. Limits or taboos of certain shapes, colours or materials
4. Meaning
Questions D o you give baskets as presents? When?

D o other people do so?
Did the people in the old days give baskets as a present?
Is the present formed by the basket itself or by its contents.

Are they new baskets or old baskets?
D o the basket have to be returned?
At what occasion do you give baskets (specify, e.g. childbirth, marriage,

death).
Are those baskets decorated?
Is it a special decoration for the occasion?
Does the decoration have a certain meaning?
Do you have baskets which belonged to your mother/grand-mother?

Are old baskets thrown away or used for something else?
For w h o m do you make baskets?
K. Traditions and enculturation
Focusing 1. Passing on knowledge

points 2. Assessment of baskets
3. Assessment of basket makers' profession
Questions W h e n did you make this basket?

D o you repair a basket which is broken?
D o you replace a broken basket by a new one?
D o you keep old baskets?
D o you still use baskets that have belonged to your (grand)parents?
Which baskets are best, old baskets or new baskets? W h y ?
H o w long can you use this basket?
Who taught you to make baskets?

Are you teaching your skill to somebody at the moment?
At what age have you learned to become a basket maker?
What age was your pupil when he/she learned how to make baskets?
D o you like making baskets?
Who makes the best baskets around here?

W h y are these good baskets?
Where are the best baskets being made?
._
L. Social position of basket makers
Focusing 1. Assessment by basket makers
points 2. Assessment by others
Questions What was your fathers job?

Did your father/mother make baskets too?
Would you like your son/daughter to become a basket maker?

or: are you proud that your son/daughter is a basket maker too?
What kind of job would your son like to do?

D o you want to continue making baskets all your life? ;
Would you like to do another job?
M . General background information
Focusing 1. Geographical position of the place of residence

points 2. Character of the place of residence (village, town)
3. Number of inhabitants
4. Local specializations
Questions H o w many inhabitants has the village?

H o w many houses are there in the village?
H o w many people are making baskets here?
APPENDIX C
VIDEO C O N C O R D A N C E
This appendix helps to connect the video and the book. While reading the book,
the connection is m a d e by the video times quoted. Throughout the book references
are m a d e to the timer in the right hand corner of the video image. This timer
indicates minutes, seconds and frames (mm:ss:ff). There are 25 video frames in
one second.
While watching the video, corresponding pages in the book, quoting a
particular scene, can be found through the index of video times (Table C-7). T h e
other images inserted in the video (symbols in the right upper corner, examples of
the analysis of action in the left lower corner and frozen images with an indication
of the passive and active systems) have been listed here, as well as the use of tools
and sequences in which the instrumental use of the body is clearly visible. For an
introduction to these aspects, see Chapter 6 (pp. 101-108).
Table C-l main sequences

Table C-2 occurrence of symbols (in the top left corner)
Table C-3 frozen moments: indication of active and passive systems
Table C-4 examples of an analysis of actions (in the bottom left corner)
Table C-5 use of tools
Table C-6 instrumental use of the body
Table C-7 index of video times (reference to pages in the book)
Table C-l main sequences
al-maqtaf plaited carrier basket 02:53-24:04
al-sabat coiled basket 24:41-32:58
al-shauwer coiled basket 33:34-41:33
al-masballa twined sacking 42:10-47:34
al-serir bil habl woven bed matting 48:05-58:09

Table C-2 Occurrence of symbols in the top left corner
'palm tree' symbol
properties of the raw material

A .
06:46 hardness of the veins of date palm leaflets
08:59 suppleness and strength of a plait m a d e out of date palm leaf strands
16:01 toughness of the leaf sheath fibre of the date palm
27:27 tension strength of soaked date palm leaf strands of 6 m m wide
36:49 tension strength of soaked d o a m palm leaf strands of 2 m m wide
42:29 strength and roughness of date palm fibre string
48:20 dry grass, in contrast to wetted grass
'sitting man' symbol
& working position
03:54 sitting cross legged

12:30 sitting cross legged with left foot inside the basket
19:48 sitting with right leg stretched, left leg bent
31:47 sitting with right leg bent, left leg stretched in front
43:25 sitting, knees pulled up against the chest
53:22 standing upright with one foot on the bed, while beating the weft
55:32 standing bent over the entire width of the bed while inserting the weft
'eye' symbol
S * eye contact with the work
05:13 paying attention to plaiting, looking up while putting in a n e w leaflet

12:02 paying attention to needle in edge, looking up while pulling string
25:43 paying attention making knot, looking away while tightening knot
45:09 looking up while inserting a new twining string
'foot' symbol
L L instrumental use of the body
07:22 holding start of palm leaf string with toes

12:20 holding the basket with one foot inside
20:12 holding palm fibre rope with toes
25:00 shredding palm leaf with teeth
49:15 holding string by putting the foot on top
Appendix C, Video Concordance 451
'hammer' symbol
use of tools
09:20 needle: stitching into sides of the plait (holding needle in the middle)
17:54 needle: stitching into sides of the plait (holding needle in the middle)
21:04 needle: stitching through basket wall (holding needle in the middle)
22:41 knife: cutting off string
23:02 needle: making running stitches (pushing needle from the back)
26:58 needle: stitching through previous bundle
29:37 needle/awl: piercing the previous bundle, to prepare hole for needle
37:33 awl: piercing the previous bundle, to prepare hole for palm leaf tip
56:36 wooden peg, used to secure the edge
56:47 wooden peg, used to beat the weft
57:17 wooden peg, used as an awl, to make space in the weft
Table C-3 frozen m o m e n t s : indication of active and passive systems
al-maqtaf
04:39 plaiting active / active

07:39 making string active / active
12:16 sewing the plait active / passive
16:14 making string active / passive
as-sabat
29:30 wrapping . stitching into previous bundle active / passive
as-shauwer
39:37 wrapping, stitching into previous bundle, active/(active)/passive

while changing colours
al-masballa
44:28 twining, the warp is passive, the active system active / passive
consists of two active strands.
al-seru bil habl
55:23 weaving, the w a r p is passive, the double active / passive

weft strand is active
Table C-4 Examples of an analysis of actions within the production process
o = occurring action
r = recurring or repetitive action
05:47-06:15 rolling up the finished plait

rl rolling up the plait
o2 picking off bits of palm leaf
08:05-10:50 start of sewing the plait with string
o3 threading the needle
o4 letting drip the plait
o5 arranging the plait
06 picking up the plait
o7 making the plait more supple
08 pushing the needle along and into the side of the plait and
pulling through the string
o9 opening up the loop at the end of the string and pulling the
string through the loop
olO moulding the centre of the basket
rl 1 picking up the plait edge with the needle and pulling the
string
29:40-31:37 wrapping and sewing the coils
rl2 piercing with the awl, positioning the strand, pushing the
needle through the hole and tightening the strand
ol3 inlay of new date palm leaflet
40:19-41:12 wrapping and sewing the coils
ol4 shredding the fibre of the fruit stem of the date palm
ol5 wetting the foot of the fibre bundle
ol6 insertion of new bundle material
43:52-46:14 twining
rl7 twining
ol8 insertion of a new length of string
ol9 making the side edge
Appendix c. Video Concordance 453
53:09-56:21 weaving with two parallel strands

o20 making the edge, securing it with a peg
o21 beating the weft
r22 inserting the weft
o23 change of orientation in the pattern
o24 measuring the pattern
o25 pulling tight the weft
rl . . 335, 336, 341, 342 olO . . . 337, 338, 344, ol9 372
o2 . 335, 336. 341, 342 345, 355. 356 o20 379, 385-387
o3 337, 343 rll 337-339, 345 o21 380, 385-387
o4 337, 343 rl2 105, 355, 356 r22 . 377, 379, 385-387
o5 337. 344 013 ... . 105. 355. 356 o24 387
06 337, 344 ol5 364 o25 . . . . 379, 386, 387
o7 337, 344 ol6 364
08 337, 344 rl7 372
o9 .... 337, 338, 344 ol8 372
Table C-5 Use of tools
tool action video sequence remarks
large threading 08:04-08:17 ready made string

needle 11:07-11:41 ready made string
(mesalla) 13:26-13:31 string made on the spot
18:20-18:54 ready made string
running 09:08-09:20 through edge of plait

09:33-09:35 through loop of string
18:56-18:58 through loop of string
20:57-21:01 through loop of handle
22:15-22:39 through plaited fabric
22:49-23:44 through plaited fabric
sewing 09:48-10:51 sewing with kneading

10:58-11:07 basket
11:41-12:47 basket
13:37-13:53 basket
16:32-16:54 basket
17:27-18:19 basket
19:04-19:40 rim
piercing 20:46-20:50 through basket wall

21:01-22:12 through basket wall
small threading 25:59-26:01

needle 29:08-29:13
(ibra) 30:42-13:48
piercing 26:12-27:46 without awl, through bundle

28:35-29:03 without awl, through bundle
29:13-30:24 with awl, through bundle
awl piercing 29:13-30:24 needle used as awl

30:56-31:37 needle used as awl
31:55-32:55 needle used as awl
33:52-33:53 part of sewing machine
36:13-41:12 awl with wooden handle
splitting 40:52 awl with wooden handle
knife cutting 22:30-22:44 cutting string

23:47-23:48 cutting string
trimming 23:48-24:00 removing irregularities
peg securing 50:06-52:42 peg to prevent the loops at the

54:13-54:19 edge from becoming undone
55:47-56:02
56:35-56:41
spacing 57:09-57:11 using the peg as awl

57:17-57:23 using the peg as awl
tightening 56:30-57:00 peg beaten with stick
stick beating 53:11-53:16

53:21-53:50
56:02-56:21
56:30-57:00 beating against peg
Appendix C, Video Concordance 455
Table C-6 Instrumental use of the body
tool action video remarks
foot holding with toes 07:22-07:58 string held between 1st and 2nd toe of
12:52-13:32 right foot, sitting cross legged.
14:37-15:48 string held between 1st and 2nd toe of bent

right foot, left foot stretched in front.
15:49-16:27 string held between two feet, right bent,

left stretched in front, between 1st and 2nd
toe.
19:47-20:22 string held between 1st and 2nd toe of

right foot, stretched in front (making
handle).
holding with foot 12:00-12:47 holding basket, no shoes

13:33-14:00
16:32-16:54
17:27-19:40
17:07-17:23 holding plait under left foot, wearing shoes
49:15 holding string under left foot, barefooted
mouth wetting 18:45 wetting end of string (threading needle)
biting off 18:28 biting fibre from string (?)
shredding 24:46-25:05 shredding date palm leaf strands
splitting 27:59-28:03 splitting fruit stem fibres of the date palm

28:14-28:18
nails cutting 35:59-36:01 cutting doam palm leaf strand in two
splitting 40:36 splitting fruit stem fibres of the date palm

INDEX VIDEO C O U N T E R
02:53 269, 331 09:49-13:02 334

02:53-24:04 269, 331 09:49-18:20 332
03:32-03:46 331 10:58 339, 340, 345
03:34-03:47 347 10:59-11:50 347
03:35 296 11:03 339,340
03:38 341 11:07 332,339
03:47 296. 332, 347 11:26-11:28 348
03:47-04:10 332 11:42 339
03:47-04:20 296 11:46 340
03:47-04:55 347 11:50 339,347
03:54 341 11:51 271
04:11 332, 334, 341 11:53 340
04:11-06:35 332,334 11:58 340
04:12 335 12:00 345
04:16:28 296 12:02 339, 345
04:26 335 12:03 340
04:47 335 12:07 340
05:10 341 12:08 345
05:13 341 12:11 340
05:25 341 12:15 340
05:29 341 12:20 345
05:33 335 12:22 340
05:34 342 12:26 340
05:47-06:15 335, 336 12:29 340
06:07 342 12:30 345
06:19 296,342,369 12:32 340
06:36-06:56 332 12:35 340
06:36-07:58 336 12:38 340
06:45-06:53 347 12:42 340
06:46 296, 336 12:44 339
06:46-06:56 336 12:47 296, 332, 340
06:49 300 12:52 345
06:53 342, 347 13:00 346
06:57-07:24 332 13:02 332, 334, 336
06:57-07:58 332, 347 13:27 346
07:00 342 13:33 339
07:06-07:58 300 13:36 340
07:07 342 13:41 340
07:15-07:58 332. 366, 374 13:43 339
07:18 342 13:45 340
07:22 342 13:50 340
07:39 271, 332 13:54 340, 393
08:05-10:50 332, 337 13:54-14:01 393
09:20 302, 344 14:00 340
09:48-10:51 347 14:25 332, 346
Index Video Counter 457
14:25-15:04 332, 346 20:41 303, 347

14:37-15:04 347 20:41-22:12 347
15:05 271. 333, 341, 346, 347 20:41-23:50 303
15:05-15:34 341 21:04 303,348
15:05-16:28 333. 347 21:18 348
15:18 346 22:15-22:40 347
15:47 346 22:49 348
15:49 300 23:48-24:00 347
15:49-16:27 300 24:04 269, 331. 333
16:01 296 24:41-25:07 350
16:02 346. 347, 369 24:41-32:58 269, 350
16:02-16:08 347 24:45-24:52 351
16:14 271, 311, 333 24:54 353
16:32 340, 345, 347 24:57 353
16:32-16:54 347 24:59-25:05 351
16:33 302, 333, 334 25:00 353
16:33-16:55 333, 334 25:07-25:11 353
16:35 339, 340 25:07-28:28 350
16:40 340 25:08-25:45 270, 306
16:45 340 25:11-25:32 353
16:49-17:05 347 25:22-25:32 353
16:49-17:23 333 25:26 353
16:50 340 25:32-25:45 353
16:56-17:06 346 25:43 353
16:57 332 25:46-25:49 353
17:06-17:22 347 25:58-26:02 354
17:07-17:24 333 26:09-26:21 354
17:27-17:54 347 26:10 270, 368
17:54 302, 346, 347 26:24 354
17:54-18:18 346, 347 26:46 354
17:59-18:20 333, 334 26:58 306, 354
18:20 332-334, 346 27:00 354
18:20-24:04 333 27:27 303, 354
18:28 346 27:36 354
18:45 348 27:39 354
18:50 348 27:54-27:58 350
19:02 340 28:03-28:06 354
19:15 340 28:15-28:25 270
19:22 340 28:23-28:28 354
19:25 348 28:35-31:37 350
19:27 340 28:40 353
19:32 340 29:20 355
19:47-20:31 333 29:22 355
19:48 347, 348 29:25 355
19:48-20:32 347 29:30 350
19:55-20:13 300 29:37 271, 306, 355
20:12 348 29:40-30:04 356
20:12-20:23 348 30:04-30:05 356
20:13-20:22 300 30:05-30:06 356
30:06-30:24 356 38:10-38:34 364

30:20-30:24 356 38:11 364
30:24-30:27 356 38:34-38:42 364
30:27-30:29 356 38:37-38:41 358
30:29-30:32 356 38:42-38:46 364
30:33-30:55 356 38:46-39:03 364
30:51-31:11 307 38:48-38:50 307
30:55 354, 356 39:03-39:34 364
30:57-31:05 356 39:34-39:43 270, 307
31:05-31:08 356 39:35-39:44 364
31:08-31:37 356 39:45-40:14 364
31:42 350, 356 40:19-40:35 364
31:42-32:53 350 40:25 362
31:47 356 40:32 362
31:52 357 40:35-40:36 364
32:39 357 40:36-40:54 358
32:54 357 40:42 364
33:34-33:39 358 40:43 364
33:34-41:33 269, 358 40:49 364
33:35-33:55 393 40:52 364
33:39-34:10 358 40:54-40:57 364
33:52 363 40:55 364
34:05-35:37 362 40:57-41:04 364
34:09-34:10 290, 358 41:04-41:07 364
34:10-36:12 358 41:08-41:12 364
34:11-34:28 362 41:10 362
34:11-35:46 306, 307 42:10-46:14 366
34:28-34:50 362 42:10-47:34 76,269,366
34:50-35:04 362 42:14-42:20 372
35:47-36:12 358 42:20-42:22 372
35:50 362 42:22 320, 372
35:52 363 42:23:21-42:26:10 368
35:53 362 42:23-42:26 372
35:58-36:00 270 42:24 270, 368
35:59-36:01 358 42:26:10-42:28:08 368
36:06 270 42:28:08-42:30:02 368
36:07 363 42:30:02-42:31:13 368
36:12-36:28 363 42:31:13-42:33:00 368
36:13-41:12 358 42:33:00-42:34:14 368
36:28-37:04 363 42:34:14-42:36:16 368
36:49 303, 363 42:36:16-42:37:22 368
37:04-37:06 363 42:37:22-42:39:08 368
37:09-37:21 363 42:39:08-42:41:00 368
37:21 358, 363 42:41:00-42:42:24 368
37:21-37:23 363 42:42:24-42:44:05 368
37:26-37:31 363 42:44:05-42:50:03 368
37:33 271,306, 363 42:50:03-42:51:22 368
37:45-37:56 363 42:51:22-42:53:08 368
37:56-38:07 363 42:53:08-42:54:14 368
Index Video C o u n t e r 459
42:54:14-42:55:18 368 48:20-48:34 381

