Production and Domination

Afghanistan, 1747-1901

Ashraf Ghani

Submitted in partial fulfillment

of the requirements for

the degree of Doctor of Philosophy

in the Graduate School of Arts and Sciences

Columbia University

1982
 @ 1984

ASHRAF GHANI

ALL RIGHTS RESERVED

 ABSTRACT

PRODUCTION AND DOMINATION AFGHANISTAN

1747-1901

My objective in this work is to provide an anthopo1og-

ica1 analysis of the historical processes through which

structures of production and domination were reproduced

as a totality from 1747 to 1901 in the present territory

of Afghanistan. To bring the changing articulation of

structures in the total system into sharper relief I have

opted for a combination of temporal and spatial frameworks.

The period from 1747 to 1901 forms the unit of long

duration for the first part of the study. Within that

period, depending on the changes in one or several of the

elements within a structure, I have chosen an appropriate

conjunctura1 unit - such as 1840-1901 for the changes in

the system of inter-territorial communication - to high-

light the relevant changes. In part two f I deal with two

main units of long-durat~~n, roughly 1500-1722 and 1747-

1901. The choice of the conjunctura1 units within each

of these periods is, again, determined by changes in the

systematic relations among elements composing the struc-

ture. Thus, for changes in the patterns of long-distance

trade, the years 1500-1622 are demarcated as being signi-

ficantly different from the following one hundred years.

The second unit of long-duration, being the main focus of

analysis and an era of rapid restructuring and destructur-

ing, is broken into four conjunctural units: 1747-1818;

1826-1839; 1843-1878; and 1880-1901.

Spatially, instead of taking the whole territory of

the state as the unit of analysis I have focused on six

regions. Within each of these regions I distinguish be-

tween central and marginal ecological zones and then ex-

plore the interrelations among these elements within re-

gional, inter-regional and inter-territorial units.

In the first part of the study I provide an analysis

of geography and ecology; labor-process and technical or-

ganization of production in agriculture; social organiza-

tion of production; and circulation, transportation, and

markets. In the second part I first provide an histori-

cal overview of the social relations and then analyze the

formation and reproduction of the Afghan state. I:n each

chapter the relevant theoretical literature is discussed

and in the conclusion the interrelation between structures

of production and domination is viewed from a comparative

perspective.
 Table of Contents

Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

Dedication. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. iii

Preface................................. ............. iv
PART I

Chapter I:
Geography and Ecology: The Historical Setting ••••••• 1

Chapter II:
Labor Process and Technical Organization of
Production in Agriculture ..............•.....•..•... 32

Chapter III:
Social Organization of Production •.•••••••••••••.••• 82

Chapter IV:
Circulation, Transportation, and Markets ..•••.•••••• 158

PART II

Chapter V:
Social Relations - An Historical Overview •••..•••.•• 226

Chapter VI:
The Afghan State: Formation and Reproduction .••••.•• 292

PART III:

Chapter VII:
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427

i
 ACKNOVlLE DGEMEN'l'S

As my research on this work has taken almost a decade

my debts to individuals and institutions are numerous, but
I will only name a few. I am grateful to the members of

my committee: Professors C. Arensberg, R. Bulliet, M. Cohen,

E. Embree, M. Fried, R. Murphy and J. Vincent who were all
generous with their time and patient with my style of
thinking and writing. I am also deeply indebted to
Professors Marvin Harris, Richard Lee, Jane and Peter
Schneider, Gerald Sider and other members of CUNY seminars
for intellectual stimulation. Many of my ideas were ham-
mered in discussion with my fellow students at Columbia.
I will only name A. Barstow, L. Brock, G. Mescher, B.
Miller, and D. Nugent as a sample.

The straint imposed on Rula Saade Ghani since 1975

has been continuous. I am grateful to her for not com-
plaining more than she did. My daughter Mariam was more
cheerful than she could have been. I cannot thank Susan

Mescher adequately for typing the final version of this

work.

ii
To the memory of Fouad Saade

iii
 PREFACE

In his discussion of history and dialectics, Claude

Levi-Strauss argues that IIscientific explanation consists

not in moving from complex to the simple but in the re-

placement of a less intelligible complexity by one which

is more SOli (1962/1966:248). Yet, in a moment of frustra-

tion, the eminent anthropologist declared that "for me as

a European, and because I am a European, Mohammed inter-

venes with uncouth clumsiness, between our thought and

Indian doctrines that are very close to it, in such a way

as to prevent East and West joining hands, as they might

well have done, in harmonious collaboration ••• The two

worlds are closer to each other than either is to the Mos-

lem anachronism. Rational evolution would have been the

converse of what actually occured historically; Islam cut

a more civilized world in two. What appears modern to it

belongs to a bygone age; it is living with a time lag of

a thousand years" (1955/1974:466-467).

I have no desire to examine the motives that led

Levi-Strauss to his proclamation. At issue are his use of

the notion of structure and the relation between anthro-

pology and history. Social structure, in his view, has

IInothing to do with empirical reality but with models which

are built up after it" (1963:279). Social relations pro-

vide the raw material out of which the model is constructed

iv
but the social structure can not be reduced "to the ensem-

ble of the social relations to be described in a given so-

ciety" (Ibid). A model must meet several requirements.

Exhibiting the characteristics of a system, the elements

in the structure are related in such a way that changes in

one effects changes in others and, for any given model,

there should be a possibility of ordering a series of trans-

formations resulting in a group of models of the same type.

These properties should make it possible to predict the re-

actions of the models if one or several elements are sub-

mitted to change (Ibid). Lastly, it should be possible to

"experiment" on models (Ibid:280).

Experimentation is closely linked to observation. He

affirms that "on the observational level, the main -- one

could almost say the only -- rule is that all the facts

should be carefully observerd and described, without allow-

ing any theoretical preconception to decide whether some

are more important than others, This rule implies, in turn,

that facts should be studied in relation to themselves (by

what kind of concrete process did they come into being?)

and in relation to the whole (always aiming to relate each

modification which can be observed in a sector to the glob-

al situation in which it first appeared)" (Ibid).

From the last observation, it might be inferred that

time should constitute an important element of the defini-

tion of structure, but this is not so. Levi-Strauss in-

v
sists that "anthropology uses a 'mechanical' time, rever-
sible and non-cumulative .•• On the contrary, historical
time is 'statistical'; it always appears as an oriented and
non-reversible process" (Ibid:286).
Fernand Braudel, in an essay on history and social
sciences, states that "for historians, a structure certain-
ly means something that holds together or something that
is architectural; but beyond that it means a reality which
can distort the effects of time, changing its scope and
speed. Certain structures live on for so long that they
become stable elements for an indefinite number of genera-
tions: they encumber history, they impede and thus control
its flow. Others crumble away faster. But all operate si-
multaneously as a support and an obstacle" (1958/1972:17-
18). Responding specifically to Levi-Strauss's notion of
non-temporal structure, he writes that "the essential thing
is to define the role and limits of the model (which cer-
tain ventures run the risk of over-extending), before set-
ting up a joint programme for the social sciences. Hence
the need to confront models, too, with the idea of duration;
for their depth of meaning and explanatory value depend, in
my view, largely on the duration they imply" (Ibid:27).
Furthermore, Levi-Strauss's emphasis on the non-reversibil-
ity of historical time misses the essential point that, in
fact, it is history that provides the greatest examples of
transformation as well as breakdown of social structures.

vi
 My acceptance of Braudel's notion of duration as the

defining criterion of structure has determined the order

of presentation in this work. By focusing on movements of

concomitant structures, I have attempted to isolate the

systemic relations among the changing or non-changing ele-

ments that combine to form a structure, as well as the ar-

ticulation of different structures in a total system of

reproduction. The period from 1747 to 1901 forms the unit

of long duration for the first part of the study. Within

that period, depending on the changes in one or several

elements within a structure, an appropriate conjunctural

unit -- such as 1840-1901 for changes in the system of in-

ternational communications is chosen to highlight the

relevant changes. In part two, I deal with two main units

of long duration, roughly 1500-1722 and 1747-1901. The

choice oftheconjunctural units within each of these per-

iods is, again, determined by changes in the systemic re-

lations among the elements composing the structure. Thus,

for changes in the patterns of long-distance trade, the

years 1500-1622 are demarcated as being significantly dif-

ferent than the following one hundred years. The second

unit of long duration, being the main focus of analysis

and an era of rapid restructur~ng and destructuring, is

broken into four conjectural units: 1747-1818; 1826-1839;

1843-1878; and 1880-1901. Avoidance of years, such as

1818-1826, 1839-1843, and 1878-1880, is deliberate. These

vii
periods were times of intense violence and, although I dis-

cuss their impact on the process of reproduction of the

total system, greater concentration on the events would not

have enriched the analysis

In general, structures analyzed in part one show great-

er stability through the period 1747-1901 than those exa-

mined in part two. This is one of the reasons for the or-

der of presentation. Had the changes been equally rapid

in the two sets of structures, I would have opted for a

different order of presentation.

But the method of presentation also aims at building

an increasingly more complex model of the interrelationship

between structures. In treating relations of production

and domination as relations between people, I have followed

Karl Marx who wrote that "the exchange-value of things is

a mere expression, a specific social form, of the produc-

tive activity of men, something entirely different from

things and their use as things" (1863/1971:181). Levi-

Strauss, in a comment on a passage from Marx on the issue,

asserts that "as soon as the various aspects of social

life -- economic, linguistic, etc -- are expressed as re-

lationships, anthropology will become a general theory of

relationships. Then it will be possible to analyze so-

cieties in terms of the differential features characteris-

tic of the systems of relationships which define them"

(1963:95-96). In another essay, he elaborates on this no-

viii
tion by stating "anthropology considers the whole social

fabric as a network of different types of orders" (Ibid:

312). He contrasts orders such as kinship, socialorgani-

zation, economic relations, etc, defined as "lived-in" or-

ders that can be studied from outside, to those orders de-

fined as "thought-of" orders that can be analysed only in

relation to "lived-in" orders. "The 'thought-of' orders are

those of myth and religion. The question may be raised

whether, in our own society, political ideology does not

belong to the same category" (Ibid:313).

Two important questions are raised in these remarks:

the degree to which "horne-made" models, constructed as

explanations of the "thought-of" orders, are helpful in

the grasping of a system by an analyst, and the degree to

which the categories employed in the analysis by an anthro-

pological observer are the "home-made" models of another

culture. Levi-Strauss argues that the study of "home-made"

models is indispensable (Ibid:282), but he does not reflect

on the imposition of cultural biases in the guise of anthro-

pological analysis. His Eurocentric view of Islam, des-

cribing it as a "barrack religion" that has "no economic

or social character" (1955/1974:460), should serve as a

warning on the adequacy of a structuralism that refuses to

confront its constructs against the test of duration. Levi-

Strauss's remarks on Islam are an extreme case and their

relevance may be considered dubious in the assessment of a

ix
"non-temporal" structuralism. The more formalistic exer-
cises of his structural-marxist followers, however, reveals
the sa1"ne inadequacy. Having inherited the categories of
infra-structure and super-structure, these analysts (e.g.,
Friedman,J.1975:46-63iGodelier,M.1977:3-l3,13-29) through
a series of mental gymnastics, attempt to demonstrate that,
in one synchronic instance, it is religion that forms the
infrastructure, in another, kinship, and yet in another,
politics.
Edward P. Thompson's observations on Louis Althusser's
work describe, in fact, the methodological orientation of
the whole field of structural-marxism. He notes that, be-
fore analysis starts, Althusser's categories have "already
been de-socialized and de-historicized ••• we are off.ered an
arbitrary selection of categories -- as 'economics', 'poli-
tics', 'ideology' - and neither the principle of selection
nor the categories themselves are examined •.• If we scarce-
ly hear about the state or about class, we can not expect
to hear about particular state formations or about which
classes or about alternative and conflicting beliefs with-
in 'ideology'. The talismanic concepts are 'relative au-
tonomy' and 'in the last instance determination'" (1978:
287). In a passage which comes close to Braudel's notion
of structure, Thompson insists on "defining 'determine'
in its sense of 'setting limits' and of 'exerting pressures'
and of defining 'law of motion' as 'logic of process'"

x
(1978:351) .

The object of this work has been to provide an analy-

sis that mediates between theoretical discourse and de-
tailed historical investigation. It is an ethnography.
The endeavor, in more than one sense, has involved a
journey to the origins. A large body of current notions
in the social sciences, notwithstanding the unawareness of
some practitioners, have been inherited from the Classical
Pol tical Economy and Jurisprudence of the eighteenth and
nineteenth centuries. I have tried to make the nexus ex-
plicit and, although I have been considerably influenced
by the works of Marx, I have treated his notions with the
same degree of critical scrutiny as those of any other in-
fluential thinker in the social sciences. I find his treat-
ment of the labor-process inspiring, but consider his occa-
sional pronouncements on relations of domination not to
form a coherent theory and his notions of Asiatic mode of
production and Asiatic society as relevant as Levi-Strauss's
portrayal of Islam. The works which have conditioned the
direction of the inquiry have been discussed at the begin-
ing of each chapter.
As an analysis of a society in which I grew up and
spent most of my life, this analysis has involved yet an-
other journey to the origins. At the outset, I must make
it clear that, while I consider the grasping of the "home-
made" model of a culture to be an important part of anthro-

xi
pological theory, I do not think that the participant in a

culture is in any privileged position to provide an account

of either the genesis or the contradictory process through

which the system is reproduced. The merits of this study

if any -- are based on the relevant archival research and

fieldwork as well as the comparative framework which has

guided the inquiry.

The archival research, on which this study is based,

began with reading through the extensive collection of

British official documents on the Middle East, Central and

South Asia that were available in the library of the Ameri-

can University of Beirut, where I acquired my B.A. and M.A.

degrees. Between 1973 and 1977, I conducted archival and

field work in Afghanistan. Most of the historically rele-

vant material in the Afghan state archives had been, in

the course of the 1929 civil war, stolen or destroyed. In

addition to the National Archives of Afghanistan, I have

had, therefore, to rely on the available private collections.

My most helpful instructors in learning about and acquiring

access to the relevant manuscripts and printed works of the

period were the booksellers in the old section of the Kabul

bazar. It was through these individuals that I acquired

the valuable set of archives from the valley of Kunar which

prov:;.des the greatest amount of details on the interrela-

tions of state and society in an area fairly remote from a

central place. The most valuable printed work on the per-

xii
iod is the three-volume account of Mulla Faiz Muhammad en-

titled Saraj ul-Tawarikh (The Lantern of History) written

on the orders of Amir Habibullah (1901-1911) and covering

the period from 1749 to 1896. The third volume -- 862

pages of print -- is entirely devoted to the reign of Amir

Abdur Rahman (1880-1901) and contains an enormous amount

of information on the implementation of state policies in

every region of the country as well as the resistance to

these measures. The printed edition is arranged in chrono-

logical order, but the original manuscript from which the

author worked, and which only recently became available to

me, discusses the events in each major region separately.

The author had been given access to all state archives and,

as such, his work can be viewed as a compilation rather

than strict history.

Next to archives and manuscripts, the most valuable

source of information on the period is the collection of

contemporary autobiographical, biographical, and historical

works, that have been published by the Historical Society

of Afghanistan, either as pamphlets or as series in Aryana

and Afghanistan, the two journals published by the society.

Of works published outside Afghanistan, biographical, auto-

biographical, historical works, and collection of original

documents published in India, Iran, Pakistan, and the So-

viet Union, contain valuable information on all aspects 'of

life in various regions of Afghanistan. I have systemati-

xiii
cally read most of the material available in libraries in

Afghanistan, Pakistan, Great Britain, and the United States,

and although only a limited number are referred to in the

text, my generalizations have been based on information con-

tained in the larger set.

I have had to exercise the same degree of selection in

the use of accounts of contemporary travellers and British

archival material. The former set has been invaluable in

the determining of commonly travelled routes in each con-

juncture, and in revealing the identity of groups that were

able to impose costs of protection on the roads. Malcolm

Yapp's essay entitled "The India Office Records As a Source

For the Economic History of the Middle East" (1970:501-513),

has been a very useful guide on the type of material avail-

able on different periods of Afghanistan. The collection

in the American University of Beirut as well as the collec-

tion in London were consulted between 1979 and 1982. Host

of the available information was digested in a series of

Gazetteers. Yapp claims that the "Gazetteer is probably

the most important single work on Afghanistan for the per-

iod before 1914, and in many ways comparable to the great

Description de l'Egyptf" as a source" (Ibid:508). I have

consulted the original editions which were issued between

1908 and 1914 but, for the volumes already printed, I have

quoted from the recent edition prepared by Ludwig Adamec.

Despite the value attached to the Gazetteer by Yapp,

xiv
its use entails two problems. Editors, including the re-
cent one, often do not give the precise point in time in
which the information was collected and, for regions which
did not fall under British occupation during the two Anglo-
Afghan wars, the accuracy of the information has to be
checked against Afghan sources. Indeed, during the first
two decades of the twentieth century, the British compilers
of Gazetteers on Afghanistan and India, two sets of which
were issued in 1885 and 1907-1908 containing invaluable
information on the relations between regions in Afghanistan
and neighboring regions under British occupation, were still
quoting the figures on prices and wages that were provided
by Elphinstone which are an indispensable source for the
eighteenth century. In order to overcome these obstacles,
I have, in as many instances as possible, tried to consult
the original copies of the reports or determined the period
through other means. The Gazetteers are most useful for
the period from 1826-1878 and, with some exceptions, rather
inadequate for periods either before or after these dates.
The two general works in English that cover the whole
length or a major section of my unit of long-duration are
both based on very inadequate research. Louis Dupree's
(1973/1980) work is largely based on secondary sources in
English and the utility of his work is largely confined
to his discussion of geography and archeology of the coun-
try. Vartan Gregorian (1969), whose work is far more ser-

xv
ious in intention and who attempted to use sources in Per-

sian, was unfortunately not given help in locating even

the published sources -- such as the third volume of Saraj

ul-Tawarikh. Consequently, his generalizations on internal

developments are usually unfounded. He, however, provides

a useful survey of the foreign relations of the state. Both

of these authors have largely based their works on the pub-

lished material on Afghanistan but have neglected most of

the relevant data contained in the British archives. A

symptomatic sample of the unwarrantable trust in some secon-

dary material is provided by Gregorian's assertion regarding

wages of laborers in Afghanistan. He quotes the relevant

figures from the 1908 edition of the Imperial Gazetteer of

India thinking that they applied to the year in question.

The same figures had been quoted in the 1885 edition and,

in fact, had not been updated since the time of Elphinstone

(1839,vol.I:391).

The only historical work covering the entire length

of the period of long-duration for the northern region is

Robert D. McChesney's unpublished Ph.D. dissertation on

waqf at Balkh. He has consulted all the relevant litera-

ture available at the time of his writing and his work is

a very serious attempt at providing a dynamic interpreta-

tion of one of the most important institutions of Muslim

societies.

Hasan Kakar's two volumes (1971,1979) on the reign of

xvi
Amir Abdur Rahman, (1880-1901), are based on an extensive

survey of the relevant Afghan and British sources. The

utility of the bocks is, however, confined to the de scrip-

tion of the relevant events and institutions, facts and

figures. The most useful part of the work for my purpose

has been an appendix in which he compiled the data contained

in British sources on the prices of commodities in various

regions of the country.

Value of historical works published in Afghanistan

mainly lies in their utility as source material and as docu-

ments of the ideological tendencies prevailing in any par-

ticular period. Thus, Fofalzai (1958,1976) who reproduced

an extremely useful collection of documents for the period

1747-1818 from his family archives, attempts in his own

writings to imitate the style of court scribes of the period.

The only work which aims at providing a systematic survey

of the social forces in society and the state is Mir Ghulam

Muhammad Ghubar's book entitled Afghanistan Dar Masir-i-

Tarikh (Afghanistan in the Course of History, 1967) and is

an attempt to provide an interpretation of the history -~

VL

the area from mythical times to the rebellion of 1929. Un-

til the coup of 1978, the book w~s considered prejudicial

to the interests of the royal lineage and officially banned

in Afghanistan. Yet, it was probably the single most widely

read book in the country. My initial interest in the his-

tory of Afghanistan was mainly fostered by this work, but

xvii
its relevance to the present project has been marginal.

Having realized the meagerness of secondary historio-

graphical literature on the period, I soon was forced to

gather the relevant primary material. It is research on

those sources and my own fieldwork that have provided the

empirical foundations of the present study. The impetus

for fieldwork stemmed from disenchantment with the fashion-

able literature in political science which I found totally

inadequate for the understanding of Afghanistan, and from

the theoretical stimulation provided by works of Perry An-

derson on the absolutist state (1974), Marc Bloch on feudal

society (1971), Barrington Moore on social origins of dic-

tatorship and democracy (1966) and Eric Wolf on peasantry

in general and peasant wars in particular (1969). The

attractive feature of these works was their attempt to link

the processual changes in macro-structures of society with

the system of relations and institutions prevailing in peas-

ant society. I was, therefore, gradually led to formulate

a project on the investigation of the total system of re-

production of state and society. My choice of variables

to be investigated was determined by Max Weber's definition

of pre-conditions of a politically centralized state. In

his view, the pre-conditions are: n(l) monopolization of

the means of domination and administration based on: (a)

the creation of a centrally directed and permanent system

of taxation; (b) the creation of a centrally directed and

xviii
and permanent military force in the hands of a central,

governmental authority: (2) monopolization of legal enact-

ments and the legitimate use of force by the central au-

thority; and (3) the organization of a rationally oriented

officialdom, whose exercise of administrative functions is

dependent upon the central authority" (Bendix,R.1962:383).

Weber was offering a list of the formal characteris-

tics of the absolutist state. In his account on the genesis

of that structure, Perry Anderson has singled out the threat

of peasant unrest and the pressure of mercantile or manufac-

turing capital within western Europe (1974:23-24). He also

points out that the new state apparatus was capable of

"breaking or disciplining individuals and groups within the

nobility itself. The arrival of Absolutism was ••• never a

smooth evolutionary process for the dominant class itself:

it was marked by extremely sharp ruptures and conflicts

within the feudal aristocracy to whose collective interests

it ultimately ministered. At the same time, the objective

complement of the political concentration of power at the

height of the social order, in a centralized monarchy, was

the economic consolidation of the unit of feudal property

beneath it" (Ibid:20).

In the conclusion of his comparative study, Barrington

Moore argues that although there was no universal connection

between the replacement of subsistence farming with produc-

tion for the market and the creation of a strong central

xix
government, it was, nonetheless, the combination of the two

processes that "has yielded modernization in various parts

of the world since the fifteenth century" (1966:468). He

points out that, in these changes, three aspects were politi-

cally important: "The character of the link between the

peasant community and the overlord, property and class divi-

sions within the peasantry, and the degree of solidarity,

or cohesiveness displayed by the peasant community" (Ibid).

The state/society is the effective analytic unit for

both Anderson and Moore. Wolf, on the other hand, focuses

on the disruptive role of the world market in areas of pe~

ripheral capitalism and links the differences in response

to these changes to the eco-systems in which peasants worked.

Three aspects of my fieldwork (1973-1977) have been

relevant to the purposes of this work. Through archival re-

search and interviews, I managed to gather the relevant ma-

terial affecting the military, political, and fiscal organ-

ization of the state. Secondly, focusing on these institu-

tions, I soon discovered that the state did not monopolise

the control of these functions; other institutions in the

society had considerable autonomy in these spheres. The

religious stratum and the clans were the most important of

these groups. Thirdly, attempts at determining production

for markets led me to consider patterns of cooperation and

conflict engendered in a system of agriculture dependent

on irrigation. It was here that Robert Murphy's (1971)

xx
emphasis on Georg Simmells interpretation of conflict as

a social relationship and the importance of ambiguity in

social relations, was useful.

By 1977, I had managed to amass a large body of his-

torical and modern material on different aspects of social

life in Afghanistan. Although I have travelled in all re-

gions of the country and gathered the initial data for more

intensive subsequent examination, most of my fieldwork was

concentrated on the eastern region. Instead of lessening

my analytic confusion, the amassing of data only managed

to increase it. It finally dawned on me that the problem

was a theoreticnl on; and I was, again, in need of a com-

parative framewo:ck to make sense of a massive body of facts

that could not be related to each other systematically.

Between 1977 and 1982, most of my readings have been

on other areas of the world and, although limitations of

time and space have prevented me from making all the com-

parative links explicit, in all relevant junctures the ar-

gument has been made possiblf~ by a comparative concern. As

I have discussed the relevar't theorE"tical material in the

context of each chapter, I do not need to dwell on those

points here. Yet, before closing this preface, I must

make explicit my choice of presenting the material through

a combination of multiple spatial and time frameworks.

In 1936, ChIao Ting Chi presented an analysis of the

political power in China that focused on regional relations

xxi
(1936/1970:2). He argued that the "unity or centralization

of state power in China could only mean the control of an

economic area where agricultural productivity and facilities

of transport would make possible the supply of a grain tri-

bute so predominantly superior to that of other areas thCl.t

any group which controlled this area had the key to the

conquest and unity of all China" (Ibid:5). Chi defended

such an area as a key economic area. Moving from the des-

cription of key economic areas in the course of Chinese his-

tory to the character of the Chinese state, he wrote that

"a state of this nature is entirely different from the modern

state. The looseness of its internal organization and self-

sufficient character of its regional divisions greatly mag-

nified the importance and difficulty of the problem of re-

gional relationships and the vital necessity of a Key Eco-

nomic Area as a material basis for unity. It is a state

which properly regards its public works for control as a

weapon, and its policies are consciously or unconsciously

guided •.. for strengthening its Key Economic Area" (Ibid:

149) .

Peter Schran, in his recent reassessment of inland

cornrnuni,:ations in Ch'ing China, argues that "wherever pos-

sible, movement by junk was the cheapest form of transpor-

tation by far, followed by the cart, various pack animals,

the wheelbarrow, and the human carrier" (1978:37-38). Wa-

ter transport had yet another advantage. The boats men-

xxii
tioned ranged in size from those carrying 16 cwts to "large

junks of 40-50 or even 70 tons dead weight, which could be

found on the major arteries" (Ibid:37). Schran notes that

there seems to have "been "very limited involvement of the

Ch'ing government in matters of transportation" (Ibid:41),

and affirms that sections of the gigantic waterway known

as the Grand Canal were used by different dynaties to move

the surplus of grain from the Yangtze region to their capi-

tals (Ibid:43).

As the data in chapter four will make clear, the con-

trast with Afghanistan could not have been greater. My

choice of regions has been conditioned by the fact that move-

ment of grains to the central place of each region, in years

of ordinary harvest, is far less costly than to central

place(s) in other region(s). I have argued that the struc-

ture of transportation formed one of the strongest obstacles

to the emergence of centralized states that were found on

a statewide economy of grain production. If one can accept

Chi's description of Chinese state as a loose organization

when such enormous quantities of grain were carried for such

long distances, the degree of looseness of polities in the

territory that became Afghanistan can hardly be contested.

Yet, the looseness of the state, as I have tried to demon-

strate, did not mean a corresponding weakness of cultural

institutions that mediated the relations among the regions

of the country. Nor did the constraints imposed by the re-

xxiii
gional geography and ecology of the area prevent the rulers

of states from attempting to forge more oentralized insti-

tutional means of government. It should be cle~r that these

states displayed different characteristics from those of

absolutist and post-absolutist states of Europe. It might

have been simpler to refer to them as empires but, as there

is no general agreement on the usage of that term, I have

used the term state consistently hoping that my descriptions

will make the difference clear.

xxiv
PART I
 1

CHAPTER ONE

Geography and Ecology: The Historical Setting

Every inquiry in historical anthropology must, at its

very inception, corne to grips with the challenge of dura-

tion. The choice is between uniform and multiform consid-

erations of time. An eminent historian associates the first

modality with the approach of social scientists and the se-

cond with that of historians. Fernand Braudel (1958/1972:

37) clarifies his stand through an example:

"Economic cycles, the ebb and flow of material life,

can be measured. A crisis in social structure must
likewise be observed in time, and in its passage
through time; it must be possible to locate it exact-
ly, in itself, and even more important, in relation
to the movement of concomitant structures. The his-
torian's real interest is in the meeting points of
these movements, their interaction and their breaklng-
points: and these are all things which can be re-
corded in relation to the uniform time of histor-
ians, which is the general measure of all such phe-
nomena; they cannot be recorded in relation to multi-
form social time, which is merely the particular mea-
sure of each taken separately."

Perfect synchrony, "a momentary stop, suspending of all dura-

tion~ is considered totally artificial (Ibid:26). Instead,

multiplicity of time is fully recognized and three frame-

works are delineated: short term, focusing on the individual

and the event; conjunctures, inquiring into cycles and semi-

cycles with a duration of ten, twenty, or fifty year periods;

and long duration, embracing hundreds of years (Ibid: 13-21) .

The depth and semi-mobility characteristics of long dura-

tion detach the analyst from the requirements of historical

 2

time in order that (s)he may return to them with new in-

sight.

Analysis over long duration is especially helpful

in revealing the operation of constraining structures.

Geography is one of the most enduring of such constraints

in Braudel's framework (Ibid:18). The conjectural unit,

by contrast, permits an investigation of structures that

undergo rapid change. The combination of the two frame-

works of time, by focusing on the simultaneous movement of

concomitant structures, their meeting - and their breaking-

points, allows for an analysis of social reproduction that

reveals both the constraints and potentials under which

historical actors shape their destiny.