42:55:18-42:57:05 368 48:20-48:53 375
42:57:05-42:58:14 368 48:34-48:56 381
42:58:14-42:59:19 368 48:47-48:53 318, 375
42:59:19-43:02:00 368 48:56-49:13 382
43:25 372 49:13-49:57 382
44:04 368, 372 49:28-49:13 375
44:04:07-44:06:17 368 49:58-50:11 376, 382
44:06:17-44:08:13 368 49:58-50:42 376
44:08:13-44:13:12 368 49:58-50:58 375
44:13:12-44:16:00 368 50:11-50:28 376, 382
44:16:00-44:22:17 368 50:28-50:42 376, 382
44:20 375 50:34-50:42 270
44:22:17-44:28:10 368 50:51-50:54 376, 382
44:28:10-44:32:00 369 50:55-51:31 270
44:39:19-44:41:12 369 50:56-51:32 383
44:41:12-44:45:01 369 50:56-52:09 378
44:45:01-44:48:04 369 50:58-57:30 375
44:48:04-44:54:19 369 51:00-51:04 382
44:54:19-44:56:15 369 51:06 383
44:56:15-45:01:10 369 51:11 383
45:09 372 51:39 318
45:12:10-45:16:02 369 51:40 383
45:16:02-45:19:24 369 51:43-51:54 383
45:19:24-45:23:17 369 51:48 383
45:23:17-45:27:24 369 51:54-52:00 384
45:27:24-45:31:10 369 52:00-52:09 385
45:31 369, 372 52:09-53:24 385
45:36:17-45:39:03 369 52:10-52:24 386
45:39:03-45:41:15 369 52:24-52:43 386
45:41:15-45:44:19 369 52:46-53:08 385
45:44:19-45:48:04 369 53:09-53:11 385
45:48:04-45:50:10 369 53:11-53:16 385
45:50:10-45:53:05 369 53:16-53:21 385
45:53:05-45:56:06 369 53:21-53:50 385, 386
45:56:06-45:58:17 369 53:22 385
45:58:17-46:06:19 369 53:50-53:56 385
45:58-46:06 372 53:50-54:04 378
46:06:19-46:09:24 369 53:50-54:20 377
46:09:24-46:12:24 369 53:57-54:19 385
46:12:24-46:14:17 369 54:04-54:20 378
46:15-46:26 373 54:11-54:19 270, 318
46:15-47:19 366 54:19-54:37 385
46:27-47:21 373 54:27-55:02 382
46:28 270 55:05-55:14 387
46:33-47:20 320 55:14-55:15 387
47:19-47:34 366 55:15-55:39 387
48:05-58:09 269, 374 55:18-55:39 386
48:20 375, 381 55:32 385
55:39-55:51 387
55:52-55:56 386, 387
55:56-56:02 386
56:02-56:21 387
56:03-56:16 386
56:21-56:28 385
56:36 387
56:43-57:00 386
56:47 387
57:06-57:15 382
57:06-57:36 378
57:08-57:26 387
57:17 271, 318, 387
57:26-57:56 388
57:31-57:59 375
58:03 388
APPENDIX D
GLOSSARY
T h e glossary consists of three different parts:

D.l English Glossary
D.2 Arabic and Nubian terminology
D.3 S o m e Ancient Egyptian Basketry T e r m s
D.l English Glossary

In the descriptions the words highlighted in italics are also incorporated in the glossary.
References to the text can be found in the Index.
active system The strands in a basketry technique which cause the coherency in a
technique. A basket maker handles the active strands more frequently than
those that make up the passive system. Usually the active system is flexible.
while the passive system is rigid, or fixed in a frame.
anchoring fabric
The previous rows of stitches onto which each subsequent row of a basketry
fabric is added.
aspect A n aspect is a variable property of an object, such as its length, diameter,
colour, raw material, or decoration. The basic structure of a basketry
technique is also expressed as a combination of aspects, such as number of
systems, number of elements, orientation and activity. The variations which
occur in a certain aspect, are indicated as attributes.
attributes The variaties of an aspect: 10, 20, 30 ares attributes of the aspect 'length':
the aspect material has, for instance, the attributes grass, palm leaf, sedges
and rushes. Brown, green, yellow, are attributes of the aspect 'colour'.
basic structure The composition of a basketry technique as found in the fabric of the base or
sides of a basket, as well as in the middle of objects with a linear form
(brushes and rings). The basic structure is described by the interaction of the
systems (an abstracted grouping of the basketry strands).
basketry Artefacts in which vegetable materials with a limited length or a plant-
specific shape have been incorporated. The strands or systems of strands
interact in order to form a fabric or linear form. Basketry comprises baskets,
bags, mats, brushes, hurdles, screens, furniture, pot stands, head rings, hats,
sandals and belts.
border A n edge of a mat or basket, at which the strands are fastened off.
bundle The passive element in coiled basketry, also indicated as 'foundation' (see:
winder, wrapping strand, coiling).
cataract Rapids in the river Nile caused by blokkades where granite basement rock,
rather than Nubian sandstone lies at the surface.
coiling A technique which involves two systems: a passive bundle which is fastened
in a coil to form a fabric. The winder wraps around the bundle and stitches
through the previous row of coiling (anchoring fabric).
date palm fibre
Leaf sheath fibre of the date palm tree
date palm mid rib
Petiole of the date palm, or date palm branche
direction Specific term for the orientation of one system
edge The edge of a basket or mat can either be a border (the line at which strands
are fastened off) or a selvedge (a line at which the strands continue, but turn
back into the fabric). The edge of a basket is usually called rim.
element Basketry systems are made up of one ongoing element or of several parallel
elements. Elements m a y consist of one, two or three members.
english randing
A stake-and-strand technique of which the basic structure is a form of
weaving with one strand at a time (cf.french randing).
essence The indivisible combination of measurable and immeasurable aspects (of
basketry)
fabric Surface or three-dimensional object produced by the interaction of strands
producing. A fabric can be made in a continuous spiralling movement, which
follows one direction, or a fabric can be worked alternately from left to right
and from right turning at the edges.
fingers Divisions of a brush.
form Term to describe the result of a basketry technique, which can be linear or
fabric A head ring, or brush is a linear form, a basket or mat is a fabric
(classification criterion in Chapter 9).
frame 1. an external structure which is used to fix the passive strands of a basketry
technique. By holding the strands under tension the passive system can be
made rigid even though it consists of flexible materials. Examples of frames
are a loom for weaving and twining and a bed frame for furniture webbing.
2. the smallest unit in the video timer. There are 25 frames in one second.
framing (cadrage) The cut-out of the image while filming (general view, total, half-
total, close-up).
trench randing A stake-and-strand technique which is a from of weaving, in which the
active strands are all layed in and worked up at the same time (cf. english
randing).
granny knot A knot similar to the reef knot, but made of two half knots in the same
orientation (SS or Z Z ) . The granny knot is much less secure than the reef
knot and is in fact a wrong application of the latter.
grommet A rope ring, made by twisting a strand a number of times around itself.
Small grommets are often used for strengthening an opening. Large
grommets are used for padding, as head rings or supports for pots.
half knot A simple knot in two strings (or in the two ends of one string). String A runs
over and under strand B. This can be done in S or Z direction.
knotless netting
A one-system basketry technique in which a row of loops is fastened to a
previous row (anchoring fabric).
Appendix D, Glossary 463
knotting A connection of strands which involves more than one crossing. In knotting
a strand moves in three planes.
lazy basket makers' stitch
A fast variation of coiling: a combination of winding a strand around a
bundle a number of times and then making a stitch.
leaflets Small side leaves of the large feather-shaped date palm leaf (which consists
of a stalk or midrib with small leaves branching off at the sides).
netting General term for all widely spaced techniques, usually made of yarn or
string. Netting techniques can be one-system techniques (knotted netting with
a netting needle, knotless netting made of palm leaf) or two-systems
techniques (netting made of overhand knots, twined netting).
member Part making up the active element of a basketry technique: a basket is made
of interacting systems. The systems consist of elements (either one long
strand or a number of parallel strands). The active elements consist of one,
two or three members. The members of an active element move in certain
way, thereby creating the fabric. The mobility of the members differs and of
the members is said that they are moving in one, two or three planes.
multi-system technique
A technique made up out of more than two systems
N e w Nubia (Egyptian Nubia) The area in Egypt around the city of K o m O m b o where the
Nubians have been resettled.
Old Nubia Nubia, before 1960 (the year of the resettlement following the building of
the A s w a n High D a m ) .
one plane Indication of the mobility of the active system. Mobility in one plane is a
simple up-and- down movement as occurs, for instance, in sewing, weaving
and plaiting.
one-system technique
A technique made with one ongoing strand. A one-system technique can be
made of many short strands (e.g. strips of palm leaf of 500 m m long each)
which are inserted after the previous one is exhausted. Knotless netting,
knotted netting and grommets are examples of one-system techniques.
Textile techniques such as knitting and crocheting are also one-system
techniques, but with a long continuous yarn, rather than strands of limited
length.
orientation The predominant direction that the strands of a basketry system follow.
overhand knot Overhand knots are simple knots made by twisting a strand over and under
itself. This can be done in two orientations: S and Z. W h e n an overhand
knot is made in two parallel strands, it is called overhand bend. This knot
was used in Amarna to make a carrier net.
palm fibre Leaf sheath fibre of the date palm tree.
passive system The strands in a basketry technique which form the body of that technique.
but do not cause the coherency. A basket maker usually handles the passive
system(s) less frequently than the active system. Often the strands of the
passive system are less flexible than those of the active system, or made
rigid artificially by fixing them in a frame (such as a loom).
plait Three or more strands interlaced. The variety of plaits includes round plaits
(plaiting around a core), plaited strips and continuous plaiting. A feature
typical for plaits is that all strands are active.
plane Indication of the mobility of the active system. An active basketry strand
moves up and down, weaving through the strands of the passive system; or a
strand can m o v e spirally around passive strands; or the active strands makes
knots around the passive strands (see: one plane, two planes, three planes).
plying twisting two or more yarns around each other to form a string. Plying can be
done in S or Z direction.
Qasr Ibrim plaiting rule
Plaited strips of which the edges are orientated perpendicular to each other
(SZ) can be used for making fabrics, while strips with parallel edges (SS)
cannot. From this rule it follows that:
Plaits made in a \1/1\\1 pattern can be sewn into a fabric when made with an
odd number of strands, but plaited strips made of an even number of strands
cannot.
Plaits made in a \2/2\\l pattern can be sewn into a fabric when made with 9,
13, 17, 21, 25 etc. strands.
reef knot A knot made of two half knots in opposite orientation. The reef knot is the
symmetrical variety (SZ or Z S oriented) of the granny knot (SS or Z Z
oriented).
rim The edge or border of a basket
S Clockwise twist: indication of the orientation of spinning, plying, twining or
waling.
strand 1. Non-technical term for basketry materials. A strand can be a strip of palm
leaf, a string or a willow rod.
2. The active element in a stake-and-strand basket
selvedge Edge of a mat or basket, where the strands are not fastened off, but turn
back into the fabric.
set of elements
A number of strands of the same material, with the same orientation,
performing the same task in the technique.
stake The passive element in a stake-and-strand basket.
stake-and-strand basketry
A collective term for techniques in which the rigid passive elements (stakes)
are fastened at a straigth angle by more flexible active elements (strands).
The techniques used are weaving (e.g. trench or english randing), twining or
waling.
system Basketry strands forming a basic structure. All strands that are made of the
same material, follow the same orientation and have the same function in a
technique are considered to belong to the same system. The basic structure
of a basketry technique is described as systems that interact (Chapter 5)
tabby A simple plaiting or weaving pattern of which the shift is the same as the
number of strands that are crossed (e.g M / 1 W 1 or \2/2\\2). Also known as
plain weave. See also: twill
textiles Fabrics which consist of raw material or half products of unlimited length
and uniform in shape.
three planes Indication of the mobility of the active system. Mobility in three planes is a
movement that goes up-and-down, left-right and front-back. The result is a
movement which crosses its o w n path, such as eight-shapes or knots.
Appendix D. Glossary 465
three-systems techniques
Basketry techniques made with three systems. These systems can all have
different orientations (this is well known, for instance from fine chair
matting, which in origin is an Asian technique). In Egypt, however, only
three-systems techniques in two directions occur. T w o systems are oriented
in the same direction, while the third one is inserted at a straight angle.
twill A plaiting or weaving pattern which is staggered and looks like a herring
bone: the shift is smaller than the number of plaiting strands crossed (e.g.
Y2/2W1, \3/3\\l, \6/6\\2). See also: tabby or plain weave.
twining A stake-and-strand technique which consists of a set of passive strands which
are held in place by two strands which are held in place by rows of active
strands. The active system consists of two strands which are twisted
alternately behind and in front of the passive strands. Twining can be done in
S or Z orientation.
two planes Indication of the mobility of the active system. Mobility in two planes is a
movement that goes up-and-down as well as left-right. The result is a
circular or spiral movement as occurs, for instance, in wrapping, coiling,
twining, waling and looped matting.
two-systems techniques
Basketry techniques made with two system. These two systems can follow
the same direction (as in coiling) or are inserted at a straight angle (as in
weaving).
unit A n archaeological feature such as a layer or deposit (used in Amarna and
Qasr Ibrim)
waling A stake-and-strand technique, similar to twining but made with more than
two active members which twist around passive strands. The pattern of
waling depends on the number of strands involved. Three strands move in
front of two passive strands, behind one passive strand. Four waling strands
m o v e in front of three, behind one. Waling can be done in S or Z
orientation.
wickerwork A stake-and-strand baket made out of willow rods. The term is often used
loosely as an alternative for all stake-and-strand basketry.
winder The active element in coiled basketry (cf. bundle, wrapping strand).
wrapping strand
(wrapping system) The active system in coiled basketry or in wrapping
brushes or rings (synonym of winder, or winding system)
wrapping Production method for brushes or pot stand rings: a (passive) bundle of
material is wrapped with an (active) wrapping strand (synonymdinear
wrapping).
Z Anticlockwise twist: indication of the orientation of spinning, plying,
twining or waling.
D.2 Arabic and Nubian terminology

The transcription of the Arabic words does not follow the Arabic spelling
consistently, but tries to convey the middle Egyptian pronunciation. Sometimes the
qof is transcribed as "g", the "z" as "s" and the gim as "j".
There are separate w o r d lists for basketry types and basket makers'
terminology in both southern middle Egyptian and Nubian. In the Egyptian list, the
words are in order of the English alphabet, in which the 'ayin and the h a m z a are
ignored. W h e r e possible, links have been m a d e with Badawi's dictionary of
Egyptian Arabic, and the work of Woidich o n Egyptian dialects. T h e Nubian lists
have separate columns for Kensi/Matoki and Fadidja Nubian. Therefore, there is
no consistent alphabetic order. References are m a d e to two dictionaries. B H stands
for Badawi and Hinds 1986; B W refers to Behnstedt and Woidich 1994.
References are also m a d e to Henein 1988.
T h e list given here is related to the terms encountered during field work. It
has no pretention to give an overview of the basketry of present day Egypt which
shows a m u c h greater variety than what is presented here. T h e Nubian basketry
terms are not consistent. Variations and discrepancies occur between Fadidja and
Kensi villages, but also between one Kensi village and the next or even within the
same village. In future I hope to m a p these variations in more detail.
(Southern middle) Egyptian terms for basketry types:

'aldga 1 sewn plaits carrier basket made of a plait with a length of 7 ba'at. (cf.
Henein 183 "8-9 ba'at", B W 321)
2 also used as general word for "carrier basket", referring for instance to
the glossy, light yellow stake and strand basket, made of split Arundo
donax culms (see sella).
'angareeb wooden bed with string webbing, mostly used as couch ( B W 327)
bagula plur. bagutdt. small sewn plaits basket, made from a plait with a length
of 3 or 4 ba 'at
(cf. B H 91, B W 30: baquti, in Cairo dialect this is a small basket to put
hand picked vegetables in while shopping. The baguta or gauta is shown
in B H , plate B5, as a ghalag "small palm basket with handles" (BH
628). B W 18-19 mention "ba'uuti" (Delta) for a large basket.)
barda'a 1 twined donkey saddle ( B W 17)
2 wooden donkey saddle
bursh plur. burush rectangular mat made of plaits sewn parallel to each other,
or a round mat which is sewn in a spiral (cf. W B 18-19; B H 65).
afrdd plural of fard. In Egyptian Nubia this arabic term is used for two
baskets on the back of a donkey (cf. B H 647, "large guffa"; B W 348, in
upper Egypt: carrier bag for a camel).
gauta see baguta.
Appendix D, Glossary 467
gotwiyya sewn plaits transport basket to be carried on both sides of a donkey or

camel. The two baskets are either sewn to a mat (Henein 185) or
connected with a wooden cross beam ( B W 387).
guffa 1 plur. guffaf, large sewn plaits basket made of a plait with a length of 15
ba'at, upper Egyptian term. (BH 711: "largeflexibletwo-handled
basket of palm stalks, similar to maqtaf but larger"; B W 389: largest
carrier bag).
2 round basket with straight sides and two handles, made of old car tyres,
in general smaller than the guffa made of palm leaf (cf. zanbiit).
hasir woven matting from grass or juncus/sedges ( B W 86; B H 208)
maqtaf plur. maqdtif large sewn plaits basket, with two rope handles, made of
a plait with a length of 10 ba'a (cf. Henein 183: 5 ba'at). B H 709, plate
B, 7; B W 386, in north middle Egypt: middle size carrier bag; in south
middle Egypt: basket, smaller than 'aldga, in upper Egypt: large basket
on donkey back, like the mazbala).
masbala large twined sack made of two mats sewn together, for the transport of
earth and dung on donkey back, cf. mazbala
matraha plur. matdrih, a bat-shaped implement to shake a ball of dough into a
large flat bread and throw it in the oven. The matraha comes in two
sizes: length 75 cm, diameter 55 c m for the large unlevened bread types
(e.g. rugaag and 'esh durra) and a small one used for levened bread
(esh shamsi), with a length of 50 c m and a diameter of 30 cm. B H 535,
plate B4, B W 282).
mazbala B W 181: one large bag, or two baskets on both sides of a donkey, for
transport of earth and dung, cf. masbala
meshanna plur. meshanndt, shallow stake-and-strand basket, without handles and
with a rounded base, originally made of Ceruana pratensis, which is
now rare. Instead, split stems of several plants are used: cotton bush,
pomegranate. Henna branches and Egyptian willow (Salix subserrata
(Tackholm; B W 250-25lbread basket; B H 825, plate B, 9 ). The
baskets are used throughout Egypt on the markets and in the houses, but
they are mainly produced in the Delta.
qafas plur. aqfds, crate or cage made of jerid, the split mid ribs of the date
palm leaf ( B H 711, B W 389).
rigleen also: rigleen Iff, twined doormat, made of the leaf sheath fibres of the
date palm..
sabat plur. sebita, isbata, coiled basket with high rim and straight sides. (BH
394, carrier bag; B W 197-198, in upper Egypt: general term for basket,
but also bread basket from halfa grass).
sella plur. silal, general term for basket (BH 426, cf salldl. B W 214-215,
referring to several baskets in stake-and-strand technique, mainly in
Delta dialects.
salldl plur. sallaldt B H 426: "basket woven from split cane" (twined and
woven stake-and-strand basket made of Arundo donax). This basket is
also referred to as 'aldga in middle Egypt. Produced mainly in the
Delta.
sebata screen or hut made of rigid stalks tied with string into a mat ( B W 197-
198).
seniyya tray; also term used for a round shallow coiled basket with two small
handles.
serir bil habl wooden bed frame with rope webbing
shinda cheese mat, made of the fibres of the date palm fruit stem, held together
by widely spaced rows of twining with string ( B W 249, in north middle
Egyptian sidda is used for this type of mat).
tabaq plur. atbaq, shallow coiled basket in different sizes. The most c o m m o n
are 55, 36 and 18 c m in diameter (BH 531; B W 280).
tala 'a small longitudonal flat sewn plaits basket with a long rope handle, used
during the pollination of the female date palm tree, to carry the male
pollen (cf. B W 4-288, 289).
zanbil or zambil, plur. zanabil, large sewn plaits basket made of a plaited
strip of at least 15 ba'at (BH 381, large basket made of palm leaves;
B W 191, south middle Egyptian: carrier basket for earth and dung or
two-part transport bag (cf. masbala). upper Egypt: large storage basket.
a round basket with straight sides and two handles, made of old car
tyres, in general much smaller than the zanbil made of palm leaf (cf.
guffa).
(South middle) Egyptian terms related to basketry production

'arjun Theflowerbunch of the female date palm. In upper Egypt this term
is also used to indicate the fibres from the fruit stem, on which the
arjun proper is growing, cf. zabata (cf. Henein, 181, 189; B W 305).
ba'a plur. ba'at, the length measured accross the breast between two arms
which are fully extended sideways, approximately 1.50 m (cf. Henein
183 1.60-1.80).
boos culms of large grasslike plants, such as reeds and sugar cane ( B W
40).
dees sedges, Cyperus species (cf. samaar. sa'ad)
dhafira plur. dhafdyir, plait (BH 523; B W 275).
gesh el-gamh wheat straw (BH 701; B W 403)
girdb plur. girbdn, sheath in which the date flowers grow, strips of which
are used to decorate the rim of the large tabaq baskets (cf. B W 370,
in the oases this term is used to indicate the dry fibres of the zabata).
gourbah fruit stem of the date palm. Term used in upper Egypt, while in south
middle Egypt zaghauwa is used (cf. B W 370 in Kharga: dried shoot
of the zabata).
halfa "halfa grass", tall grass species, 1 - 1,50 m long.
jerid midrib of the date palm leaf (BH 154; BW59-60; Henein 181).
khus date palm leaf (BH 269; B W 126).
khus abyad is the young, white palm leaf from the centre of the
crown, khus akhdar are the more mature, green leaves (cf. B W 126:
khusa beyda for young date palm leaf).
Iff 1 the fibre from decayed leaf sheaths at the base of the crown of a date
palm tree (BH 807; B W 443).
2 plastic fibre scouring pads or wire wool .
mesalla large needle used for making sewn plaits basketry (BH 426; B W 215
masalla, plur. masallaat).
makhraaz awl (BH 246, mukhraaz; B W 111 makharraz: term used in the
Dakhla oasis for a large needle.
qalb el-nakhla lit. "heart of the date palm", the young white leaves from the middle
of the palm crown (cf. galb el-nakhla B H 713; B W 390).
sa'ad sedges. Cyperus species (cf. samaar, dees)
sa'af date palm leaf (BH 414;). In the Kharga oasis za'af indicates the
leaves on the tree, before being dried ( B W 187: za'af, B W 207
sa'af), cf. khus.
samaar sedges, Cyperus species (Dakhla oasis, cf. B W 216), cf. Dees, Sa'ad
samaar morr rushes, Juncus acutus and Juncus rigidus
sasha curved knife
shibr plur. ashbdr, width from thumb to small finger of a hand with
stretched fingers, approximately 18 cm.
sharmukh plur. sharamikh, pollen of the male date palm, also sometimes used
for its stem, or 'sword'. It is not used as synonym for zaghauwa,
which is a comparable feature, but in a female tree ( B W 236).
shoka thorn at the base of the date palm mid rib, used as awl for basket
making.
sikh metal bars of the mat loom (BH 144, B W 224)
wated wooden peg which hold the metal bars of the mat loom (cf. B H 922;
B W 497,
zabata the flower bunch of the female date tree, which grows at the end of
the zaghauwa. Also used as a synonym for zaghauwa or gurbah ( B W
298), cf. arjun.
zaghauwa fruit stem of the date palm, the 'sword' of the zabata (south middle
Egypt), cf. arjun, gurbah ( B W 188).
Nubian terms for basketry types

Kensi / Matoki Fadidja
falan nagandi coiled lid for a flat basket
karadj .' karadj 1 coiled smallflatbasket, also used as lid.
coiled large, d o m e shaped basket with a foot
adda karadj duwal large d o m e shaped basket, used for winnowing
karadj kinna small karadj
karen konta coiled largeflatbasket used at weddings
'omrah konta coiled large flat basket, often used in combination
with a waliil
konteega coiled flat basket
konteen bode coiled basket containing harvest products, similar to
kubush
kubush or coiled basket, bowl shaped, used as measure for
kubushema flour
tagdiah shauwer coiled largeflatbasket, used as a tray
shibr coiled or plaited large basket, which holds the hot
bread when taken out of the oven.
walR coiled smallflatbasket, used as lid or cover
bani kaldiyya coiled small basket to present dates
kutayya coiled large, deep, d o m e shaped basket, used as a
lid.
sidra small tagadii. coiled lid to cover a plate
tagadii/tagdiah coiled flat basket in several sizes, similar to the
Egyptian tabaq .
shaloob sernun lauwer plaited or knotted hanger for coiled baskets or plates
bursh nibid mats made of sewn plaits
sellan bursh sellan nibid baskets made of sewn plaits
gamali large sewn plaits carrying basket
(used on camel back?, cf. Badawi 173).
mughraaf sewn plaits carrying basket, with straight sides, for
shopping and storing goods
(cf. Badawi 620, ladle, scoop)
shanta Arabic "bag", in Kensi used for a specific small
sewn plaits basket, made of two circular sides, sewn
together.
shibr sewn plaits carrying basket, which holds hot bread
which is taken out of the oven, is used on the market
and as container for goods
Nubian terms related to basketry production

Kensi Fadidja
mufraz ryen awl
shoka shoka thorn of date palm tree used as awl (Badawi 487).
gourbah arm fruit stem of the date palm, used as bundle material
for coiled basketry.
ambu nig imbe n Fg doam palm leaf (coll.)
sa'af el-Nakhla ninten nig date palm leaf (coll.)
brubi far'a wheat straw (cf. Egyptian gesh el-gamh)
shen alum used with boiling the doam palm leaf
senag addoka natural colorant
takata "crossing", plaited start of Nubian coiled basketry
D.3 Some Ancient Egyptian Basketry Terms
This brief list includes the basketry terms that featured in the previous chapters and
are mainly found in the Deir el-Medina texts. T h e terminology on ancient Egyptian
basketry is m u c h m o r e extensive than the few words listed here.
irgs, irks sack made out of a twined mat P-413,415

jsr rushes P-286
•nb halfa grass P-282
'4 sack (also of leather)? P-416
wnr reed or giant reed P-283
wd papyrus, other sedges? P-284, 285
b'j. w doam palm leaf stem or date palm mid rib P 274
mnh papyrus P-284
m n d m + nkr (coiled?) basket with twined sieve as lid P-413,415
mbj.t papyrus P-284
mtrht twined sieve with coiled rim P-413,415
nwh string (in combination: tm3 sm 'n nwh) P.413,417
nbd twined? P-412,413,416
417
rdmt sedges (in combination: tm3 rdmt) p. 413,417
h'w cheap basket, no specifications p. 413, 416