The requirements and implications of this mode of in-

quiry are radically different from the common social science

exercises in.ahistorical model-building. Maurice Godelier' s

pronouncement on the rules of anthropological method serves

as an example of the dissimilarity. He proclaims that so-

cial relations must be analyzed as "systems" and that "the

inner logic of these systems must be analyzed before their

origin is analyzed" (1966/1972,xxij emphasis in the original).

According to him, Marxist, structuralist, and functionalist

approaches, all share these "principles."

That Godelier's stand echoes an earlier generation is

revealed in the contrast drawn by Radcliff-Brown between

functional and historical explanations. He insisted that

 3

"for any particular system as it exists at a certain time

we can make a study of how it works .•• When we succeed in

discovering the function of a particular custom, i.e., the

part it plays in the working of the system to which it be-

longs, we reach an understanding or explanation of it which

is different from and independent of any historical explana-

tion of how it came into existence" (1950:3). He defined

synchronic analysis as "dealing with a system as it exists

at a certain time, abstracting as far as possible from any

changes that it may be undergoing" (Ibid). Whereas Rad-

cliffe-Brown offered the plea that, due to the scarcity of

data, "a history of African institutions" could not be had

(Ibid), Godelier insists that only on the basis of "pre-

liminary structural research" is true history of the genesis

of structures and functions of systems possible (1966/1972:

xxiii).

Braudel's observation on model-building exposes the

central weakness of the functionalist and structural-marxist

enterprises. He affirms that the entire value of the con-

clusions "depends on the value of the initial observation

and the selection carried out in isolating essential ele-

ments from the reality observed, and determining their re-

lationships at the heart of that reality" (1958/1972:34).

As Godelier, by theoretical fiat, has already chosen his

structure and the hierarchy of the elements within it, re-

course to history merely serves the a priori function of

 4

confirmation.

The prerequisites of a truly historical anthropology

have been succintly formulated by Conrad Arensberg. He con-

firms that "the anthropologist must not only identify the

social arrangement of human interaction and organization

as invented and patterned into culture that explains eco-

nomic behaviors and motives, but he must also account for

its presence." These relations have to be accounted for

historically and geographically, functionally as well as

in terms of the values that sanction them as custom and

make for their transmission as cultural inheritance (1957:

101). Stressing the "temporal, processual view of social

arrangements" in anthropology (Ibid:l08), Arensberg insists

that "consciousness of emergence, and the combination of

temporal and formal observations which have yielded it out

of ethnographical data, is another of the empirical dis-

coveries of anthropology which make it different from the

system-building disciplines of sociology and economics"

(Ibi~:llO). From the comparative perspective of anthropol-

ogy, he argues that the record of human history is full of

examples of emergence, elaboration, and revolution (Ibid:

Ill).

That this advise has not been heeded is most glaringly

revealed in the dominance of synchronic studies in nec-func-

tionalist and neo-evolutionist approaches in ecological an-

thropology. The functionalist fallacy, ecological reduc-

 5

tionism, energetics, local population as the unit of study,

and choice of time scale, are enumerated by Benjamin Orlove
(1980:244) as the major failures of these approaches.
By locating ecological factors within global patterns
of political economy over conjunctures and long durations,
and by demonstrating the changing significance of these
elements in conjunction with different social relationships,
historical anthropology can avoid the pitfalls of a narrow
ecological anthropology. I will illustrate this contention
by reference to two studies.
Morton Fried, analysing the entry of a society with
intensive exploitative demands into an "area in which the
population pursues less promising modes of production" (1952:
392), singles out land tenure, geography and ecology for
especial attention. He demonstrates his hypothesis by data
from the contact of expanding European culture with the
aboriginal population of the New World north of the Rio
Grande as well as the meeting of expanding Chinese civili-
zation and the aboriginal culture in southwest China (Ibid:
293-412).
By conceptualizing agrarian society's organized flow,
Joan Vincent connects time and space through viewing society
"as 'men in action' and policy as control of movement."
This perspective directs attention "to who lives where, when,
and how and who goes where, when, and how as a preliminary
to the understanding of politics both as process and as
 6

historical development" (1977:64). Depending on the nature

of the problem under investigation a "square yard of turf"

is arbitrarily chosen as the unit of analysis but it is

affirmed that the boundaries of the unit "whether political

or administrative, are the outcome of a process and so may

partially reflect the decisions and purposive actions of

individuals of the past" (Ibid:57). There seems to be an

in-built test of duration in this outlook. In order to

demonstrate that boundaries are outcomes of processes, the

historical anthropologist has no choice but to turn to the

multiform time of the historian. The past in the present

and the present in the past become the simultaneous focuses

of inquiry. Yet, the anthropologist, by grounding the

separate time units of the historian in the organized fields

of action of social actors, can grasp emergence and trans-

formation, hence make a fuller analysis of reality as a

process that embodies contradictory structures.

Positing a nexus between movements of people and the

emergence of boundaries not only calls attention to politi-

cal and economic forces that condition the fields of c~tivi­

ties of the people concerned; characteristics of the space

over which movement takes place and the means of communica-

tion that determine the pace of the movement become equally

significant. Furthermore, comparison of differences in means

of communication in adjacent areas with different socio-

economic systems may allow for the comprehension of the

 7

directionality of movement and thereby reveal the degree

of conjunction and/or disjunction of different fields of

activities and policies. It is in this sense that in the

remaining part of this chapter I will describe the histor-

ical geography and ecology of the presen~ territory of

Afghanistan for the period 1747-1901.

By making use of a unit of long duration, 1747-1901,

I shall first delineate thos.e features of the spatial organ-

ization and means of communication that remained stable.

Then, though a conjuncture of fifty years, 1850-1900, I

shall chart the changing structure of communication of the

neighboring areas. The use of these units of duration in

the subsequent chapters will help in elucidating the larger

pattern of interaction that resulted in the emergence of

contemporary Afghanistan.

Even a cursory examination of maps 1-4 brings the

significance of regions as distinct areas of concentration

of population and resources into sharp relief. By the

northeast-southwest course of the Hindu Kush range and by

the Kohi-Baba and Feroz-Koh peaks from Bamiyan westwards,

the country is both geologically and geographically divided

into a northern third and a southern two-thirds (Amin and

Schilz, 1976:23). Each area is in turn divided into dis-

tinct geographical sub-areas. Johannes Hunlum (1959:l03ff),

as Map 1 shows, divided the country into ten zones. More

recently, Louis Dupree (1973:3-33), as indicated in Map 2,

 8

has offered a classification of eleven geographic zones.

No matter \vhose nomer,clature is adopted, once Map 3, which

records the elevation of different parts of the country,

and Map 4, which lists the major areas under cultivation,

are considered in conjunction with Maps 1 and 2, the ques-

ti.on of politics as control of movements of people becomes

intimately linked to the regional organization of space and

the means of communication that link the various regions

together.

Analysis of communication between regions, however,

can be meaningful only when characteristics of patterns

of settlement within and between regions have been des-

cribed. As the issue of pattern of settlement revolves

around the concept of central place, it is essential that

the usage of this term be clarified at the outset. Areas

where relatively sizeable concentrations of population are

established and where agricultural activity is underwritten

by a relatively major irrigation network are defined as

central places. Peripheral places are defined as areas

within a region where relatively smaller concentrations

of population are supported by relatively minor irrigation

networks. Intermediate places lie between the other two.

The nexus between irrigation and nature of settlement will

be addressed at greater length in chapter two. Meanwhile,

an examination of Maps 4 and 5 will suffice to underline

the importance of the linkage.

 9

The maps show an almost complete fit between the dis-

tribution of the major rivers and the number of the major

areas of cUltivation. Furthermore, large zones of popula-

tion concentration are located on or at fairly close dis-

tance from junctions between major rivers and their tribu-

taries. This observation allows us to offer a classifica-

tion of major areas that is more in keeping with patterns

of historical interaction than those based on geographical

criteria alone. I shall distinguish six areas: (1)

Badakhshan, with the central focus of Faizabad; (2) a north-

ern area with a number of intermediate centers; (3) a north-

western area with the central place of Herat; (4) a south-

western area with the central place of Kandahar; (5) an

eastern area with the central place of Kabul as well as a

number of intermediate centers; 6) and a central area with

a number of intermediate centers. The criteria of popula-

tion size and extent of irrigation network merely serve to

indicate agricultural potential. The actual facts of popu-

lation size, as the second part of this study will demon-

strate, were an outcome of the operation of larger politi-

cal forces. However, since over our long unit of duration

these places retained their importance to varying degrees,

I feel justified in referring to them as major regions. I

shall now turn to describe the salient features of one

such central place, Kandahar in the southwestern region.

A distinction between kariajat (suburbs), mahalajat

 10

(quarters of the city) I and karezat (villages dependent on

underground irrigation channels), was the organizational

principle of the pattern of settlement throughout 1747-1901.

In 1880, the political officer of the British force that

had occupied the city applied the term kariajat to all

"villages within a radius of about 20 miles round the city

with a few isolated spots of cUltivation to the northwest"

(India Army, General Staff Branch, Gazetteer of Afghanistan,

hereafter GA, Kandahar 1914/1980:228). The shape of the

central place itself was described as that of "an irregular

oblong, the length being from north to south, and with a

circuit of 3 miles, 1006 yards. It is surrounded by a

ditch, 24 feet wide and 10 feet deep, and by a wall which

is 20~ feet thick at the bottom, 14~ feet thick at the top,

and 27 feet in the height" (Ibid:239). The whole area was

watered by three rivers and five canals were brought from

one of them into the Kandahar valley. In locations which

did not have access to the canals, irrigation was carried

out by the underground networks. Hereafter is a summary

of the available information on the three types of settle-

ment based on reports of the British officers in charge

of revenue and administration in 1880 (Ibid:231-38:242-44).

Out of a total of 155 kariajat (suburbs) that are

listed, complete data on both acreage and number of dwelling

places is available only for 116. The average number of

houses was 60 -- range 0-350. The mean acreage was

 11

1,289.112 acres -- range 11-6,068.

The distribution of settlements in the karezat (villages

dependent on underground irrigation channels) indicates a

different pattern. Out of a total of 103 villages, informa-

tion is available on 93. The average number of dwelling-

places was 8.45 -- range 0-67. The difference in the range

and average between kariajat and karezat can be interpreted

as tentative evidence for a positive correlation between the

density of population and the scale of irrigation networks

in this region.

Data on the demographic composition of the residents

of mahalajat (quarters of the city) disclose the lack of

participation of the inhabitants of the central place in

agricultural production and reveal the bonds that tied them

to the surrounding areas. As the inference is drawn from

the demographic information, the evidence has to be pre-

sented. There were 4,320 houses in the 87 quarters of the

city. The average number of houses in a quarter was there-

fore 49.65 -- range 7-140. Classification of the population

on the basis of age and sex reveals a significant fact: out

of a total population of 29,244 individuals, 35% \,:ere de-

fined as children and 65% as adults, and of the latter

53.67% were female. Since the available evidence suggest

that women did not take direct part either in agricultural

production or exchange and that most of the 8.803 men en-

gaged in craft production and activities in the spheres of

 12

exchange and administration, the inference that most of

the occupants of the central place did not take part in

agricultural production can be considered justified. It

can therefore be assumed that the residents of the central

place of Kandahar depended for most of their food on the

surrounding areas.

That the surrounding areas did have grain surplus is

confirmed by reports of British officers for 1880. They

estimated that in addition to supporting the local popula-

tion the area could feed a foreign army of 19,000 men and

4,000 horses yearly (GA,Kandahar:23l).

How was the surplus moved? As far as the movements

of goods between the surrounding areas and the central

place is concerned the availability of means of communica-

tion as well as the organized nature of the carriers is

affirmed. British officers testified that "transport can

be obtained to the extent of 2,000 camels, 2,000 donkeys

and 2,000 bullocks, but owners will not take service with

troops and will only work as carriers independently" (Ibid:

228) •

If the area was capable of producing a surplus and

providing for its organized transportation for local pur-

pcses, could the surplus be moved to other places? This

question brings us back to our original concern with commun-

ication between regions, the nature of the means of communi-

cation and the consequent implications for the pace of the

 13

movement of goods and people. Before addressing these

issues, we must take account of the pattern of settlements

between regions and examine the degree to which relations

between the central place in Kandahar and its surroundings

was typical of other regions.

Since potential for settlement was related to avail-

ability of water and land, the distribution of these factors

determined to a larger degree the location of settlements.

The cultivated area forming a mere 5% of the territory and

mountains occupying a full third of the country, the pattern

of settlement in general was necessarily dispersed. The

difference in the relative concentration of population in

and around central places and in glens of large mountain

chains was therefore striking. Yet, diversity in the na-

ture of areas surrounding central places reflected itself

in the size and number of intermediate places within and

between regions. This in turn affected the relations of

a central place with its surrounding area and the orienta-

tion of each locality to central place(s) within the same

region, within the same country, and in neighboring coun-

tries.

Variation in the pattern of settlement between

Kandahar and Herat, Kandahar and Kabul, and Kabul and Nor-

thern Afghanistan can serve as illustration.

A British military party that made the journey between

Herat and Kandahar in 1838 estimated the distance between

 14

the two central places as 380.5 miles. The location which

marked the boundary of the two regions was 151.5 miles

distant from Kandahar and 229 miles from Herat. A total

of 35 stations existed between the two places. Average

distance between settlements in the Kandahar region was

12.625 miles and 10.409 miles in the Herat region. Ex-

cluding the intense cultivation in the 37 miles adjacent to

Kandahar and in a place 94 miles from it, all stations were

settlements of small size. Their only importance lay in

the availability of water and forage. But at a distance

of 80 miles and 141 miles from Herat were two districts

with fairly dense population and significant amounts of

cultivation. The area in the immediate vicinity of Herat,

despite signs of former habitation, had been largely de-

populated by the wars of the previous decade (Thornton, E.

A Gazetteer of the Countries Adjacent to India, 1844: 293-

309) .

Distance between Kandahar and Kabul was estimated at

317 miles. An average distance of 9.606 miles separated

the 33 places which were noted for any importance by the

party. Only 83 miles of the route were within the region

of Kandahar and, excepting the three miles in the proximity

of Kandahar, settlement was very dispersed. The picture

on the Kabul side of the route was, however, radically dif-

ferent. Relative density of population was commented upon

and the extent of CUltivation for a number of places was

 15

described in such words as: "the whole plain, as far as

the eye can reach, one large yield of wheat," "the whole

plain covered with green wheat and fine clumps of trees,"

"abundance of water .•. " (Ibid:312). The intensity of cul-

tivation around Kabul itself was also noted.

The first 32 miles of the 357 miles journeyed from

Kabul to Balkh in northern Afghanistan coincided with the

route from Kabul to Kandahar. Then, a number of passes

ranging from 10,000 to 14,000 feet had to be crossed for

getting into central Afghanistan and again from the central

to the northern region. In central Afghanistan, only the

valley of Bamian, at a distance of 103 miles from Kabul and

162 miles from the first large settlement in the north,

was noted for availability of supplies (L/P&S/C7l Memoran-

dum On Afghanistan, On the Road ••• ,1878:l-7). The inter-

val between the passes and the locality of Haibak in the

northern region was, in 1886, described as "for the most

par.t uninhabited and devoid of supplies" (F.O.539/44 West

Ridgeway to the Earl of Iddesleigh, Inclosure 7, No.7.,

Dec.20,1886:14). The northern plain itself was watered

by four riveL3, three of them flowing in parallel direc-

tion to each other, and a number of places had the agri-

cultural potential for supporting large numbers of people.

The actual number of settlements at any time, however, was

not a function of the potential fertility of the soil but

a resultant of politcal factors.

 16

Nonetheless, the fact that the pattern of settlement

described for the central place of Kandahar and its sur-

rounding areas repeated itself in the eastern, northern

and western regions of the country and that actual surplus

of agricultural produce in the 1880s is also reported, is

significant. The situation in Badakhshan and the central

area was different.

The organization of space into a central place(s), its

surrounding areas and dispersed or concentrated settlements

within and between regions has, until this point, been

pursued in relative terms only. In order to render the

discussion more meaningful I shall present the actual fig-

ures for 1880s. Except for the population of the central

places of Kabul and Kandahar, which were based on actual

censuses, all figures are essentially informed estimates.

The total population of the country during the l880s and

1890s was approximated at 5-6 million people (Great Britain,

India Office Records, Military Report on Afghanistan, 1906:

194-5). Of these approximately 20% were living in the cen-

tral places. As Humlum's classification of the geographi-

cal characteristics of the country helps in conveying a

picture of the nature of the terrain, his table is presented

next to that of the population distribution.

 Table I: Regional Concentration of Population of Afghanistan, 1880s-1890s

Approximate Total Name and Population Population of

Regions Area (km2) %
Eopul.ation of Central Place Surrounding Area

Kandahar
Southern 148,100 23.3 35,700
32,500

105,000 16.6
Kabul
Eastern
140,000

Faizabad
Badakshan 48,900 7.7 120,000 17,300
8-9,000

87,105 families & Balkh, Mazar, &

Northern 92,200 14.5
110,000 individuals Ackcha:15,OOO

Western and Herat

76,900 12.1 500,000 300,000
NorthWestern 9,000

Bamian
Central 163,900 25.8 500,000
3-6,000

Total 635,000 100.0 5 to 6 million

Sources: Adapted from Humlum 1959; GA,12 volumes; Military Report on Afghanistan,1906.

f--'
-.]
 Table II: The Altitude Zones of the Natural Regions

Altitude Zones Lower than 300-600 600-1800 1800-3000 Higher than Total %
(km2) 300m m m m 3000 m

South Afghanistan 29,200 118,500 400 J.48,100 23.3

Afghanistan of the
800 11,700 3,200 15,700 2.5
Monsoon

East Afghanistan 11,500 61,000 1,600 74,100 11. 7

Nuristan 2,700 6,300 6,200 15,200 2.4

Badakhshan 1,600 12,100 13,700 10,400 37,800 6.0

Wakhan 1,900 9,200 11,100 1.7

North Afghanistan 5,600 26,700 44,800 15,100 92,200 14.5

Northwest Afghanistan 200 21,300 9,100 30,600 4.8

West Afghanistan 45,000 1,300 46,300 7.3

Central Afghanistan 26,400 98,000 39,500 163,900 25.8

Total 5,600 58,500 294,000 210,000 66,900 635,000 100.0

% 0.9 9.2 46.3 33.1 10.5 100.0

Source: Hum1um 1959:17.

f---'
00
 19

The distribution of population in the regions and

their central places in the l880s and l890s was not typical
of the long unit of duration 1747-1901. Movements of popu-
lation in southern, central, and eastern regions followed
a very different trajectory than that in the regions of
Badakhshan, north and northwestern Afghanistan. While
conditions in the former regions remained relatively sta-
ble, in the latter, the basic tendency was towards depopu-
lation. A combination of political factors and epidemics
was the cause.
Fluctuations in the size of the central place of Herat
were by contemporary observers, attributed to political fac-
tors. In 1809, the estimates of Herat's population ranged
from 45,000 to 100,000 people. Before the year long Persian
siege of 1838, it had probably reached 70,000 but after the
siege it was approximated at 6,000 to 7,000 people. In 1845,
it had gone up to 22,000 and had probably reached 45,000
by 1857. The Persian siege of that year and the wars of
the subsequent years brought a reduction to the low ebb of
9,000 in 1885 (GA,Herat:174).
At times also, the political agrandizement of the
powerholder of one region through military conquest led to
forceful movements of population from one region into
another. When the chief of Kunduz, the northeastern
Afghanistan, conquered Balkh and Badakshan in the l820s,
he marched off great numbers of inhabitants from both
 20

places to the area around Kunduz (Burnes,A. Travels Into

Bokhara vol.II,1834/1973:346,351).

The impact of political factors was exacerbated by the

outbreak of epidemics as attested by the following remarks

of the compiler of the Gazetteer for northern Afghanistan:

"The population of the province is small in comparison with

the area. This is partly due to devastating wars, and to

the chaotic conditions of the country before it came under

Afghan rule, but in a great degree to famine and pestilence.

It may be stated here that what is called the 'Persian'

famine of 1872 was terribly severe in Herat and Afghan

Turkistan. It was followed by a dreadful outbreak of chol-

era, and Maitland was assured that some districts were al-

most entirely depopulated between 1871 and 1873" (1914/1979:

12) .

Thus, explanation of population fluctuations cannot

be isolated from the operation of larger political forces.

I shall resume the discussion of the implications of these

changes in the second part of this study where our unit of

duration will be conjunctural, but structural constraints

imposed by the means of communication on the movements of

people and goods inside Afghanistan remained constant

throughout the long duration and will have to be considered

here.

As indicated in our discussion of patterns of settle-

ments, communication between regions took place over land.

 21

The reason for this is rather simple. Most rivers in

Afghanistan rise in the central range of mountains and

flow northwards or southwards. There is therefore no water-

way connecting the northern third with the southern two-

thirds of the country. Furthermore, most of the rivers

lack the necessary depth for navigation and cannot be used

for the inter-regional transportation of either people or

goods, the only exception being the floating of wood on

some of the rivers.

Reasons for the absence of wheeled vehicles are, how-

ever, not clear. Their non-existence from Morocco to

Afghanistan until the nineteenth century is demonstrated

by Richard Bulliet (The Camel and the Wheel 1975:8) who

persuasively argues that camels were the standard means of

transportation in the area. Until the beginning of the

twentieth century, I have not found any evidence for the

use of wheeled means of transportation in Afghanistan ex-

cept in urban centers. The dominant position of the camel

did not mean that other animals were not used. Ponies,

mules, horses and donkeys were quite common as some of the

terrain was so rugged that even camels could not cross it.

In order to convey an impression of the degree to

which the speed of movement was conditioned by the means

of communication, I have summarized the available informa-

tion in table three.

 Table III: Communication Between Central Places in Afghanistan

Estimated Number of Time Closed

Central Places and Route Number of Number of
Distance (miles) marches by to
Days for Days by
Traffic
mili tary PCl~ty Caravans Other Means
332.5 or 2 passes:2 months
Kabul-Balkh; through Bamian 28
357 1 pass :9 months
1 pass :2 months
Kabul-Bamian 106 7
1 pass :9 months
1 pass :8~months
Kabul-Mazar: through Ghorband 290-300 26 1 pass :8 months
2 passes:no information
Kabul-Khanabad; through Khawak 237
Kabul-Faizbad; through no
Kkanabad-Khawak information
Kabul-Kandahar; through Ghazni 318 29 2-4 months
Kabul-Kandahar; through horse:8days
308 29 15 2-4 months
Kilat-i-Ghilzai courier:6days
Kabul-Herat; through
718 66
Kandahar-Farah
Kabul-Herat; through
687 62 31-33 horse: 18days
Kandahar-Girishik
Kabul-Herat; through
Kandahar-Sabzwar 697.5 62

Kandahar-Herat; through Girishik 369 33

Kandahar-Herat; through Farah 400 37 16-18 horse:10days
Herat-Kabul; through the
central region 469.5 3-5 months

Herat-Balkh; through Maimenah 430

N
N
 23

The above data, having been gathered under relatively

stable political conditions, mostly by British officers,

reflect the technical rather than the social conditions

of movements of people and goods. But mobility, as we

shall witness in other chapters, was intimately connected

with the political conditions as well.

I have, until now, deliberately treated the present

territory of Afghanistan in isolation. It is time to place

our spatial unit wi thin the wider frameworks of time and

space. By locating the chosen unit within the larger flows,

this procedure allows for a better grasp of the histori-

cally changing significance of geography and history. An

examination of changes in the internat~onal systems of com-

munication in the area will serve as an illustration of my

contention. The old system of communication which was in

operation for centuries and, depending on areas, continued

to function until 1900, is presented in Table four.

 Table IV: Inter-territorial Routes of Communication Before 1850

Estimated Number of Number of Number of

Name of Places and Routes Distance marchGs by days by days by Other Features
(miles) military party caravans other means

Kabul-Deir Ismael Khan; through horses & camels

217 39
Ghazni ?

horses & camels

Kabul-Quetta; through Ghazni 376 27 27
?

horses & mules

Quetta-Kelat 112.43 10
10
horses & mules
Kelat-kotree 182.16 14
?
horses & camels
Kotree-Sukhur 173 10
?
Quetta-Meshed 1,050 150
Excessive heat
horses & camels
Kandahar-Dadur 225.5 35 May - August
?
shortage of water
horses & camels
Dadur-Shukhur 173 11
?
horses & camels
Shukhur-Larkhana 49.6 5
?

horses & camels

Larkhana-Seh\o1an; Arul river 122.2 13
?

horses & camels

Sehwan-Karachi 16
?

N
,J::.
 Table IV (continued)

Estimated Number of Number of Number of

Name of Places and Routes Distance marches by days by days by Other Features
(miles) military party caravans other means

Kandahar-Sind; through horses & camels

692.5 61
Balouchistan ?
Kandahar-Shikarpur; horses & camels
369 32
Bolan pass ?
Kandahar-Dera Ghazni Khan; horses & camels
414
Sukhee Sarwan pass ?
Herat-Sarakhs; through
469.5
Sang Kotal and Zulfikar
Herat-Meshed 320

Meshed-Bander Abbas camels & donkeys

970 110
?

camels & mules

Meshed Teheran 560 35-40
?

Balkh-Kilif Ferry 55
Balkh-Khwajeh Saleh Ferry 68
Kashgar & Yarkand (China) ponies ~ camels
65
I3okhara; through Badakhshan-Balkh ?
passes 20-25,000 feet
Faizabad-Peshawar; through Pamir 343
closed 3-9 months
Peshawar-Kabul; through horses Oc camels Kabul-Jalalabad
191 19
Jalalabad ? closed 3 months

N
Ul
 Table IV (continued)

Estimated Number of Number of Number of

Name of Places and Routes Distance marches by days by days by Other Features
(miles) military party caravans other means

horses & camels

Kabul-Kohat 234 22
?

horses & camels

Kohat-Rawalpindi 106 16
?

tv
en
 27

Unlike the system of communication within Afghanistan

outlined in Table three, the inter-territorial system de-

lineated in Table four underwent radical shifts from 1840

onwards. Indeed, road-building was undertaken only about

1840 and the compiler of the Imperial Gazetteer of India

offered the following reason for the late start: "The

level plains of India, scoured by streams which, for eight

months or more in each year, are passable without difficulty

by the conveyances generally used in the country, offer

so small an obstacle to intercourse between different lo-

calities that, up to the end of the eighteenth century,

there was no demand for prepared tracks even for military

purposes, transport being chiefly effected by pack animals

travelling along village pathways, while travellers could

ride or be conveyed in palanquins" (1907,vol.III:402).

Communication between India and its naighboring coun-

tries, as the British army sent to conquer Afghanistan in

1838 found out, was not easy. En route from Sukkur in Sind

to Kandahar in Afghanistan, the army had to cross the Bolan

pass and in attempting to do so lost 38,000 camels in one

day (L/P&S/C71,1878:17). The expansion of the British do-

main in India, as shown in Maps 6a and 6b, made the radical

alteration of the existing structure of communication an

economic and military necessity. Road-building was signifi-

cant but the introduction of railways was the most momen-

tous component of this restructuring.

 28

The first comprehensive plan for building railways in

India was drawn up in 1853 but actual construction only

started in 1859. By 1905, the total mileage of the Indian

railway, as displayed in map 7, amounted to 28,054 miles

(Imperial Gazetteer of India, vol.III,1908:366).

While most of the important lines in the interior of

India were completed between 1860 and 1870, extension of

lines to territories adjacent to Afghanistan took place

between 1880 and 1905. The Afghan war of 1878-1880 pro-

vided the first impetus. The marches from Sind to Quetta

had resulted in the "waste of treasure and life" and demon-

strated the necessity of a railway. Not surprisingly, "all

the railways and the best of roads" in Baluchistan had

"their origin in strategic needs" (Ibid,vol.VI:3l2). That

the modern system of rail communication had not yet reached

the north-western frontier of India in 1879 is affirmed

by a report of a military officer. He stated that "all

travelling has to be performed on horses or camels, and

during the rainy season, owing to the alluvial character

of the soil which becomes after rain almost as slippery as

ice, travelling is excedingly difficult and laborious" (L/

P&S/18/A134,1879:l3).

In 1885, Russian advances in Central Asia and the

possibility of war between Great Britain and Russia looming

large on the horizon provided the second impetus. A

special defense committee in India called for the extension

 29

of railway lines to points "which are open to attack or

where future offensive operations must be initiated" (LjP&Sj

Al17,1885:1). Baluchistan and the Peshawar district in the

North-West Frontier received especial attention under the

terms of the plan. The total length of the railway in

Baluchistan increased from 277 miles in 1891 to 399 miles

in 1901 and 481 miles in 1905. Sind and Baluchistan were

connected by the railway and, by 1892, one section of the

line had already reached the Afghan frontier at Chaman

(Imperial Gazetteer of India,vol.VI,1908:312-13). By 1885,

a railway linked the city of Peshawar with the main network

of the Indian railways and a metal road joined Peshawar to

the Punja~) (Imperial Gazetteer of India,1886,vol.XI:155).

The district possessed 157 miles of metal roads and 672

miles of unmetalled road in 1905 and the railway line had

;)een extended to the Afghan frontier (Imperial Gazetteer

of India,1908,vol.XX:120).