htp twined matting, container? p. 413, 416, 417
hnd woven? p.412-414,417
ht.ti specification of basketry: coarse p. 413, 415
sdr woven grass sleeping mat? p. 413, 416, 417
swy hay (in combination: tm3 sm 'n swy) p. 413, 417
sm' specification of basketry: fine p. 413, 415, 417
skr, skr' twined or woven item? p. 413, 417
krht coiled or twined basket for fruit p. 413, 415
kbs twined grain basket p. 412, 413
kskst large twined bag p. 413, 416
g3s" reed or giant reed p. 283
V3w doam palm leaves p. 274
rwfj papyrus p. 284
dnit coiled or twined basket for fruit or incense p. 413-415
BIBLIOGRAPHY
Adams, W.Y.
1984 Nubia, corridor to Africa, second edition, London and Princeton: Alan Lane
and Princeton University Press;
Adams, W.Y., J.A. Alexander, R. Allen
1983 "Qasr Ibrim 1980 and 1982" in: Journal of Egyptian Archaeology 69, pp.
43-66;
Adovasio, J.M.
1977 Basketry Technology: a guide to identification and analysis Chicago: Aldine
Publishing Company;
Aldred, C.
1988 Akhenaten, king of Egypt, London: Thames and Hudson;
Almquist, H.N.
1911 Nubische Studien im Sudan, 1877-78, Hergestellt von K.V. Zettersteen,
Uppsala/Leipzig:
Anderson, R.D., W . Y . A d a m s et alii
1978 "Qasr Ibrim 1978", in: Journal of Egyptian Archaeology 65, pp. 30-41;
Arnold, D.
1971 "Ethnomineralogy of Ticul, Yucatan Potters: ethics and emics" in: American
Antiquity 36, pp. 20-40.
Badawi. S. and M . Hinds
1986 A Dictionary of Egyptian Arabic, Beirut: Librairie du Liban;
Baines, J. and J. Malek,
1984 Atlas of Ancient Egypt, Phaidon: Oxford;
Balfet, H.
1952 "La vannerie; essai de classification", in: L'Anthropologie, Vol.56 (1952),
pp. 259-280.
Behnstedt. P. and M . Woidich
1994 Die dgyptisch-arabischen Dialekte. Glossar Arabisch-Deutsch(Band 4),
Wiesbaden: Dr. Ludwig Reichert Verlag.
Bentz Hogseth, H.
1998 "Transmission of Knowledge through Action; H o w to Interpret Timber from
Archaeological Sites", in: Andersson, A-C. et alii eds. The Kaleidoscopic
Past. Arkeologiska Skrifter no. 16, pp. 231-235, Goteborg: Goteborg
University, Department of Archaeology;
Berlandini, J.
1979 "Varia Memphitica III" in: Bulletin de ITnstitut d'Archeologie Orientale 79.
pp. 249-265.
Bierbrier, M .
1982 The Tomb Builders of the Pharaoh, London: British M u s e u m Publications;
Binford, L.
1967 "Smudge pits and hide smoking: the use of analogy in archaeological
reasoning", American Antiquity 32. 1-12;
1972 An Archaeological Perspective, N e w York: Harcourt;
1977 For Theory Building in Archaeology, N e w York: Academic Press;
1980 "Willow smoke and dogs' tails: hunter-gatherer settlement systems and
archaeological site formation" in: American Antiquity 45, pp. 4-20;
Blackman, W.S.
1927 The Fellahin of Upper Egypt, their Religious Social and Indutrial Life To-
day with Special Referenceto Survivals from Ancient Times, London: G.G.
Harrap;
Bliimner, H.
1875 Technologie und Terminotogie der Gewerbe und Kiinste bei Griechen und
Romern, Leipzig: Teubner;
Bongioanni, A.
1987 "Weaving: Mats and Baskets" in: A.M. Donadoni-Roveri (ed.), Egyptian
Civilization, Daily Life, pp. 106-119, Turin: Egyptian Museum;
Borchardt, L.
1907 "Drei Hieroglyphenzeichen" in: Zeitschrift fur die Agyptologische Sprache
44, pp. 75-79;
1907-1914 "Ausgrabungen in Tell el Amarna", in: Mitteilungen des Deutschen Orient
Gesellschaft no. 34, 46, 50, 52, 55.
Borchardt, L. and H. Ricke
1980 Die Wohnhduser in Tell El-Amarna, Berlin: Gebr. Mann Verlag;
Bouriant. U.
1903 Monuments pour servir a I'etude du culte d'Atonou, Memoires de l'lnstitut
Francais d'Archeologie Orientale no. 8, Cairo: Institut Francais
d'Archeologie Orientale;
Bruyere, B.
1937a Rapport sur lesfouilles a Deir el Medineh 1933-1934, premiere partie: La
necropole de I'ouest, Cairo: Institut Francais d'Archeologie Orientale;
1937b Rapport sur lesfouilles a Deir el Medineh 1934-1935, deuxieme partie: La
necropole de I'est, Cairo: Institut Francais d'Archeologie Orientale;
1939 Rapport sur lesfouilles a Deir el Medineh 1934-1935, Cairo: Institut
Francais d'Archeologie Orientale;
Buhler, A., K. Buhler-Oppenheim,
1948 "Grundlagen zur Systematik der gesamten textilen Techniken", in: Die
Textiliensammlung Fritz Ikle-Huber im Museum fitr Volkerkunde und
Schweiz. Museum fur Volkskunde Basel, Band 78 Denkschrift der Schweiz.
Naturforschenden Gesellschaft, Zurich: Schweiz. Naturforschenden
Gesellschaft;
Burckhardt, J.L.
1819 Travels in Nubia, London: John Murray
1981 Entdeckungen in Nubien 1813-1814, Herausgegeben, bearbeitet und
eingeleitet von Helmut Arndt, Tiibungen: Erdmann;
Caminos, R.A.
1968 The shrines and rock-inscriptions of Ibrim, London, Egypt Exploration
Society:
Bibliography 475
Clarke, D.L.
1978- Analytical Archaeology, second edition revised by B. Chapman, London:
Methuen (first edition 1968);
Collier, J.and M . Collier
1986 Visual Antrhopology: photography as a research method, Albuquerque:
University of N e w Mexico Press (first edition 1967, N e w York: Holt,
Rinehart and Winston);
Cooney. J.D.
1965 Amarna reliefs from Hermopolis in American collections: Brooklyn Museum;
Costin, C L .
1991 Craft Specialization: Issues in Defining, Documenting, and Explaining the
Organization of Production', in: M.B. Schiffer, ed. Archaeological Method
and Theory Vol. 3, pp. 1-56, Tucson: University of Arizona Press;
Crowfoot, Grace, M .
1954 "Textiles, basketry, and mats" in: C. Singer, E.J. Holmyard, A.R. Hall, A
history of technology, Vol.1 pp. 413 - 447, Oxford: Clarendon Press;
Curto, S.
1965 Nubien; Geschichte einer ratselhaften Kultur, Miinchen: Wilhelm Goldmann
Verlag;
Dauer, A . M .
1980 Zur Syntagmatik des ethnografischen Dokumentationsfrlms, (Acta ethnologica
et linguistica 47, Series General is 6, Wien: Fohrenau;
Davies. N. de G.
1903-1908 The Rock Tombs of Tell el Amarna I-VI, London, Egypt Exploration Fund;
1937 Paintings from the tomb of Rekh-Mi-Re' at Thebes, N e w York: The
Metropolitan Museum of Art;
1941 The tomb of the vizier Ramose, London: Egypt Exploration Fund;
1943 The tomb of Rekh-mi-Re at Thebes, N e w York: The Metropolitan Museum
of Art;
Driskell, B.N.
1981 "Basketry Analysis at Qasr Ibrim, Egypt" in: Nyame Akuma 18, pp. 9-10.
1986 "Qasr Ibrim - 1986 season. Field Report" in Newsletter of the American
Research Centre in Egypt 135, pp. 3-10.
Driskell, B.N., N.K. Adams, P.G. French
1989 "A newly discovered temple at Qasr Ibrim, preliminary report", in:
Archeologie du Nil Moyen, no. 3, pp. 11-54.
Douglas, F.H.
1935 Basketry decoration techniques, Denver Art Museum Leaflet 68;
Dunnell, R.C.
1971 Systematics in Prehistory, N e w York: Free Press;
Edwards, D.
1998 "The 1998 excavations in the Trench 10/14 area", in: Sudan and Nubia 2,
pp. 65-73.
Emery, I.
1980 The Primary Structures of Fabrics: an illustrated classification, 2nd edition
(1st edition 1966) Washington D.C.: The Textile Museum:
Emery, W . B .
1938 The royal tombs of Ballana and Qustul, 2 vols., Cairo: Government Press;
Emery, W . B . and L.P. Kirwan
1935 The excavations and survey between Wadi es-Sebua and Adindan 1929-1931,
Cairo;
Erman, A. and H. Grapow
1971 Worterbuch der Aegyptischen Sprache, Berlin: Akademie Verlag;
Esau, K.
1960 The Anatomy of Seed Plants, N e w York: Wiley;
Evans, Robert.K.
1978 'Early Craft Specialization: A n Example from the Balkan Chalcolithic', in:
Charles.L. Redman ed. Social Archaeology: Beyond Subsistance and Dating,
N e w York, London: Academic Press;
Fernea, R.
1973 Nubians in Egypt: peaceful people, Austin Texas;
Fernea, R. and A. Rouchdy
1987 "Contemporary Egyptian Nubians" in T. Hagg. ed. Nubian Culture Past and
Present. Stockholm: Almqvist and Wiksell;
Firth
1912 Tlie archeological survey of Nubia, Report for 1908-1909 2 vols, Cairo:
Government Press;
Fleisher, R.M.
1956 Measures and containers in Greek and roman, unpublished thesis, N e w
York.
Norton, R.E.
1990 Florian, M.-L.E., D. P. Kronkright and R.E. Norton, Tlie Conservation of
Artifacts M a d e from Plant Materials, Princeton: the Getty Conservation
Trust;
Forbes
1964 "Sewing, basketry and weaving" in Ancient Technology, Vol. IV, pp. 175-
195, Leiden: Brill;
France, C. de
1982 Cinema et Anthropologic Paris: Editions de la Maison des Sciences de
I'Homme;
Frankfort, H., J.D.S. Pendlebury e.a.
1933 The city of Akhenaten II, London: Egypt Exploration Society;
French, C.A.I.
1984 "Geomorphology and prehistory at El-Amarna" in: B.J. K e m p e.a. Amarna
Reports I, London: Egypt Exploration Society;
Frend, W . H . C .
1974 "The podium site at Qasr Ibrim", in: Journal of Egyptian Archaeology 60,
pp. 30-59;
Gardiner, A.H.
1911 Egyptian Hieratic Texts, Series I part 1, Leipzig;
1937 Late Egyptian Miscellanies, Bibliotheca Aegyptiaca VII, Brussels:
Geirnaert-Martin D.C.
1992 The Woven Land ofLaboya. Socio-cosmic ideas and values in West Sumba,
Eastern Indonesia, Leiden: C N W S publications;
Geiser, P.
1986 Tlie Egyptian Nubian; a study in social symbiosis, Cairo: The American
University in Cairo Press;
Bibliography 477
Germer, R.
1985 Flora des pharaonischen Agypten, Sonderschrift 14 Deutsches
archaologisches Institut Abteilung Kairo, Mainz am Rhein: Philipp von
Zabern;
Gould, R. A.
1978 "Beyond Analogy in Ethno-archaeology". in: Gould, R. ed. Explorations in
Ethno-archaeology, University of N e w Mexico Press: Albuquerque;
Gould, R A. and M.B. Schiffer, eds.
1981 Modern Material Culture: the archaeology of us. N e w York: Academic
Press;
Gourlay, Y.J.-L.
1981 Les Sparteries de Deir el-Medineh I, catalogue des techniques de sparterie,
Cairo, Institut Francais d'archeologie orientale;
Les Sparteries de Deir el-Medineh II, catalogue des objects de sparterie,
Cairo, Institut Frangais d'archeologie orientale;
Greiss, E.A.M.
1957 Anatomical Identification of Some Ancient Egyptian Plant Materials,
Memoirs de 1' Institut d' Egypte 55, Cairo.
Giddy, L.L.
1987 Egyptian Oases, Bahariya, Dakhla, Frafra and Kharga during Pharaonic
Times, Warminster: Aris & Phillips;
Guss, D.M.
1989 To Weave and Sing: art, symbol, and narrative in the South American rain
forest, Berkely: University of California Press;
Hari, R.
1984 "Quelques remarques sur l'abandon d'Akhetaton", Bulletin de la Societe
d'Egyptologie (Geneve), no. 9, pp. 113-118
Hanke, R.
1978 Amarna-Reliefs aus Hermopolis: neue Veroffentlichungen und Studien,
Hildesheim;
Heider, K.G.
1976 Ethnographic Film, Austin & London: University of Texas Press;
1983 Films for Anthropological Teaching, special publication no. 16, Washington:
American Anthropological Association.
Helck,
1961 Materialien zur Wirtschaftsgeschichte des Neuen Reiches Teil I-IV (1961-
1969), Wiesbaden: Akademie der Wissenschaften und der Literatur in
Mainz. Abh. der Geistes- und sozialwiss. Klasse. With an index by I.
Hofmann (1970).
Henein, N.H.
1988 Mart Girgis, village de haute-Egypte, Cairo: Institut Francais d'Archeologie
Orientale;
Herzog, R.
1957 Die Nubier; Untersuchungen und Beobachtungen zur Gruppengliederung
Gesellschaftsform und Wirtschaftsweise, Berlin: Akademie-Verlag;
Hodder, I.
1982 77jt? Present Past; an introduction to anthropology for archaeologists,
London: Batsford;
1992 Theory and Practice in Archaeology, London: Routledge;
1995 " Material culture in time" in: Hodder e.a. eds., Interpreting Archaeology;
finding meaning in the past, London: Rougledge.
1999 The Archeological Process, an Introduction, Oxford: Blackwell;
Hodder, I.ed.
1989 The Meaning of Things; material culture and symbolic expression, London:
Unwin H y m a n
Hodder, I.e.a. eds.
1995 Interpreting Archaeology; finding meaning in the past, London: Rougledge.
Hoffmann, M.A.
1979 Egypt before the pharaohs, London: Michael O'Mara Books;
Hofmann, I.
1986 Nubische Worterverzeichnis, Nubisch-deutsches und deutsch-nubisches
Worterverzeichnis nach dem Kenzi-Material des Samuel All His'en (1863-
1927), Berlin: Dietrich Reimer Verlag;
Horton, M
1988 "Report on the 1988 season at Qasr Ibrim" in: Journal of Egyptian
Archaeology 74, pp. vii-ix;
1991 "Africa in Egypt: new evidence from Qasr Ibrim" in: W . V . Davies ed. Egypt
and Africa, pp 264-277, London: British Museum Press;
Ikram, S.
forthcoming "Mud Matters: domestic silt technology in Upper Egypt" in: W.Z. Wendrich
and G. van der Kooij eds. Moving Matters: Proceedings of the Seminar on
Ethno-archaeology in the Middle East, Publications of the Netherlands
Flemisch Institute in Cairo no. 4, Leiden: Research School C N W S
publications;
Irving, J.
19783 77;t? World According to Garp, London: Victor Gollancz;
Janssen, J.J.
1975 Commodity prices from the Ramesside Period, Leiden: Brill;
Kemp, B.J.
1980 "Preliminary report of the el-'Amarna expedition, 1979", Journal of
Egyptian Archaeology, no. 66, pp. 5-16.
1981 "Preliminary report on the el-'Amarna expedition, 1980", Journal of
1983 "Preliminary report on the el-'Amarna expedition 1981-2", Journal of
1985 "Tell el-Amarna" in: W . Helck, E. Otto eds. Lexicon der Agyptologie,
Band I, pp. 309-319, Wiesbaden: Harassowitz;
1987 "The workmen's village in retrospect". Journal of Egyptian Archaeology 73,
pp. 21 - 50.
1989 Ancient Egypt, Anatomy of a Civilization, London, N e w York: Routledge;
Kemp, B.J. et alii
1984 Amarna Reports I, London: Egypt Exploration Society;
1985 Amarna Reports II, London: Egypt Exploration Society;
1986 Amarna Reports III, London: Egypt Exploration Society;
1987 Amarna Reports IV, London: Egypt Exploration Society;
1989 Amarna Reports V, London: Egypt Exploration Society;
Bibliography 479
Keppel, S.
1984 Primaire textiele technieken van de Mentawai-eilanden; systematiek van het
lus- en vlechtwerk, Amsterdam: VU-uitgeverij;
Kuhn, T.S.
1970 The Structure of Scientific Revolutions, second edition, Chicago, London:
University of Chicago Press;
Larsen, J.L. and Betty Freudenheim
1986 Interlacing; the Elemental Fabric. Tokyo: Kodansha International;
Lehmann, J.
1907 Systematik und geographische Verbreitung der Geflechtsarten, Abhandlungen
und Berichte des Koniglich Zoologischen und Anthropologisch-
Ethnographischen Museums zu Dresden Bd XI nr.3.
Leroi-Gourhan, A.
19712 L'homme et la matiere, (2nd edition. 1st edition 1943), Paris: Editions Albin
Michel;
Lesko, L.H.
1982 A Dictionary of Late Egyptian, Vol. I, Berkely: Scribe;
1990 (Vol II 1984; Vol III 1987; Vol IV 1989; Vol V 1990)
Lichtheim, M .
1973 Ancient Egyptian Literature, a Book of Reading. Volume I: the Old and
Middle Kingdom, Berkely, University of California Press.
Lucas. A.
1959 Ancient Egyptian Materials and Industries, (reprint of 3rd edition of 1948,
1st ed. 1926), London: Edward Arnold;
Lucas. A. and J.R. Harris
1962 Ancient Egyptian Materials and Industries. (4th edition), London: Edward
Arnold;
Martin, G.T. e.a.
1985 The tomb-chapels of Paser and Ra 'ia at Saqqara, London:
1989 The Memphite tomb of Horemheb. commander-in-chief of Tut'ankhamun,
London:
1991 The hidden tombs of Memphis, London: Thames and Hudson;
Mason, O.T.
1904 "Aboriginal American Basketry: studies in a textile art without machinery",
in: Annual Report of the Board of Regents of the Smithsonian Institution,
Report of the U.S. National Museum for the year ending June 30 1902, pp.
171-548, plate 1-248, Washington: Government Printing Office;
Metcalfe, C.R.
1960 Anatomy of the Monocotyledons. Volume I: Gramineae. Oxford: Clarendon
Press;
Metcalfe, C.R. and P.B. Tomlinson
1961 Anatomy of the Monocotyledons, Volume 2: Palmae, Oxford: Clarendon
Press
Metcalfe, C.R. and M . Gregory
1971 Anatomy of the monocotyledons, Volume 5: Cyperaceae, Oxford: Clarendon
Press;
McGregor, R.
1992 Prehistoric Basketry of the Lower Pecos. Texas, Monographs in World
Archaeology no.6, Madison Wisconsin: Prehistory Press;
Mileham, G. S.
1910 Churches in Lower Nubia, Philadelphia: University Museum;
Mills, A.J.
1982 Tlie cemeteries of Qasr Ibrim; a report of the excavations conducted by W.B.
Emery in 1961, London: Egypt Exploration Society;
Newberry, P.E.
1893 Beni Hassan, Part II, Archaeological Survey of Egypt 2, London: Egypt
Exploration Society;
Nicholson. P.T.
1987 "The Potters of Deir Mawas, an Ethnoarchaeological Study" in: Kemp e.a.
Amarna Reports IV, pp. 279-308, London: Egypt Exploration Society;
Nijland, D. J.
1989 Schaduwen en werkelijkheid; cultured bepaald kennen, waarnemen,
nonverbaal communiceren en onderzoek daarvan via de terugvertoning van
de etnografischefilm"TOBELO MARRIAGE", Leiden: Doctoral thesis
defended at Leiden University;
North, R. E.
1990 "Technology of Plant Materials Used in Artifacts", in: M-L. E. Florian,
D.P. Kronkgrass, R.E. North The conservation of artifacts made from plant
materials, pp. 83-138, Princeton: J. Paul Getty Conservation Trust;
Ochsenschlager, E.L.
1998 "Viewing the Past: ethnoarchaeology at al-Hiba" in: Visual Anthropology 11,
pp. 103-143;
Peet, T.E. and C L . Woolley,
1923 The city of Akhenaten I; excavations of 1921 and 1922 at el'Amarneh,
London: Egypt Exploration Society;
Pendlebury, J.D.S. e.a.
1951 The City of Akhenaten HI, London: Egypt Exploration Society;
Petrie, sir W.M.Flinders
1894 Tell el-Amarna, London:..;
Pinsky, V. and A. Wylie, eds.
1989 Critical Traditions in Contemporary Archaeology; essays in the philosophy,
history and socio-politics of archaeology, Cambridge: Cambridge University
Press;
Plumley, J. M.
1964 "Qasr Ibrim 1963-1964", in: Journal of Egyptian Archaeology 50, pp. 3-5;
1967 "Qasr Ibrim December 1966", in: Journal of Egyptian Archaeology 53, pp.
3-5;
1975 "Qasr Ibrim, 1974", in: Journal of Egyptian Archaeology 61, pp. 5-27;
Plumley, J. M and W.Y. Adams
Plumley, J. M , W.Y. Adams and E. Crowfoot
Procopius
Procopius, translation by H.B. Dewing, 1914-1935, London: Heinemann;
Reeves, N.
1990 The Complete Tutankhamun, London: Thames and Hudson;
Bibliography 481
Reinisch, L.
1911 Die sprachliche Stellung des Nuba, Kaiserliche Akademie der
Wissenschaften, Bd III, Schriften der Sprachenkommission, Wien: Alfred
Holder:
Reisner, G.A.
1910 Archaeological Survey of Nubia. Report for 1907-1908, Cairo: National
Printing Department;
Renfrew, A C .
1973 Social Archaeology, Southampton: Southampton University Press;
1982 Towards an Archaeology of Mind. Cambridge: Cambridge University Press;
Roeder, G.
1969 Amarna-Reliefs aus Hermopolis : Ausgrabungen der Deutschen Hermopolis-
Expedition in Hermopolis. 1929-1939 hrsg. von R. Hanke, Hildesheim :
Gerstenberg;
Rose, P.J.
1992 77ii? aftermath of the Roman Frontier in Lower Nubia, unpublished PhD
dissertation, University of Cambridge;
1995a "Report on the handmade sherds" in: S.E. Sidebotham, W . Z . Wendrich,
Berenike 1994, Preliminary Report of the Excavations at Berenike (Egyptian
Red Sea Coast) and the Survey of the Eastern Desert, pp. 41-43, Leiden:
C N W S publications:
1995b "Report on the 1995 season at Qasr Ibrim" in: Journal of Egyptian
Archaeology 81, p. 11;
Archaeology 82, p. 19-22;
Arcliaeology 83, p. 14-15;
1998a "Report on the 1995 season at Qasr Ibrim" in: Journal of Egyptian
Archaeology 84, p. 20-22;
1998b "Qasr Ibrim 1998" in: Sudan and Nubia 2, pp. 61-64.
Rose. P.J.. and P. Rowley-Conwy
1989 "Qasr Ibrim Regional Survey: Preliminary Results", in: Archeologie du Nil
Moyen, no. 3, pp. 121-130.
Rouchdy, A.
1991 Nubians and the Nubian Language in Contemporary Egypt: a case of
cultural and linguistic contact, Leiden: Brill;
Rowley-Conwy, P.
1989 "Nubia A D 0-550 and the 'Islamic' agricultural revolution: preliminary
botanical evidence from Qasr Ibrim, Egyptian Nubia", in: Archeologie du
Nil Moyen. no. 3, pp. 131-138.
Saitta, D.J.
1983 'The Poverty of Philosophy in Archaeology', in: J.A. Moore, A.S. Keene
eds. Archaeological Hammers and Theories, pp. 299-304, N e w York,
London: Academic Press;
Salmon, M . H .
1982 Philosophy and Archaeology, N e w York, London: Academic Press;
Schiaparelli, E
1927 Relazione sui lavori della missione archeologica italiana in Egitto (anni
1903-1920). Volume secondo: La tomba intatta dell' architetto Cha nella
necropoli di Tebe, Turin;
Seiler-Baldinger, A.
1973 Systematik der Textilen Techniken, Basler Beitrage zur Ethnologie, Basel:
Pharos-Verlag;
1979 Classification of Textile Techniques, Ahmedabad, India: Calico M u s e u m of
Textiles;
Shanks, M .
1992 Experiencing the Past; on the character of archaeology, London: Routledge;
1996 Classical Archaeology of Greece; Experiences of the discipline, London:
Routledge;
Shanks, M . and I. Hodder
1995 "Processual, postprocessual and interpretive archaeologies" in: Hodder e.a.
eds. Interpreting Archaeology;findingmeaning in the past, London:
Rougledge.
Shanks, M . and C. Tilley.
1987 Social Theory and Archaeology, Cambridge: Polity Press;
19922 Re-constructing Archaeology, London: Routledge;
Skeat, T.C.
1977 "A letter from the king of the Blemmyes to the king of the Noubades",
Journal of Egyptian Archaeology 63, pp. 159-170
Smith, R.W.
1967 "The Akhenaten Temple Project", in Expeditions 10. no. 1 pp. 24-52,
University M u s e u m of Pennsylvania;
Smith. R.W.. D.B. Redford,
1976 The akhenaten Temple Project. Vol I "Initial discoveries", Warminster;
Struever, S.
1995 "The Role of Film in Archaeology" in: Paul Hockings ed. Principles of
Visual Anthropology, pp. 193-199, Berlin/New York: Mouton de Gruyter;
Tackholm, V
1974 Students' Flora of Egypt, Beirut: Cairo University;
Tackholm, V. and G. Tackholm
1941 Flora of Egypt I, Fouad I University, Bulletin of the Faculty of Science,
Vol. 17, Cairo.
Tackholm, V., M . Drar
1950 Flora of Egypt II, Fouad I University, Bulletin of the Faculty of Science,
Vol. 28, Cairo.
Tackholm. V., M . Drar
1954 Flora of Egypt III, Cairo University, Bulletin of the Faculty of Science, Vol.
30, Cairo.
Tackholm, V., M . Drar
1969 Flora of Egypt IV, Cairo University, Bulletin of the Faculty of Science, Vol.
36, Cairo.
Teeter, E.
1987 "Techniques and Terminology of Rope-making in Ancient Egypt", Journal of
Egyptian Archaeology 73, pp. 71-77;
Bibliography 483
Tilley, C.
1998 Methaphor and Material Culture, London: Blackwell;
Timme, P.
1917 Tell el-Amarna vor der deutschen Ausgrabung im Jahre 1911, Leipzig;
Trigger, B.G.
1965 History and settlement in Lower Nubia, Yale University Publications in
Anthropology no. 69, New Haven: Department of Anthropology Yale
University;
1989 A history of archaeological thought. Cambridge: Cambridge University
Press;
Valbelle, D.
1985 Les Ouvriers de la Tombe, Deir el-Medineh a I'Epoque Ramesside, Cairo:
Institut Francais d'Archeologie Orientale;
Vandier, J.
1964 Manuel d'Archeologie Egyptienne, tome IVBas-reliefs et peintures, scenes
de la vie quotidienne, Paris:Editions A.et J. Picard;
1969 Manuel d'Archeologie Egyptienne, tome V Bas-reliefs et peintures, scenes de
la vie quotidienne, Paris:Editions A.et J. Picard; Manuel d'Archeologie
Egyptienne. tome VI Bas-reliefs et peintures. scenes de la vie agricole a
I'ancien et au moyen empire, Paris:Editions A.et J. Picard;
Vogt, Emil
1937 Geflechte und Gewebe der Steinzeit, Basel: Schweiz. Gesellschaft fur
Urgeschischte;
Waddell, H.
1936 The Desert Fathers, Translations from the Latin with an introduction by
Helen Waddell, New York: Holt.
Wallert, I.
1962 Die Palmen im Alten Agypten; Eine Untersuchung ihrer praktischen,
symbolischen und religiosen Bedeutung. M A S 1, Berlin: Verlag Bruno
Hessling.
Watson, A.
1977 "The Rise and Spread of Old World Cotton", in V. Gervers ed.. Studies of
Textile History, pp. 355-368, Toronto: Royal Ontario Museum;
Wehr, H.
1980 A Dictionary of Modern Written Standard Arabic. Lebanon: Librairie du
Liban (reprint of the 1974 edition: Wiesbaden. Harrassowitz)
Weatherhead.F.
1987 "Report on the 1986 Excavations; Fragments of a Painted Royal Figure with
Artist's Grid From West Street 2/3" in: Kemp, Amarna Reports IV, pp. 17-
29, London: Egypt Exploration Society;
Wendrich, W . I.
1988 Bescherming of Verstrikking, de magische betekenis van knopen in het oude
Egypte, Unpublished M A thesis University of Amsterdam, Department of
Theology.
1989 "Preliminary report on the Amarna basketry and cordage" in: B.J. Kemp,
Amarna Reports V, London: Egypt Exploration Society;
1990 "Vlechtwerk, zelfkant van het specialisme textiel?" in: Oudheidkundige
Mededelingen van het Rijksmuseum van Oudheden no. 10, pp. 211-216.
1991a "Basketry, a textile technique?" in: Archaeological Textiles Newsletter no.

12, pp. 18-20.
1991b "Basketry versus textiles: on terms" in: Archaeological Textiles Newsletter
no. 13, pp. 9-10.
1991c Who is afraid of basketry ? a guide to recording basketry and cordage for
archaeologists and ethnographers, Leiden: Centre of Non-Western Studies;
1995 "Basketry and Cordage" in: S.E. Sidebotham, W . Z . Wendrich eds. Berenike
'94; Preliminary Report of the Excavations at Berenike (Egyptian Red Sea
Coast) and the Survey of the Eastern Desert, pp. 69-84, Leiden: C N W S
publications;
1997 "Manden en Touw" in: D e Oudheid Verpakt, catalogue of an exhibition on
ancient packing materials in the Allard Pierson Museum, Amsterdam,
Voorhout: Topa Holding.
1998 "Basketry and Matting" in: S.E. Sidebotham and W . Z . Wenrich eds.
Berenike 1996, Report of the Excavations at Berenike (Egyptian Red Sea
Coast) and the Survey of the Eastern Desert, pp. 253-164, Leiden: C N W S
publications;
1999 "Basketry" in: P.T. Nicholson, I. Shaw eds: Ancient Egyptian Materials and
Technology, pp. 254-267, Cambridge: Cambridge University Press;
Wipszycka, E.
1986 "Les aspects economiques de la vie de la communaute des Kellia" in: Le site
monastique copte des Kellia. Sources historiques et explorations
archeologiques, Actes du Colloque de Geneve, Geneve: Mission Suisse
d'archeologie copte de I'Universite de Geneve;
Wylie, A.
1985 "The reaction against analogy", in: M . Schiffer (ed.) Advances in
Archaeological Method and Theory 8, pp. 63-111, N e w York: Academic
Press;
Whitley, D.S.
1998 Reader in Archaeological Theory; post-processual and cognitive approaches,
Routledge: London.
Worp, K.A
1977 "P.Vindob.Boswinkel 8 + 9" in: Bulletin of the American Society for
Papyrologists, pp. 95-97.
Worp, K.A.
1988 "Ein addendum lexicis in P.Soterichus 4*", in: Tyche, band 3, pp. 277-278.
INDEX
cvverircAoKov 214, 260 Badawi and Hinds 181, 273

AeuKOipuAAwv 279 Balfet
7TevraTc[A]6Kov 214, 260 25, 26, 29-32, 34, 35, 38, 44-46, 55, 56
bardi 150, 284
active basic structure
28, 35, 41, 43-45, 47, 48, 60, 66, 69, 28, 35, 56, 58-60, 62-65, 69, 70, 72, 81,
82, 84-92, 95. 97-100, 107, 153, 155, 82, 91, 92, 98, 99, 153, 161, 185, 217,
157-168, 171, 172, 174. 176, 177, 179, 263, 269, 271, 289, 291, 292, 295, 302,
182. 184, 186. 189-192. 194, 198, 199, 303, 306-309, 311, 312, 317, 318, 324,
201-204, 206-209, 217, 220, 221, 225, 326-328, 330-333, 346, 347, 350, 351,
227, 230, 231. 236, 237, 241, 243-247, 366, 389,419-422
249, 256-258, 260, 269-271, 300, 302, ba'a 180, 181, 375, 397
303, 307, 310, 314, 323, 324, 332, 333, Behnstedt, Woidich 181, 273, 279, 280
335, 356, 358, 369, 372, 387, 420, 421 benti nig 277
Adams 124, 126-128, 130, 134-136, 152 Bentz Hogseth 391
additional features 58, 65 Berlandini 121
Adovasio Bierbrier 120
25-27, 29. 30. 33. 35, 38. 40-43. 55. 64 Binford 7, 18. 19
Aldred 112, 116, 120, 124 Blemmyes 126, 135
Amarna boos 149, 283
1-4, 20, 27, 56, 66-68, 72, 75, 79, 83, Borchardt 114
87, 109-112. 114-124. 145. 147, 148, Bouriant 114
150-152, 155, 163, 168-170, 175-180, branch 262, 350. 392
182-184. 186-189. 191. 192, 194-196, brubi 284
198-201, 203, 204, 206, 209, 217, 225, brush
226, 231-233, 237, 249-256, 258-266, 31, 120, 169, 170, 186-190, 198, 200,
273, 275, 282, 283, 285-287, 289-293, 225, 232, 309-312, 321, 415
295, 303, 304, 306-309, 311, 312, Bruyere 109, 110, 120
315-321. 323-328, 330, 406, 409, 410, Buhler 26, 29-31, 38, 46, 52
412-417, 420, 425 Burckhardt 127
ambu nig 274
anchoring fabric Caminos 125, 130, 131
85, 98. 155, 157, 217, 271, 290, 292, carrier basket .... 181, 214, 259, 398
302 carrying
Anderson 128, 130 53, 136, 186, 202, 204, 205, 215, 258,
'arjun 279, 280 259, 327
Arnold 420 cataract 137
arm 277, 280, 282 Clarke 7
Arundo donax 148, 149, 283 coiling
awl 33, 39-41, 43, 44, 47, 49, 52, 62-64, 85,
271, 304, 306, 350, 355-359, 362-364, 86, 90, 97, 98, 159, 160, 166, 169. 170,
377. 378, 387, 392 172, 174, 182, 184-186, 204, 217, 218,
220, 245, 249. 250, 254, 263, 265, 270, doam palm
271, 303, 304, 308, 309, 328, 331, 350, 62, 66, 92, 145, 146, 157, 176, 177,
359, 392, 413, 414 182, 186-189, 192, 195, 198, 199,
Collier 72, 101-104 208-210, 213-216, 221-224, 226-228,
common reed 149 231, 232, 237-239, 241, 242, 246, 247,
composition 252, 254-256, 260-263, 274-277, 282,
11. 43, 59, 81, 91, 131, 171, 194, 209, 284, 289, 290, 298, 301, 303, 304, 306,
213 307, 310, 311, 317, 321, 330, 358, 359,
concentration 362,409, 411,414, 416
70, 107, 353, 364, 390. 391, 393, 422 Dongola 126, 127
continuous plaiting Douglas 64
60, 158, 165, 169, 170, 203, 237, 241, Driskell
250, 326 110, 111, 125, 128, 130, 136, 423
Cooney 114 dyeing 284
Crowfoot dynamic
26. 27, 29, 30, 32, 34, 35, 38, 52, 128 19, 20, 72, 145, 266, 269, 270, 421
Curto 136, 138
Cyperus alopecuroides 151, 285 Edwards 131. 132, 207
Cyperus rotundus 151, 285 element
Cyperus schimperianus 151, 285 28, 34, 41, 42, 44, 48-52, 81, 91-93, 95,
97, 99, 162, 167, 182, 186, 199, 241.
date palm 256, 258, 333, 356
10, 62, 66, 105. 122, 133. 145-147, el-Amariyya
181, 185-187. 189, 201, 202, 208, 213, 3, 112, 122, 124, 185, 201, 204, 269,
214, 216, 221-223, 225, 231, 232, 234, 329, 350, 355. 366, 371. 373. 390, 401.
237-239, 243, 244, 246, 247, 252, 406
254-256, 260-262. 264, 265, 270, 274, el-Hagg Qandil
277-282, 291, 295, 296, 298, 303, 304, 112, 122-124, 181, 185, 187, 204, 262,
306, 319, 324, 326. 327, 331-333, 336, 269, 273, 278-280, 331, 343, 401, 424
341, 346, 350, 351, 353, 358, 359, 362, el-Hawata 112, 122, 124
363, 366, 367, 392, 398-401, 404, 411 el-Till
date palm fibre 112, 122-124, 185, 204, 262, 279, 400
187, 189. 201. 208, 225, 232, 237, 243, Emery
247. 256, 260. 281, 291. 298, 303, 319, 26, 29-35, 37, 38, 48, 49, 51, 55, 130,
333, 341, 346, 366 131, 157
Dauer 104 essence 391, 392
Davies 119, 307,416 ethno-archaeology
decoration 2-4, 10, 16, 18-20, 71,425
20, 40, 41, 43, 52, 58. 63-65, 72, 134, experience
166, 170, 220, 222, 263, 264, 302, 303, 9, 16, 23, 68, 77, 105, 122, 285, 341,
308, 328, 334, 347, 351, 358, 363, 391, 391,419
392, 402, 422, 424 'eye' symbol 107
dees 151,285
Deir el-Medina fabric
52, 109, 110, 120. 121, 309, 409, 412, 5, 16, 17, 21, 22, 28, 33, 34, 46, 58,
413, 415 59, 62, 69, 81, 84, 85, 92, 97-100, 105,
Desmostachya bipinnata 155, 157, 159, 165, 171, 172, 176, 179,
147. 148, 177, 182, 282 182, 186, 190, 192, 194, 201, 203-206,
dnit 413-415 208-213, 217-219, 227, 230-232,
Index 487
236-238, 241, 243-246, 252, 259, 271, Germer . . 146-152, 277, 283, 284, 286
290-292. 294, 302, 303, 306. 316, 318, ghaab 149, 283
319, 326, 327, 333, 345, 348, 351, 359. giant reed 149, 425
364, 373-375, 387, 388, 391, 408, 421 gid 282
Fadidja 'good' side of the work 393
137, 140, 141, 146, 215, 216, 223, 273, Gossypium arboreum 152, 287
274, 276, 277, 280, 282-284, 304, 306, Gould 19
358, 392, 403, 404, 414, 423 Gourlay
fast forward 104, 352, 390 26, 27, 30, 31, 33, 34, 38, 52, 54, 55,
fellahin 226, 259, 264, 265 109, 110, 254
Fernea 136, 138, 140, 141 graminae 147, 282
fibre Greiss 66
66, 67, 85. 145. 147, 152, 164. 187. guffa 181, 215, 217, 399
189. 201, 208, 214, 222, 225, 232, 237,
243, 247, 255, 256, 260, 277, 280, 281, halfa (grass)
287, 291, 296, 298, 300, 303, 319, 323, 130, 137, 140, 141, 147, 148, 276, 282,
328, 333, 341, 346-348, 351, 353, 354, 283,401. 403, 405,416
364, 366, 367, 397 'hammer' symbol 107
fingertips 300, 336, 341 handles
fingers 52, 53, 60, 65, 67, 69, 70, 92, 153, 186,
187-189, 225. 312, 341. 342, 344, 363, 187, 204, 215, 225, 259, 295, 296, 300,
382 302, 303, 309. 327, 328, 332, 333, 335.
Firth 128 346-348, 351
'foot' symbol 107 Hanke 114
Forbes . . 26, 27. 29. 30, 33. 34. 38, 52 Hari 121
frame head ring 181, 222, 292,402
5. 11. 33, 74. 90. 91, 98, 103. 104, 106, Heider 73, 101, 102
192, 227, 228, 271, 317, 318, 324. 327, Helck 273
367, 375, 377, 379, 381-388, 392. 404, Henein
405 Ill, 181, 260, 264, 273, 279, 280
framing 102, 103 hermeneutic 9, 22
France 2, 101-104. 106 Herzog 136-138
Frankfort 114 Hodder
French 25. 36, 38. 92. 102, 112 6, 8-10, 15. 17-19, 21, 22,419,423
Frend 128, 130 Hofmann 136, 273, 279
front horizontal loom . . . 319, 392, 409, 410
106, 118. 192, 255, 294, 300, 313, 314, Horton 2, 128, 130
343, 345, 348, 349, 356, 357. 363, 364, Hyphaene thebaica . 146.176.182,274
367, 371, 372, 389, 390,409
frozen ibra 350, 354
107, 269, 289, 332, 350, 369, 372, 387 Ikram 255
full-time professional 405, 406, 409, 424 imbi nig 274
furniture imbrication
2, 53, 147, 148, 170, 188, 189, 192, 64, 65, 88, 150, 166, 170, 223, 245
193, 227-229, 250, 261, 262, 276, 287, Imperata cylindrica 148, 177, 182,282
293, 317,404,410, 412
Janssen 273, 412-417
Gardiner 276 jerd 404, 405
Geiser 136, 137 jert 274, 277, 279,281,295
Juncus rigidus 152, 286 34, 50, 51, 106, 156, 157, 161, 165,
jurbah 277, 279, 282 168, 170, 171, 177, 186, 189, 203, 207,
225, 236, 241, 247, 266, 331, 390, 422
Kemp Linum usitatissimum 152, 287
1, 112, 114-116. 118-121. 175, 184, loom
409, 410 31-33, 47, 70, 85, 90, 92, 161, 269,
Kensi 271, 312-321, 323, 366, 370-372, 374,
137, 140, 141, 146, 215, 216, 222, 223, 392, 404-406, 409, 410. 422
273. 274. 277, 279, 282-284, 304, 307. Lucas .... 26, 27. 29, 30, 34, 38, 52
392, 403, 404, 408, 414, 423
Keppel 26, 27, 46, 174 macro-analysis 103, 104
khus 15,279,281,400 Mahasi 137
kneading .... 337-339, 345. 352, 356 Martin 20
knife Mason . . . 25-28, 30, 31, 34, 38, 40-42
275, 281, 286. 292, 296, 303, 311, 312, mat maker
317, 323, 348, 349. 392 76, 252, 286, 313, 314, 316, 318, 323,
367, 370-374, 379, 404-406, 410, 424
Larsen.. 26,30,34,38.39.48-51,55 matraha 201, 202, 324, 407, 408
leaf McGregor .... 19, 25-27, 38, 40, 43
15, 53, 54, 60, 62, 66, 67, 83-85. 87, meaning
88, 91, 92, 105, 122, 133, 145-147, 2. 3, 8. 9, 18, 20, 21, 55, 56, 69, 71,
157, 159. 160, 176-178, 182, 184-192, 88, 111, 145, 175, 192, 264, 274, 389,
194, 195, 198, 199, 208-210, 213-219, 391, 413, 414, 422, 423
221-228. 231, 232. 234, 237-239, member
241-244, 246, 247, 252, 254-256, 93, 95, 153, 161, 176. 177, 179, 182,
260-264. 270, 271, 274-277, 279-282, 186, 189, 190, 207, 208, 217, 225, 227,
284, 289-291, 295, 296, 298, 300, 301, 272
303, 304, 306, 307, 309-311, 317, 318, Meroe 125, 134
320, 321, 326, 327, 330-333, 335, 336, Meroitic
341, 342, 345, 347, 349-354, 356, 358, 125, 126, 128, 131. 133-136, 213, 215.
359, 362-364, 367, 389, 397, 399, 400, 218, 219, 227, 261, 263-265, 410, 411,
403,409,411, 414, 416, 425 425
leaf sheath fibre 237, 277, 281, 296, 333 mesalla 300, 301, 337, 339
leaf strips metaphor 9, 20
54, 60, 176, 187, 189. 198, 213. 223, micro-analysis .... 103, 104, 106, 389
239, 241, 276, 281, 290, 291, 296, 298, mid rib . . 244, 277, 282, 324, 326-328
350, 362 Mileham 128
leaflets Mills 128, 130, 131, 134
34, 90, 92, 147, 237-239, 243, 244,
264, 277. 279, 281, 296, 298, 326, 327, nakhla 146, 277, 278, 397
335, 336, 341, 342, 350, 351, 389, 393 Napata 125, 134
Lehmann narrative
25,26, 28, 30, 31, 34, 38, 39,41,46 2, 8,9, 11,22,73, 101, 102,419,426
Leroi-Gourhan 25, 26 needle
Lesko 274 98. 105, 215, 271, 293-295, 300-304,
Lichtheim 371, 374, 425 306, 332, 337, 339. 342-350, 352-357,
lid 222, 301,402,413,415 362, 392,408,411,422,425
linear net20, 79, 157, 202, 204, 205, 208, 259,
289, 290, 292-295. 327, 407, 409, 424
Index 489
New Nubia 247, 249, 256-258, 260, 269-271, 282,