Strategic considerations, however, were not the only

reason for the alignment of the communication system. Even

in the North-Western Railway, encompassing both Baluchistan

and the Peshawar district, only 1,042 miles were built for

strategic considerations. The remaining 2,986 miles owed

their realization to economic causes. The nexus between

production, communication, and domination was cogently ar-

ticulated by the compiler of the Imperial Gazetteer of

India. In his discussion of the beginnings of the railway

 30

he wrote that "the Directors [of the East India Company]

begun to realize that, without the material appliances

which facilitate and cheapen the means of communication

and production, there could be DO rapid progress in the

country either normally or materially, or in the efficiency

of the administration" (1908,vol.III:365).

The completion of the railway network and the opening

in 1869 of the Suez Canal, which reduced the length of the

voyage from India to Europe from 100 to 25 days, did indeed

bring about a transformation in the system of agricultural

production in India, and directly linked it to the trends

in the cosmopolitan economy. The Indian wheat crop, with

its main center of production in the Punjab, was almost

all shipped to Europe, the main port of export being Karachi

in Sind. Karachi was also the railway port of the North-

Western State Railway and was connected through the branches

of the line to places as far as Chaman on the border of

Afghanistan. In 1904 alone, this line carried 25,500,000

people and 6,500,000 tons of goods -- mainly wheat (Ibid:

262,285,398-99).

India was not, however, the only place where system-

atic changes in communication and production were occurring.

During the course of the nineteenth century Russia, as

shown in Map 8, advanced to the borders of Afghanistan in

Central Asia. Between 1864 and 1885, the Russians methodi-

cally pursued a policy of conquering, organizing and colo-

 31

nizing the conquered territories. The expansion of modern

communication was one aspect of this undertaking and, by

1890, the Russian railway had reached at Sarakhs, the

northern boundary of Afghanistan (MSS Eur. D727 A).

In Iran, the old and new systems of communication co-

existed (Issawi,C. ed. The Economic History of Iran 1800-

1914 1971:l96ff.). Only in Afghanistan did the old system

completely prevail and about 800 miles of Afghan space pre-

vented a nexus between India and Europe by railway.

The intrusion of rail communication and the intro-

duction of fundamental changes in the political economy of

some of the neighboring countries had considerable implica-

tions for the articulation of relations between the various

regions of Afghanistan and the relations of these regions

with forces outside the country. In marked contrast to

the direct involvement of surrounding areas, the country

was only indirectly becoming involved in the world political

economy. In this regard, the case of Afghanistan provides,

in two essential respects, a contrast to the cases analyzed

by Fried. It thus gives us an opportunity to examine the

social relations of people whose system of production is

not transformed by conquest while that of neighboring areas

is undergoing such a transformation.

As production seems to be playing such a crucial role

in the structuring of social relationships, I will now turn

to examining it in chapter two.

 32

CHAPTER TWO

Labor-Process and Technical Organization of

Production in Agriculture

In an article in 1938 that foreshadowed most of the

features of the type of analysis that has subsequently be-

come associated with his name, Karl Wittfogel called atten-

tion to the combination of land and water in setting the

agricultural process of production in motion. He asserted

that "differences in the type and productivity of the land,

the soil, determine to a large extent the settled social

stage of the resultant labor-process; they lead, however,

only to variations, not to a fundamentally divergent pattern

of the labor-process. This effect proceeds from the second

of the chief factors in the means of work -- from water"

(1938/1969:183). Wittfogel outlines the possible tendencies

in a table that is copied below.

 33

TABLE: Impact of Irrigation on

Social Organization
Rain for Specific
Rivers, Ground-Water,etc.
variations Agricultural Types
Time- Pre-
of the Water- Tendency to
Sufficiency sent? The Task:
Situation liness
al Drainage, Patterns of
a ++ a+ Protection Water-Control
+
1 a2 --
b b- b --
Rainfall
A + agriculture.
2 + Insignificant
In case of
temporary
a + a-
3 + nomadism
b b+ Transition to transition to
rainfall
B 1 a
+- agriculture
plus cattle-
raisinq.
1 a + - a+
Supplementary Irrigation
But enough Irrigation Agriculture
for pasture
b b- Nomadism
Irrigation Irrigation
2 a - + a+ for the sake Agriculture
B of Insurance
b b- Nomadism
- a - + Irrigation to
Irrigation
3 a - But enough Make agriculture
Agriculture
for pasture possible
b b- Nomadism
Irrigation to
-- Irrigation
a a+ Make Agriculture
No Rain at Agriculture
possible
C - All

b b- Desert!

The sign + means: Sufficent or present in considerable quantity

The sign ++ means: Presnet in excess
The sign - means: Present only in slight, insufficent quantity
The sign -- means: Not present at all
 34

Although in his major work on the subject Wittfogel

differentiates between hydro-agriculture and hydraulic

types of societies as well as compact and loose varieties

of the latter (1957/1962:3,166), the posited nexus between

the factors of production and the labor-process has disappeared

from the analysis. As the subsequent debates in anthropology

and other disciplines regarding the vailidity of propositions

advanced by Wittfogel have been largely structured around

his thesis of "oriental despotism," the specificity or lack

of the labor-process in a system of agricultural production

based on irrigation has been neglected. An adequate evalu-

ation of the larger hypothesis, however, is not possible

without a prior examination of the labor-process (es) in

such systems. Clarification of the notion requires turning

not to Wittfogel but Marx, under whose impact Wittfogel's

1938 article was written.

Marx wrote that "the determinate social form of the

worker's labour corresponds to the form which the conditions

of labour -- that is, in particular, the land, nature, since

this relationship embraces all others -- assume in respect

of the worker. But the former is in fact merely the objec-

tive expression of the latter (Theories of Surplus-Value

part III,1861-63/1971:415). He later phrased this idea in

a more general proposition on the determination of "organi-

zation of labour ... by means of production" (Selected Cor-

respondence July 7,1866,to Engels).

 35

Marx's contribution to the investigation of the labor-

process should not be sought in the generality of its claim

alone. An even more important dimension is the forging of

a series of concepts that can be historically tested.

His juxtaposition of social wants and social time as

mediated by labor-time is unique. He argued that "if society

wants to satisfy some want and have an article produced for

this purpose, it must pay for it .•. society buys it with a

definite quantity of its disposable labor-time" (Capital

vol.III,1894/1977:l87). Through a series of determinate

definitions, he further clarified his notions of labor and

time distinguishing between general labor, labor-power, sim-

ple average labor and complex labor.

General labor is viewed as an inherent attribute of

humanity; "it is an eternal nature -- imposed necessity,

without which there can be no material exchange between

man and Nature, and therefore no life" (Capital vol.I,1867/

1975:42-43). Labor-power is defined or capacity fcr la-

bour is defined as "the aggregate of those mental and physi-

cal capabilities existing in a human being, which he exer-

cises whenever he produces a use-value of any description"

(Ibid:167). But it is the concept of simple average labor

that allows for comparison across time and space. It is

defined as the expenditure "of the labour-power which, on

an average, apart from any special development, exists in

the organism of every ordinary individual" (Ibid:44). Marx

 36

immediately added that IIsimple average labour, it is true

varies in character in different countries and at different

times, but in a particular form of society it is given ll

(Ibid). Complex labor -- involving skills is defined as

"simple labour raised to a higher power" (Contribution to

the Critique of Political Economy 1859/1970:31).

The concept of necessary labor allows for the establish-

ment of a nexus between labor and time. Marx's use of this

concept is two-fold and has, therefore, given rise to con-

fusion. On the one hand, it designates "the time necessary

under given social conditions for the production of any

commodity;" on the other hand, it also refers to the IItime

necessary for the production of the particular commodity

labour-power" (Capital vol.I:217). It is the former sense

of the usage that interests me here. Although Marx's unit

of analysis is a society, there is no logical reason for

confining the utility of the concept to that framework.

Juxtaposition of differing ratios of necessary and surplus

labour may in fact provide a better understanding of the

totality of social relations than an abstract notion of nec-

essary labour that society as such has to perform. As far

as agricultural production is concerned, utilization of con-

crete units, such as domestic organizations, villages, lo-

calities, etc., by revealing the seasonal distribution of

the necessary labor brings the role of the natural factors

into sharp relief.

 37

Analysis of necessary labor in concrete units permit

the determination of the impact of different factors of the

labor-process on the ratio of necessary and surplus labor-

time for the units concerned. Marx listed three items as

elementary factors of labor. These are: "I, the personal

activity of man, i.e., work itself, 2, the subject of that

work, and 3, its instruments" (Capital vol.I:178). He

stressed the fact that soil, including water "exists inde-

pendently of him, and is the universal subject of human

labour" (Ibid). An instrument of labor is defined as "a

thing, or complex of things, which the labourer interposes

between himself and the subject of his labour, and which

serves as the conductor of his activity" (Ibid:179). The

primary significance of instruments of labor is their effect

on the ratio of necessary and surplus labor and the conse-

quent creation of additional disposable time. In Marx's

scheme, expansion of the productive capacity of society

is an inherent aspect of changes in instruments of labor.

The impact of two instruments of labor -- plough and hoe --

on systems of agricultural production and a series of con-

comittant associations in social relations have recently

been argued in anthropology (Goody,J.1976:1-22).

In this chapter I shall attempt to describe the various

details of the system of production with reference to the

concept of necessary labor and time. In the second part, I

shall attempt to test some of ·the broader correlations be-

 38

tween a system of production and aspects of social relations.

But I must begin the description of the labor-process by

focusing on the impact of natural factors on the duration

of time during which labor could be undertaken.

* * * *

Climatic variation in Afghanistan, in general, is a

function of altitude. On this basis the country can be

divided into five zones. Amin and Schilz (1976:53-54)

offer the following classification.

A. Desert Type: This zone is located mostly below

900 meters, covering the plains in the south, southwest,

west, north and east. It covers 1/5 of the total area

of the country.

B. Desert with vegetation, or Steppe: Situated on

the plains near the foothills, this zone include regions

of warm or hot and humid climate. In the eastern region,

it embaces the upper part of the valley of Nangarhar, the

valleys of Kunar, Alingar, Alishing, and areas around Khost

-- in each case up to 1,500 meters. In the southern and

western regions, the eastern part of Kandahar, the northern

part of Helmand, parts of Farah and Herat, with elevations

between 1,000 and 1,600 meters, fall within this zone. In

the northern region, it encompasses such areas as Balkh,

Khulm and Kunduz.

 39

C. Sub-Humid climate, or Cool Steppe: Most of the

pasture lands and grasslands of the country are contained

in this zone. Its distribution in the eastern and southern

regions is scattered. But in the north it includes the

Taloqan valley, the eastern side of the Kunduz valley,

Baghlan and Doshi up to 1,000 meters, southern parts of

Balkh, and western parts of Maimana.

D. Humid climate: This zone coincides with the for-

ested areas of the country. It comprises the forests of

eastern, central and northern Afghanistan as well as those

of Herat, Badghis, Jawzjan, Faryab, and Badakhshan.

E. Very Cold Humid climate: It covers all the areas

above 3,000 meters. Polar type vegetation can grow in the

lower parts of this zone but the upper part is too cold

for vegetation.

Furthermore, there seems to be an inverse relation

between temperature and altitude. The "average temperature

for the year decreases with altitude, with latitude being

an associated factor" (Ibid:49). The length of the growing

season, as the following table shows, follows the same pat-

tern.
 40

Table II: Altitude, Growing Season, and Average Temperature (Co)

For 28 Stations*

Last Last .
Lati- Elevation Freezing F . Grow~ng
Region Station reez~ng S Temp
tude (meters) Day, Day, eason CO
Spring (Days)
Fall

Eastern Gardez 33.5 2,350 11 Apr. 18 Oct. 189 9.3

Ghazni 33.5 2,183 6 Apr. 12 Oct. 188 9.5

Jabul Seraj 35.0 1,630 24 Feb. 6 Dec. 289 15.0

Jalalabad 34.5 580 22 Jan. 4 Dec. 315 21.5

Kabul 34.5 1,791 29 Apr. 26 Oct. 210 11. 7

Khost 33.3 1,146 13 Feb. 25 Nov. 284 17.0

Laghman 34.5 770 6 Feb. 9 Dec. 305 19.6

Logar 34.0 1,935 22 Apr. 10 Oct. 198 10.7

Moqur 33.0 2,000 10 Apr. 12 Oct. 193 11.2

S. Salang 35.2 3,172 17 Mar. 4 Oct. 137 2.5

Central Barnyan 35.0 2,550 11 Apr. 18 Oct. 189 6.8

Ghalmin 35.0 2,070 26 Apr. 23 Sept. 166 7.8

Lal 34.5 2,800 7 June 27 Aug. 80 3.3

Panjab 34.3 2,710 8 May 5 Sept. 119 6.8

Sharak 34.0 2,325 20 May 28 Aug. 99 4.6

Southern Bost 31.5 780 19 Feb. 14 Nov. 269 19.5

Kalat 32.0 1,565 28 Mar. 2 Nov. 218 14.6

Kandahar 31.5 1,010. 21 Feb. 20 Nov. 271 19.3

Western Farah 32.3 660 22 Feb. 20 Nov. 270 19.0

Herat 34.3 964 19 Mar. 3 Nov. 228 16.0

*Adapted from Amin and Schilz, 1976:49-50

 41

Table II ( continued)

Last Last
Lati- Elevation Freezing Freezing Growing Temp
Region Station Season CO
tude (meters) Day, Day,
(Days)
Spring Fall

Northern Baghlan 36.0 510 12 Mar. 17 Nov. 236 14.7

Kunduz 36.6 433 13 Mar. 27 Nov. 258 16.5

N. Salang 35.2 3,366

Maimana 36.0 815 19 Mar. 11 Nov. 236 14.4

Mazar 36.6 378 7 Mar. 16 Nov. 253 16.9

Shebergan 36.6 360 9 Mar. 29 Nov. 264 16.4

Talooqan 36.6 804 27 Mar. 4 Nov. 224 14.7

Badakhshan Faizbad 37.0 1,200 24 Mar. 4 Nov. 224 13 .0

Examination of the table not only reveals a significant

range of variation in the season of growth -- 99 to 315

for the country as a whole; it also documents important dif-

ferences between as well as within regions. For those in-

dividuals and groups of population that led

life in the rural areas, the length of the season of growth

indicates the theoretical maximum necessary labor time that

could be directly spent on agriculture. But the data con-

tains even more significant information.

Differences in the beginning, end, and duration of the

season of growth within and between regions discloses the

fact that the peak periods of demand for agricultural labor

and availability of agricultural produce were distinctly

 42

different. This possibility was ideally suitable to the

needs of those individuals and groups who, for various

reasons, had to undertake seasonal migrations. without

precise information on the nature and direction of move-

ments of groups and individuals within a community it is,

therefore, not legitimate to infer the duration of the

necessary labor from the length of the growth season. How-

ever, information on necessary labor time for the production

of various crops in a locality, in conjunction with data on

system(s) of communication to and from other places and the

labor requirements of other areas, can lead to the possi-

ble discovery of the range of time during which movement

was an actual possibility. I shall return to the issue

of movement of people and its articulation with larger

political and economic relations in the subsequent chapters.

Knowledge of the growth season is mainly helpful in

indicating the potential readi~ess of one factor of pro-

duction, soil, to be subjected to labor. Availability of

water is equally crucial and I will now turn to a discussion

of this factor. Systematic data on precipitation for the

period of our investigation is not available. The following

tables summarize the information for a more recent period.

 Table IIIa: Precipitation Data*

Average Median % % Varia- Years

Region Station Precip- Precip- r·1ax. Above Min. Below tion Ob-
itation itation Average Average Extent served

Eastern Gardez 316 314 564 +78 184 -42 130 10

Ghazni 296 278 555 +87 69 -77 164 15

Jabul Seraj 510 449 730 +43 125 -76 119 13

Jalalabad 172 147 390 +126 31 -82 208 15

Kabul 346 343 524 +51 176 -50 101 15

Khost 448 461 677 +51 185 -59 110 12

Lagham 275 277 526 +91 130 -53 144 9

Logar 227 213 392 +72 107 -53 125 7

Moqur 200 164 348 +74 129 -36 70 8

S. Salang 1096 1147 1741 +59 699 -35 94 12

Central Bamyan 182 100 503 +176 48 -74 250 5

Ghalmin 214 228 284 +32 133 -38 70 9

Lal 296 299 370 +23 137 -54 77 9

Panjab 334 261 502 +50 166 -51 101 9

Sharak 348 254 501 +44 71 -80 124 6

Southern Bost 90 97 158 +75 44 -51 126 14

Kalat 260 205 556 +113 104 -60 173 7

Kandahar 134 158 222 +65 57 -58 123 10

western Farah 84 88 202 +140 20 -77 217 14

Herat 207 214 401 +93 113 -46 139 16

Zaranj 30 34 55 +83 5 -84 167 5

*Adapted from Amin and Schilz, 1976:52

 44

Table IlIa (continued)

Average Median 0_
-0 % Varia- Years
Region Station Precip- Precip- Max. Above Min. Below tion Ob-
itation itation Average Average Extent served

Northern Bagh1an 280 285 413 +47 117 -59 106 16

Kunduz 335 313 477 +42 204 -40 82 16

N. Salang 1060 1212 1349 +27 432 -60 87 14

Maimana 376 375 482 +29 214 -43 72 15

Mazar 183 193 289 +57 57 -69 126 15

Sheberghan 227 188 364 +60 110 -52 112 11

Talooqan 642 559 873 +36 430 -34 70 5

Badakhshan Faizabad 548 551 703 +28 289 -48 76 14

 45

Table IIIb:

Seasonal Distribution of Precipitation and its Percentage in Comparison

to the Total Amount for the Year 1976-77*

SPRING SUMMER FALL WINTER

Total
Station Amount Amount Amount Amount
(rom)
(rom) % (nun) % Jl)1IlI) % (nun) %

Eastern Region

Ghazni 44.6 21.5 29.2 14.0 17.7 8.5 115.9 55.9 207.4

Jalalabad 59.7 32.5 67.7 36.9 0.3 0.16 55.7 30.4 183.4

Kabul 55.3 36.6 0.0 0.0 14.5 9.6 81.3 53.8 151.1

Khost 108.0 27.4 220.2 56.0 0.5 0.13 64.8 16.5 393.5

S. Salang 183.9 28.5 0.0 0.0 94.8 14.7 367.3 56.9 646.0

Southern Region

Kandahar 15.8 7.2 38.3 17.4 1.6 0.73 164.3 74.7 220.0

western Region

Herat 97.2 38.8 0.0 0.0 41. 2 16.4 112.2 44.8 250.6

Northern Region

Baghlan 149.6 48.6 0.0 0.0 26.7 8.7 131.8 42.8 308.1

Kunduz 84.2 31.6 0.0 0.0 25.8 9.7 156.3 58.7 266.3

N. Salang 276.6 33.5 1.5 0.18 165.0 20.0 383.6 46.4 826.7

Maimana 176.4 54.8 0.0 0.0 46.5 14.4 99.0 30.7 321.9

Mazar 44.2 31.4 0.0 0.0 19.4 13.8 77 .2 54.8 140.8

SlEberghan 60.2 30.9 0.0 0.0 33.1 17.0 101.3 52.0 194.6

*Statistical Information of Afghanistan 1976-77, 1977:53-54

Note: The duration of precipitation has been computed from the 10th day
of the first month of each season. Spring begins on the 10 of
Hamal (March 31) •
 46

The timing, type, and variations in the extent of pre-

cipitation calls for comment. The major period of precipi-

tation falls between the months of November to May. The

melting of the winter and spring snows accounts for the ex-

istence of late spring and summer water in the rivers. The

fluctuations in the amount of precipitation is fully docu-

mented in the above tables. with a range of variation that

is between 70% to 250%, 18 stations out of a total of 29

have a rate of 100%. During a drought year, the gap be-

comes even wider. The average yearly precipitation for the

whole country being 5415.7mm, the decline during the agri-

cultural years of 1969-70 and 1970-71 -- during which the

country was struck by a severe drought -- was 37% and 42.8%

respectively (Statistical Information of Afghanistan 1976-

77:50). But even in ordinary years, the change in the

yearly total precipitation is significant; the totals for

the years 1973-4, 1974-5, 1975-6, and 1976-7 were suc-

cessively 38l2.1mm, 4800.9mm, 5600.lmm, and 3831.2mm (Ibid).

While some of the variation might be due to the low number

of years recorded, two geographers have noted that "such

.variations are typical of semi-arid and desert climates"

(Amin and Schilz, 1976:51). Without giving the historical

basis for their judgement, they have also asserted that

the drought of 1970 and 1971 "may be repeated in 33 or 34

years" (Ibid). Irrigation under such conditions is, there-

fore, a necessity. In terms of Wittfogel, scheme conditions

 47

in Afghanistan fall into categories Band C, which imply

tendencies towards nomadism and irrigation agriculture.

Despite the general dominance of irrigation, the ex-

tent of climatic diversity in Afghanistan is such that rain-

fall agriculture is practised in most of the regions as

a supplementary activity. In some localities, it is even

the dominant form of agricultural production. Its contri-

bution to the total production of the country is, however,

relatively small as shown in table four.

Table IV: Irrigated and Rainfall Agricultural Production, 1977*

Total Agricultural Area Total Yield of Grains

Region Irrigated Rainfall Irrigated Rainfall
(jeribs) * % (jeribs) % (sirs) * % (sirs) %

Eastern 2,239,970 62.28 1,356,460 37.72 92,557,200 94.85 5,028,600 5.15

Central 1,113,000 67.03 547,570 32.97 40,774,800 89.52 4,775,000 10.48

Southern 1,715,900 85.51 290,620 14.49 59,392,800 96.47 2,177,400 3.53

Western 2,502,540 37.42 3,329,490 62.58 66,042,900 72 .24 25,373,600 27.76

Northern 4,643,320 45.33 5,601,030 54.67 187,986,300 78.04 52,814,400 21.96

Badakhshan 308,800 33.05 600,800 66.05 258,700 45.07 304,100 54.03

Total
for the 12,523,530 52.28 11, 429, 270 47.72 447,012,700 83.16 90,473,100 16.84
country

*Source: Statistical Information of Afghanistan, 1976-77:72,78-79.

One jerib is a square unit of measurement, the length of one side of which
is 44.183 meters; it is roughly equal to 0.5 acreas. One sir is roughly
equal to 7 kgms or 16 lbs.
 48

It can be easily seen that even though the area

covered by the two methods of production is roughly the

same, the difference in yields is great. However, the

total yield from the area under rainfall cUltivation in

the northern and western regions, in comparison to the

yield from the irrigated areas in other regions, is large

enough to make a significant difference in the overall food

supply in these regions in particular and in the country

as a whole. As far as can be determined, the significance

of rainfall in the agricultural production of the northern

and western regions is not recent. Ibn Hugal, the tenth

century Arab geographer, who had visited these regions and

noted the nature of the water supply for all the major

settlements of the period, wrote that the "best lands of

Khurasan [a region covering a large part of the present

Afghanistan, Central Asia, and Eastern Iran] are the irri-

gated fields of Nishapur and the area under rainfall that

is located between Herat and Mervrud" (1966:186).

Irrigated agriculture, however, is not a unitary

category. Since differences in the source of irrigation

have definite implications for the amount of necessary

time that is devoted to irrigation in the course of the

labor-process, different types have to be clearly distin-

guished and their labor requirements analysed. I shall

first present the necessary information regarding types

of irrigation in a more contemporary situation, and then

 Table V: Water Distribution - A Rough Estimate of the Number of Mills, and the Number of Various
units of Irrigation and the Areas They Irrigate (Base Year 1967)*

Areas Under Different Types of Irrigation (jeribs) Number of Different Types of units of
Region Irrigation and of Mills
Canal Spring Karez** Well Total Canals Springs Karez Wells Mills

Eastern 1,665,610 234,250 311,550 28,560 2,239,970 3310 2210 3203 2167 7673

% 74.35 10.45 13.90 1.30

Central 349,685 308,190 87,760 3,130 748.760 542 566 84 510 1917

% 46.70 41.16 11.72 0.42

Southern 1,345,770 108,120 257,350 4,660 1,715,900 705 1149 1650 460 1272

% 78.43 6.30 15.00 0.42

\\1estern 1,922,970 164,080 177,778 14,560 2,279,380 1731 869 632 3077 2695
o.
'0 84.36 7.21 7.79 0.64

Northern 4,516,980 103,330 4,050 8,960 4,633,320 1322 683 972 2227 4288

% 97.49 2.23 0.08 0.2

oJ::>
I.D
 Table V: (continued)

Areas Under Different Types of Irrigation (jeribs) Number of Different Types of Units of
Region Irrigation and of Mills
Canal Spring Karez** Well Total Canals Springs Karez vlells Mills

Badakhshan 289,160 19,190 0 450 308,800 212 82 0 54 730

% 93.64 6.21 0.15

'rota1s 10,090,170 937,160 837,160 60,320 11,926,130 7822 5559 6541 8495 18475

% 84.61 7.86 7.02 0.51

*Source: Statistical Information of Afghanistan, 1976-77:76-77.

**Karez is an underground channel.

U1
o
 51

address the issue of the labor requirements. The latter

question will be discussed on the basis of sources from the

period of long duration that is my unit of analysis in this

chapter.

The crucial importance of rivers to the agriculture

of the country is underlined by the 84.61% of the irrigated

land that derives its water from canals. The northern and

central regions with 97.49% and 46.70% respectively form

the extreme opposites of the range. Yet, if canals provide

the major source of water in all regions, other sources are

also important determinants of patterns of settlement and

food resources. In the central region, the percentage of

area irrigated by springs is a close second, while in the

eastern, southern, and western regions, underground channels

are the second most important source of water. In

Badakhshan, water from springs supplements that from canals.

The following figures portray the extent of area

covered by vurious u,nit.s of irrigation. A canal in ·the

eastern region provides water to 503.25 jeribs of land, the

lowest porportion for the country. In the central region,

the average area is 645.175 jeribs, in the southern region

1908.89 jeribs, in the western region 1110.90 jeribs, in

the northern region 3416.77 jeribs, and in Badakhshan

1363.96 jeribs.

The northern area with 4.16 jeribs per karez has the

lowest proportion for that unit. The average in the eastern

 52

region is 97.27 jeribs, in the central region 104.76 jeribs,

in the southern region 155.97 jeribs, and in the western

region 281.29 jeribs. Badakhshan has no underground chan-

nels.

The southern region with 94.09 jeribs per spring has

the lowest ratio for that unit. The average in the eastern

region is 105.99 jeribs, in the central 544.50 jeribs, in

the western region 188.81 jeribs, in the northern region

151.28 jeribs, and in Badakhshan 234.02 jeribs.

\vith an average of 1.48 jeribs per well, the west has

the lowest ratio for that unit. The proportion in the

eastern region is 13.17 jeribs, in the central region 6.14

jeribs, in the southern region 10.13 jeribs, and in

Badakhshan 8.33 jeribs. The low ratio of wells to the

area under irrigation is a function of the fact that wells

are generally the source for drinking water.

with the exception of the western region where wind

power has been used to operate mills, water mills are the

dominant form in the country. Their number in a region is

a reflection of the velocity and length of the units of

irrigation.

But the regional averages, by hiding important intra-

regional variation, might be misleading. The average area

per canal for instance in the western region is 1110.90

jeribs. The range of intra-regional variation per canal

is from the lowest figure of 347.73 jeribs in Ghur to the

 53

the highest figure of 2646.52 jeribs in the valley of Herat.

In between, the ratio is 843.92 in Badghis, 1423.65 in

Farah and 1547.62 in Nimroz.

The significance of the ratio of a unit of irrigation

to the area it waters lies in the manner it affects the

proportion of necessary and surplus labor-time of social u-

ni ts that are brought into interaction because of irrigation.

As elucidation of some of the critical features of the ar-

gument requires historical depth, I have opted in this sec-

tion for a longer time framework.

The fertility of the valley of Herat, in the western

region, has been a constant theme in the writings of local

and foreign observers alike. An English officer affirmed

in 1885 that "the whole of Herat Valley, from Obeh to

Kushan, or a distance of nearly 120 miles by a width of per-

haps 14 miles, is cultivated like a garden. Everything

grows with the greatest luxuriance and in the greatest pro-

fusion. Wheat and barley enough to feed man and horse

of a large army are always procurable, while fruits, such

as peaches, apricots, grapes of many varieties, plums of

many sorts, walnuts, and almonds are grown in plenty" (Col.

Stewart to Marquis of Salisbury, F.O. 539/44 Inc. in no. 27,

July3,1885:42).

A description of the valley by a fifteenth century

Herati geographer has a more direct focus on irrigation.

He wrote that "the dependencies and districts of Herat are

 54

are Bulukat and Willayat. Those that are in the valley of

Herat and close to the city are called Bulukat. Between

Herat and Fusjanj, which is approximately fifteen farsangs

[unit of length approximately equal to 18,000 feet] in

length and five farsangs in width, the length of the valley

is from east to west. Bulukat are situated both on the

northern and southern side of the river, which is in the

middle of the valley. From both sides of it canals have

taken and villages and fields have been constructed on it.

Evey juybar (main canal) is called a Buluk" (Geographia-i-

Hafiz Abru,1349/1970:15).