3, 56, 62, 65, 74, 75, 78, 111, 136-140, 292, 300, 302-304, 307, 310, 314, 323,
142, 155, 168-171. 193, 207, 208, 215. 324, 332, 333, 356, 372, 387, 420, 421
222, 223, 226, 230, 231, 235-237. 241, Peet
243-245. 247-249, 251, 252, 254-256, 110, 114, 116-118, 120. 175. 177-180,
258-266, 270. 271. 275-277, 282. 289, 184, 186-188, 192. 193, 195. 200, 203,
291. 292. 295. 300-304, 306, 307. 309, 204, 410
311, 312, 317, 318, 323-328, 330, 392, Peg
399, 400, 403, 404, 406-408, 411, 420, 271, 310. 318, 371, 375, 378, 381-387
423-425 Pendlebury 114, 116
Newberry 293 period
Nicholson 406 25, 26. 73. 74, 76, 78, 106, 109-112,
nintin nig 277 115. 116, 120, 121, 125-128, 131,
Nobatae 126, 135 133-136, 145-147, 152, 170, 180, 204,
nonprofessional 405. 406, 409-411, 414 213-215. 218, 219, 221, 225, 230-232,
occurring action 236, 239, 245, 247. 249, 254, 255. 260,
105. 335. 336, 351. 356, 364, 372, 379, 261, 263. 265, 272, 275-277. 281, 283,
380, 387, 422 287, 293, 296, 307, 312, 323, 332,
Ochsenschlager 18, 20, 21 334-336, 351, 390, 411, 412, 414, 421,
Old Nubia 424, 425
140. 142. 264, 276. 277, 363, 365. 399, Petrie 114, 119
403 Phoenix dactylifera 146, 277
one plane Phragmites australis 149. 283
95. 97. 161, 162. 164. 179, 190, 192, pierced
203, 208, 227, 230, 237. 241, 247 98, 163. 169, 170, 174. 189, 201. 202.
orientation 249, 250, 258, 259, 261. 279, 281, 324,
38, 43. 44, 59, 75. 81-84, 88, 90. 91. 407, 408
93. 95, 97. 106, 131, 153, 157. 167, plain weave 35, 36, 312
171. 172, 174, 179, 190, 194, 195, plait pattern
199-201, 208, 210. 227. 231, 290, 291, 43, 51. 164, 181, 209, 212, 237, 241,
301, 325. 367. 372. 377. 379. 383. 247, 261, 296, 302, 335
385-387 plaited strips
pairing 95 98, 157, 158, 164, 165, 169, 170, 180,
palmae 145, 274 203, 204, 209-212, 215. 216. 241, 243.
'palm tree' symbol 106 249, 260, 264, 296, 302, 326, 335. 411
papyrus plaiting
150, 151, 188, 191, 204, 214, 252-254, 33, 39. 41, 43, 47. 49. 60. 70, 87, 90,
259, 261, 279, 284, 285, 298, 315, 320, 91, 97, 158, 165, 168-171, 174, 180,
321, 323, 327, 409.417 181, 203, 204, 209-212, 214-216, 236,
part-time professional . . 405, 406, 423 237, 239-241, 247-250, 265, 270, 296,
passive 298, 326, 330, 332, 334, 335, 338, 341,
28, 35, 41, 43-45, 47, 48, 66, 69, 82, 342, 346, 347, 358, 389, 393
84-86, 88-92. 95, 97-100, 107, 153, plane
155, 157-163, 166-168, 172, 179, 182, 32, 33, 95, 97, 153, 161, 162, 164, 179.
184, 186, 189-192, 194-196, 198, 199, 190, 192, 203, 208, 227, 230, 237, 241.
201, 202, 206, 208, 209, 217, 220, 221, 247
225, 227, 230. 231. 236, 241, 244, 246, Plumley 125, 126, 128, 130, 131
positivist 6-11,21, 22, 102
post holes 16, 392, 410
pot stand 291, 292 67. 145, 150, 151, 161, 186, 253, 284,
Primis 126 285, 413, 416, 417
processual archaeology 7, 9, 10, 18, 22 sewn plaits
Procopius 135 2, 54, 62, 69, 73, 87, 88, 90, 92, 97,
professional 98, 122, 157, 165, 169, 170, 179-181,
72, 121, 122, 136. 181, 292, 367, 391, 204, 208-211, 213-217, 223, 235,
397, 399, 401, 404-406, 408-411, 414, 249-252, 254, 255, 258-261, 264, 265,
423. 424 269, 295, 296, 301, 331, 333, 335, 337,
338, 356, 392, 393, 397, 399-401, 403,
qafas 406-408, 411, 412, 414, 425
122, 170, 324, 325, 407, 408, 424 seyf el-zabata 279
qalb el-nakhla 277. 278, 397 Shanks 6, 8-10, 15, 22, 419, 420
qass el-qamh 284 shibr 185, 215, 355, 399
qutun 152 'sitting man' symbol 106
Skeat 126
recurring action skill
105, 335-339, 351, 355, 356, 359, 364, 20, 103, 107. 122, 353. 390, 391, 401
367, 372, 377, 379, 387 slow motion 104
reed . . . 148, 149, 201, 262, 412, 425 Smith 134
Reeves 254 shoka 279
Reinisch 136, 137 static
Reisner 125. 127, 128, 134, 265 3, 19, 20, 57, 72, 145, 153, 249, 264,
Renfrew 8 266, 269, 270. 421
repetitive Stipa tenacissima 148
105, 295, 331, 334, 387, 389-391. Struever 101
421-423 symbol
Roeder 114 20, 106, 107. 341, 342, 344-346, 348,
rolling movement 353-356, 363, 364, 372, 375, 381, 382,
298, 300, 336, 342, 346, 381, 382 385
roof system
118. 133, 152, 170. 171, 176, 246, 247, 5, 7. 9, 14, 28-30, 34, 38, 39, 43, 44,
255, 259, 330, 346, 349, 410 46-48. 50. 51, 59, 60, 66, 69, 81-93, 95,
Rose 78, 130, 135, 140 97-100, 110, 111, 153, 155-161, 163,
Rouchdy 136. 138-140 166-168. 171, 172, 176, 177, 179, 182,
rushes 184, 186, 189-192, 194, 196, 198, 199,
59, 67, 78. 88, 89, 99, 100. 145, 151, 201, 202, 206-209, 217, 220, 221, 225,
152, 161, 192, 230, 231, 236, 237. 252, 227, 230. 231, 236, 244-247, 249, 256,
286, 312. 314, 315, 405, 425 257, 260, 269-271, 282. 290, 292, 302,
304, 307, 310, 316, 323, 332, 333, 372,
sabata 122, 150. 201 381, 411, 420. 421
sacaf en-nakhla 277 sZ2 string 182
saddle 232, 234, 236, 404
Salmon 7 tabby 35, 36, 38,45,411
samaar 151, 152, 285, 286, 405 Tackholm 66, 146-149, 151, 152
samaar morr 152, 286 Taharqa 125, 128
sa'ad 285 technique
sa'af 281 7, 15, 22, 27, 28, 30, 32, 35, 37-39, 41,
sedges 43, 46. 47. 49. 51, 53-55, 58, 59, 62-64,
66, 68-72. 81. 83-93. 95. 97, 99. 103.
Index 491
105, 110, 122, 145, 150, 152, 153, two planes

155-158, 160, 162-166, 168-172, 95, 97, 155, 156, 158, 160, 162, 167,
174-177. 180, 182, 184. 185, 189-192, 176, 177, 182, 186, 194, 203, 206, 207,
194, 195, 198. 201, 202, 204, 206, 207, 217, 225, 231, 236, 244
209-211, 217-220, 223, 227, 230. 233,
236, 245, 247-250, 252, 259, 262-265, unit
269, 290, 291, 323, 324, 328, 372, 27, 50, 104, 121, 131, 175, 177. 195,
390-392, 408, 412-415, 417, 420-422, 196, 198, 204. 207, 367
425
technology Valbelle 120
2, 3, 25-27, 29, 32, 38, 39, 44, 46, 49, veins
51, 53, 55, 56, 58, 66, 109, 152, 153, 91, 237, 275, 295, 296, 298, 332, 336,
249 341, 342, 345-347
Teeter 325 video
test 8 2-4, 20, 58, 69-71, 73, 74, 76, 78, 79.
three planes 101-107, 269-271, 296, 301, 303, 318,
95, 97, 156, 157, 161, 165, 167, 179, 331-337, 340, 341, 345, 347, 349-351,
189, 204. 206, 208, 243, 246, 420 354, 356, 358, 359, 362, 364, 366, 367,
Tilley 6, 8, 9, 20, 420, 423 369, 370. 372, 374, 376-378. 385-388,
time 390. 392. 405. 419, 421, 426
2-4, 6. 11. 15. 16, 20-23, 37, 43, 57, video timer 104, 369
66. 69-71. 73. 77. 78. 92, 99, 101-104, Vogt .... 26-29, 31, 32. 34, 38-41, 52
109-112, 120-123, 126-128, 131, 133,
135-137. 155. 172, 182, 209, 214, 215, Waddell 296, 334
219, 222, 236, 240. 249, 253, 259, 261, waling
264, 265, 269-276, 279, 285-287, 289, 45, 59, 60, 63, 65, 95, 97, 99. 100, 161,
291, 292, 295. 296, 298, 300, 303, 162, 169. 170. 232, 234, 236, 250, 323
308-312, 317-319, 323. 324, 326-328, Wallert 145, 273, 277
330-336, 339. 341, 344-346, 349, 351, Watson 152, 287, 293
352, 354, 356, 359, 362. 364, 366, 367, Weatherhead 121
370. 371, 373, 375, 377, 379, 380, 383, Wehr 273
385, 389-391. 393. 399, 401. 403-406, Wendrich
408-411, 419,423-425 1, 27, 58, 66, 72, 83, 165, 170, 182,
timer 2, 101, 102, 104, 367-369 191, 195, 209, 214, 253, 261, 301, 327
tomb painting 312, 313, 315 wheat straw
tradition 147, 150, 166, 221-223, 284, 304, 315
3, 5, 18, 38, 60, 62, 102, 125, 136, 216, Whitley 6
223. 227, 249, 250, 253, 260, 261, 263, Woolley
264, 307, 334, 391. 399, 403. 421, 423. 110, 114, 116-118, 120, 175, 177, 179,
425 180, 184, 186-188, 192, 195, 200, 203,
Trigger . . 1, 6, 21, 124, 125, 134, 135 204, 410
twill work position 371
38, 43, 45, 211-213, 296, 297, 326, 411 work songs 105
twining work space
33, 39. 41, 43-45, 47-49, 52, 59, 60, 95, 70, 104, 291, 292, 298, 316-318, 331,
97-99, 161, 162, 169, 170, 194-196, 349, 357. 365, 374. 388, 392, 406, 410,
198-201, 204. 231-236, 249-252, 254, 423
258, 264. 265, 270, 318-321, 323,
366-373. 404, 406, 413-416, 424, 425
working rhythm
70, 104-106. 331, 334, 339, 351, 353,
356, 359, 362, 367. 368. 375, 377, 379,
390. 391, 393,419.422,423
workmen's village
1, 110, 112, 115-117, 119, 120, 187,
254,256,409,410,412,425
work-eye contact
106, 107, 341, 345, 372, 390, 393
Worp 214, 279
wrapping
42, 43, 47, 49. 63, 64, 85, 86, 95, 97,
156, 159, 160, 164, 166, 169, 172, 177,
178, 182, 184-189, 217-222, 225, 245,
249, 250, 257, 258, 261, 263, 269, 270,
303. 304, 306, 307, 309-311, 320, 331,
333, 351, 352, 354-356, 358, 359,
362-364, 389, 392, 411,413
wrapping strand
85, 86, 95, 160, 166, 182, 184, 218,
221, 222, 245, 257, 258, 269, 270, 303,
304, 306, 307, 309, 331, 351, 352,
354-356. 358. 359. 362, 363
wrist
298, 300. 336, 341, 342, 346, 381, 382
Wylie 17
zabaa 277
zacabaa 277
zaghauwah
185, 277, 279. 280. 282, 350, 351, 354.
401
NEDERLANDSE SAMENVATTING
Hoewel manden tot de meest vergankelijke materialen behoren, worden ze, dankzij
de unieke omstandigheden in Egypte (een extreem droog klimaat) in opgravingen toch
aangetroffen, voomamelijk gemaakt van palmblad en gras.
Aan de hand van het materiaal van twee verschillende archeologische
vindplaatsen, Tell el-Amarna in Midden Egypte (rond 1350 voor Christus) en Qasr
Ibrim in Nubie (voornamelijk derde eeuw voor tot zesde eeuw na Christus) behandelt
dit proefschrift de productie van manden in het oude Egypte, daarbij gebruik makend
van de studie van hedendaagse mandenmakers in Midden Egypte en Nieuw Nubie
(het gebied dat de Nubiers werd toegewezen omdat hun land werd overstroomd door
de bouw van de Aswan High D a m ) .
Het textuele, het Audio-visuele en het Tactiele
Het verslag over oud-Egyptische en hedendaagse mandenmakers wordt gedaan door

middel van woorden. O m onze waarnemmgen in taal uit te drukken is er een mentale
ordening nodig. In dit proefschrift verandert de invalshoek van gedetailleerd
(beschrijving van de technieken) naar het brede spectrum van de wereld van de
mandenmakers zelf. D e "world according to basketry" is samengesteld uit textuele,
audio-visuele en tactiele aspecten. D e eerste twee komen aan bod in, respectievelijk,
het boek en de video-band. Ze verwijzen op een aantal manieren naar elkaar. In de
rechter benedenhoek van het video-beeld loopt een tijdsaanduiding mee, waarheen
regelmatig wordt verwezen in de tekst van dit proefschrift. Er komen ook
symbooltjes en aanwijzingen in beeld die via de Appendix C gevonden kunnen
worden.
D e tactiele aspecten zijn ook een belangrijk deel van het onderzoek, maar het
medium van boek of video-band is niet geschikt o m deze kennis door te geven.
Tactiele kennis wordt overgebracht door het delen en verwerven van ervaring
(leerlingwezen). In mijn onderzoek heb ik niet alleen gekeken naar hedendaagse
mandenmakers, maar ook les genomen. Deze ervaring kan niet via papier of beeld
gedupliceerd worden. "Ik zal het je laten zien" betekent in deze context namelijk "Ik
zal het je laten voelen". Ergens een 'gevoel' voor krijgen is tegelijkertijd concreet
(hoe ga je met de materialen o m ) en abstract (wanneer je tracht deze ervaring in tekst
te vertalen).
Een hulpmiddel o m het visuele of tactiele te vertalen naar tekst, is het gebruik
van numerieke gegevens. Een van de nineties van meten (tellen, wegen) is o m een
concrete situatie op papier te kunnen vastleggen. Het werkritme (visueel) wordt
daarom uitgedrukt in eenheden van tijd of beweging (hoofdstuk 16). O o k tactiele
aspecten, zoals de treksterkte van materialen, kan in meetbare eenheden
worden weergegeven. Door middel van ingewikkelde standaardiseringen via een

laboratorium opstelling kan worden uitgedrukt, wat een 'ervaringsdeskundige'
zonder nadenken in zijn of haar vingers heeft.
In het algemeen wordt een dergelijke vertaling van het tactiele naar het
textuele (cijfermatige) door academici hoger gewaardeerd dan praktische kennis.
Toch ontstaat er in de academische wereld langzamerhand het besef dat er andere
soorten kennis zijn, die wel degelijk waarde hebben.
Classificatie en Terminologie
D e eigenschappen van archeologische manden (ook grotendeels visueel en tactiel)
worden in woorden gepresenteerd met behulp van classificatie. In dit proefschrift
wordt beargumenteerd dat classificatie iets anders is dan het indelen van
voorwerpen volgens hun eigenschappen (hoofdstukken 2 en 3). Dat is namelijk
niets meer dan het opbergen van een voorwerpen in hokjes, een soort gedachteloze
archeologische opruimwoede. Voor een 'echte' classificatie wordt eerst bepaald
welke aspecten er van belang worden geacht voor een specifieke vraagstelling.
Deze aspecten worden vervolgens in een schema met elkaar verbonden waardoor
er een aantal op duidelijke criteria gebaseerde klassen ontstaan. D e voorwerpen
worden vervolgens in deze klassen geplaatst. In deze opzet blijven er klassen leeg
wanneer er zijn geen voorwerpen zijn die aan een bepaalde combinatie van criteria
voldoen. Door te verklaren waarom bepaalde klassen wel, anderen niet gevuld
zijn, wordt classificatie een krachtig middel o m tot nieuwe inzichten te komen.
In dit proefschrift wordt, door middel van classificatie, twintig verschillende
technieken onderscheiden (hoofdstuk 15). Het werk van de hedendaagse
mandenmakers wordt aan deze indeling gekoppeld, zodat er uitspraken kunnen
worden gedaan over het productieproces dat ten grondslag ligt aan de diverse
technieken.
O o k de video is het resultaat van een strenge ordening. D e beelden en
geluiden zijn door verschillende selectieprocessen gegaan. Ten eerste werd er
tijdens het filmen al bepaald wat werd opgenomen en hoe dat gebeurde.
Vervolgens werd deze keuze nog veel meer toegespitst tijdens de montage.
Beelden lijken objectiever dan tekst, maar vertellen evenzeer een zorgvuldig
geregisseerd verhaal.
Er zijn veel termen voor manden, die niet alleen functionele, maar ook
regionale en beroepsmatige verschillen uitdrukken. Terminologie is vaak
inconsequent, en bij nadere beschouwing blijkt dat de namen van manden gelieerd
zijn aan verschillende verzwegen classificaties. Het ontwerpen van een
consequente manden terminologie kan alleen op basis van een classificatie, die
eenduidige criteria geeft op basis waarvan w e verschillende soorten kunnen
onderscheiden.
Toch is daar in dit proefschrift van afgezien. Het ontwerpen van een
consistente terminologie is misschien wetenschappelijk gezien wel bevredigend (als
m e n tenminste van archeologische bureaucratie houdt), maar dient geen duidelijk
doel. Het mist de aansluiting met bestaande terminologieen en is bovendien
Nederlandse Samenvatting 495
afhankelijk van een bepaalde vraagstelling. In dit proefschrift zijn alleen de criteria
die ten grondslag liggen aan de twintig onderscheiden technieken expliciet
gemaakt, zonder dat daar een term op wordt geplakt. Het centrale begrip in de
classificatie is de basic structure, de techniek die gebruikt wordt voor het bouwen
van de b o d e m of wand van een mand, omdat dat deze structuur van de techniek bij
zelfs hele kleine fragmenten nog opgetekend kan worden.
D e classificatie is gebaseerd op een combinatie van binaire opposities (passief
/ actief), driedelingen (elementen die bewegen in een, twee or drie vlakken) en
nummertellingen (aantal systemen, aantal orientaties, aantal elementen). Dit is
maar een bepaalde manier o m het materiaal te ordenen, namelijk een westerse
academische. Hedendaagse Egyptische mandenmakers hanteren een hele andere
classificatie, die minder consequent is, maar gebaseerd op criteria die voor hen
relevant of functioned zijn. O m erachter te komen wat die criteria zijn (ze zijn
namelijk vaak onuitgesproken of onbewust) kunnen w e naar de hedendaagse
terminologie kijken.
Is het mogelijk o m de oud-Egyptische classificering van het materiaal te
achterhalen? Dat zou ons namelijk een blik gunnen in de denkwereld van de oude
Egyptenaren. Materiele cultuur is namelijk niet alleen een weerspiegeling van het
cognitieve systeem en de sociale gewoonten, maar spelen een actieve rol in het
formeren en structureren van deze gewoonten. Sterker nog, het oud-Egyptische
cognitieve systeem, dat gedeeltelijk gevonden kan worden in de mandenmakers
wereld, houdt direkt verband met het wereldbeeld. Hierbij moet direkt opgemerkt
worden dat er in een maatschappij een complex van meerdere (en vaak
tegenstrijdige) noties van ordening van de wereld bestaan.
Systemen, Activiteit and Handelingen

O m de zichtbare en tastbare aspecten van manden en mandenmakers handelingen
te beschrijven, worden de termen "systeem" en "activiteit" gebruikt. In hoofdstuk 5
wordt uitgelegd dat een systeem een groepje bladrepen of grassprieten is dat
dezelfde rol speelt in het tot standkomen van de techniek. Door de interactie van
verschillende systemen te beschrijven, begrijpen w e alle technieken volgens
dezelfde criteria.
Belangrijker is nog, dat w e door naar de activiteit van de systemen te kijken
(ze kunnen actief of passief zijn), een verband kunnen leggen met de handelingen
van de mandenmakers. In dit proefschrift wordt op twee plaatsen een definitie
gegeven van de activiteit, waarin een geleidelijke accentverschuiving van structuur
naar handeling duidelijk wordt. O p pagina 85 wordt gezegd dat de (rigide) passieve
elementen het geraamte van de techniek vormen, terwijl de (flexibele) actieve
elementen de structuur verbinden en bij elkaar houden. O p pagina 269 luidt de
definitie van actief en passief dat elementen van het actieve systeem vaker en
langer door de mandenmaker worden gehanteerd dan de elementen van het
passieve systeem. Beide definities zijn nuttig, omdat ze verschillende aspecten
benadrukken. De eerste kijkt naar de structuur van de mand als resultaat van een
proces, terwijl de tweede zich richt op de handelingen van de mandenmaker.
D e video sequenties worden uitgebreid in de tekst beschreven (hoofdstuk 16).
Een dergelijke gedetailleerde beschrijving helpt bij het observeren van het proces
en het bestuderen van de bewegingen van de mandenmaker. Hoewel elke
producent andere handelingen verricht, die samengevat kunnen worden in een
meer algemene beschrijving van het proces, brengt de gedetailleerde beschouwing
aspecten aan het licht die bij een veralgemenisering niet opvallen.
D e handelingen van de mandenmakers worden verdeeld in handelingen die
achter elkaar worden herhaald (repetitive actions) en handelingen die af en toe
nodig zijn (occurring actions).1 Een breiwerk, bijvoorbeeld. bestaat uit een
heleboel steken (repetitive actions) en af en toe moet de pen worden omgedraaid,
of een nieuw bolletje wol aangehecht (ocurring actions).
D e repetitive actions bepalen het gezicht van een productie proces, terwijl de
occurring actions ondersteunend zijn en veel minder opvallen. O o k de resultaten
van de occurring actions zijn vaak onzichtbaar: aan- and afhechtingen worden
bijvoorbeeld zoveel mogelijk verborgen in de structuur van een mand.
D e occurring actions nemen vaak weinig tijd in beslag, het zijn snelle
handelingen die moeilijker te volgen zijn dan de repetitive actions, die zo vaak
herhaald worden dat het voor de toeschouwer op den duur duidelijk is wat er
gebeurt. Alleen als een mandenmaker duidelijk toont of uitlegt wat hij precies
doet, kan de toeschouwer (leerling) de handeling begrijpen en imiteren. Vandaar
dat tradities of scholen geidentificeerd kunnen worden door bestudering van details
als de aan- en afhechtingen, het begin en de rand van de m a n d en de afwerking.
D e productie fases van het begin en de afwerking van de m a n d worden
gedomineerd door occurring actions. Het opbouwen van de b o d e m en wand (de
basic structure), daarentegen, wordt gekenmerkt door repetitive actions. D e
kenmerken die een mandenmakers traditie tonen nemen in het geheel van het
productieproces maar een geringe tijdsperiode in beslag. Het begin van een vlecht
neemt bijvoorbeeld maar 24 seconden in beslag.
D e bewegingen van de repetitive actions zijn zeer regelmatig, omdat het van
groot belang is een werkritme te ontwikkelen o m dezelfde beweging eindeloos te
kunnen herhalen. Toch wordt er ook in de occurring actions een werkritme
aangetroffen, als is dit moelijker te ontdekken. Het werkritme verschilt niet alleen
per techniek, maar zelfs per mandenmaker.
Kundige mandenmakers werken in een zeer regelmatig ritme en intense
concentratie. Dat wil niet zeggen dat hun bewegingen het meest economisch zijn.
Veel mandenmakers hebben bijvoorbeeld de gewoonte o m voor elke herhaalde
handeling even een klopje tegen de mand te geven. Dit heeft geen enkele functie
Aangezien de Nederlandse vertaling van deze termen ('herhaalde handelingen' en

'voorkomende handelingen') niet goed uitdrukt wat er bedoeld wordt, handhaaf ik in het
vervolg van deze pagina de engelse terminologie.
voor de productie, maar helpt in het volhouden van het werkritme. Veranderingen
in het werkritme treden op wanneer het productieproces een andere fase ingaat.
Tijdens de overgang naar een volgende fase laat de mandenmaker ook vaak zijn
ogen even wegdwalen van het werk. Het oogcontact is over het algemeen sterk,
niet alleen tijdens de occurring actions, maar ook tijdens de bijna automatische
bewegingen van de repetitive actions.
In oud-Egyptische grafreliefs en -schilderingen komen veel afbeeldingen voor
uit het dagelijks leven. Vaak wordt van een productieproces een soort
'samenvatting' gegeven, bestaande uit een of meerdere scenes die kennelijk als
kenmerkend werden beschouwd. Voor veel activiteiten, bijvoorbeeld de
graanoogst, vlasoogst, weven, bier brouwen, leerbewerking, worden
productiefasen afgebeeld die gekenmerkt worden door repetitive actions. Voor
mattenmakers scenes ligt de situatie een beetje anders. Een unieke scene uit het
Middenrijk toont een mattenwever die zich vooroverbuigt o m met een brede
houten balk de inslag aan te slaan (zie p. 313). Dit is een occurring action. D e
enige andere mattenmakers scenes zijn die van het maken van gepaarde matten.
Deze komen vanaf het Ouderijk regelmatig voor te midden van andere scenes uit
het dagelijks leven en zijn nogal raadselachtig. Ze lijken het begin en einde van het
gehele proces te tonen. Het begin wordt gekenmerkt door afbeeldingen van het
uitleggen van papyrus Stengels en soms het maken van touw. Afbeeldingen van het
eind van het proces tonen twee tegenover elkaar zittende mattenmakers, die de
voltooide mat bewerken met een onduidelijk werktuig. Mogelijk hebben de
mannen in beide handen een borsteltje, waarmee overtollige vezels worden
weggeveegd. In ieder geval representeert de afbeelding de laatste fase van de
productie.
Context en Betekenis
Context en betekenis kunnen beschouwd worden als tactile aspecten, in de zin dat
ze betrekking hebben op dingen die aangeraakt kunnen worden (materiele context)
of be-grepen moeten worden (betekenis). D e betekenis van manden en matten in de
Egyptische maatschappij is in feite de immateriele context. Dit is meer dan de
sociale status van de mandenmaker. Productie en gebruik van mandwerk is
gerelateerd aan groepsidentiteit en persoonlijke uitingsvormen.
H o e kunnen w e grip krijgen op de betekenis van mandwerk? Betekenis is
gewoonlijk geen bewuste constructie. Het kan niet gescheiden worden van het
object of andere dimensies van de productie (economisch, sociaal, ritueel,
magisch, politiek, etcetera).
D e Nubische mandenmakers zeggen dat hun kleurrijke versieringen alleen
een decoratieve functie hebben en verder niets betekenen. In zowel de interviews
die Boyce Driskell gehouden heeft, als mijn eigen onderzoek, bleek dat met name
de Kensi Nubische mandenmakers zichzelf beter vinden dan de vorige generatie,
omdat ze meer kleuren en nieuwe patronen gebruiken. Ondanks deze wil tot
innovatie, kan mandwerk van de verschillende Kensi dorpen onderscheiden
worden. Er is helemaal een groot onderscheid tussen de Kensi en Fadidja dorpen