The magnitude of the labor required for maintaining

the irrigation system was noted by an English observer in

1885: "The irrigation-works are certainly one of the won-

ders of this country. The valley here is a perfect net-

work of canals and juis, as they are called, varying in

size from some 30 feet in breadth and 2 in depth to the

smallest cut of barely a foot in breadth. The annual labor

expended in the repair alone of the canals is very great;

but for all that, the people apparently prefer canals to

any system of well irrigation, which is here unknown" (Yate,

1888:17).

Thanks to the survival of the handbook that regulated

the labor-requirements of the valley, the issue can be

discussed in quantified terms. The work was written at the

end of the fifteenth ceritury or the beginning of the six-

 55

teenth. It refers to all the previous regulations on

irrigation and it has been the basis of all subsequent re-

visions until the 1970s [Mayil Hirawi, Introduction Tariq-

i-Qismat Ab-i-Qulb ("Method of Division of Water of Qulb")

1969: ]. I have summarized the relevant information into

the following tables.

Table VIa: Labor-Requirements of Irrigation in the

Bulukat of Herat

Number Average
Length
of Number of number
Number of main Number
Name of of main canal (s.) men-days of men-days of men-
per settle- per days per
Buluk canals (zara=1.5ft)
main secondary settle-
ments
canals canal ments

Anjil 2 38,000(7.19mi) 351.5 59 505.5 14.53

Alinjan 5 20,000(3.78mi) 244.0 42 269.0 12.21
19,000(3.6 mi)
11,OOO(2.08mi)
Guzara 3 49,000(9.28mi) 248.5 117 441.0 5.89
17,000(3.22mi)
19,000(3.6 mi)
Khiaban 1 37,000(7.0 mi) 208.0 71 409.0 8.69
Sabqar 3 116.0 20 164.75 14.04
Adwan and 114.5 41
1 206.0 7.82
Tizan
Turan and 20.0 24.0
1 9 4.88
Tunian
Ghurwan and 35.0 32.0
1 13 5.15
Pashtan
Kambraq 2 151.0 26 296.0 17.19

Totals 18 210,000(39.77mi) 1488.5 398 2347.25 9.64

 56

Table VIb: Labor-Requirements of Irrigation in the

Wil1ayat of Herat

Number Average
of Number of number
Length Number
Number men-days men-days of men-
Name of of main of
of main per per days per
Wilaya cana1(s) sett1e-
canals main secondary sett1e-
(zara=1.5ft) ments
canals canal ments

Kardbar 1 irregular 60 16 49.0 6.81

Fushanj 7 230 39 510.5 18.98

Shafilan 115 33 165.0 8.48

Karoukh 1 100 8 124.0 28.0

Totals 9 505 96 848.5 14.09

Source: Qasim Bin Yusuf Abu-Nasr Hirawi Tariq-i

Qismat-iAb-i-Qulb: 16-83 for Bu1ukat;
and 85-97 for Wi11ayat.

The figures of 3835.75 man-days of labor for the Bulukat

and 989.5 man-days for the Willayat represent the very mini-

mum amount of labor that had to be expended on the mainten-

ance of the canals. Since the task varied from year to

year, the handbook does not go into further details. How-

ever, the context makes it clear that, depending on the

duration of the repair of a main or secondary canal, the

settlements were to provide labor in the same ratio until

the completion of the tasks. Under the unlikely assumption

of one day for the completion of the task, a settlement, on

 57

average, had to contribute the labor of 9.64 men in the

Bulukat and 14.09 men in theWillayat. On a more reason-

able assumption of 10 days of cleaning and repairing canals,

the result would be an expenditure of 38357.5 man-days for

the Bulukat and 9895 man-days for the Willayat. The corre-

sponding averages for the settlements would have been 96.4

and 140.9 man-days respectively. The calculation is en-

tirely based on the work that has to be performed in the

beginning of the year. Any work that had to be performed

in the course of the agricultural year would then increase

the total amount of necessary labor.

The high demand for labor of irrigation agriculture

in the Herat valley comes to full light in the Spring.

with a growing season of 228 days the last freezing

day of Spring being March 19 -- this time of the year

is one of peak season of labor activity. Therefore,

the greater the time spent on the maintenance of canals,

the lesser the time available for activities such

as plowing that had to be performed at the same time. In

comparison with rainfall agriculture not only is the amount

of necessary labor in irrigation agriculture greater; the

time spent on irrigation may significantly reduce the social

time for the undertaking of activities which due to natural

conditions have to be performed in limited periods.

Production under conditions of irrigation seems to

impose specific patterns of intra- and inter-community

 58

cooperation. In the valley of Herat for instance, the

tasks had to be distributed among the settlements along

every main canal. Subsequently, for every outlet along

the secondary canals, the tasks had to be distributed again.

In both areas, the specifics of labor allocation had to be

worked out by the merr,bers of the settlements concerned.

That the process may have generated a certain amount of ten-

sion can be easily imagined. But if the tasks were to be

successfully performed and the working of the entire system

insured, cooperation at various levels of the community was

a must. The political implications will be taken up in the

second and third part of this study. As far as the techni-

cal aspects of the maintenance of canals and the distribu-

tion of water were concerned, the details were worked out

in such a manner that they were easily understood.

The cycle of fallowing and the velocity of water rela-

tive to the location of land were the basic variables in

determining the number of laborers and allocating water"

The soil in the valley of Herat was cultivated either every

year, every other year, or every three years. Depending

on the ratio of the three types of land in a settlement,

the grouping of man-days would have fallen on the same or

different individuals in different years. As this informa-

tion was most readily available at the level of the settle-

ment, it was logical that an organization there would work

out the details.

 59

Determination of speed of water involved a more com-

plex series of operations. Lands were categorized as up-

stream, midstream and downstream with respect to the sub-

canal from which they were watered. The sub-canals diverted

water from a canal through the construction of a flat darn,

called nutra, which was to insure even distribution to all

the channels. The opening through which measurable amounts

of water were to pass was called a qulb and it was con-

structed according to fixed specifications. This was clearly

distinguished from an irregular fissure which was called a

rakhna. The latter was usually not allowed on a properly

maintained canal or subjected to different requirements of

labor and material.

Evey qulb was divided into 18 units. Location of land

relative to the course of water directly affected the defini-

tion of the area and the labor requirements of a piece of

land. While the unit of land that was worked by a man with

two oxen was the basic unit of measurement, its actual size

varied from plot to plot. The reason was simple: in a plot

of land located upstream the qulb was supposed to be 80

jeribsi midstream the unit was 50 jeribsi and downstream

it was a mere 30 jeribs. As the speed of water decreased

while spreading over the land, the ratios of water followed

a course opposite to the amount of land. Plots upstream

received 4 units of water, those midstream 6 units, and

those downstream 8 units. The ideal unit of cooperation

 60

was visualized as a team of three share-croppers who cul-

tivated an area of 160 jeribs watered from one qulb (Ibid:

12-15) .

This was the picture at the micro level. In order to

insure an even flow of water in the course of main and

secondary canals, the whole course of the canal was accur-

ately measured. As it was known that the flow of water in

the month of Saratan (June 22-July 23) was uniform, the

amounts in other months were adjusted accordingly.

Water supply along the course of the river helped in

insuring access of the various settlements to water. As

the fifteenth century geographer Hafiz-i Abru noticed,

IIthis river has the characteristic that when its water is

dammed and diverted to a canal, only a short distance after

the place that has been rendered dry water is once again

found in plenty II (1349/1970:15). It seems that springs

were fairly evenly distributed along the course of the

river. Furthermore, excess water from the Willayat was

regularly diverted from those Willayat which were located

upstream to the Bulukat that were located downstream. The

water from Karoukh, for instance, was taken three times a

year -- 7 days out of every 21 days -- in fixed portions

to the Bulukat (Hirawi Tariq-i. Qismqt-i Ab-i Qulb: 96-97) .

Efficient working of the system required that a number

of individuals devote their full time to the mastery of the

rules and the supervision and routine maintenance of the

 61

works. Every qulb was entrusted to a Sar-qulb (literally

head or cover) who was paid a fixed sum in money. Canals,

depending on their length, were supervised by one or several

Mir-Ab (head or master of water) who were paid a fixed

amount in kind by every cultivator (Ibid:20ff).

From the twelfth century on, literature describing the

devices that accurately marked the volume of water and the

passing of time is abundant (Khazni, Abu AI-Fatah, Mizan

Ul-Hikmat "Scales of Reason" 1346/1967:18-21). There is

no contemporary description of the system for the period

under study for the valley of Herat. But a detailed des-

cription of the system in another locality in the western

region demonstrated that despite its complexity, the opera-

tion of the system was fairly easy. Ward, an irrigation

officer attached to the Sistan Mission of 1903-1905, des-

cribed the irrigation system of the lower Helmand river

in the Sanjarani tract of country in the following terms:

"The distribution of the water from the canal to the

branches is made very scientifically by weirs so that each

branch may take a discharge proportionate to the number of

Bazgar [cultivators] on the branch. The place where several

branches take off is called an aubakhsh or regulator. The

weirs are made by spreading tamarisk mats on the bed and

sides of the branch at its head; the tamarisk mats on the

sides are held up by stakes, while a piece of tamarisk cut

by the village carpenter to a rectangular shape serves as

 62

a board to form the sill of the weir. If two or more

branches take out side by side one continuous piece of board

forms the sill of them all so that any tampering with the

head of one may be detected by an alteration on the heads

of the other branches.

"The sill is kept at the same level and the same

depth of water passes over the sill into all the branches.

The irrigating capacity of each branch is altered by al-

tering the width of the sill ... The correct lengths of

the weirs are obtained tentatively; the Kadkhuda [village

headmen] accompanied by the village carpenter makes the

weir the length that he believes will be correct after

allowing for the number of Bazgar on each branch, the length

of the branch and the height of the land to be irrigated

with reference to the supply level in the canal. As a rough

rule, one to two nakhun (finger's breadths) are allowed for

each Bazgar. (It is said that the long branch of Chahar

Burjak canal for Deh Ghulam Haidar is allowed l~ times its

share as determined on the count of the cultivations to

compensate for the greater distance the water has to travel.)

"If there is any doubt, the water is made to flow in

the branches and the area of land watered by the different

branches in a day is compared to see if it is correct. If

not, the weirs requiring alteration are attended to. The

area is noted in Langar or the area that can be sown by

one plough in one day. So far as is possible all their

 63

canals are so made that all the water-courses may run con-

tinuously, and this is generally possible as the canals on

the river take out from branches of the river which them-

selves come from above natural weirs so that by closing

the branch of the river by a dam at the place where the

canal leaves it a good supply is usually obtained. If the

river is abnormally low a weir or even a dam is built

across the river at the place where a branch feeding the

canal leaves it. (From July to September 1902 all the

water of the Helmand was turned down the Rudbar canal

leaving the river dry below the dam made at the head of the

canal. )

"If more land has for some reason been given out on

the canal than its volume can irrigate a rotational closure

of the branches is instituted, but this is very seldom

needed.

"The supply in the branches among the Bazgar is made

by rotational turns; each Bazgar usually gets the water for

two days and two nights. The supervision of the weirs of

the canal is entrusted to a petty official called the Kotwal

and he guards the weirs to see that they are not tampered

with." (GA,Farah 1914/1973:262-63).

A major difference between the system in Sistan and

the Herat valley and most of the rest of the country was

that the annual silt clearance was done after the harvesting

of the autumn crops. This was due both to the longer grow-
 64

ing season and to the fact that the headwork of the canals

was secure. When the rise in the river was enough to damage

the headworks "the cultivators [had to] turn out and make

good the damage or they [would have not gotten] water to

finish their sowing or mature their Spring crops" (Ibid:264).

What impressed Ward most was the fact that the elabor-

ate system was worked by formerly nomadic Baluchis who had

settled in that tract only about 100 years prior to his

visit (Ibid). It seems that mastery of the system had

posed no major obstacles.

The secure nature of the headworks in Sistan had a

direct effect on the amount of the labor required to main-

tain it and poses the question of the ratio of necessary

and surplus labor for different social units of a community

in a new light. Being surrounded by a desert until the

fourteenth century Sistan had one of the most sophisticated

systems of irrigation in Central Asia and the Middle East.

Tarnerlane (1370-1405) having met strong opposition from

the inhabitants of the area, destroyed the irrigation net-

work and moved most of the population to other parts of

his domains.

Fortunately, an anonymously written history of Sistan

of the eleventh century provides an account of the revenue

and expenditure of the local state. Out of a total annual

revenue of 3,597,000 dirhams only 139,000 dirhams -- a mere

3.8% -- was spent on the maintenance of the agricultural

 65

system. The breakdown is revealing: 4,000 dirhams for re-

pair of damages done by floods to the farmland; 25,000 for

the dams with more available if need arose; 30,000 to the

sand dammers; 50,000 for the protective earth embankments;

30,000 for the care of bridges, canals, streams, and the

boat routes along the course of the Helmand river (Tarikh

Sistan 1976:21-22).

Most of the revenue of the state was undoubtedly ex-

tracted from the direct producers in the form of taxes.

However, the assumption of activities related to the main-

tenance of the agricultural system could not fail to affect

the ratio of necessary and surplus time among the other

organizations in society. Removal of the labor requirements

dictated by the maintenance of the system by the settlements

would have considerable impact on the forms of cooperation

that this aspect of the system of production required. This

in turn may have influenced the patterns of solidarity of

the communities and thereby affected the political balance

of pmler between the state and the localities.

Analytically, the coordination of activities dealing

with maintenance of a system of irrigated agriculture has

to be distinguished from activities related to the creation

of such a system. I shall examine these issues in greater

depth later. Meanwhile, forms of irrigation other than

canals have to be explored.

As far as the construction and maintenance of under-

 66

ground channels (karez or qanat, pl. Karezat or qanawat)

was concerned, the work had to be performed by specialists

and did not directly affect the ratio between necessary

and surplus labor of the direct producers. However, as the

specialists had to be paid in cash, in kind, or in a com-

bination of both, expenditure on construction required a

fairly large initial outlay and yearly expenses of mainten-

ance, by affecting the net amount of the produce that re-

mained at the disposal of the producers indirectly entered

into the balance of necessary and suplus labor time.

The actual cost of the construction was determined by

the distance of the underground water from the fields, the

nature of the obstacles in the terrain, the type of soil

through which the channe1(s) had to be constructed, and

the skill of the workers. While most of the work in

Afghanistan is still being performed by travelling specia1-

ists who belong to a number of known kinship groups, avai1-

able literature from as early as the twelfth century offers

detailed ana~ysis of types of underground water, methods

of discovering it, and detailed instructions on the con-

struction and maintenance of the channels (Al-Karaji, Abu

Bakr Muhammad Bin al-Hassan al-Hasab Istikhraj Ab-ha-i-

Pinhani "Extraction of the Hidden Waters" 1345/1966:15-127).

If a channel was properly constructed, maintenance

chiefly involved the cleaning of the accumulated mud.

According to Karaji, the basic criterion for the payment

 67

of the cleaners was the hardness or softness of the soil of

the channel. They were to be paid for work on a unit of

length that was equivalent to 1.5 ft (Ibid:126). As Karaji

clearly recognized, what made investment in a karez a suc-

cess or failure in the long run was whether the underground

source of water was continuously renewed from above the

ground or from other underground sources or whether it was

an isolated supply that would be exhausted after some time

(Ibid:15-l8).

The nature of the terrain at times significantly re-

duced or eliminated the human labor input. As the channel

had outlets to the surface at fixed distances, in case of

blockage or water at any given point the whole channel

did not need to be cleaned.

Although canals were the dominant form of irrigation

in the country as a whole, the share of the underground

channels around central areas and in places where no other

sources of irrigation were available was significant. The

following table summarizes the role of the channels in

the irrigation of a number of districts and dependencies

of Herat in the fifteenth century.

 68

Table VII: Forms of Irrigation Herat, the Western Region

Name of Buluk Location Major Expression

oe Wilayat with respect form of used in
to the river cultivation describing the
source of water

Anjil North of the River The city;gardens From the river

Alinjan North of the River Gardens and fields From the river

Guzara South of the River Fields River and some

Karezat

Khaiban North of the River Fields; graveyards River and Karezat

of the city but more from river

Sabqar North of the River Many gardens;fields From the river

Adawan & South of the River Gardens; fields River and Karezat
Tizan

Turan & North of the River More fields than From the river
Tunian gardens

Ghurwan & North of the River More fields than Many Karezat
Pashtan gardens

Kambraq South of the River More fields than River and Karezat
gardens

Parwana & Mostly fields;few Mostly Karezat

Haudshtag gardens
/

Fushanj Mostly gardens From the river

especially grapes

Shafilan Mostly gardens;

also wheat

Karoukh 89 villages in the River and Karezat

register; used to
be very fertile

Source: Geographia-i-Hafiz Abru

 69

The author does not, unfortunately, provide any infor-

mation on the length of the channels or the number of cul-

tivators that made use of them. There is no systematic

data on the extent of the channels in other regions either.

But there is information on the number of channels and

families living in settlements dependent on them for the

area surrounding the central place of Kandahar in the

southern region.

In 1879-80, at the foot of the mountain close to the

central place of Kandahar, there were 62 karezat in one

cluster. There were 550 houses in this area. With a range

of 0-70 houses per karez, the mean size of houses dependent

on a karez was 8.3. As only in three cases the number of

house~ per unit of irrigation w~s greater than 20, it can

be assumed that the volume of water from an average karez,

which conditioned the size of the cultivable area, was not

very large. The issues of access to water, rights to land

adjacent to a karez and construction of new karezat, being

intimately connected with the social organization of pro-

duction, will be addressed in the next chapter.

For the people whose irrigation water was derived from

springs, the ratio of necessary and surplus labor was direct-

ly affected by the relative ease of access. Volume of water,

nature of terrain, relative size of population and land in

the settlements dependent on a spring, and the distance from

the source of water have all been significant variables in

 70

determining the labor requirements of a system. I will

illustrate the degree of complexity of such systems through

a description of a unit in the valley of Logar in the

eastern region. The relevant information was collected

during fieldwork between 1973 and 1977.

The spring, in this case, was located in the middle

of solid rock in a mountain the peak of which was snow-

clad most of the year. An indication of the velocity of

the water is provided by the fact that even in the last

settlements along its first 20 miles above the ground, it

can turn a mill around the clock. Close to the last above

ground settlement, an underground channel of approximately

20 miles takes the water to a settlement of approximately

2,000 houses. Settlements deriving their water from its

above ground course are dispersed and vary in size from

5 to 50 houses.

The spring itself does not bear any signs of past la-

bor and requires no annual labor to keep it functioning.

Labor spent in bringing the water to the fields and annual

maintenance of the system has differed in the two sections

of the channel leading from the spring. Most of the above

ground course seems to have been shaped by the rapid flow-

ing action from the mountain to the valley. The annual

labor spent on maintaining the system in its upper part

is fairly small and, without requiring the coordination

of all the settlements, is carried by residents of the

 71

various settlements.

The single large settlement at the end of the second

portion of the channel is not located along the natural

course of the water and required the construction of an

underground channel to guarantee its access to water. As

the length of the channel is approximately 20 miles, the

labor spent on its construction must have been considerable.

Pre-Islamic and early Islamic pottery has been continuously

found during plowing and the system might be more than a

thousand years old. However, there is no information on

the way the underground channel was constructed. The rapid

velocity of water prevents the accumulation of unwanted

mud and thereby lowers the amount of labor related to main-

tenance. But when something does go wrong with a part of

the channel, it requires immediate attention from special-

ists.

The most significant form of annual labor expenditure

has not been related to the physical maintenance of the

system but with guaranteeing access to water. By ,custom,

the dispersed settlements along the above ground section

have the right to the water during the day and the large

settlement at the end of the underground channel during the

night. Since the course of the water has innumerable out-

lets in the above ground section, anyone upstream can di-

vert the water to his fields. The occurrence of the event

during periods of water shortage is frequent between the

 72

the various settlements along the upper course itself and

almost universal in regard to the nightly use of water by

the residents of the large settlement. Therefore, a spe-

cial mechanism for guaranteeing the delivery of the allo-

cated water has been devised. It is called pay-aba (liter-

ally feet for water) and involves the policing of the whole

upper portion of the channel by members of the large settle-

ment during the night and by those located'upstream during

the day. During the period of my fieldwork, the number of

men on nightly assignment from the large village came

close to 200. They were recruited from families who were

supposed to have access to the water at that time.

Irrigation from wells, by contrast, was more of a

limited enterprise. The construction and maintenance of

wells was a job for professionals. The mechanical device

for lifting water to the surface known as the Persian

wheel requires animal-power -- horses or donkeys -- for

operation. The usual size of an area irrigated by such a

unit varied from 5 to 20 acres. This type of irrigation

has been commonly used in providing water to walled gardens

and vegetable fields. Since such operations have required

very regular schedules of watering, the investment in wells

has freed the labor of the producer from the uncertainties

of water-sharing arrangements. While the labor needed for

maintaining such facilities is small in comparison to that

required by other types of irrigation, the initial cost of

 73

investment was quite large and could only be undertaken

by a minority of the population.

Regardless of the source of water, the needs of the

cultivators for water was not fixed. It varied with the

type of the soil, the cycle of fallowing, and the needs

of different plants. That all these factors were taken

well into account by the cultivators is revealed in a hand-

book of agriculture written in 921 A.H. (1515 A.D.) by

Qasim Bin-Yusuf Abu-Nasr Hirawi. He compiled a handbook

on irrigation on the basis of detailed conversations with

the cultivators of Herat (Irshadual-Ziraat "Guide to Agri-

culture" 1346 A.H./1967:45) and as such, his handbook re-

flects the practices prevalent in the western region.

The author describes a number of methods for deter-

mining the quality of the soil. One of these involved

taking some soil from a depth of two to three yards below

the surface of the ground, pounding it into powder, mixing

it with rain water, and waiting until the water was clear.

If the taste of water had not changed, the soil was of good

quality but if it had become salty the soil was saline.

A bad smell from the soil was interpreted as a sign of poor

quality (Ibid:54-55). The author offered an eleven-fold

classification of the types of soil and their associated

characteristics. I will quote a number of these for illu-

stration.

"Z amin-i-Rigbum (Sandy earth) is of two types: in one

 74

type there is more sand than earth and plowing it twice is

useful. When cultivated it grows well and since it retains

the moisture, it does not require much water. It matures

quickly but the product is poor. In the second type, there

is more earth than sand and plowing it three times is appro-

priate. Whatever is sown grows well and the product is ex-

cellent. It reaches maturity ten days before the [crop

grown on] Shakhriq (hardgrained sand). As it retains water,

it does not require much watering. The more it is ferti-

lized the more the profit. In this type of soil all pro-

ducts grow well.

"Shakhriq (hardgrained sand): plowing it four times

is best. When cultivated it grows well and the product is

reasonable. Large amounts of water and fertilizers are

profitable to it. Vines grow well in such soils.

"Siahriq (black sand): What is sown on it grows well

and does not require much water but the product is not

good. However, trees and vines grow well on such soils

and give good yields.

"Shakh (hard soil): plowing it four times is appropri-

ate. All of what is sown on it does not grow. Since it

does not hold moisture it requires a lot of water and ferti-

lizer but its product is excellent. Plowing and harrowing

it should be undertaken when the soil is moist. Grapes grow

well on such soils.

"Zard Khak (yellow earth): plowing it four times is

 75

best but the soil lacks strength and what is cultivated

does not all grow. It requires little water and a lot of

fertilizer but the product is poor.

"Siah Khak (black earth): it is plowed four times and

as the soil is strong all that is sown grows very well. If

a lot of fertilizer is available it should be used and its

water requirements are moderate. The product is excellent

and most cultivators are of the opinion that everything

grows well on such soils and that such land is the best

of all and its utitlities are countless ... (Ibid:55-57).

It is obvious that regardless of the nature of the

system of irrigation, the ratio of necessary and surplus

labor of the producers had to reflect the nature of the

soil which was subjected to labor. The suitability of

the soil to crops in turn affected the seasonal intensity

of the labor of the producers. The issue is well illu-

strated in Hirawi's discussion of the timing of agricultural

operations that the cultivation of different ~rops and

varieties of the same crop entailed. Out of hundreds

of items that Hirawi provides information on, I have cho-

sen four commonly cultivated crops for illustration.

a. Vines. The timing of the various operations is

summarized in the following table.

 76

Table VIII: Time-Table of Tasks for Vines and Trees

Period of the Year Tasks to be performed

Mid-Farvardin to ivlid-Ordibehesht Vines have to be watered once and

(April 5 - May 5) another time after budding; the
soil around the vines has to be
dug to help maintain moisture.

Mid-Ordibehesht to mid-Khordad Pruning of the vines. Beginning

(May 5 - June 7) of watering trees except fig trees.

Mid-Khordad to mid-Tir The soil under the vines has to be

(June 7 - July 7) dug again but not too deeply so as
not to expose the roots to heat.

Mid-Tir to mid-Mordad Pruning of vines. Pruning of ex-

(July 7 - August 6) cess of grapes from young vines.
Trees are grafted in this month.

Mid-Mordad to mid-Sharivar Branches of vines marked for

(August 6 - september 6) planting are lubricated with a
mixture of tar and oil.

Mid-Sharivar to mid-Mehr Vines are fertilized and almonds

(September 6 - October 7) and figs planted. Fruits to be
kept for winter are collected.
Land is tilled.

Mid-Mehr to mid-Aban Vines should be planted in case of

(October 7 - November 6) rain; this speeds up maturation to
the extent of a full year.

Mid-Aban to mid-Azar In some places vines are planted.

(November 6 - December 6) Other trees planted mature quickly.
Trees cut for their wood.

Mid-Azar to mid-Dey Pruning of vines by some people but

(December 6 - January 6) to prevent freezing it should not be
done during first and last 3 hours of
daylight. Cutting trees for wood.
Planting of almond, peach and apricot
trees. Pruning of trees other than
vines on unwindy days.
 77

Table VIII (continued)

Period of the year Tasks to be performed

Mid-Dey to mid-Bahman Transplanting of two and three year

(January 6 - February 5) old vines; one year vines not to be
transplanted. Planting of flowers.

Mid-Bahman to mid-Esfand Frequent watering of vines and

(February 5 - March 5) trees. Some grafting. Pruning of
three year old vines.

Mid-Esfand to mid-Farvardin Planting of olives, pruning of

(March 5 - April 5) other trees, grafting of some
trees; transplanting fig tress.
Some pomegrantes.

Source: Hirawi Irshadu al-Ziraat: 76-78.

Hirawi provides detailed description of the schedule of tasks

that one year, two year, and three year old vines involved.

As there were more than eighty varieties of grapes in the

valley of Herat, the time they became ripe for picking

ranged from June till September (Ibid:105-ll6).

b. Barley. Hirawi provides a list of 14 types of bar-

ley, their dates of cultivation and the time of harvest.

Except for one variety, that matures towards mid-May, har-

vest time for most varieties was during the month of Jawza

(May 22 - June 22). Planting took place during the following

months: 1st of Sunbulah (August 23) for one variety; Mizan

or Hut (Sept. 23 - Oct. 22; Feb. 23 - March 20) for three

 78

varieties; Hut (Feb. 23 - March 20) for four; and Dalwa

(Jan. 21 - Feb. 23) for five (Ibid:88-90). Harvest time

for the most commonly cUltivated variety during the l880s,

according to British observers, was about the 20th of June

(GA Herat:7) .

c. Wheat. Hirawi recorded 16 varieties of wheat cul-

tivated in the valley of Herat and distinguished between

those suitable to irrigation and those suitable to rain-

fall agriculture. Harvest time for all of them fell in

the month of Saratan (June 22 - July 22). Spring and Fall

were the two main seasons of cultivation. The planting

of the Spring crop took place between February and mid-

March. The cultivation time of the ten Fall varieties

was during the month of Mizan (Sept. 22 - Oct. 22). Hirawi

recommended that seeds be changed every three years and

that utmost attention be paid to the proper timing of irri-

gation (Irshad ul-Ziraat:79-80) .

d. Cotton. As Hirawi's description of the performance

of tasks and watering arrangements for this crop is very

succint, I will quote him in detail.

"Time of planting is the lOth of Saur (May 1st). Wa-

tering the fields at the beginning of winter when they are

froz8n will kill any worms that may be there. The the fields

have to be plowed to loosen the roots of the weeds. Expo-

sure of the clods to rain and snow makes the earth strong

and prevents the buds of cotton from falling in case they

 79

do not receive sufficient water during the summer. The

fields should be harrowed in the winter and fertilizer

spread over them so that rain and snow mix with the earth.

In Hamal and Saur [March 21st to May 21st] the fields

should be plowed seven fold, the clods broken with a hatchet

and left for three days to warm up. Whenever each row has

the proper amount of soil and the crop is planted regularly,

the harvest is good. If the field is full of clods and

lacks sufficient soil, the crop will not grow properly. To

prevent the occurrence of worms, the seeds of cotton should

be mixed with the seeds of Hanzal [Cucumis melo] in a solu-

tion of water and left for a day. After that, the fields

should be watered and on the second day 2 mans [24 Ibs] of

seeds per jerib [0.5 acres] be planted. On the third day,

it should be given one finger of water and left until the

crop sprouts. Then it should be weeded and when it becomes

like a twig and the stem which is green tends to red up-

wards it should be watered in the manner of daghab. This

means that the water be four fingers away from the stem of

the cotton but moisture will reach the roots. If it is

watered fully it will not grow well. After that, it should

be irrigated four times in such a manner that in each week-

ly round of irrigation the water should not fallon the

cotton buds, as it will obstruct their growing, but the wa-

ter should be in motion. Weeding should be carried on as

the need arises. On a flat surface, water should be spread

 80

from down upwards. Until all the sterns are of the same

height, the fields should be watered uniformly ... Then,

whenever the soil appears yellowish it should be watered ...