in Nieuw Nubie. Het antwoord dat de decoraties niets betekenen is wellicht waar
op een bewust niveau, maar daaronder ligt een laag waarin via het mandwerk wel
degelijk betekenis wordt gegeven aan identiteit.
Er is een duidelijke gender verdeling tussen de personen die manden maken
en gebruiken. Dit is het sterkst aanwezig in Midden Egypte, waar de vrouwen de
manden produceren die ze zelf in huis en voor het doen van boodschappen
gebruiken. D e mannen maken de manden die op straat en in de velden gebruikt
worden. In Nieuw Nubie maken alleen vrouwen manden voor gebruik in huis en
voor boodschappen (vervoer naar huis). Het grove mandwerk voor het werk op de
velden, wordt gekocht van (mannelijke) Egyptische mandenmakers.
In Midden Egypt zijn er professionele, niet-professionele en part-time
professionele mandenmakers. Het blijkt dat hedendaagse professionale
mandenmakers een grote vaardigheid bezitten, maar zich daarenboven vooral
onderscheiden doordat ze een specifieke werkplek nodig hebben. D e
mandenmakers werken geconcentreerd, in een zeer vast ritme. Het resultaat van
hun werk vertoond een grote uniformiteit in uiterlijk en maten.
W e kunnen stellen dat de professionele mandenmakers vaardig zijn, maar dat
argument kan niet worden omgedraaid: een vaardige mandenmaker is niet per
definitie professioneel. Veel niet-professionele mandenmakers vertonen een grote
vaardigheid en kunnen evenzeer lange periode met grote concentratie aan het werk
zijn. Als w e in een opgraving een zeer regelmatig gevormd mandje vinden, kunnen
w e dus veronderstellen dat de producent had een grote vaardigheid, maar niet dat
het zijn of haar beroep was.
Professionaliteit is veeleer verbonden met de fysieke (materiele en
lichamelijke) vereisten van het werk. Als er een werkplek en instrumentarium
nodig zijn, is de producent waarschijnlijk professioneel. Er kan echter niet gesteld
worden dat zware, vermoeiende taken een relatie hebben tot professionaliteit: alle
mandtechnieken veronderstellen lange periodes van herhaalde handelingen.
Alle vaardige mandenmakers werken in een regelmatig ritme, waarbij opvalt
dat de niet-professionele mandenmakers langzamer en preciezer werken dan de
professionele mandenmakers. Voor de interpretatie van het archeologische
materiaal betekent dit dat consequente onregelmatigheden een duidelijkere
aanwijzing zijn voor professionaliteit dan een zeer regelmatig gemaakte mand.
O p dit moment zijn de enige full-time producenten de mattenwevers, de
mattenmaker die gepaard werk maakt en de producent van qafas (kratjes van
palmribben). Dezen hebben alien werkplaatsen en maken gemiddeld meer gebruik
van gereedschappen dan de niet-professionele mandenmakers. D e oud-Egyptische
afbeeldingen in grafreliefs en -schilderingen laten precies die technieken zien, die
tegenwoordig door professionele mattenmakers worden uitgevoerd. O o k
nettenmakers, touwslagers en textielwevers worden getoond. Afbeeldingen van de
productie van de gespiraalde huishoudmand k o m e n in het geheel niet voor. Dit is
opmerkelijk, omdat er per graf tientallen gespiraalde manden worden afgebeeld in
offerscenes. Dit kan wellicht gezien worden als een argument dat gespiraald
mandwerk niet door full-time professionals werd gemaakt, maar waarschijnlijk

door niet-professionele vrouwen.
D e sociale positie van de hedendaagse mandenmakers in Midden Egypte en
Nieuw Nubie verschilt. In Midden Egypte kijken de stadsmensen neer op de
boeren bevolking die ook manden maakt. Het is echt een dorpse activiteit. D e
vrouwen die manden maken zijn echter trots op hun vaardigheden. Zelfs de
vrouwen uit de rijkste familie van het dorp maakt manden (net zoals ze nog zelf
brood bakken en kaas maken). D e professionele mannen zijn minder gelukkig met
hun beroep. Ze proberen hun positie te verbeteren en als dat niet lukt, streven ze
er in ieder geval naar dat hun kinderen beter werk krijgen.
D e Nubische vrouwen, daarentegen, zijn zonder meer trots op hun werk. D e
oudere vrouwen klagen erover dat de jonge meiden het handwerk niet meer leren
en geven de televisie daarvan de schuld.
Over de sociale positie van mandenmakers in het oude Egypte is niet veel
bekend. Er is een tekst die aanwijzingen biedt dat de oud-Egyptische mattenwevers
er zowel sociaal als economisch slecht aan toe waren. Aangezien algemeen wordt
aangenomen dat dit een satyrische tekst is, kan deze tekst niet zonder meer als
getuigenis worden aangenomen.
Continuiteit en Verandering
In de hoofdstukken 10 en 11 wordt aangetoond dat er zowel sterke indicaties zijn
voor een grote continuiteit, als grote veranderingen. D e continuiteit is
voornamelijk regionaal: mandwerk opgegraven in Qasr Ibrim heeft veel
overeenkomsten met het moderne Nubische mandwerk, terwijl het materiaal uit het
arbeidersdorp van Tell el-Amarna duidelijke overeenkomsten vertoont met dat uit
het hedendaagse Midden Egypte.
Gespiraald mandwerk uit Amarna en Midden Egypte is ongedecoreerd en vrij
grof. Het Nubische mandwerk is daarentegen veel fijner en gedecoreerd met
patronen en gekleurde vlakken. Het gebruik verschilt ook. In Midden Egypte zijn
de manden vooral bedoeld o m brood, meel en andere goederen in te bewaren. In
Nubie worden de manden gebruikt als deksel, waaronder de maaltijden geserveerd
worden. N a afloop van de maaltijd worden de manden aan de muur gehangen en
hebben ze een decoratieve functie.
In het oud-Egyptische Amarna komt gepaard mandwerk het meeste voor. In
de Ptolemaische periode (derde eeuw voor Christus) werd een nieuwe techniek
geintroduceerd (manden gemaakt van spiraalsgewijze genaaide vlechten). Deze
techniek heeft het gepaarde mandwerk in Midden Egypte grotendeels vervangen.
In Qasr Ibrim en Nieuw Nubie wordt dit genaaide vlechten mandwerk gebruikt
voor draagtassen en vloermatten.
W a a r op het ene niveau sprake is van continuiteit, is er volgens andere
criteria juist sprake van een verschil in traditie. Dit brengt de waarde van het etno-
archeologische werk geen moment in gevaar. Etno-archeologie is bij uitstek niet
bedoeld o m oud met modern te vergelijken op basis van een veronderstelde
continuiteit. Ook de video is uitdrukkelijk niet bedoeld om een 'beeld te geven van
het leven in het oude Egypte'. Een dergelijk nadruk op een geconstrueerde
continuiteit ontkent de historische en maatschappelijke ontwikkelingen die plaats
hebben gevonden. Het etno-archeologische onderzoek concentreert zich
daarentegen op het begrip van het productieproces en is vooral van belang voor de
realisatie dat er vele immateriele aspecten kleven aan materiele cultuur.
M a n d w e r k kan beschreven worden als een combinatie van eigenschappen.
Als w e ons daar toe beperken doen w e het belangrijkste aspect schromelijk te kort.
Naast technische en functionele aspecten zijn namelijk de specifieke voorkeuren en
tradities van (sub)culturen van groot belang voor de interpratatie van materiele
cultuur. D e betekenis van ongedecoreerd mandwerk in het tegenwoordige Midden
Egypte is bijvoorbeeld neutraal. Deze functionele ongedecoreerde gespiraalde
manden k o m e n in elk huishouden voor en hebben een neutrale waardering. In
Kensi Nubie, daarentegen, worden ongedecoreerde manden als minderwaardig
beschouwd. Ze worden alien op onopvallende plaatsen gebruikt en nooit aan
bezoekers getoond. Dit staat in sterk contrast met de gedecoreerde manden, die
een ereplaats hebben aan de muur van de gastenkamer. D e tegenwoordige
felgekleurde decoraties doen het versierde mandwerk uit Qasr Ibrim verbleken.
Tradities zijn niet statisch, maar worden voortdurend opnieuw uitgevonden en
veranderen van betekenis.
D e interpretatie van het oude en moderne mandwerk kan uitgaan van een
keuze voor het benadrukken van de continuiteit of de verandering. Z o kan de
komst van het genaaide vlechten mandwerk in de Ptolemeische periode
geinterpreteerd worden als de culturele invloed van een buitenlandse bevolking, die
een breuk met de traditie heeft veroorzaakt. D e continuiteit van het
ongedecoreerde gespiraalde mandwerk in Midden Egypte kan als (politiek)
argument gebruikt worden dat de Egyptische boerenbevolking de waarlijke
erfgenamen van de farao's zijn. Beide interpretaties munten uit door eenzijdigheid.
Een maatschappij is niet monolithisch en de introductie van nieuwe ideeen wordt
vermengd met plaatselijke uitvindingen en aanpassingen aan de locale situatie.
Via 'the world according to basketry' kunnen w e een beeld krijgen van een
segment van de oud-Egyptische maatschappij die normaal gesproken buiten ons
bereik blijft. Dorpsvrouwen en ongeletterde boeren worden zelden gehoord in
officiele documenten. Het mandwerk geeft ons een ingang, al moet daar bij
aangetekend worden dat het een beperkt zicht biedt op de maatschappij als geheel.
Desalniettemin levert het een bijdrage aan de gezamenlijke inspanning van
archeologen, egyptologen, papyrologen en historici o m verhalen en beelden uit de
oudheid te constueren.
C U R R I C U L U M VITAE
Willemina Zwanida (Willeke) Wendrich

Haarlem, 13 September 1961
Education
1973 - 1979 Atheneum A, Rijksscholengemeenschap Texel (the
Netherlands)
1979 - 1981 Greek, Latin and Ancient History at the Theological Faculty
of the University of Amsterdam (the Netherlands)
1981 - 1988 Study History of Religions at the University of Amsterdam.
Doctoral exam ( M A ) on 26 August 1988, subjects: ancient
Egyptian religion, archaeology, Egyptian languague,
philosophy of science. Judicium: c u m laude
1989 - 1993 P h D research at the Centre of Non-Western Studies at Leiden
University (the Netherlands). Subject: the social context of
Ancient Egyptian basketry
Archaeological Experience
1987, '89, '92 Expedition to Tell el-'Amarna (excavations of the Egypt
Exploration Society, directed by B.J. K e m p ) .
1990, '92 Expedition to Qasr Ibrim, (excavations of the Egypt
Exploration Society, directed by dr.M.C. Horton).
1990-1993 Expedition to A b u Sha'ar (excavations of the University of
Delaware, U S A , directed by prof.dr.S.E. Sidebotham).
Professional Experience
1993 - present Co-director excavations at Berenike (Egyptian Red Seacoast).
1995 - 1998 Assistant professor of archaeology at Leiden University,
working at the Netherlands Institute for Archaeology and
Arabic Studies in Cairo (Egypt)
1996 - 1999 Project leader of the Eastern Desert Antiquities Protection
Project ( E D A P P ) , a project of the Royal Netherlands
Embassy in Cairo for the preservation of the cultural heritage
in the Eastern Desert of Egypt
1998 Director of the Baynun Mapping Project (Yemen)
STELLINGEN
behorend bij het proefschrift

"The World According to Basketry"
1. Het gebruik van videobeelden in wetenschappelijk werk is alleen

verantwoord wanneer zij vergezeld gaan van een expliciterende beschrijving,
zij spreken nooit voor zichzelf (dit proefschrift).
2. Hoewel ogenschijnlijk in tegenspraak, zijn de beweringen: "Egyptisch
m a n d w e r k kent een grote continuiteit" en "Egyptisch mandwerk kent grote
regionale en temporele verschillen" beide waar (dit proefschrift).
3. Bij het produceren van manden zijn het de telkens herhaalde basishandelingen
waarmee de wand wordt opgebouwd die resulteren in het algemene
kenmerkende uiterlijk van een bepaalde mandsoort, terwijl de aan- en
afhechtingen een weerspiegeling vormen van de lokale tradities (dit
proefschrift).
4. O m d a t terminologie alleen consistent kan zijn wanneer zij wordt gebaseerd
op classificatie en omdat classificatie afhankelijk is van een bepaalde
vraagstelling, is het ontwikkelen van een definitieve consistente terminologie
per definitie onmogelijk (dit proefschrift).
5. Het wordt tijd dat de Egyptologische en de archeologische wetenschappen het
oud-Egyptische vermogen van het moeiteloos accepteren en integreren van
een aantal parallelle werkelijkheden overnemen (dit proefschrift).
6. Het gevlochten mandwerk, gevonden in de "Cave of Letters" in Israel is niet
gemaakt van een palmblad dat rond een koord gevlochten is, maar, net als
het Egyptische materiaal, van een spiraalsgewijze genaaide vlecht (tegen
Y . Yadin 1963, 137).
7. Het bedrijven van archeologisch onderzoek en het geloof in de vooruitgang
van de mensheid is een vrijwel onmogelijke combinatie.
8. Het is mogelijk o m als archeoloog in Egypte te werken zonder vuile handen
te krijgen.
9. Kunst is zoeken naar de grenzen van het medium: het hoogst bereikbare in de
moderne dans is het dansen van stilstand, terwijl het hoogst bereikbare in de
beeldende kunst het vastleggen van beweging is.
10. Zoals aan heilige oorlogen zelden religieuze principes ten grondslag liggen,
zijn de principes die in geding zijn bij wetenschappelijke conflicten zelden
van wetenschappelijke aard.
11. D e term "verloren generatie" is volstrekt misplaatst voor een groep
onderzoekers die niet is ingelijfd bij en daardoor niet gehinderd wordt door
de vergaande bureaucratisering van het universitair bedrijf.
12. Het is zeer geruststellend voor een academicus o m in enige mate een
handwerk te beheersen.
O n the basis of two different archaeological sites, namely Tell el-'Amarna
in Middle Egypt (about 1 350 BC), and Qasr Ibrim in Nubia (mainly third
century BC to sixth century AD), this book deals with the production of bas-
ketry in ancient Egypt. Use is also made of a study of contemporary basket
makers in Middle Egypt and N e w Nubia.
The book first deals with the technical aspects of basketry production,
subsequently with the wider world of basket makers. The book is accom-
panied by a video tape which illustrates the findings presented in the writ-
ten presentation.
Willeke Wendrich is at present co-director of the excavations in Bereni

(Egyptian Red Sea coast). Between 1 987 and 1 993 she participated in the
excavations at Tell el-'Amarna, Qasr Ibrim and at Abu Sha'ar.
ISBN: 90-5789-035-6
iber Procyon lotoris

-1

The World According To Basketry An Ethno

Uploaded by

Copyright:

Available Formats

The World According To Basketry An Ethno

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The World According To Basketry An Ethno

Uploaded by

Copyright:

Available Formats

Ail

An Ethno-archaeological Interpretation of Basketry Production

ter verkrijging van

Co-promotor: Dr. P. van dcVclde

Referent: Dr. L. Bender J0rgensen

Vitenskapsmuscct, Trondhcim, Noorwegcn

Ovcrige leden: Prof.Dr. C.C. Bakcls

Prof.Dr. H.D. Schneider

Dr. R. van Walscm

John Irving, 1978 p. 176

Co-promotor: Dr. P. van deVelde

Referent: Dr. L. Bender J0rgensen

Vitenskapsmusect, Trondhcim, Noorwcgcn

Overigc leden: Prof.Dr. C.C. Bakcls

Prof.Dr. H.D. Schneider

Dr. R. van Walscm

John Irving, 1978 p. 176

Preface and Acknowledgements

Chapter 3 O n the classification of basketry

Chapter 4 Field Work Methods

Chapter 5 Basketry Analysis

Chapter 6 And Action

Chapter 7 Sources along the River Nile

Chapter 8 Leaves and Branches

Chapter 9 Basketry Techniques, a Classification

Chapter 10 Basketry from Amarna and Middle Egypt 175

10.2.2 Passive and one active system, two directions 190

Chapter 11 Basketry from Qasr Ibrim and New Nubia

Chapter 12 The Status of the Static

Chapter 13 The Introduction of Time 269

Chapter 14 Climbing the Tree

Chapter 15 Making Mats, Bags and Baskets

Chapter 16 The Choreography of Work

16.2 The Sabat

Chapter 17 The Dance of the Dynamics

Chapter 18 Producers at Present

Chapter 19 Producers in the Past

Chapter 20 Summary and Conclusion

Appendix A Fibre Identification Criteria 429

Methods and Materials

The world according to basketry is an introduction to basketry analysis, focusing

'Basketry is defined as a class of artefacts made out of vegetable fibres of limited

remarks are made in the conclusion in respect to the meaning of basketry in

THE FABRIC OF ARCHAEOLOGY

polysemous, located along open systems of signified-signifiers" (Shanks and Tilley

2.2 Pigeons and Pots

Descriptions and designations are based on classification and consequently

relevant dimensions and attributes is inspired by k n o w n objects, and specific

cylindrical globular flat Form

Figure 2-2 A hierarchical classification of objects found in a hypothetical kitchen area

When a white hand made pot is found, however, this classification is

that those terminologies are usually incompatible, in spite of their internal

2.3 Comparing Apples and Oranges?

not an objective exercise, because it involves judgement and selection. Making

Ethno-archaeology focuses on processes instead of artefacts. It brings the

impressionistic rendering of ancient basket makers which is a possible

2.3.3 Meaning and Metaphor

Technological processes are used metaphorically also. The creation and

Lucas 1926'5' (English)