When the crop is in bloom, it should be irrigated weekly

and when the buds have become firm it should be irrigated

every three days. For cotton to be soft and plentiful

there is need for a lot of water •.. Its maturation is from

the 15th of Asad until the 15th of Mizan [August 7 - October

8]" (Ibid:15l-52).

Throughout his discussion of agricultural tasks,

Hirawi takes the existence of the plow as the basic instru-

ment of production for granted. Indeed, throughout

Afghanistan, during our unit of long duration, the basic

units of measurements were defined in terms of the area

of land that a cultivator using a team of oxen could plow.

While oxen were preferred for plowing, there is at least

one reported instance where, in the absence of oxen, horses

camels and donkeys were yoked together to a plow in the

northern region (Yate,C.E. Northern Afghanistan 1888:132).

Comparison of the growing seasons of the four crops

reveals the periods of planting and harvesting as the peak

periods of labor activity. By contrast, the rest of the

growth season seems to have been a period of uniform labor

expenditure. Whereas most of the tasks related to the

raising of these crops would have been the same if watering

arrangements were not taken into account, the water require-

 81

ments of each crop involve crucial differences in the

amount of necessary labor as well as socially necessary

arrangements to render the water available. Focus on labor-

time l therefore, can meaningfully reveal the specificities

of a system of production based on irrigation agriculture.

But the unit of analysis has to be shifted from so-

ciety considered abstractly to more concrete social entities.

Juxtaposition of any social entity's labor profile with in-

formation on labor demands generated by the totality of the

production system -- irrigation, soil, crops, length of

growing season -- will allow for concrete examination of

the ways in which the technical requirements of production

have been mediated through the existing social relations.

Recognition of the differences in the combination of these

factors will allow us to understand the significance of

regions as preliminary units of analysis as well as en-

abling us to place the regions in the larger social flows

which are characteristic of society defined as people in

motion. It is time to move from the determinate technical

requirements of the labor-process to a consideration of

the social organization of the labor-process.

 82

CHAPTER THREE

Social Organization of Production

In his attempt to elaborate an economic theory of the

feudal system, witold Kula asserts that "the task of every

economic theory of a system consists in formulating the

laws governing the volume of the economic surplus and its

utilization ••. , and that these problems have to be ex-

plained in the short-term and in the long-term" (1962/1967:

16-17). Kula does not define the concept of economic sur-

plus himself but refers the interested reader to the work

of Paul Baran who differentiates between actual and poten-

tial economic surplus. Actual economic surplus is defined

as the "difference between society's actual current output

and its actual current consumption" (1957/1968:22). Poten-

tial economic surplus is defined as the difference between

the "output that could be produced in a given natural and

technological environment with the help of the employable

productive resources, and what might be regarded as essen-

tial consumption" (Ibid:23). Baran offers four criteria

for estimating the extent of potential economic surplus

that drastic reorganization of production might bring about.

These are: 1- society's excess consumption -- predominantly

associated with the upper income groups; 2- output lost to

society through the existence of unproductive workers; 3-

output lost because of irrational organization of produc-

 83

tive apparatus; and 4- output lost due to the existence of

unemployment.

Baran recognizes the fact that rationality, produc-

tivity, and wastefulness in the utilization of resources

are determined by the structures of social orders and that

they display different features in different social orders.

His inquiry, however, being conducted in terms of "compara-

tive statics" requires that "paths of transition from one

economic situation to another" be ignored and situations

considered as ex post (Ibid:22). Furthermore, as the pri-

mary focus of his analysis is on monopoly capital, little

effort is spent on testing the concept historically.

In anthropology, however, the debate has revolved

around the utility of the concept of surplus in the under-

standing of the evolution of stratification. In 1957,

Harry Pearson sharply challenged a then widely held assum-

tion that the existence of surplus is a necessary precon-

dition for the emergence and development of stratification

and forms of complex social organization. He asserted

that "the surplus theorem is useful only where the condi-

tions of a specific surplus are institutionally defined"

(1957:321). Furthermore, he argued that postulation of

surplus as either a cause of change or as a necessary but

insufficient condition of change was logically inadmissable.

The reason for this stand was Pearson's belief that it was

impossible to determine the subsistence needs of a society

 84

and that "man, living in society, does not produce unless

he names it such, and then its effect is given by the man-

ner in which it is institutionalized" (Ibid:326).

One response to Pearson's argument came from Marvin

Harris who held that Pearson's claim that "there are al-

ways and everywhere potential surpluses available" (Pearson,

1957:339) amounted to elevating the power of choice "to

astonishing levels" (1959:188). Harris proposed that if

subsistence level was defined as "the amount of energy

necessary to do nothing except what is biophysically neces-

sary to satisfy the metabolic requirements of the popula-

tion concerned" (Ibid:189) , it could be measured. He af-

firmed that equating surplus with excess has resulted in

a failure to "make explicit the need which was satisfied

and ••. to define the time period during which the state

of satisfaction persisted" (Ibid:191).

In his review of the Pearson-Harris controversy,

George Dalton ignored Harris' suggestion and chose to em-

phasize the popular usage of the word as excess -- "an

amount over and above what is really useful or necessary"

(1960:485). He asserted that, as an analytical device

the concept is used as "a deus ex machina which allegedly

explains complex social structure or some unobserved socio-

economic development" (Ibid:485). In a subsequent publi-

cation, Dalton claims that "exploitation and surplus are

prejudicial words used by some social scientists (perhaps

 85

unintentionally) to condemn only those systems of exploita-

tion they dislike and disapprove of" (1974:559). Positive-

ly, he called for definition and illustration of surplus

in concrete situations as well as specification of returns

from non-producers to producers (Ibid:556).

Although Dalton's views have been severely criticized

(Newcomer,1977iDerman and Levine,1977:115-25), the confu-

sion surrounding the concepts of surplus and exploitation

has not been cleared up. Participants in the debate have

expressed their approval or disapproval of Marx' stand on

the issues but they have largely ignored the writings of

the Classical Political Economists. As the contribution

of Marx to the subject is primarily a gloss on the works

of those economists, in order to assess the concept of

surplus, we have to take account of the broader context of

inquiry.

Dalton needs to be reminded that the distinction be-

tween productive and unproductive labor was made by the

advocates of capitalism and not its critics. Adam Smith

was categorical in his pronouncement that lithe labor of

some of the most respectable orders in the society is,

like that of menial servants, unproductive of any value .•. "

(1954:295). His illustrations ranged from sovereigns to

opera dancers. Following Smith's lead, John Stuart Mill

defined productive labor as "only those kinds of exertion

which produce utilities embodied in material objects"

 86

(1864:76). However, he considered the labor of the offi-

cers of government "indispensable to the prosperity of

industry" and therefore indirectly productive. He also

offered an unambiguous criterion for distinguishing types

of labor by defining unproductive labor as "labor which

does not terminate in the creation of material wealth;

which, however largely or successfully practiced, does not

render the community, ana the world at large, richer in

material products, but poorer in all that is consumed by

the laborers while so employed" (Ibid:77).

Classical Political Economists clearly took the pri-

macy of production for granted while defining their cate-

gories, but they never lost sight of the impact of con-

sumption on production. David Ricardo expressed the issue

very succintly: "when the annual production of a country

more than replaces its annual consumption, it is said to

increase its capital; when its annual consumption is not

at least replaced by its annual production, it is said

to diminish its capital. Capital may therefore be in-

creased by an increased production, or by a diminished un-

productive consumption" (1821:162-63).

Smith and Ricardo had no qualms in their beliefs that

their categories were applicable to all times and places.

This assertion was, however, questioned by Rev. Richard

Jones, successor to Malthus at the chair of Political

Economy at King's College of London, who declared that

 87

"the mixt of causes which concur in producing the various

phenomena with which the subject is conversant, can only

be separated, examined, and thoroughly understood by re-

peated observation of events as they occur, or have occurred,

in the history of nations" (1831:xix).

Out of a projected four volume work on rents, wages,

profits, and the revenue derived by the state from each,

Jones only finished his volume on rent. He distinguished

between peasant's and farmer's rents. He used the term

peasant "to indicate an occupier of the ground who depends

on his own labor for its cultivation" (1831:11). The pay-

ment made by the peasant to the landlord for access to the

soil was called peasant rent. The distinguishing feature

of farmer's rent, to Jones, was the mediation of a capi-

talist between the laborer and the landlord. The farmer

takes "charge of cultivation" of the land "by the labor of

others" and pays a sum to the landlord for using the land

(Ibid:12). Jones believed that "the circumstances which

determine the amount of peasa.nt rents are much less com-

plex than those which determine the amount of farmer's

rents" (Ibid:13). The reason for the difference was the

interrelationship of rents with other aspects of produc-

tion in the complex economy of England at the time of his

writing.

Jones called attention to the investigation of cir-

cumstances which affect the peasants' "power of production,

 88

or their share of produce" and "the circumstances which

distinguish one set of peasant tenantry from another" (Ibid:

156). An important variable to him was the "mode in which

their rent is paid, whether in labor, produce, or money"

(Ibid). He illustrated the different types of peasants'

and farmers' rents through historical and contemporary

examples.

He argued that "rent has usually originated in the

appropriation of the soil, at a time when the bulk of the

people must cUltivate it on such terms as they can obtain,

or starve" (Ibid:ll). But the necessity was not an out-

come of any property of the soil "and would not be removed

were the soil all equalized" (Ibid). However, he argued,

peasant rents prevented "the full development of the pro-

ductive powers of the earth" (Ibid:157). An important con-

sequence of this stunted development was the small number

of non-agricultural classes. "For it is obvious, that the

relative numbers of those persons who can be maintained

without agricultural labor, must be measured wholly by the

productive powers of the cultivators" (Ibid:159-60). Jones

argued that with an expanding system of production the in-

terests of landlords, tenants, and community would be in

general compatible. But he recognized the fact that the

landlord could extract more out of the peasant "by en-

croaching on the tenant's share of the produce, while the

produce itself remains unaltered" (Ibid:162).

 89

In his introductory lecture on political economy,

Jones went beyond descriptions and offered a conceptual

frame of analysis. The concept was that of the "economic

structure of nations," which he defined as "those relations

between the different classes which once established in the

first instance by the institution of property in the soil,

and by the distribution of its surplus produce, afterwards

modified and changed (to a greater or less extent) by the

introduction of capitalists, as agents in producing and

exchanging wealth, and in feeding and employing the la-

boring population" (1833:22).

He claimed that the key to the past of different peo-

ples could be provided by the knowledge of their economic

structure, for this structure alone displays "their eco-

nomical anatomy" (Ibid). He declared that "we must learn

the circumstances vlhich divide them into classes, and the

value and influence of each, as component parts of a state

or agents in producing its wealth" (Ibid).

After quoting in full the above passage, Karl Marx

asserted that "the main point .•. in Jones' work is that

the whole economic structure of society revolves around

the form of labor, in other words, the form in which the

worker appropriates his means of subsistence, or that part

of his product upon which he lives" (Theories of Surplus-

Value, Part 111:414). Subsequently, he restated the point

in more general form by claiming that "the specific econom-

 90

form, in which unpaid surplus-labor is pumped out of direct

producers, determines the relationship of rulers and ruled,

as it grows directly out of production itself and, in turn,

reacts upon it as a determining element" (Capital,III:791).

He then proceeded to illustrate his assertion through a

discussion of labor-rent, rent in kind, and money-rent.

As labor-rent was considered the simplest form, its trans-

formation to rent in kind and money-rent was related by

Marx to the development of more complex forms of social

organization. He claimed that rent in kind was bound to

an "indespensable combination of agriculture and domestic

industry, through its almost complete self-sufficiency

whereby the peasant family supports itself through its

independence from the market and the movement of produc-

tion and history of that section of society lying outside

of its sphere, in short owing to the character of the na-

tural economy in general, this form is quite adapted to

furnishing the basis for stationary social conditions as

we see, e.g., in Asia" (Ibid:796).

For Marx 1 rent in kind was a representation of sur-

plus-labor but he discussed two forms of its extraction

involving very different consequences for the total econo-

mies of the societies concerned. In the first form, the

rent does not "fully exhaust the entire surplus-labor of

the rural family" and thereby provides the possibility of

greater differentiation "in the economic position of the

 91

of the individual direct producers" (Ibid:795-96). In the

second form, the extraction of rent "may assume dimensionns

which seriously imperil reproduction of the conditions of

labor, the means of production themselves, rendering the

expansion of production more or less impossible and reducing

the direct producers to the physical minimum of means of

subsistence" (Ibid:796).

By focusing on the nexuses between forms of labor and

property, and conditions of production and extraction,

Jones and Marx placed the notion of surplus in the context

of total production of the social relation of societies

over time. In order to define economic structures in terms

of class relations, it is essential that inquiries on the

extraction of surplus take account of the distribution of

political forces between classes; for it is in this politi-

cal relation rather than some abstract and arbitrary de-

finition of excess that one can seriously attempt to chart

the contours of the relations between producers and non-

producers.

The appeal of the holistic framework of analysis es-

tablished by Jones and Marx should not obscure the basic

conceptual difficutlties of their inquiry. Since they

take, as the basic condition of all societies, the exist-

ence of unequal access to basic resources that is asso-

ciated in anthropology only with stratified societies

(Fried,1967:186), their notions are of little direct use

 92

in understanding the emergence of stratification and the

role that the existence of a surplus may have played in

that process. But even in situations where class relations

have been fully established, Jones's emphasis on "property

in the soil" as the determining feature of class relations

may distort the role of other factors, such as the rela-

tions between landowners and government, merchants and

markets, and the maintenance and reproduction of these re-

lations. Furthermore, emphasis on the "economic structure

of nations," which is presumably identified with the terri-

torial states, seriously underestimates the significance

of regional differences in the class relations within such

states.

Marx's equation of form of labor with the economic struc-

ture of society reduces the totality of a system to one of

its constituent elements and is, therefore, _methodologically

unwarranted. Since society is conceived abstractly, delin-

eation of concrete units of analysis is eschewed. This

neglect contributes in turn to the unstated assumption

where forms of domination are assumed as stable and their

complexity reduced to the form in which unpaid surplus la-

bor is pumped out of the direct producers. Furthermore,

the posting of a self-sufficient peasant family under con-

ditions of natural economy forming the basis of stationary

formations, ignores the many-fold relations between pea-

sants and markets and disregards the diverse characteris-

 93

tics of domestic organizations and their impact on the or-

ganization of labor in agrarian structures.

Marx's three-fold classification of rents -- taken

over from Jones -- and his belief that circulation of land

as a commodity was a practical development of capitalism

(Capital,III:811) reveals a Eurocentric view of the world.

Reasons for this attitude are related to his methodologi-

cal commitment to the study of capitalism in its "pure

form" and the prevailing ignorance regarding conditions of

production in non-European societies. Jones, who closely

followed the available literature, claimed that outside

Europe the form of tenancy known as "metayage" was only

found in Afghanistan. He wrote that "in Afghanistan, a

race of tenants is found called Buzgurs, who seem to dif-

fer in no respect from the metayers of Western Europe.

This is a singular instance in Asia, where this tenancy,

although sometimes partially engrafted on Ryot rents, is

perhaps in no other spot to be found coexisting in its pure

form" (1831: 100) •

That Jones had heard of Afghanistan was due to a re-

markably detailed study of the country by Mountstuart

Elphinstone who visited the winter capital of the Afghan

Empire in Peshawar as the first envoy of the East India

Company. His book, first published in 1815 then revised

in 1839, is an indispensable source of information for

conditions in eighteenth and early nineteenth century

Afghanistan. By checking and supplementing his information

 94

from other sources, I hope to draw a coherent picture of

the social organization of production for my unit of long

duration 1747-1901.

In order to relate my theoretical synopsis to the

ethnographic material available for Afghanistan, I shall

first discuss the main forms of organization of labor and

property and then illustrate the relation between classes

of direct producers and non-producers through concrete

cases of extraction of surplus. As I will discuss, later

in the course of this study, taxation by the state and

the role of the state in the reproduction of different

classes, I have confined the ~nalysis in this chapter to

the organization of production.

The direct producers in Afghanistan were not a homo-

geneou:,'i group; they were differentiated into classes.

Elphinstone succintly affirmed that "there are five classes

of cultivators in Afghanistan: 1st, Proprietors, who cul-

tivate their own land; 2nd, Tenants, who hire it for a

rent in money, or for a fixed proportion of the produce;

3rd, Buzgurs, who are the same as the Metayers in France;

4th, Hired laborers; and 5th, Villains who cultivate their

lord's land without wages" (1839:389).

What were the conditions of property that confronted

these classes of direct producers and to what extent was

the "peasant family" the basic social unit of production?

Before attempting to analyze the social organization of

 95

labor, it is essential to perceive clearly the diversity

of types of property which, for the period, are private,

corporate, and state property.

contrary to Marx's belief that circulation of land as

a commodity was a product of capitalism, land indeed was

a commodity in most areas of Afghanistan for t~e period

1747-1901. In fact, the existence of land as a commodity

long pre-dated this period.

By reproducing deeds of sale and purchase of land,

Aziz-ul-Din Wakili Fufalzai, an Afghan historian who has

compiled an impressive collection of documents from the

eighteenth and early nineteenth centuries, has provided

the evidence that land was a commodity (1337 A.H./1958

A.D.:280-83). However, the most comprehensive evidence

on this point comes from the records of courts of sharia

(Islamic Law) in the valley of Kunar, Eastern Afghanistan.

The valley, which today forms one of the twenty-six pro-

vinces of the country, was in the l880s and 1890s an ad-

ministrative district of the province of Kabul. It was

divided into twelve sub-districts and, although an economi-

cally marginal area compared to other parts of the country,

it had a fairly dense pattern of settlement. The northern

part of the valley was until 1896 inhabited by non-Muslim

people who were forcefully converted to Islam in that year.

The eastern boundary of the valley marked the limit of

the sphere of competence of the Afghan state and the be-

 96

ginning of a series of local states. In 1893, the area

was formally recognized as the line of demarcation of

spheres of influence of the Afghan and the British Indian

governments.

Although court records cover a wide range of subjects,

in this section I shall basically make use of cases in-

volving land transactions and inheritance. The data I

shall refer to is derived from three volumes covering the

period 1303-1307 A.H. (1885-1890 A.D.).

The standard formula of sale leaves no ambiguity re-

garding the private nature of land ownership. To take an

example:

"Deed of sale in the name of the seller Hamid resi-

dent of Badyali and in the name of the buyer Shuja resi-

dent of Islampur.

"Mention of boundaries: One piece of land located

in the place of Badyali, specifically in the field known

as Natdasar, connected to the east to the channel of run-

ning water; to the west to the land of Mir Waez son of

Muhammad Qayum son of Muhammad Hanif; to the north to the

land of Saman son of Abdullah Khan son of Hawas Khan; and

to the south to the land of Hamid son of Mir Muhammad Khan

son of Muhammad Fidar of the group of Sarwal resident of

the Badyali village of Ghlam Khail.

"On the sixth of Muharam al-Haram (Sept. 3 , 1880, [a

man] of sound reason and in possession of all his faculties,

 97

stating his name and descent as Hamid son of Mir Muhammad

Khan son of Muhammad Didar, acknowledged and attested in

the correct shara'i manner that: 'I have sold, in a binding

manner and without corrupting conditions, my right and pro-

perty, namely the said piece of land, which is free from

claims from anyone, with all its rights and accessories,

for a "price of 96 rupees, half of which is 48 rupees, to

Shuja son of Istalish son of Suliman resident of Islampur.'

The buyer accepted the binding agreement on behalf of him-

self and the two parties attested to having received what

was due to each in the agreement. The said seller and

buyer absolved each other from any subsequent dispute and

both accepted the mutual acquittal.·

"All this took place in the presence of the following

witnesses: ..• Fee for writing 288 dinars [600 dinars =1

rupee]. Only."

More than one set of standard phrases expressed the

total transfer of rights that an act of sale conveyed.

Some of the qudat (judges) wrote the formula in a more

abbreviated form, but the private nature of ownership

and the binding character of the sale were always clearly

stated. Significantly, a deed of sale from 1168 A.H. (1754

A.D.) reproduced by Fufalzai uses the same formula for a

transaction that took place in the court of the capital

city of Kabul. Although existing information needs to be

supplemented by data from other areas of the country for

 98

the unit of long duration, it can be reasonably asserted

that land was a commodity in circulation during 1747-1901.

In Karl Wittfogel's view, however, alienation alone

is not the defining feature of private property. "Only

when the proprietor" he states, "has the right both to hold

his land indefinately and to alienate it to persons outside

his social group do we encounter what, in conformance with

established usage, can be called full private landownership"

(1957/1976:275). To him, private property stands in sharp

contrast to the semi-complex pattern of property, the char-

acteristic marks of which were strong development of ac-

tive property in "industry and commerce but not in agricul-

ture" (Ibid:231). He argues that the semi-complex pattern

has been prominent in India and the Near East for most of

their history, and in China for part of its history (Ibid:

260). Did the ownership of land in Afghanistan, during the

period under discussion, amount to private property in the

sense defined by Wittfogel? As already mentioned, the

formula for the sale of land is quite explicit on both the

complete rights of the owner over the land and the total

transfer of these rights, as a result of the binding char-

acter of the sale, to the buyer. But we do not need to

get entangled in the complexity of legal terms; data from

the court records from the the valley of Kunar allow us to

answer the question empirically:

Out of a total of 272 cases of property transactions,

 99

236 cases involved agricultural land. Analysis of these

cases reveals that alienation and acquisition of land was

not limited to any legally or socially defined group. While

160 cases of transactions between people residing in the

same named locality, 70 were between people residing in dif-

ferent localities at the time of the sale. People residing

in different sub-provinces or provinces transacted business

in 6 cases. Neither religion nor gender barred people from

the acquisition or disposal of land. 17 Hindus bought land

from Muslims and 3 from other Hindus while 5 Hindus sold

land to Muslims. As for women, 16 sold land but only 6

bought. All the named groups both sold and bought land

from each other. Pashtuns, who formed the majority of the

population, bought land from other Pashtuns in 52 cases

and from non-Pashtuns in 30 cases. In 29 cases, Pashtuns

sold land to non-Pashtuns.

Islamic law does give the persons whose land is ad-

jacent to a piece of property to be sold the first option

on that property, and this right may be construed as a

limitation to private property. However, as the law re-

quires that neighbors match the price offered by any other

prospective buyer, the seller is insured to get the highest

price for his commodity. Furthermore, as the ability to

buy land is a function of access to money, in practice

neighbors are not always in a position to benefit from

their precedence. In the case of the Kunar valley, out of

 100

236 transactions on land only 37 feature a neighbor pur-

chasing the land. Since in 101 of these cases the land

sold was in more than one parcel, the number of neighbors

in a position to exercise their rights was rather small.

Clearly Islamic law vested an owner of private pro-

perty with wide powers over its use and disposal during

his or her lifetime. An owner could also, by turning his

private property into an endowment (waqf) change it into

corporate property and thereby affect its utilization long

after he was dead.

Upon the owner's death and in order of precedence,

the property is to be used for the following purposes: to

cover the funerals' expenses; pay the debts; pay legacies

up to one-third of the remaining estate; and divide the

other two-thirds among the inheritors (Baillie,1832:1).

In principal, legal heirs include both ascendants

and descendants of the ego. In cases where no individuals

in those categories are to be found, collateral kindred

are considered eligible. Some heirs receive fixed shares

of the inheritance and are to be distinguished from other

heirs who have claims only on what remains of the property

after the fixed shares have been allotted. Twelve classes

of persons -- four male and eight female -- form the group

of inheritors but all of them do not succeed simultaneously.

The exclusion of would-be inheritors is based on two gen-

eral principles. First, a person related to the deceased

 101

through another, except for stepbrothers and stepsisters

in the case of a mother, has no right to the patrimony

during the life of that other. Second, the nearer relative

to the deceased excludes the more remote (Ibid:58-59ff.).

The demographic composition and the sex ratios of the

family of the deceased by marriage and/or birth has a criti-

cal impact on the number of those who are eligible for in-

heritance as well as the amount of the inheritance. The

existence of a son or a father, for instance, absolutely

excludes the full sisters. A wife receives a fourth and

a husband one-half of the property in cases where there

are no children; the existence of any child reduces their

shares to an eighth and a fourth respectively. During

their lifetimes, however, through testament or declaration

of sale, spouses can transfer property to each other. The

sum of money undertaken by a husband to be paid to his wife,

mahr, is considered a debt and is to be paid before the

property is divided.

A statistical analysis of 499 cases of inheritance

from the valley of Kunar reveals that the children of the

deceased were their main inheritors. Out of 462 cases

where marriage could be ascertained, only 32 were without

children. The average number of children among the married

families were 3 and the average number of inheritors, based

on all cases, was 4 persons. The number of inheritors

being 2136 persons and that of the children 1392 persons,

 102

the latter formed 65.16% if the total inheritors. As Is-

lamic law favors the children by excluding most classes of

inheritors when the deceased has children and allotting

them the largest shares from among those -- e.g. husband,

wife, father -- who are not excluded by their presence,

the children's actual share of the inheritance was even

larger than the above percentage would indicate.

What are the possible implications of the operation

of Islamic law of inheritance for the social organization

of production? In order to answer this question we have

to take account of the prevailing pattern of stratifica-

tion.

Unequal access to strategic resources in the valley of Kunar

can be readily seen from the following table. The data is

based on the cases of inheritance. A monetary value was

placed on every item of the inheritance in every case and

since the relative meaning of the figure can be readily

grasped in terms of an abstract unit of measurement, I

have presented the data in terms of their monetary equiva-

lence.
 103

Table I. Stratification in the Valley of Kunar

Number Number
Amount in of Amount of of
Rupees Cases Rupees Cases

0-49 44 900-999 10
50-99 64 1000-1499 17
100-199 133 1500-1999 9
200-299 94 2000-2499 2
300-399 39 2500-2999 o
400-499 33 3000-4000 4
500-599 l7 4600 1
600-699 12 6000 1
700-799 6 6400 1
800-899 11 12500 1
Total Number of cases = 499

The value of all the goods left by the 499 deceased

amounted to 209,780.5 rupees. Although the mean amount
of inheritance was 420 rupees, 78% of the cases fell below
this average.
The monetary value, however, only portrays the rela-
tive access of different families to the strategic resources in
the valley. To relate these figures to the organization
of production, we have to translate them into their equiva-
lence in land -- the single most important item in the in-
heritance cases.
The most widely used unit for the measurement of land
was the weight of seed necessary for its cultivation.
Since factors related to fertility and location affected
the price of land, this measure does not easily lead to
inferences on acreage. From instances where information
 104

on acreage, amount of seed, fertility and location sin-

gle or double crop, irrigated or rainfall, etc. and

prices could all be checked against each other, I have ex-

tracted the following information. The price of one jerib

(roughly = 0.5 acre) of inferior land in the valley of

Kunar was 40-50 rupees, that of average land 70-90 rupees,

and that of good land 150-250 rupees. In terms of quality

these would roughly correspond to rainfall, single crop,

and double crop land, and the respective amount of seed

used on each type was about 9, 13, and 18 kgs. per jerib.

Thus, the sum of 209,780.5 rupees, which is the total

value of the goods left by the 499 deceased, would amount

to 2622 jeribs of average land, and 420 rupees, which was

the average size of inheritance, would amount to 5.25

jeribs of average land. The assets of 78% of the families

in our sample being below this figure, only about 20% of

the population in the valley of Kunar had access to 5.25

or more jeribs of average land. Division of such a pro-

perty among three sons would have left each with 1.75

jeribs of average land. Division of the same property be-

tween a son and two daughters according to Islamic law,

where a son's share is twice that of his sister, would

have left the boy with 2.625 and each of the girls with

1.31 jeribs of average land. As the number of the child-

ren per family varied from 0-11, the degree of fragmenta-

tion of property was directly related to the demographic

 105

compostion of the family.

The yield of a jerib of irrigated land for cereals in

the valley of Kunar was 282 kgs. in 1977 (Statistical In-

fo.crnation 1977:78-81). Assumption of nineteenth century

amounts of seeds would imply a ratio 21 to one and 16 to

one between amounts harvested and sown for average land

and good land respectively. Information on yields of

corn from property of exiles managed by state-appointed

agents gives a 9 to one ratio for good land. If the pro-

portional difference between average and good land was

the same in the nineteenth century, the yield on average

land would have been 12 fold. Depending on the ratio of

the amount of seeds sown to that harvested, the harvest

of a family owning 5.25 jeribs of average land relying

exclusively on family labor would have been 600-800 kgs.

of cereals. Of this amount, 68 kgs. would have to be set

aside for the seed of the coming year leaving the family

with 532-732 kgs. One rupee in 1889, on the average,

bought 20 kgs. of corn or rice and 13 kgs. of wheat. The

monetary value of the harvest of the family would have,

therefore, been 26.6-36.6 rupees for corn and 41-56 rupees

for wheat. As such, depending on the type of crop, 9 to

15 years of the value of the crop would have been required

to pay the initial price of the land.

But before thinking of marketing its produce, the fam-

ily had to meet its own needs. As already mentioned, the

 106

average family was composed of two adults and three

children. The standard food rations of the "native" sol-

diers in the British army was 170 kgs. of grain a year

(Gazetteer, Kandahar:231); the same ration was given to

agricultural laborers by landowners in Afghanistan until

the 1950s. Positing the above figure as the consumption

requirement of an adult and half of that amount as that

of a child would mean that the average family in our sam-

ple required 595 kgs. of cereals for its food for one

year. However, since the amount of property owned by 78%

of our sample was below the average of 5.25 jeribs, the

majority of those who owned private property would have

been unable to meet their own consumption needs. And,

as we shall see in greater detail later, the payment of

taxes to the states, which ranged from 1/10 to 1/3 of the

produce at various times, would have even further reduced

the number of people who could provide for their own con-

sumption needs.

The ratio of adults to children in each family also

had a direct bearing on the size of its consumption needs.

I have summarized the relevant information on the families

in the valley of Kunar at the time of the death of one of

the parents in table II. In the records, age groups are

distinguished from each other only by being "minor" or

"of age" and I have had to follow that usage.

Minors accounted only for 17.8% of the children in the

 Table II. Size and Composition of Landowning Families in the Valley
of Kunar

Number of
children 1 2 3 4 5 6 7 8 9 10 11
per family

Number of
families 62 105 94 69 50 26 5 5 0 1 1

Total
number of 62 210 282 276 250 216 35 40 0 10 11
children

Total
number of 34 99 136 156 121 76 16 20 0 3 6
girls

Total
number of 28 111 146 120 129 140 19 20 0 7 5
boys

Total
number of 13 32 59 42 53 30 4 15 0 0 0
minors

Total
number of 25 10 11
49 178 223 234 197 186 31 0
those
"of age"

Source: Diwan-e-Qaza-Kunar ("Court Records of Kunar") 1303-1307/1885-1980

f-'
0
~
 108

sample. Since the average consumption of an adult was

twice that of a child, the higher ratio of adults in the

sample implies that a family's average consumption needs

might have been even greater than our earlier estimation

allowed for. It is therefore highly probable that most

landowners who cultivated their own land not only were

not in a position to market any of their produce but

would have to undertake other activities to meet their

own consumption needs.

How could a family increase its assets? In order

to answer this question, I shall analytically distinguish

the family estate -- the plot of land which it owned and

cultivated through the employment of its own labor -- from

l
the family income. Since land was a highly valued com-

modity, a family could not increase the size of its es-

tate in accordance with the changing ratio of its adults

and children without resorting to other activities that

would increase its sources of income and its purchasing

power. Labor not needed on the family estate could be

employed outside to enhance the position of the family

economy.

Yet there already was a strong cultural stress that

women not engage in work outside the domestic context and,

although no family can be viewed as a complete producing

lPor elaboration of this distinction see Cohen, M.

1976:58ff.
 109

unit without taking account of the work performed by wo-

men inside the house and in rearing children, women as a

group did not directly participate in agricultural pro-

duction. Moreover, and contrary to Marx's assertion re-

garding the unity of agriculture and manufacture as a

universal feature of peasant family, there was no domes-

tic manufacture of significance in the Kunar valley. This

fact is clearly revealed in a document of the court in

1889. The government wanted to purchase headwear, foot-

wear, and other types of garments from the local producers

for the soldiers stationed in the valley. Government a-

gents and local headmen gave written testimony that none

of these items was being produced for commercial use in

the valley. Although women did produce some kinds of

scarves, their number did not exceed one or two hundred

a year. Significantly, they stated that whenever clothing

items were needed in the valley they were brought from

the large market towns of Peshawar and Jalalabad. Thus,

any surplus labor that could be utilized outside the fam-

ily estate was essentially male labor.

But females at the age of puberty or older were a

valuable economic asset to their families. Women of that

category were highly sought after in marriage and in most

situations, a form of payment to the family of the girl

took place. Although Islamic law is clear on the point

that the payment of mahr -- the sum of money in return

 110

for which a woman accepts to marry a man -- is to be paid

to the woman herself, the practice is more complex. The

mahr is normally divided into two named parts: an amount

which is paid before the consumation of marriage and an

amount which is considered a debt of the husband to the

wife. While the marriage contract may specify a date by

which the second portion is to be paid to the wife, it is

usually left outstanding. In case of divorse, the husband

has to pay first this amount, and in case of the husband's

death, it is considered a debt and has to be paid in full

before the division of the estate among the inheritors.

Analysis of 1095 cases of marriage, with complete

information on mahr, from the valley of Kunar illustrates

a number of linkages between marriage and access to re-

sources. The total amount pledged in mahr was 164,948

rupees, and 61.8% of it had in fact been paid by the

time of the marriage. In a large number of cases, the

sum paid was the full amount of mahr and although in most

cases there is no information as to whom the payment had

been made, in a number of cases it is clearly stated

that the payment was made to the family of the bride.

Evidence from fieldwork carried during the 1970s leads

me to the conclusion that, in most cases, sums paid, or

their equivalent in goods, actually passed to the control

of the bride's family and not to the brides. The sums

left in suspended payment may be interpreted as a form of

 III

concern for the future welfare of the bride by her family.

The amounts paid in mahr, however, were not uniform

and, as table III shows, bear a highly stratified charac-

ter.

Table III: Marriage, Mahr, and

Stratification

Amount pledged Number of cases Amount pledged Number of cases

0-49 245 700-799 5

50-99 223 800-899 2
100-199 373 900-999 2
200-299 128 1000-1499 15
300-399 43 1500-1999 5
400-499 27 3000 1
500-599 22 5000 1
600-699 2 6000 1

The average amount of mahr pledged was 150.6 rupees

which was the equivalent of 1.6 joerib of average land.

The sum of 102,039 rupees transferred to the brides and/

or their families would have bought 1275.5 jerib of aver-

age land. Through the transfer of a daughter(s) in mar-

riage, a family could reduce its number of consumers and,

by investing in land the money received in mahr, adjust

the size of the estate to the number of the available

male producers.

But this option was only open to families where

 112

spouses had been married for at least fifteen years and

where the number of daughters exceeded the number of sons.

Indeed, the family was responsible for the marriage ex-

penses of the sons and, in families where the number of

boys and girls was even, the sums of money realized for

the girls' mahr might just balance the expenses entailed

by the marriage of the son(s). Families with only male

children, or with more male than female children, and be-

longing to the 78% of the sample owning less than the

average 5.2 jerib of average land, could have been easily

forced to sell part or all of the family estate to pay

for the marriage of their sons. The status of the family

would have thereby changed from owner-cultivator to that

of share-cropper or agricultural laborer. Whereas delayed

marriage and/or search for other resources might have

postponed the prospects for families with a high ratio of

adults to minors, death of the father in a family where

all the children were either girls or minors could have

only hastened the process of alienation of the family

estate. There are a number of cases in the records of

the valley of Kunar where the qadi (judge), who in such

cases had to appoint a guardian for the minors, author-

ized the sale of their land to pay for their living ex-

penses. At the time of puberty, those children would

have only had their labor power to rely on.

Even when children were not minors, the death of the

 113

father could cause the break-up of the family estate and

economy by putting in motion the process of the division

of the family estate. Although there is no legal compul-

sion to divide the estate, the majority of the inheritors

did opt for division. Since 93.2% of the inheritance

cases feature the deceased as a male, the resulting de-

gree of property fragmentation can be readily seen. In

an average family of three children, owning less than 5.2

jerib of average land, each of the inheritors would have

found himself/herself with a fragment of the land.

Although the daughters' share in the inheritance

might be interpreted as delayed dowry, especially in cases

where they were sole inheritors, in general this was not

the case. The decades of 1880s and 1890s, as we shall see

later, were a period of strong centralization of power by

the central government and the rights of women to inheri-

tance was backed by the state. During most of the years

of our period of long duration, according to the norms

of local customs, the women were denied their share of

the inheritance. Elphinstone, without the least hesita-

tion, categorically stated that "the Afghans purchase

their wives" (Vol.l:236). Thus, despite the predominance

of plow agriculture and the clear prescription of Islamic

law, whenever that law was not sanctioned by the power of

the state, the form of the transfer of goods entailed by

marriage was closer to bride-price than dowry. This in-

 114

terpretation is reinforced by the fact that even in cases

where large sums of money were pledged in mahr -- a char-

acteristic of the large landowning families of the valley

of Kunar -- the sums paid at the time of the marriage con-

stituted the bulk of the pledge.

I have argued that expenses of consumption, lack of

necessary adult manpower to cultivate the estate, and ex-

penses in the marriages of sons were contributing factors

towards the alienation of land among those who owned it

as private property. Indebtedness and payment of taxes

to the state were also contributory causes. That the im-

pact of these factors was not limited to small landowners

is revealed in the stratified character of the sales and

purchases as displayed in table IV.

Table IV: Sale and Purchase of Land in the Valley of Kunar

Amount in Number of Amount in Number of

rupees cases rupees cases

4-49 35 500-599 5 "..

50-99 65 600-699 5
100-199 49 700-799 1
200-299 40 800-899 1
300-399 20 900-999 0
400-499 12 1000-1290 3

How did the buyers gain access to the money that they

invested in land? We have already seen that the mahr of

 115

daughters could have enabled some families to buy land.

Families with excess adult male labor but owning their own

land could have increased their assets by renting out land

or hiring out their labor-power. In addition, Elphinstone

identified factors which operated through our unit of long

duration. He affirmed that "purchasers are found among

those who have been enriched by the King's service, by

war, and by successful agriculture or commerce" (Vol.I:

390). It was the ope\ation of these factors that criti-

cally linked the process of reproduction of systems of

stratification at the local levels with the conjunctures

of larger forces. The political implications of this

articulation will be discussed in part two of this study.

Meanwhile, I shall examine the implications of the scale

of transactions compiled in table IV.

The total value of the transactions in land in the

valley of Kunar amounted to 45,223 rupees which were the

equivalent of 565 jeribs of average land. The labor of

an adult male was considered sufficient for the cultiva-

tion of 7-10 jeribs of average land. If the families

that bought land fell within 78% who owned less than 5.2

jeribs of average land, they could easily cultivate the

newly acquired land by their own labor. However, if the

size of land owned by them prior to purchase was already

absorbing all their energies, they would have had to look

for outside labor. Those who invested in land but were

 116

not using their own labor-power for its cultivation would

have had to seek arrangements whereby the land could be

cultivated by the labor of others. It can, therefore,

be seen that even in the case of owner-cultivator the fam-

ily of the direct producer may not be the sole unit of

production and, depending on the exigencies of its demo-

graphic composition and the amount of land at its disposal,

it would be in need of extra land and/or labor. As a group,

owner-cultivators could not exist in isolation and had to

interact with other forms of social arrangements under

which property and labor relations were organized.

All private property was not of small size. The deed

of sale reproduced by Fofalzai for the year 1168 A.H./1754

A.D. referred to easlier, was for 110 jeribs of land in

the vicinity of the city of Kabul. Since the total price

was 5019 rupees, the price of a j"erib, on the average,

1
should have been around 45 rupees. However, the price of

land around the city of Kabul was not stable; the same

property sold for 6,000 rupees in 1183 A.H./1769 A.D.

(Fofalzai, 1967 vol.II:605). But the landed property of

the new owner, Abdullah Khan Fofalzai, was not limited

to the land that he had purchased in that year. Fofalzai,

lOne rupee exchanged for two shillings and four pence in

the first decade of the nineteenth century (Elphinstone,
vol.I:391). If the same rate of exchange prevailed dur-
ing the second half of the nineteenth century, the value
of the 110 jeribs would have been around 585 British
pounds.
 117

a seventh generation lineal descendant of Abdullah Khan,

has assembled a number of unique documents which, in con-

junction with data from other works by Afghan and other

historians, enable us to examine the extent of the estates

of large landowners in the second half of the nineteenth

century, as well as the subsequent trajectory of this

class in the nineteenth century.

By 1773, the estate of Abdullah Khan Fofalzai had

reached 190 kulba (plows) (Ibid,vol.I:159-61i181). Since

the type of kulba owned by or granted to an official to

the state was calculated as 63.75 acres (0.5 acres roughly

= 1 jerib) (CA, Kandahar:230), the total extent of his land

would have been 7012.5 acres. The geographical distribu-

tion of the estate was even more remarkable: of the 110

kulbas of land, only 35 were concentrated around the city

of Kabul or localities close to it (Fofalzai,1958:357).

The rest of the estate was located in various places in

Kandahar, Jalalabad, Kashmir and other regions of India

which, at the time, were tributaries of the Afghan state

(Ibid,l967,vol.II:605-12) .

Details provided by Fofalzai on the other leading

families of the period (1967 1 vol.II:596-726) clearly in-

dicate that the case of Abdullah Khan Fofalzai was more

of a typical rather than an exceptional example of the

class of large landowners. The means which enabled this

groups of landowners to acquire their landed estates, the

 118

political relations which charted their course of conduct,

and the extent to which the same families were able to

maintain and reproduce their economic and political sta-

tus will be addressed in the context of a conjectural anal-

ysis of the relevant time periods in the second half of

this study. Here it is sufficient to point out that

throughout the period of our unit of long duration, pri-

vate property in the form of small owner-cul"tivated plots

coexisted with medium range as well as very large estates

of land. The holdings of these large proprietors, due

to a variety of political and economic factors, were geo-

graphically dispersed.

Purchase of land was not the sole means for acquiring

it as private property. Grants from the state and claim-

ing" of waste land -- called mawat (literally dead) in

Islamic law -- by bringing it under irrigation were other

legal avenues of acquiring land. The harim (sphere of

proprietary rights) which fell within the area of an irri-

gation facility depended, according to Islamic law, on

the purpose for which the facility was built. The harim

of a well for watering animals was defined by a radius of

40 zara (60 feet) around the well, that of a well used

for irrigation by a radius of 60 zara (90 feet). In both

cases, if the length of the rope used for drawing water

was longer than the specified radius, the harim was to

be defined by the legnth of the rope. The harim of a

 119

of a spring specifically built for bringing mawat land

under irrigation was defined by a radius of 500 zara (750

feet) (Karaji,1967:44). As for the harim of a karez

(underground channel), the determining of its radius var-

ied a~cording to concrete circumstances, mainly the quali-

ty of the soil through which the channel was constructed

and the relative location to mawat land of already exist-

ing karezat (Ibid:46-52).

In places where agriculture was an established prac-

tice, the main issue was the availability and location

of such land. An application granted by the king in 1187

A.H./1773 A.D. was for only two jeribs but the land was

located in a fully inhabited neighborhood in the city of

Kabul (Fofalzai, 1967:675). A request granted in 1172 A.H./

1758 A.D., however, was for the repair of an abandoned

underground channel in the region of Kandahar. The area

of land is not specified but it is stated that the chan-

nel would be built by "the man and money" of the appli-

cantsj that it should be cultivated every yearj and that

it would be subjected to taxation after one year of exemp-

tion (Ibid:640-41). The interest of the state in enhanc-

ing its revenue through additional taxation explains its

positive attitude towards such undertakings.

Examination of the maintaining of such units through

time throws light on the social organization of societies

practicing irrigation agriculture. The terms of the grant

 120

clearly indicate that the reclaimed land was to be the

private property of the investor. As such, it could be

sold in parts or in whole or passed on in inheritance.

The inheritors could reduce or augment their inherited

parts by selling or buying; they could also transact

business within the network of inheritors or with people

outside this network. Whether the process would be the

result of inheritance or sale alone and/or both factors

combined, in time the system would be characterized by a

group of shareholders in an irrigation system, whose

shares may be evenly or unevenly distributed but whose

cooperation in maintaining the system would be essential.

A larger system of irrigation would be more complex in

the necessary scale of cooperation as well as in the num-

ber of shareholders but the necessity of cooperation in

maintaining the system would remain. Depending on the

degree to which people outside the kinship and marriage

networks of the original owner(s) may have gained access

to the water and land in the area: the group of property

owners might be characterized by some analysts as an open

or closed corporate community.

However, recourse to such imagery misses the defining

criterion of corporations, possession of joint property

by a juristical person (Von Savigny, F.e., 1884:181).

Radcliffe-Brown, who popularized the notion of corporation

in anthropology, considered the establishment of "joint

 121

rights" as its hallmark (1935/1952:44). But he was mis-

reading the legal meaning of the notion in Roman law.

"The essential quality of all Corporation," wrote Von

Savigny, the nineteenth century German authority on Roman

law, "consists in this, that the Subject of the Right does

not exist in the individual Member thereof (not even in

all Members taken collectively), but in the ideal Whole:

a particular, but specially important, result thereof is,

that by the change of an individual Member, indeed even

of all the Members, the Essence and Unity of a Corporaton

is not affected" (1884:181). In order to avoid any con-

fusion, Von Savigny contrasted the concept of corporation

to that of joint ownership. He stated that "the Immov-

able Estate of a Corporation may be used either for the

purposes of the Corporation, by Farming it out or by dir-

ect Administration; or for the purposes of the individual

Members; and lastly, a sort of mixed enjoyment of the

Estate might also happen •.• From all these cases, however,

we must carefully distinguish those in which the Right

appertains to particular individuals or to certain classes

amongst the Members of the Corporation: for then there is

simply Joint Ownership, and no longer the Ownership of

the Corporation" (Ibid:214). Furthermore, he was quite

explicit that the association of a number of people or

the will of a founder were not the acts that constituted

a corporation. "For this purpose the sanction of the

 122

Sovereign Power of the State is necessary, which may be

conferred not only expressly but also tacitly, by a con-

scious toleration or by an actual recognition" (Ibid:204).

It was on the basis of these criteria that classes and

tribes, which were all important political entities, were

not recognized as juristical persons in Rome (Ibid:177).

The reason for my preference for Von Savigny's suc-

cint legal definition of corporation over Radcliffe-Brown's

commonsensical usage is not an outcome of theoretical pur-

ism. During the period of our unit of long duration there

was a form of property in Afghanistan that corresponds

closely to the strict legal definition of corporation.

Our discussion should help us in avoiding confusion be-

tween the two forms of property.

Emergence of property in the form of waqf (pl.awqaf,

endowed) could be the result of an initiative on the part

of the state, of owner(s) of private property, or of the

combined action of the two. The singularity ofwaqf, in

theory, lay in its corporate character: the sanctioning

of the state was obtained through the official witnessing

of the deed of endowment in a court of Islamic law and

once the property was definitely committed it was consid-

ered mahbus (detained; lit. imprisonned) and could not be

alienated. Furthermore, through provisions dealing with

the selection of trustees and beneficiaries under differ-

ent circumstances, the survival of the status of property

 123

as wafq was made independent of the survival of the origi-

nal membership of beneficiaries.

The purpose for which a waqf was set up, the terms

of its endowment, and the fluctuations of relations of

domination had a strong impact on the continuation of any

waqf over time. As the identity of the initiators played

a significant role in this regard, I will differentiate

between those awqaf where the descendants of the endower

were to be given priority and those where the beneficiar-

ies were to be other people. I will illustrate the dif-

ference by refering to concrete cases.

C.D.Chekovich, a Soviet scholar, has published two

volumes of waqfnama (deeds of endowment) where the initia-

tive for conversion was taken by individuals. The volume,

published in 1965, is a reproduction of one main waqfnama

from the fourteenth century setting up a waqf in a district

of Bokhara, Central Asia, and supplemented by other deeds,

the latest of which is from the nineteenth century. The

second volume (1974) is essentially the record of the

transformation of part of the properties of one man,

Khodja Ahrar, into awqaf in various regions of Central

Asia and Afghanistan between 1470 and 1490 A.D. These

waqfnama are supplemented by a series of earlier deeds

of purchase and by other documents showing the relation-

ship of the awqaf to political authorities.

Even though the first waqfnama encompassed a very

 124

large area of land, "so large that it can not be counted"

is the expression used in the document (1965:39), there

was no comparison between it and the waqfnama appearing

in the second volume. Chekhovich describes Khodja Ahrar

as "the largest landowner in Central Asia at the time.

He possessed thousands of hectares of the best irrigated

land in the regions of Tashkent, Samarkand, Bukhara,

Kashkadria, and in other regions of Central Asia as well

as in Afghanistan. Immovables in his possession, accord-

ing to the documents under discussion, included 64 vil-

lages, 30 out-of-town orchards, 11 town estates, scores

of commercial establishments and artisan workshops ... "

(1974:630). Khodja Ahrar's estates in Afghanistan were

located in and around the city of Kabul and although

Centra: Asia and Afghanistan were not under one system

of political authority, the courts in both areas adhered

to the same school of Islamic law.

The terms of the waqfnama bear strong testimony to

the power of the ini,tiator in defining the structural

conditions under which the awqaf were to function. These

powers can be best grasped through an analysis of simi-

larities and differences in the two main waqfnama re-

corded by Chekhovich.

That the property converted to waqf was to be taken

completely out of circulation is either clearly stated

or strongly implied in all waqfnama. In the waqfnama

 125

of 1326 A.D. for instance, after enumerating all the pro-

perties to be changed intowaqf, it is stated that "what

has been mentioned in this waqfnama, its status as waqf

is detained in this deed. May God bless it. Therefore,

it should not be sold, given as a gift, inherited, become

the property of another person, or wasted for any reason

whatsoever" (1965:87).

The powers of the endower during his lifetime are

unambiguously defined in all deeds. In the waqfnama of

1489 for instance, it is declared that "he made it a

condition that he, himself, be the trustee of this waqf

and control its properties and products in any manner

that he deems fit. Should he desire it, he can replace

some of the properties of the waqf, delete a condition

or add one. But these powers regarding replacement, addi-

tion and deletion are limited to the person of the en-

dower only. His descendants and others are without such

guardianship" (1974:173).

Conditions of selection of trustees, designation of

beneficiaries and disposal of the products of the awqaf,

are different in the two wa"qfnama.

In the waqfnama of 1326, the endower made it a con-

dition that only those of his descendants living in the

Fatehabad district of Bokhara should have priority in

being appointed trustees of thewaqf. Among them, the

"high born" were to be given preference over the "low

 126

born". Even the style of life of the trustees was

strictly defined. The endower declared that he had nei-

ther worn expensive clothes, nor bought a slave for his

own service; should any of the future trustees deviate

from this style of life, he was to be removed from the of-

fice (1965:98-99). In case of extinction of his own line

of descent, through males and females, he stipulated that

the line of one of his paternal uncles be given preference

over that of other kin (Ibid:100-10l).

Khodja Ahrar, by contrast, opted for a more diverse

series of options. In the ~eed. of 1489, he designated

one son as the trustee of the waqf covered by the deed,

and in the deed of 1490, he named the above son and a sec-

ond son as trustees, each to be in charge of half of the

produce and fruits of that waqf. In the deed of 1546,

no individual is specifically named but it is stated that

after the death of the endower his descendants should suc-

ceed him "generation after generation and century after

century. Males are to be given preference over females.

If they are equal in learning, judgement, and capacity,

he who is closer to the endower in kinship is to be pre-

ferred, and among those of the same degree of kinship

the oldest is to become the trustee" (1974:338). In the

event of extinction of his line, Khodja Ahrar imposed

different conditions for the different endowments. Awqaf

in Central Asia were to become part of the waqf around

 127

the tomb of his grandfather, but the waqf in Kabul was

to be used for the support of the school that was set up

there (Ibid:339).

Properties attached to a waqf were ostensibly be-

stowed on a building -- e.g., a school, a mosque, a tomb,

etc. After the payment of taxes to the state, in case

the waqf had not been exempted from this obligation, the

income from the properties had to be first used for the

maintenance and repair of those premises. The number of

religious scholars, students, and other personnel of a

waqf were enumerated in the waqfnama, and the percentage

of the income of the properties to be devoted for their

upkeep specified. But it is in the rules regulating the

disposal of the remaining income of a waqf that the goals

of the endower are best revealed.

In the deed of 1326, the conditions make it clear

that the endower aimed at serving members of Sufi orders,

orphans, poor and needy members of his community, through

the waqf. The produce of the waqf was to be converted

into cooked food and served to the members of those groups

on the premises of the waqf. The spirit of the endowment

is best caught on the delineation of circumstances where

his own descendants and those of a named uncle were to

receive help from theW'aqf. He stated that whoever from

these individuals "whether male or female, living in

Fatehabad, follows the persuation of the sound people

 128

[implying membership in the Sufi order to which he be-

longed], does not have any wordly possession, is poor,

does not have any means of work, and does not derive any-

thing from agriculture, co~~erce, or other goods and pos-

sessions, should be given cooked food [lit. bread and

soup]" (1965:95). If any of his direct descendants or

kin were to cultivate the lands of the waqf as sharecrop-

pers, they were to be given the same share of the pro-

duce as any other sharecropper (Ibid:96).

Details of the waqfnama make it clear that the en-

dower had not converted all of his property into waqf

(Ibid:45,5I,67,etc.), and that as such his inheritors

were presumably provided for. But Khodja Ahrar imposed very

different conditions for the disposed of the awqaf that he set

up. In one waqfnama, he stipulated that after the cost of the

upkeep of the waqfwas met the remaining produce "in ac-

cordance to the principles of inheritance were to be

divided among the inheritors" (Ibid:173). In another

deed, he specified that the excess income of the endow-

ment should be used for augmenting the property of the

waqf (Ibid:29l). As far as the treatment of his descen-

dants was concerned, he made it obligatory that in case

they were in need, .the trustee of the waqf should help

them in amounts that he deemed necessary. Furthermore,

should any of his descendants not own a house, he was to

be provided with accomodations in the houses that were

 129

part of the waqf (Ibid:340). Khodja Ahrar's endowments

benefited a large number of religious scholars, their stu-

dents and attendants; but it is clear that insuring the

future well-being of his own descendants was his dominant

goal.

Thus, Islamic law, by providing the owners of pri-

vate property with such a wide range of latitude in the

determination of terms ofoawqaf, allowed these individuals

to wield considerable influence on their properties long

after their own death. The degree to which these provi-

sions were in fact carried out depended on the relative

stability and fluctuation of relations of domination. As

far as the estates of Khodja Ahrar are concerned, a docu-

ment recorded by Chekhovish makes it clear that by 1543

parts of theawqaf had been seized from his descendants.

The decree issued in that year by a ruler of Central Asia

restored all the properties and, in addition, exempted

those awaqf from taxation by the state (1974:311-313).

A degree issued much later (1825) by a ruler of Bokhara,

warning one of his nobles that an attendant of the said

noble should cease interfering with the awqaf of Khodja

Ahrar, indicates that until then the endowments were still

existed.

Continuity over time was even more pronounced in the

case of some of theawqaf that carne into existence through

actions of the state and were augmented by the offerings

 130

of individuals. The waqf formed around the alleged tomb

of Ali Ibn Abi Taleb, the fourth caliph of Islam, in the

vicinity of the city of Balkh in the northern region of

Afghanistan can serve as an example. A collection of ori-

ginal documents, compiled by Hafiz Nur Muhammad, an Afghan

court official, and published in 1949, unravel the story

of the waqf from 1480 to 1889. As Robert McChesney has

offered an exhaustive historical analysis of these docu-

ments in relationship to the institution of waqf in his

Ph.D. dissertation (Princeton, 1973), I shall confine

my remarks to some of the salient features of this insti-

tution.

As the main beneficiaries of the awqaf founded by

the state were the members of the religious establishment,

the motivation for the endowments has to be analyzed in

the context of the relation between rulers and Muslim

scholars and mystics. The ideological and political exi-

gencies of any period were therefore of particular impor-

tance in founding, extending, or curtailing the size of

the properties attached to a particualr waqf. At the

time of the founding of the waqf in Balkh, Islamic his-

tory was marked by "the emergence and rapidly growing in-

fluence of highly organized Sufi order" (McChesney,1973~

110). By granting the administration of the waqf to mem-

bers of the Naqshbandi Sufi order, the ruler was, on the

one hand expressing his open support for their doctrine

and, on the other, by giving them a stake in the economy,

 131

linking their interests to the stability of the state (Ib-

id: 117) .

Since the original waqfnama has not been published,

the conditions under which the trustees were to be ap-

pointed can not be ascertained. The historical evidence

presented by McChesney gives me the impression that un-

til the mid - seventeenth century the ruler retained the

right of appointing the trustee. But from 1668 to 1901,

the trusteeship was retained within one family (Ibid:361-

364). The family's claim that the office was bestowed on

them by the ruler who founded the waq£ in 1480, although

historically ambiguous, has been widely accepted.

The ruler who set up the waqf bestowed large amounts

of property on the shrine; its mainstay being an irriga-

tion canal with its attached lands. The canal was speci-

fically restored in order to reclaim the land bestowed

upon the shrine (Hafiz Nur Muhammad, 1949:65). No addi-

tion to the endowment took place in the sixteenth cen-

tury, but in the seventeenth century the size of the land-

ed properties of the shrine was considerably augmented

and, with some fluctuations, remained considerable until

1889. The new endowments came from rulers and officials

of the state(s) that succeeded one another in the area.

But unlike the initial endowment, the single major sub-

sequent bestowal of lands on the' waqf by a ruler was made

from lands specially bought for the purpose (Ibid:66-67).

 132

McChesney's description of the area of the waqf provides

us with an idea of its landed holdings. He writes that

"the maximum extent of the waqf properties ... was twenty-

two miles from east to west and seven or eight miles from

north to south. Seen as a rectangle, the area would have

been approximately 154 square miles or 98,560 acres in

the middle of the seventeenth century" (1973:281). The

size of the cultivated land, being a function of the irri-

gation network, was, of course, smaller than the size of

the area would indicate. In 1889, the landholding of the

waqf was reduced to only 2, 000 acres (Ibid).. Yet, com-

pared to the size of land controlled by other awqa·f or

even by most private landowners, the property of the waqf

of Balkh was amongst the largest in the country.

Throughout our unit of long duration, the state was

one of the large landowners of the country; land owned by

the state was known as khalisa (lit. unmixed or pure) .

Our examination of the waqf in Balkh has already shown

that the state could acquire land both by purchase and

by reclaiming waste land and bringing it under irrigation.

It could also obtain land by seizing it in wartime or con-

fiscating it from rebels and actual or alleged enemies of

the state. Areas depopulated by epidemics, large scale

emigrations, and/or wars, and the properties of those who

had died without leaving any heirs, could also revert to

the state.
 133

The area of the land owned by the state was subject

to constant fluctuations, and was reduced whenever the

state granted parts of it, in possession or private pro-

perty, to officials, local power-holders, members of the

religious establishment, groups of immigrants, or in be-

stowal on awqaf. The state could also lose property in

war or due to the secession of provinces and regions.

Available evidence, unfortunately, does not permit

a concrete determination of the relative distribution of

the size of different forms of property over time and

space. The issue has to be settled by future research.

Close reading of the evidence at hand lead to the impres-

sion that, during our time-unit of long duration, private

property might have been the dominant form of property

~elatio.~s. But regardless of the dominant form of pro-

perty relations, it is obvious that if the large property

owners, whether individuals,awqaf, or the state, were

to make productive use of their land, they had to till it

by hiring other people's labor. It is to a consideration

of the relationship between forms of labor and property

that I must turn now.

The simplest instance of such relations was the case

of the owner-cultivator in which the head of the domes-

tic unit combined the function of owner of the family

estate with that of manager of the family economy. In

the valley of Kunar, as already shown, the size of the

 134

estate controlled by 78% of such families was below 5.2

jeribs of average land. As an adult male was considered

capable of cultivating from 7 to 10 jeribs of such land,

most of the owner-cultivators had an excess amount of la-

bor-power at their disposal. The point of reference for

the size of the estate being the death of the male head

of the family, the resultant division of the estate a-

mong inheritors would have further reduced the size of

the holdings. Payment of taxes, meeting of consumption

needs and/or marriage expenses might further reduce the

size of the estate(s) of such families or force them to

part completely with it and be only left in pos.session

of their labor-power. The same end-result could also be

an outcome of the division of a small or medium estate

among a number of inheritors in the course of several

generations. In short, at any given time, owner-culti-

vators with surplus labor-power, as well as producers

possessing only their labor-power, would have been looking

for land to rent or to work on.

Those in control of property, whether large, medium,

or small, who avoided direct labor or whose own labor-

power was not sufficient for meeting the labor demands

of their estates, would have, on the other hand, been in

need of people with the required labor-power who would

cultivate their lands. Hence, the necessary interaction

between the two groups. The form of the relationshop and

 135

the resultant share of the parties in the total produce

varied according to where was the locus of managerial de-

cision-making regarding the use of labor-power as well as

the number of intermediaries between owners of the proper-

ty and direct producers.

Elphinstone's observation on the existence of five

classes of cultivators was quoted in this chapter. These

were: owner-cultivators, tenants, buzgurs, hired laborers,

and villians (1839,vol.I:389). The case of "villenage"

was an outcome of special political circumstances in an

area that lies outside the present boundaries of Afghani-

stan and will be analysed in the context of the discussion

of "tribes" in the second part of this study. The cate-

gory of buzgur, which according to Ephinstone were the

same as metayers in France (Ibid), calls for clarification.

Richard Jones who claimed, on the basis of Elphin-

stone's description, that their existence in Asia was

limited to Afghanistan (1831:100), provided a very suc-

cinct definition of this class of cultivators. He wrote

that "the Metayer is a peasant tenant extracting his own

wages and subsistence from the soil. He pays a produce

rent to the owner of the land, from which he obtains his

food. The landlord, besides supplying him with the land

on which he lives, supplies him also with the stock by

which his labor is assisted. The payment to the landlord

may be considered, therefore, to consist of two distinct

 136

portions: one constitutes the profit of his stock, the

other his rent. The stock advanced is ordinarily small.

It consists of seed; of some rude implements; of the ma-

terials of others which the peasant manufactures; and of

such materials for his other purposes as the land itself

affords; building timber, stone, &c. and occasionally of

some draft animals" (Ibid:73).

The classes of cultivators described by Elphinstone

may, in practice, have existed either in pure or mixed

forms. The difference would have been a function of the

degree to which other members of a family of an individual

cultivator, who fell under a certain category, belonged

themselves to that category, as well as the degree to

which that individual participated in the family econolny.

Analysis of the owner-cultivator class made it clear that

one of the main reasons for the mixed character of classes

of cultivators resulted from the employment of surplus

labor-power at the disposal of a family manager bent on

the enhancement of the family economy. The economic uti-

lization of this surplus family labor may not have always

been in a productive activity but may have involved inter-

action with non-productive sectors of society. A hypo-

thetical example will help in illustrating the mixture of

these forms.

A family, consisting of a husband, wife, a daughter,

three sons of working age, and owning twojeribs of aver-

 137

age land and a pair of bullocks -- not an uncommon situ-

ation in the valley of Kunar in the 1880s -- would have

been a family of owner-cultivators in possession of sur-

plus labor-power. One of the sons could be hired out to

a family of owner-cultivators where one adult male culti-

vator owned more land than he could attend to by himself.

The second son could volunteer for the army -- a large

establishment throughout our unit of long duration -- or

become the follower of one of the numerous religious dig-

nitaries in the region, living off charity while hoping

to become a preacher in a mosque in time. The remaining

son could acquire six jeribs of land as a buzgur from

a family composed of women and/or minor children, the

state, a waqf, a medium or large landowner. Last but not

least, the father could rent six jeribs of land from one

of the above categories of landowners or sub-rent from

someone who had rented larger plots of land. Furthermore,

if the fru~ily was located in an area close to British

India, the two sons could move there as seasonal laborers

during the winter and work in agriculture or construction.

The mixed forms in which the surplus labor of the family

was used can also serve to illustrate the utility of the

analystical distinction regarding the loci of management.

The employer of the agricultural laborer would have made

the decisions regarding the way the latter's labor-power

was to be used. By contrast, regarding the family estate

 138

and rented land, the father would have made the necessary

decision; but he would have had to closely coordinate with

his buzgur son the timing of the use of the pair of bul-

locks. The latter would have probably had no voice on

the choice of crops and the system of rotation of the land

on which he worked as a buzgur, but would have been rela-

tively free in the determination of his daily schedule and,

depending on the pressure of work, might have been either

able to help his father or be in need of his father's help.

As long as the sons left the father in control of

their earnings, the family economy would continue and the

family could meet its consumption needs, provide for the

marriage of the sons, and hope to increase the size of

the family estate. Disruption of the family economy, pre-

cipitated by the father's death and the subsequent inabil-

ity of the brothers to cooperate, or by the refusal of

one or several sons to turn over their earnings to the

father, would have brought a shift in the positions of

the sons.

I have chosen this example, a situation that I have

frequently witnessed during my fieldwork, to show the

possible pitfalls of analysis based on ideal types and

false abstractions regarding the economic autonomy of

the peasant family and its alleged isolation from the

wider social relations of society. This emphasis should

not, however, obscure the fact that large numbers of

 139

cases throughout our unit of long duration did concretely

embody the class categories distinguished by Elphinstone.

Their existence as tenants, bUzgurs, and agricultural la-

borers is best documented in their dealings with large

property owners, an issue to which I will now turn.

It should be pointed at the outset that the large

size and/or geographically dispersed character of the

holdings necessitated that relations between owners and

cultivators of the soil be mediated by other people. In-

terposed between the two classes were a group of overseers

(nazir) appointed by the landlords, and a group of renters

(ijaradar) who rented the land from its owners for a fixed

sum of produce and/or money for a fixed period of time.

In anyone locality, depending on circumstances, one or

several classes of cultivators enjoying different degrees

of autonomy of direct producers in the use of their labor-

power, could be employed in the course of the production

process by the landlords' overseers or the renters. Large

property, as shown earlier, was divided into private pro-

perty, waqf, and state property but, as far as the organ-

izational features of the labor process was concerned,

all three forms displayed similar features. A number of

concrete examples will clarify these points.

In the valley of Kunar, the state, in addition to

khalisa lands, also controlled lands, water-mills and

other property confiscated from fugitives. The court re-

 140

cords enable us to examine the types of arrangements a-

dopted for the exploitation of these properties. The un-

dertaking encompassed renting out of property as well as

hiring tillers.

Rents were concluded for a period of one year; the

payment had to be made in cash and/or kind; the number

of renters was one person or more. In March of 1886, for

instance, two individuals undertook to make a payment of

4928 kgs. of grains -- wheat, corn and barley -- for the

rent of two confiscated water-mills in the locality of

Kunar-i-Khas. At the same time, one of these individuals

also rented some khalisa lands for a sum of 8 rupees in

cash and 1936 kgs. of produce wheat, corn, rice. Both

contracts were for a duration of one year. Since one of

the renters was a village headman, it seems unlikely that

he would have had the time, even if he had the inclination,

to farm the lands himself. The land must have been cul-

tivated by another person. Its size being rather small,

only one person was needed and the tilling might have been

undertaken by a family member, an agricultural laborer

hired out for the purpose, a buzgur, or sub-leased to a

tenant. For the operation of the water-mills, the renters

could rely on hired laborers, sub-lease to millers, or

enter into a partnership with the millers. In all these

cases, the renters would have had enough time to super-

vise the direct producers.

 141

The state hired buzgurs in four districts of the

valley of Kunar, and the financial officer of the govern-

ment appointed a numbei of oversees for their supervision.

A single agent supervised work at two of the districts

but different individuals were responsible for the spring

and autumn crops. The other two districts were simultan-

eously administered by two overseers. Two of these were

headmen from the valley of Kunar but the other two were

clerks who came from outside the valley. The overseers

received the needed seed from the financial officer of

the government and distributed it to the buzgurs. The

types and quantities of seed given to the buzgurs reveal

that they were assigned different amounts of land and

while some had to CUltivate only one crop others had to

attend to two or more. The overseers would have had to

check routinely on the work of the buzgurs and, at the

time of the harvest, would have to pay close attention

to its gathering, weighing, and dividing among the direct

producers and the state. The recording of these opera-

tions in the courts gave the arrangements the character

of legal contracts and provided for easy auditing and

reference.

There were 34 buzgurs who worked the land in the four

districts: as double crops were cultivated in two of the

districts, 24 buzgurs worked around the year. The terms

of contract of the remaining 10 were for nine months.

 142

On November 17, 1886, buzgurs of the districts of

Pashad and Kutki were given 83.6 kgs. of cotton seeds, 154

kgs. of local rice, 380.6 kgs. of thick rice, and 1331 kgs.

of corn for the spring cultivation of 1887. Out of the

1948.6 kgs. of seeds, 893 kgs. were used for the cultiva-

tion of the confiscated lands and the remaining 1055 kgs.

for the cultivation of khalisa·. The seeds used for the

fall crop of the same year had consisted of 28 kgs. of

lentils, 39.6 kgs. of barley, and 1562 kgs. of wheat. In

some years, the overseers were given seeds for some crops

and cash for the procurement of the seed of other crops.

In March 1887, for instance, the overseers were given 96.8

kgs. of wheat, 193.6 kgs. of barley, and 30 rupees in

cash for other seeds by the financial officer of the go-

vernment for cultivation of the land in the two districts.

In terms of my earlier estimates of the ratio of seed to

land, these figures seem to indicate that the area in

question must have not exceeded 150 jeribs.

The arrangements in Kunar did not involve the inter-

position of a large number of intermediaries between the

state as landlord and the tiller of the soil. In this

regard, the situation in the Garmsel district, southern

Afghanistan, was more complex. A British official who

visited the area in 1903 wrote that "all the land in the

Garmsel tract of the Helmand Valley belongs to the State.

Each canal taking off from the Helmand [river] is under

 143

the control of a Hakim who appoints an official called

the Mirabi as his deputy. Under him again is a Kotwal.

Each main canal is divided into a number of subsidary ca-

nals, each under a Kadkhuda. He, in consultation with

the Khan or headman, appoints a number of Kashtigars (sort

of zamindars) who control an allotted portion of the sub-

sidiary canal. The Kashtgirs do not actually till the

land but hand it to Bazgars, to cUltivate. The Kashtgirs

supply ~hem with seed, plough and implements and pay them

an annual sum of 20 krans [10 rupees] termed Kafshi, for

which consideration they must clean and repair the canals"

(GA, Farah: 86) .

The distinction between kashtgirs and buzgurs was

not confined to the Garmsil district or to the nineteenth

century. Henry Rawlinson, who served as the political

officer in Kandahar during the Anglo-Afghan war of 1839-

42, differentiated between the two classes in his detailed

survey of the agrarian and political situation of the

southern region from 1747 until his own time. He wrote

that during the reign of Ahmadshah, 1747-1772, "the inde-

pendent and lucrative occupations of cUltivating the

lands -- that is of providing the seeds, procuring the im-

plements of husbandry, keeping up the necessary cattle,

and realising the produce -- revolved" around different

people from those who undertook the "actual manual labor

of tilling the ground, tending the plough, etc." (GA,

 144

Kandahar:5l6). Rawlinson's description leaves little

doubt that the system was prevalent on the lands owned by

the state as well as on the large estates owned by pri-

vate individuals.

Elphinstone's remarks on another category of land-

lords in the area of Kandahar may provide us with a clue

to the identity of the kashtgirs. In a revealing compari-

son, he stated that there were landlords in Kandahar who

"have their lands cUltivated by Buzgurs, by hired laborers,

or by slaves. They act themselves as superintendants,

often putting their hand to any work where they are wanted,

like middling farmers in England" (1839/1972,vol.II:109).

The managerial expertise and the financial ability of

these landlords best fitted them to act as kashtgirs for

the state and large landlords, thereby increasing the

size of their own landholdings. It has to be noted that

the process of division of property set in motion by the

death of a large landlord, when not counteracted by fur-

ther acquisition of land by his inheritors, could in the

course of several generations reduce the heirs to the sta-

tus of middling farmers. The degree to which this pro-

cess of social reproduction was linked to the mode of

domination as well as the degree of prevalence of slavery

will be explored subsequently.

All large landlords, however, did not have to rent

their lands; they could enter into partnership with other

 145

landowners and set up a bureaucratic apparatus of manage-

mente A document reproduced by Fofalzai (1967,vol.II:665-

66) provides an instance of a very interesting sort of

partnership between a large landowner and three owner-

cultivators.

The document in question is a deed of sale from 1857,

in the village of Kalakan, north of the capital city of

Kabul. A large landowner sold, for a sum of 300 rupees,

two shares out of five in a property using 888 kgs. of

seed, to three individuals. But the transaction did not

involve any transfer of cash. The buyers, in return for

the land that they were to receive, committed themselves

to build a wall around the remaining lands of the owner,

to plant these lands with vines, and to carryall the

operations of irrigation and CUltivation for a period of

five years, at the end of which they would turn over the

vineyard to the owner. This form of cooperation was fully

sanctioned by Islamic law and was commonly reverted to in

connection with manufacturing enterprises (Halil Inalick,

1969: 101) •

In the region of Sistan, western Afghanistan, large

landowners delegated the supervision of direct producers

to a number of intermediaries. Supervision of irrigation*l

and cultivation was considered part of the same process.

*lThe system of irrigation in the Sistan was described

earlier in chapter II.
 146

A number of districts in the Sistan were directly owned

by members of one family who all carried the title of

sardar. An English official described the following a-

gents as consituting "the establishment usually employed

by each of the Sardars. The Niab, who makes the arrange-

ments for cultivation as well as for canal and band [dam]

repairs. He is allowed two ploughs of cultivation .•. of

which he pays no share of the produce to the Sardar. The

Mirab is his subordinate and he takes from him half the

grain he receives from the cultivators of the plough.

Each Sardar has his own Niab.

"The Nazir, who keeps the account of the grain re-

ceived from the cultivators and submits to the Sardar.

He takes over the grain at the division of the produce .

.•. The Mirza writes up the revenue received from each

plough and at the time of dividing out the produce attends

on the the spot. The Mashrif guards the crops in the in-

terest of the Sardar from the time that they ripen until

the produce is made over to the Nazir.

"The Mirab works under the Niab for canal and band

repairs. The grain which he obtains from the plough is

divided equally between him and the Niab. (On these ca-

nals the water baliff is usually called Mirab, seldom

Pakr. )

"The Kot';",al also works under the Niab and guards the

distribution of water into the branches from the canal.

 147

"Qail gardan. This is the man who does the labor of

measuring out the grain. "(GA,Farah:261-262).

The lands of each of the sardars was distributed

through the niab to the direct cultivators, who also de-

termined the quantity of the seed that each buzgur re-

ceived. The number of buzgurs on each of the branches of

a canal was described on by the sardar and in general

there were from 8 to 12 buzgurs on each branch. That the

buzgurs had to coordinate their work as a group is stated

clearly by the English official. He wrote that lIeach

Bazgar's cultivation is called a Kulba, and each plough's

cultivation is called a juftgao or tak; usually four

Bazgars unite to work the plough, but in some years the

syndicate consists only of two or three menll (Ibid:262).

Most of the forms of employment of labor actually

adopted by the state and by private landowners were also,

theoretically, open to the trustees of awqaf. Unfortun-

ately, I have found no concrete description of such re-

lations. Some information, however, allows me to ascer-

tain the main tendencies; the main sources are those pro-

visions recorded in waqfnamas and dealing with labor re-

lations. In the two main waqfnamas presented by Chekhovich

and described earlier, the maximum period that any part

of the land could be rented out to the same individual was

two years in one case and three years in the other (1965:

107;1974:340). One of the endowers expressly forbids the

 148

trustees of the waqf to rent any part of the land to the

governors of the city, the nobles or officials of the

court or their servants and children (Ibid,1965:105-06).

These provisions reveal that renting out land was a wide-

ly utilized vehicle for making profitable use of the

awqaf properties. That nobles and other men in power en-

gaged in renting land indicates that they must have in

turn been sub-leasing these lands to direct producers or

cUltivating them through appointed agents. Furthermore,

as slaves formed part of theawqaf (Ibid,1965:108;1974:

172), they could be put to direct work on the awqaf pro-

perties or entrusted with the supervision of the direct

producers who may have been hired as agricultural laborers

or buzgurs by the trustees of theawqaf. Indeed, one of

the waqfnamas expressly specifies that the slaves be used

for direct production, service, and supervision (1965:108).

As far as the waqf of the shrine of Balkh is con-

cerned, the only available information is that, upon its

foundation in 1480, a bureacratic staff of about 120 in-

dividuals, who were officially referred to as "the reli-

able stewards" were 3ent there by the ruler to see to the

day-to-day management of the properties of the waqf

(McChesney,1973:123). With the acquisition of trustee-

ship as a hereditary office by one family, this staff

probably became increasingly subordinated to that family.

That the trustees paid careful attention to the upkeep of

 149

of the waqf properties is revealed by the fact that they

managed to keep the main properties of the shrine "in

seemingly excellent condition" during the political tur-

moil of the eighteenth and nineteenth centuries (Ibid:252).

The task of managing awqaf properties that were in-

itiated by the state was, at times, made easier as the

state granted the trustees rights to the labor of members

of neighboring communities. The yearly labor required

for the cleaning of the irrigation network on a canal, in

the central place of Kandahar, which watered land that

had been turned into waqf by the state, had to be per-

formed by the inhabitants of four specified districts lo-

cated in the vicinity of the canal. The amount of labor

necessary to cultivate the lands of this waqf was conse-

quently considerably reduced. This "privilege" was con-

firmed in a series of edicts by Afghan rulers of the

eighteenth and nineteenth centuries (Fofalzai 1967:63,69).

Although the trusteeship of this waqf was to be confirmed

by every new ruler and upon the death of every trustee,

it in fact remained within one family during the eigh-

teenth and nineteenth centuries. From a document recorded

by Fofalzai, it appears that the sons of at least one

trustee took an active part in supervising the work of

the hired laborers and protecting the crops from the herds

of large owners (Ibid:53).

In short, during our unit of long duration, the same

 150

set of options regarding the employment of laborers for

cultivating the land was available to all large landowners

in Afghanistan and, depending on concrete circumstances,

one form was preferred over another or several forms were

combined.

How was the product distributed among the landlords

the intermediate groups and the actual tillers of the

soil? Since the mode of distribution was conditioned by

the type of relationship between owners and producers,

each type has to be taken into account separately.

The case of renters and agricultural laborers is

analytically the simplest. In one case, the share of the

landlord and, in the other, that of the laborer, are fixed

prior to the actual yield of the land. The renter had to

pay a fixed amount of produce and/or money to the land-

lord for the duration of the contract. This payment, un-

less provisions for the impact of natural conditions were

made part of the agreement, had to take in the specified

amount regardless of the yield of the land. The amount

of rent was fixed in accordance with concrete circum-

stances of time and space and could vary from 1/10 to 2/3

of the average produce (Elphinstone, l839,vol.I:390).

The common terms of lease were one to two years, the

longest contract being for five years (Ibid).

The wages of a laborer were fixed at the time that

he was hired and were not affected by the size of harvest

 151

produced by him. The usual contract of agricultural la-

borers was for nine months (Ibid:391). At the time of

Elphinstone's writing, the wages per nine months of an

agricultural laborer varied from 91 kgs. of grain and 1

rupee to 363 kgs. of grain and 2 rupees. But laborers

were also provided with food and clothes (Ibid). By com-

parison, the monthly pay of a laborer in the towns was

about 17 rupees per month (Ibid). No figures on the

wages of agricultural workers are available for the sub-

sequent period but the monthly pay of manual workers in

the government workshops in Kabul was 8 rupees and that

of skilled workers 20-30 rupees per month in the l890s

(Kakar,1979:239).

Determination of the renumeration of buzgurs poses

a different set of problems. Since buzgurs received a

fixed portion of the harvest, their actual share of the

produce along with that of their landlords varied with

the productivity of the soil. According to folk theory,

productivity of the land was a function of five elements:

land, water, labor, seeds, and draft animals; the provi-

sion of the 5 elements by those entering into agreement

for the cultivation of a piece of land should be reflected

in their respective share of the produce. But in prac-

tice, the distribution of the harvest was a function of

the interposition of intermediate agents as well as the

economic and political position of the landlord and his

 152

ability to dictate terms to the tillers.

In table V. , I have swnmarized the relevant inforrna-

tion on agricultural productivity in 28 provinces of the

country for the year 1977.

Table v. Agricultural Productivity

Region Province AVerage Yie1d/Jerib of

Irrigated Land

Eastern Ghazni 318 kgs./jerib

Kabul 333
Kapisa 311
Kunar 296
Laghman 407
Logar 296
Ningarhar 370
Paktia 333
Parwan 370
Wardak 296

Central Bamyan 311

Urizgan 318

Southern Helrnand 327

Kandahar 333
Zabu1 281

Western Badghis 305

Fa:!:"ah 274
Ghurat 311
Herat 355
Nimruz 281

Northern Baghlan 333

Balkh 355
Faryab 318
Jauzian 355
Kunduz 370
Samangan 333
Takhar 333

Badakhshan Badakhshan 281

Source: Statistical Information of Afghanistan (1976-1977) :78-81.

 153

Although the figures are in terms of average and

therefore hide the variation within each of the provinces,

the differences within and between regions can be readily

discerned. No similar figures exist for all the regions

of the country during our unit of long duration; and for

districts for which figures are available the use of in-

comparable land units makes the task of comparison diffi-

cult. The following information may still provide some

general idea of the situation.

The ratio of yield to seed quoted for the western

region in the 1890s varied from 7 to 16 times in normal

years (GA,Farah:48,163), but in bad years it fell as low

as 3.23 times the amount of seeds sown (Ibid:163). The

ratio of return in 1737 for the central place of Kandahar,

on the other hand, was 25 fold (GA, Kandahar: 509) . As

this ratio was the result of actual experiments, British

officials used it as the basis for their calculation for

yields of the southern, western, and northern regions.

The harvest of a kulba (plough) of land was at that time

around 19300 kgs. (Ibid) but as the number of jeribs in

a kulba varied, the determination of the yield per jerib

is not possible.

The valley of Logar in the eastern region is one of

the few districts for which the yield is actually given

in jeribs. In the1880s, the yield of a jerib of land

of first quality in that valley was estimated as 896 kgs.

 154

of grains (GA,Kabul:348), probably among the highest in

the country. By contrast, for the same decade, the yield

of a kulba of land in Badakhshan was estimated at 1480 kgs.

(GA,Badakhshan:31). Unfortunately, there is no indication

as to how many jeribs were in a kulba at that time but

since it is assumed that the land was worked by one cul-

tivator it must have been between 7 to 10 jeribs.

The fixed share of direct producers varied from 1/10

to 1/3 of the harvest in different parts of the country

(Kakar,1979:188). However, as the area of land cultivated

by each tiller is not given, it is difficult to arrive

at a comparison of the actual amount of produce that the

direct producers could dispose of in different parts of

the country. That the size of the area tilled by a cul-

tivator did make a crucial difference in his share is

clear from an instance where unusually complete informa-

tion on the shares of the various parties to the harvest

is provided. The example is from the district of Garmsil,

in the southern region, where, as noted earlier, the

whole area was owned by the state.

"When the harvest is reaped," wrote an English offi-

cer, "the Government tahsildar takes one seventh of the

gross result. He takes 70 mans [413 kgs.] termed the

haftad. The following men get their perquisites as de-

tailed:
 155

"Mirabi (canal overseer) gets 15 mans [1 man = 5.89

kgs]
Blacksmith 12

Carpenter 7

Hajam (barber) 5

Kotwal 2

"while the man who supplies the oxen for ploughing and

grinding gets 2 mans per diem.

"The (net) amount of the grain which now remains is

divided, between the Bazgar who gets 1/6, and the Kashtgir

who gets 5/6.

liThe distribution of the harvest would work out as

follows if we take for an example a Kashtgiri producing

2,000 mans [11793 kgs.]

Mans

Government share 355

Mirabi's 15

Blacksmith's 12

Carpenter's 7

Hajam's 5

Kotwal's 2

Oxen owner's 2

(assuming that they plough 4 and

thresh 4 days) 16

Total 412 [kgs.2429]

Balance left = 2,000 - 412 = 1,588 [kgs.9363.9]

Bazgar's share 265 [kgs.1562.6]

Kashtgir's share 1,323 [kgs.7801.3]

 156

"Thus the Kashtgir in this case receives about 66, the

Bazgar about 13, and the State about 18, per cent of the

harvest" (GA,Farah:86-87).

The situation in the district of Garmsil was rather

special. While its soil was considered "extremely fertile

and its command of water unlimited" (Ibid:85), it had only

a settled population of 5440 families and a nomadic popu-

lation of 11,000 families (Ibid:86). The tillers were,

therefore, in a position to cultivate the maximum extent

of land they wp.r~ ~~pable of attending to. However, the

circumstances in the valley of Kunar, as shown earlier,

were rather different. A family of owner-cultivators own-

ing 5.25 jeribs of average land could only produce a har-

vest of 600-800 kgs. of grains. This contrasts sharply

with buzgur's income of 1562.6 kgs. in the Garmsil al-

though that represented only 13% of the harvest.

Such an outcome can only be surprising to analysts

like Pearson who divorce the institutional definition

of the surplus from its embeddedness in the reproduction

of the technical and social process of labor.

But property relations as Richard Jones argued are

political relations and not mere technical arrangements.

The significance of this observation is clearly borne by

the respective shares of the buzgur and the kashtgir in

the produce in the district of Garmsil. The kashtgir,

whose claim to the produce was derived from his relation-

 157

to the state, and whose actual participation in the labor-

process was minimal, ended up with 66% of the harvest.

He was clearly in possession of more grains than he and

his family could directly consume. The kashtgirs were not

alone in this position. Depending on the concrete circum-

stances of production, other groups, e.g. large landowners

and sometimes even samller ones, were in possession of a

surplus. How did they dispose of this surplus? The fol-

lowing chapter is an attempt at answering this question.

 158

Chapter Four

Circulation, Transportation, and Markets

In discussing costs of circulation under conditions

of capitalist production, Marx asserted that "the circula-

tion i.e. the actual locomotion of commodities in space,

resolves itself into the transport of commodities. The

transport industry forms on the one hand an independent

branch of production and thus a separate sphere of invest-

ment of productive capital. On the other hand its distin-

guishing feature is that it appears as a continuation of

a process of production within the process of ciruclation

and for the process or circulation" (Capital,II,1885/1967:

152). Being mostly concerned with a synchronic descrip-

tion of capitalist production in "pure form", except for

a passing remark on the importance of transportation in

the land of the Incas (Ibid:149) , Marx did not investi-

gate the relationship between this "continuation of pro-

cess of production"in other forms of society. The ques-

tion, however, is of critical importance in understanding

the structure of production in pre-capitalist social form-

ations. In this chapter I have, therefore, undertaken to

investigate the articulation of the technical organization

of circulation with the social and political conditions

in the context of which the production and movement of

goods took place.

 159

To elucidate the impact of the technical organiza-

tion of circulation on the spatial organization of pro-

duction we refer to the work of Johann Heinrich von Thunen.

Thunen's analysis of the formation of the price of grain

can serve as an illustration of his method which takes

account of distances and systems of production independent-

ly from other relevant variables. He assumed that in an

"Isolated State" a central place lying in the center of

a fertile plain is the only market for grain and that

"there is no navigable river or canal in the entire State,

so that all grain has to be taken to the Town by horse and

wagon" (1826/1966:l2). He then proceeded to show that as

the distance from the Town increased, the amount of fod-

der needed for feeding the horses on their way to and

from the market also increased (Ibid:16-l7). The cost of

freight, therefore, directly entered into the cost of the

grain in the Town. Assuming that the plain possessed un-

iform geographical features, Thunen postulated that "the

Town's needs can only be met if it pays a price which at

least 'compensates, for his outlay in producing the grain

and bringing it to market, the most distant produc:er

whose grain is still required" (Ibid:144).

Thunen argued that the grain price was not arbitrary

but that "continual fluctuations in demand lead to con-

tinual fluctuations in the grain price" (Ibid:145). Since

Thunen's aim was the prescription of the most suitable

 160

pattern of crops for an Austrian estate, he gradually in-

troduced different elements of the reality into his model

to assess their impact in relationship with other condi-

tions of production and circulation (Ibid:171-74). That

he was predominantly concerned with Austrian conditions

is reflected in the fact that he thought that in the real

world lIthere is no large town that does not lie on a navi-

gable river or canal ll (Ibid:171).

It is not these particular modificiations that inter-

est me but the impact of distance on the relations between

people in concrete historical circumstances anq, in this

connection, Von Thunen's formulation of the relationship

between price of commodities in central places and their

peripheires over time will be of main concern to me. In

chapter One, I showed that throughout our unit of long

duration the means of communication within the territory

of Afghanistan remained the same while, from 1850's on,

the conditions in the neighboring territories changed. In

this chapter, I shall attempt to provide an analysis of

the concrete patterns of circulation within the regions,

among regions, and with territories lying outside the bor-

ders of Afghanistan.

Although I find a regional unit of analysis a useful

intermediary between the local and large foci of analysis,

I consider it necessary to place the regional unit within

a larger framework of historically relevant set(s) of

 161

flows. This departure from the regional approach is dic-

tated by a number of considerations. Part of the problem

stems from the very way in which a region is defined.

According to Carol Smith, a region can be "defined formal-

ly or functionally, the former placing emphasis on the

homogeneity of some elements within a territory, the lat-

ter placing emphasis on systems of functional relation-

ships within an integrated system" (1976,vol.I:6). The

highlighting of the importance of some elements or func-

tional relationships is useful, as a preliminary step, in

placing the region within a larger set of relationships.

Misplaced emphasis when the regional orientation is the

only unit of analysis may, however, lead to the obscuring

or neglecting of significant relationships. As in many

a theoretical model, the stringent conditions of the cen-

tral-place theory are "rarely met in the real world"

(Smith,C.1974:169). The point, however, is that later

practitioners of Thunen's approach, with a few exceptions,

have neglected to introduce elements of the real world

into the model in order to better account for these con-

ditions. At ti~es, they have even turned Thunen's pro-

cedure on its head. smith makes the pertinent observa-

tion that "anthropologists, particularly archeologists,

who have attempted to predict or explain settlement dis-

tribution without data on market functions ... are going

far beyond the limits of the theory" (Ibid:171).

 162

For a theory that claims to explain the systematic

features of markets and exchange relationships, a major

shortcoming of the approach has been its failure to ad-

dress the impact of these political conditions affecting

the movement of people and goods. That these conditions

have been historically important in the determination of

pattern of movements and price of goods is reflected in

the special category that Fredrick Lane has proposed to

take account of them. He states that "for individual

enterprises engaged in international trade protection

hardly ever appears as a free good. Costs of protecttion

are vital factors in production since their variations

frequently determine profits (1942/1966:385). By examin-

ing a number of examples from European history (Ibid:373-

428). Lane attempts to demonstrate the significance of

costs of protection on the organization and development

of enterprises and states. Lane also considers protec-

tion from the perspective of organizations that specialize

in suppling it; I shall address this aspect of his hypoth-

esis later and confine myself here to an examination of

costs of protection in the determination of the direction

and price of commodities in inter-territorial trade. Be-

fore engaging on this task, however, I must provide a

description of the regions of the country, the types of

circulation of commodities, and the social organization

of the process of circulation.

 163

In chapter one and two, I argued that the dependence

of the population on irrigation accounted for the differ-

ent types of settlements. Zones of intensive concentra-

tions of population were located on or near main rivers

while more dispersed settlements were located on relative-

ly minor irrigation networks. The resultant distribution

of the population within each and all regions was, there-

fore, rather uneven. As our focus of analysis in this

chapter is on the phenomena of circulation, a correct

grasp of the placement of people and resources as well

as the location of areas void of habitation 22comes neces-

sary. In order to convey the picture adequately, I have

decided to quote brief descriptions of the major regions

of the country as viewed by late nineteenth century ob-

servers. Indeed, the structure of communication during

my fieldwork (1973-77) had changed radically from what it

had been during the period of this study, and reliance

on contemporary reports is quite essential for the illu-

stration of my subsequent argument.

Major C.J. East, Quarter-Master-General's Department,

provided in 1878 a remarkably succinct view of the eastern

region of the country. He stated that "the valley of

Kunar, near the city has an elevation of from 6,000 to

7,000 feet above the sea: its length to the plain of

Peshawar is nearly 200 miles, which plain has an altitude

of little more than 1,000 feet above sea level. In the

 164

neighbourhood of Kabul, this elevated valley is very fer-

tile; and to the south of the city is a rich district ex-

tending to Ghuznee, including the valley of Logar, whilst

to the north are the very productive regions of the Koh

Daman and the valleys of Ghurband, Punjsheer, [and Nijrow].

South-east of Kabul is the hilly district of Lughman, of

no great fertility, whilst further east again extends the

sultry but rich plain of Jelalabad.

"The small range of Khyber may be said to separate

the plain of Peshawar from the valley of Kabul river; its

highest peak is 3,500 feet above the plain. It is pierced

through at two places, on the south by the valley of the

Khyber, and further north by the Kabul river.

"The Safed Koh or Speen Ghur [white mountain] range

commences west of Peshawar, and runs in a westerly direc-

tion to within thirty miles of Kabul. It throws out two

large spurs to the north, one the Khyber, above noticed,

and the other the Kurkutcha. Its general aspect is one

of greatest sterility, but it encloses numerous well-wa-

tered and fertile valleys.

"The Solimani mountains have a general direction

parallel to the Indus; their southern limits may be taken

at the Bolan Pass [in Baluchistan], whilst on the north

they extend to near Bunoo. The highest point of this

range, that of Tukht-i-Suliman, is 11,000 feet above sea

level.
 165

"Along the northern base of Safed Koh lies the well

cultivated district of Nungerhar. To the south of these

mountains is the very extensive and fertile valley of

Koorum; to the south of which again is the Khost valley.

"To the west of Solimani range, extends a large ele-

vated plateau but poorly inhabited, affording no produce,

and but indifferent grazing grounds; it is traversed by

a few tracks only.

"The district of Ghuznee is separated from those of

•• ' A.

Kabul and Logur, by a highland stretching east and west,

and having an elevation between 8,000 and 9,000 feet a-

bove the sea. To the south of the highland, are the

valleys of the Arghandab and the Tarnak [rivers], stretch-

ing as far down as Kandahar; both are extremely produc-

tive of grain, especially wheat. The plain of Kandahar,

where watered, is fertile, and has an elevation above the

sea of about 3,500 feet" (L/P&S/18/C71:2-4).

Another British officer provided an even fuller des-

cription of the physical features of this south-western

region. According to him, the region "is divided into

two clearly marked parts by the Kadanai river which rises

in the Toba highland and joins the Dori; the combined ri-

ver again runs into the Arghandab southwest of Kandahar,

and finally joins the Helmand at Kala Bist. North of this

dividing line the country is hilly, and consists of a nurn-

ber of valleys running southwest and watered by the fol-

 166

lowing rivers (from east to west): the Arghastan, Lora,

Khushk, Tarnak, Arghandab, Kuskh-i-Nakhud, and the main

stream of the Helmand. These streams all flow into the

Kadanai-Dori-Arghandab joint stream running from east to

west, and the valleys all fall from northeast to south-

west. The dividing ranges, rising in some cases up to

9,000 feet, gradually sink down to the plain along the

north bank of the Kadanai-Dori-Arghandab stream, which

falls from about 4,000 feet at Gatai, near which the

Chaman-Kandahar road crosses the Kadanai, to about 2,500

at Kala Bist. The northern part of the province is the

cultivated and inhabitant portion.

"The half of the province south of Kadanai-Dori-

Arghandab stream consists of the Registan or sandy desert

extending from the Helmand to the Khwaja Amran and Sarlat

ranges of Pishin, with the exception of the small

Shorawak district watered by the Pishin Lora where it

leaves the hills. Nearly the whole of this tract is un-

inhabited except by nomad tribes of sheperds and consists

of sand hills rising from 200 to 500 feet in height. The

sand ridges run parallel to one another in broken billows,

with. an apparent general direction of north-northwest to

south-southeast. The northeast or lee side of the sand

hills slopes at a natural angle of 45° towards the crest,

which is often sharp. Along the borders it is not entire-

ly desert, but carries some vegetation, and after rain

 167

grown some grass. This half-desert or Nim Chol, as it is

called, affords pasturage to numerous flocks of sheep and

herds of camels and is said to extend for 5 to 10 miles

inwards" (GA, Kandahar: 4) .

The country between the Helmand river in the south-

western region and Kashrud river in the northwestern re-

gion is described by Major East as little cultivated and

containing extensive tracts of desert (L/P&S/18/C71:4).

Part of this area was more fully described in the Gazet-

teer. After stating that Girshik on the river Helmand

was 2,881 feet above the sea level and Farah on the

Farahrud river 2,400, the officer wrote that "from the

line Girishik-Farah the country has a slight general fall

towards the hamuns or lakes of Sistan, which forms a de-

pression at a minimum altitude of 1,100 feet and receives

the whole water from the [area]. The rivers running into

this depression are the Helmand, Khash, Khuspas, Farah

and Harut Ruds. Between these rivers lie dasht or gravel-

ly, flat plateaux, as a rule waterless, except in depres-

sions or beds of old streams and uninhabited; the largest

of these plateaux lies between the Helmand and Khash, and

in its southern part in a northeast to southwest direc-

tion and 80 to 90 miles in width" (GA,Farah:3).

This north-western region was geographically defined

by the compiler of the Gazetteer as "that part of Afghani-

stan drained by the Murghab, the Hari Rud, the Adraskand,

 168

and the affluents of the Farah Rud, above where it enters

the plains.

"It is bounded on the north by the chol, on the east

by the mountains of the Hazara country, on the west by

the Khorasan deserts. It is thus, to a certain extent,

physically cut off on all sides but one, the south, by

natural obstacles from the surrounding countries and pro-

vinces. On the southern side it is open, and the great

roads from Kandahar and Sistan lead to it through the

broad space between the Taimani hills and the Persian

desert.

"The most populous, fertile and flourishing part of

the province is that comprised in the districts of Herat,

Ghorian, Obeh, and Karokh. North of this fertile tract

is the Band-i-Baba or Siah Bubak, known to Europeans as

the Paropamisus. This mountain range is really a prolong-

ation of the middle branch of the Koh-i-Baba. North of

Herat, and to the eastward of that place, the hills are

of some height, the peaks rising four or five thousand

feet above the valley. North of the Band-i-Baba is the

district of Badghis, for the most part an expanse of open

rolling downs of light clay soil, covered with rich grass

during the earlier half of the year, but arid and parched

during the latter half. This district becomes more and

more hilly as one travels east, until it may be character-

ized as almost mountainous. The only perennial river in

 169

the whole of Badghis, not counting the Hari Rud on its

western border or the Murghab on its eastern, is the

Kushk; so that from July to December the country away

from these streams is almost destitude of running water.

Two important features of this part of Afghanistan are

the Murghab and Hari Rud. The former is, generally speak-

ing, deep and unfordable and, when in flood, is an impas-

sable obstacle. The latter, on the contrary, is general-

ly shallow and fordable, except of course in flood season,

when it becomes a rapid river about 200 yards wide. The

flood season is from the middle of March to the end of

July. East of the Herat valley and Badghis is a wild

mountainous country ..• This is a region of barren rugged

hills rising in places to 10,000 and even 12,000 feet.

The main axis of the mountain system comprised in this

area is the Koh-i-Baba, which is in itself a continuation

of the Hindu Kush. The Band-i-Baba, just before entering

the Herat province, breaks up into three main ranges: the

northern one is the Band-i-Turkistan, the central is gen-

erally spoken of as the Band-i-Baba, and the southern one

is called the Band-i-Baian or Safed Koh ...

"South of the Herat valley is the open country of

the Sabzawar district, drained by the Adraskand, which

lower down becomes known as the Harut Rud. One prominent

feature in the southern part of the [area] is the Do

Shakh range, which runs obliquely from the Hari Rud near
 170

Zindajan to the Persian frontier" (GA,Herat:3-4).

The following is the description provided for the

northern and Central regions: "Geographically and natur-

ally Turkistan embraces all the country draining to the

Oxus from the south, exclusive of the drainage of the

Kunduz and of Badakshan generally ...

"[Turkistan] consists of two distinct regions, viz.,

the hill country lying westward of the Koh-i-Chungur, and

the great plain stretching along the foot of the hills to

the Oxus.

"The principal features of the hill country are the

lofty Koh-i-Chungur; the great plateau extending between

this range and the Band-i-Amir river; the long straight

range at the foot of the plateau which dominates the

Hazhda-Nahr, and which is generally known as the Shadian

Koh; and in the west the great barrier of the Band-i-

Turkistan, with its long spurs and glens stretching down

to the plains.

"The above mentioned plateau extends north from the

Koh-i-Baba for 140 miles in the direction of the Oxus,

and its breadth is about 80 miles. It terminates in a

range (the Shadian Koh) whose peaks rise to 8,000 feet,

and which falls, almost perpendicularly, to the plain of

Turkistan, the elevation of the latter being little over

1,000 feet. The general elevation of the plateau is from

7,000 to 10,000 feet, and its surface is diversified by

 171

by hills, valleys, etc., but on the whole it may be char-

acterized as undulating. Its appearance, as overlooked

from a height, is barren: nevertheless there is good graz-

ing, and thanks to a large snow and rainfall, corn is

grown in some parts where valleys exist, and also on the

plateau itself. The valleys are for the most part re-

markably deep gashes rather than ordinary hollows or

depressions. There are three such valleys in the south-

east portion of the plateau, Bamian, Saighan and Kamard;

the former belongs to the Kabul province. They all run

from west to east, and their considerable streams unite

to form the Surkhab or Kunduz river. All the valleys,

particularly Kamard, are full of beautiful orchards while

the hills and plateaus are grassy in spring and early

summer.

"Descending north from the Kara. Kotal the drainage

of the Tashkurghan stream is entered. It runs northwards

through an extraordinary long succession of defiles,

walled by cliffs which rival those of Kamard. In a few

places the defile opens out to a certain extent and there

are villages with dense masses of walled orchards filling

the gorge from side to side. At a distance of about 70

miles from the Kara Kotal plateau, the defiles open into

the charming valley of Aibak (Samangan), beyond which is

the small plain of Ghaznigak. The exit from the latter

is by another defile, which terminates in a tremendous

 172

gorge, immediately at the mouth of which is the town of

Tushkurghan. The western half of the great plateau drains

to the Band-i-Amir stream. This river, whose source is

in the curious lakes of Band-i-Amir (also called Band-i-

...
Babar), is bounded in the upper part of its course by

enormous cliffs, while the river itself is so deep and

swift as to be generally quite unfordable. East of the

defiles of Band-i-Amir is the valley of Dara-i-Suf, less

deep than those hitherto mentioned and in some respects

more fertile.

"Dara-i-Suf drains to the Band-i-Amir through the

defiles of Kishindi. Not far below the junction of the

streams the valley of the river opens out, while the

stream becomes comparatively sluggish.

"Now having sketched the general features of the

great plateau, it will be desirable to turn back to the

Koh-i-Baba.

"From a point about south of Yak Walang (in the Kabul

province) this hitherto well-defined range breaks up into

three. The southernmost is called the Band-i-Baian and

continues along the south side of the Hari-Rud valley to

the intermediate neighborhood of Herat, where it is called

the Safed Koh. The centre branch (This, the Band-i-Baba,

is the main range and continues westward, though pierced

by the Hari Rud, to the Caspian.) runs along the north

side of the Hari Rud, parallel to the first and is the

 173

watershed between the Hari Rud and the Murghab. It is

known by various names. The third or northern branch

strikes northwest, enclosing the basin of the Upper

Murghab, and dividing it from that of the Band-i-Arnir.

Branching right and left, it forms a mass of mountains

which are the natural boundary of this part of Turkistan.

The western half has no one name, parts of it are known

as the Band-i-Alakah, Band-i-Badak, etc., and on the nor-

thern branch, nearly south of Sar-i-Pul, are two very

fine peaks called Khwaja Saf and Khwaja Kalsaf. Their

height is probably about 15,000 feet, while the general

altitude of the mountains is from 10,000 to 12,000 feet.

Numerous spurs run down northwards from the crest line,

enclosing among them a very large number of beautiful

glens. These spurs "soon sink into grassy down-like

ridges and undulations, the glens becoming fertile and

well populated valleys. This hill tract forms the dis-

tricts of Sangcharak, Sar-i-Pul, and Maimana. The former

is shut off from the plain country of Turkistan by a

ridge which is a continuation of that forming the face of

the great plateau, and is called the Elburz. Thus will

be seen that more than half the area of Turkistan is

mountainous or hilly, though anything but barren or un-

fertile. Indeed, much of this region is really charming.

Briefly its character changes from east to west. First,

there is the great plateau. Then the comparatively low

 174

hills and undulations of Sangcharak, backed by mountains

to the south. Next, the higher and bolder, but grassy and

downlike hills of Sar-i-Pul. Lastly, Maimana, with its

glens rapidly sinking into low downs. We now come to the

plain country of Turkistan, which is the most important

part of the province.

"There is a well-marked, and even for the most part

an abrupt, transition from the hill country to the plain.

The breadth of the latter is somewhat variable, owing to

the curves of the Oxus and its northward trend, but the

average is between 40 and 50 miles. All along the river

is a narrow arable strip. In 1886 it was by no means all

cultivated though it was so in ancient times. South of

this strip is a band of sandy desert. Its breadth varies

from 10 to 20 miles -- 15 is fair average -- though "as

many roads cross it diagonally, it may seem more to the

ordinary traveller. Also, in several places this desert

comes right up to the hills, dividing the cultivated and

populated plain into distinct portions. Thus, there is

the barren tract east of Tashkurghan: again there is a

division between the latter place and Mazar-i-Sharif and

between Shibirghan and the valley of the Maimana or

Andkhui river is a wide piece of sandy desert, rather

difficult to cross. West of the Maimana stream begins

the great chol or Turkoman desert, which now belongs to

Russia.
 175

"Four streams, descending from the hills, water the

fertile portions of the plain. These are the stream of

Tashkurghan, the Band-i-Arnir, the river of Sar-i-Pul, and

the Ab-i-Maimana. None of these rivers reach the Oxus.

At most times of the year their waters are entirely used

up for irrigation, and even during the spring floods the

surplus is checked, and eventually absorbed, by the sandy

undulations of the desert" (GA,Mazar-i-Sharif:4-6).

Turning to the northeast of Afghanistan, the compiler

of the Gazetteer writes: "the country is for the most

part a waste of sterile, rocky, snow-capped mountains, di-

vided in the east by the shallow, flat, alluvial depres-

sions known as Pamirs. The main feature in this moun-

tainous land is the mighty Oxus with its numerous afflu-

ents. Rising high up in the Pamirs and draining one of

the largest and loftiest snowfields in the world, this

river is remarkable for the force with which it has cut

its way through the deepest gorges, and for the vast mas-

ses of alluvial soil which it has carried to the plains

of Central Asia. The mountain ranges for the most part

vary from 10,000 to 20,000 feet and their general direc-

tion is from east to west, though there is one great spur

which, springing from the Tiraj Mir (25,426 feet) in the

Hindu Kush, runs north, forcing the Oxus to makes its

great northward bend, while at the same time it forms the

natural eastern boundary of Badakhshan proper. Another

 176

point to be noticed in this region is that all the drain-

age of Wakhan, Shighan, Roshan, and the Pamirs collects

in the Oxus just below Kala Wamar, while that of Badakhshan

proper is carried to the Oxus by the Kokcha. Finally, we

have that stupendous mountain range, known to us as the

Hindu Kush, which with the Himalayas, is the backbone of

Asia.

"with the regard to the country north of the Hindu

Kuch, ... , although the Hindu Kush is a single range, dis-

tinctly limited on its northern side by the Andarab valley

and the deeply sunk course of the Surkhab stream, hills

actually extend for a considerable distance northward, in

fact to within a few marches of the Oxus. These spurs,

which are rather vaguely represented on some maps as run-

ning out from the Nuristan section of the Hindu Kush and

from the hills immediately north of the Khawak pass, are

at Khost, Anjuman, Farkhar, Warsuj, etc.

"The ranges, or spurs, lose their mountainous char-

acter much sooner than was represented on the old maps.

Before reaching the Faizbad-Khanabad road they have be-

come grassy downs, high in some places, but in no case

mountains. The only mountain thereabouts is the isolated

mass of the Koh-i-Ambar. The most westerly of the high

spurs is that which separates Khost from Nahrin. West

of Nahrin the hills dividing it from Ghori are much lo-

wer and more broken, while Ghori itself and Baghlan are
 177

open plains.

"It must be explained, however, that immediately

north of Andarab, and the Surkhab river, the mountains

are as high, rugged and inaccessible as any commonly met

with in Afghanistan. But they diminish in height and

rockiness very rapidly, and north of Ghori and Nahrin,

though some high hills continue, they appear to be rather

the scarps of plateaux than distinct ranges, while the

lower hills are of soft soil, bare, smooth, and grassy in

spring.

"To the west of Ghori is the lofty Koh-i-Chungur,

rising from 5,000 to 6,000 feet above the Ghori plain,

while other high and steep hills continue along the left

bank of the Surkhab to within a short distance of Kunduz.

All these hills are the eastern scarp of the great plateau

stretching north from the Band-i-Baba to the Turkistan

plain.

"Now, havl!:g described the country between the Hindu

Kush and the Faizabad-Khanabad road, it will be desirable

to say something relative to the country lying between

the latter road and the Oxus. The main thing to remember

is that all the hills crossed by, or near, the road are

of a down-like character. The only exception, as before

mentioned, is the Koh-i-Arnbar, immediately to the north-

east of Khanabad.

"East of the Koh-i-Arnbar the downs extend to the Oxus,

 178

but west of it is the large open tract of Hazrat Imam. A

considerable portion of this is chol and waterless --

though well grassed in spring. But Hazart Imar.l itself

is in the midst of a highly productive plain, and sur-

rounded by villages whose fertile fields are watered by

canals from the Oxus. Along the whole south of the Hazrat

Imam plain are the swamps and fens of the Bangi and Kunduz

rivers. For another point to be remembered is that the

tributaries of the Oxus and the Oxus itself, as soon as

they get out of the hills, run through great reed beds and

marsches abounding in game, from tiger to snipe, but ex-

ceedingly unhealthy.

"East of Khanabad is the fertile plain of Taloqan

probably the healthiest and best part of the low country

of Khataghan, while to the west are the Kunduz fens.

Having cleared the fens, one enters on open desert, with

high hills (scarp of the great plateau) on the left. The

Tashkurghan road crosses a spur of these by the Arganak

Kotal, and at 28 miles from the Kunduz river reaches

Kairabad. This place is the first of the Tashkurghan

villages, and is the only spot on this road where water

is certainly procurable. This fact is mentioned as partly

showing how Khanabad is cut off from Tashkurghan and the

plain of Afghan Turkistan. The only important geographi-

cal feature of this western portion which now remains to

be mentioned is the Kunduz river, called also the Surkhab

 179

and Ak Sarai.

"This river is formed near Doab-i-Mekhzarin by the

junction of the Bamian, Saighan, and Kahrnard streams.

At first it runs northeast through the Goab district in

Afghan Turkistan, and then it runs north to Kunduz and

so on to the Oxus, receiving the drainage of all the dis-

tricts lying west of the Khawak pass" (GA,Badakhshan:3-4).

From these descriptions, it can be seen that the

geographical and ecological features of the country did

not facilitate communication within any of the regions,

among regions, or with territories lying outside the

boundaries of the state. The difficulties of communica-

tion, however, should not be equated with absence on in-

teraction. Movements of people and commodities inside

as well as outside the country were a reality throughout

our period of long duration. At this point, I shall ex-

amine how constraints imposed by the prevailing mode of

transportation, that is the reliance on pack-animals, con-

ditioned the circulation of the various commodities,

thereby affecting the overall organization of production.

Agricultural production being the mainstay of the economy,

I shall first discuss the circulation of agricultural

commodities.

In order to assess the impact of transportation on

the circulation of agricultural commodities, the following

variables need to be examined: timing of movement; carry-

 180

ing capacity and relative speed of different animals; and

ratio between cost of transport and price of commodities

at the points of production.

Circulation of agricultural produce can only start

after the harvest has been gathered, which is in turn

inexorably bound to the rhythm of natural phenomena. Dur-

ing our period of long duration, circulation was also de-

pendent on the accessibility of routes linking regions;

as noted earlier, these routes were closed to traffic

from two to nine months a year due to adverse weather con-

ditions. Therefore, differences in the timing of harvests

-- determined by the length of the growing season and the

types of crops -- and variations in weather from year to

year accounted for crucial differences in the span of

time during which goods could be moved from one location

to another.

The distance separating centers of production and

consumption was also a decisive factor enhancing the im-

portance of speed, carrying capacity and number of trans-

port animals available. A review of the writings of Euro-

pean travellers (Forester,G.1798jBurnes,A.1834jVigne,G.T.

1840jMasson,C.1842jBellew,H.W.1862;Duke,J.,1883jMartin,F.

1907) reveals that the distance travelled per day varied

from 8 to 25 miles.

Frank Martin, who from 1895 to 1903 was the super-

intendent of government workshops in Kabul, stated that

 181

it took pack-animals a week "to travel a hundred miles"

(1907:242). This estimate was based on the assumption

that the animals were fully loaded and that the caravan

was on the move every day -- assumptions which are con-

venient at this phase of the analysis but that will have

to be modified subsequently.

As both volume and weight of goods were to be taken

account of in the way a pack-animal was loaded, the actual

weight differed from case to case. Regulations issued

by the Afghan government defining the duties of trade a-

gents set out the following two sets of weights as stan-

dards for the mentioned pack-animals (Nizamnama-i-Wakil-

ul-Tijarat,1306/l927:5-6) :

Usual Loads (cloth,sugar, Heavy goods of

etc.) in pounds Small Volume,
in pounds

Camel 717 587

Pony or Mule 391 326

Donkey 196 163

An official British report from Persia estimated

the ordinary load of a camel at 400 Ibs,reaching 480 lbs

under favorable conditions, 200-250 lbs was considered

fair for a pony or a mule, and a donkey-load was reckoned

at about 130 Ibs (Law,E.F.,1889:10i quoted in Issawi,C.

1971:195) .

A letter from an English factory in Persia in the

 182

seventeenth century recommending that goods sent to Persia

be packed in bales of 130 lbs considered two such bales

(260 lbs) constituting a donkey-load, three (390 lbs) a

mule-load, and four (520 lbs) a camel-load (Barker to

Roe, April 1618; quoted in Steensgaard,N.,1973:32).

Steensgaard qualifies this estimate by noting that "the

size of a camel-load varied with the terrain and the size

of the animal" and considers 200 kgs (441 lbs) to have

been the norm for a load of Persian silk (1973:31).

Alexander Burnes reported that a two-humped or Bac-

tarian camel "will carry 640 lbs. English, which surpasses

by 150 lbs. the burdens of those of India and Cabool"

(vol. II: 427) .

Assuming the official Afghan estimate of 587 lbs as

the norm for a camel-load of grains and Martin's estimate

of one week per hundred miles as the speed of movement,

I shall have recourse to a hypothetical example .to illu-

strate what these numbers suggest.

In the l880s, the produce of the Hazhdad-Nahr dis-

trict in the northern region was estimated, by an Indian

employee of the British government, at 1,068,981,800 lbs

of grain (GA,Mazar-i-Sharif:264). Although the figure

seems improbably large, British officers who visited the

district during the same period claimed that the area

could easily support a population of half a million --

seven times its existing number of population (Ibid:250).

 183

If we assume that half of the estimated amount was avail-

able for sale and that there was demand for it in the city

of Herat, which is located 270 miles away from the center

of the district, it would have required 910,546 camels

close to four weeks to transport the grains from Hazhda-

Nahr to Herat. The total number of camels in all of the

northern region was, at that time, estimated at only ten

thousands (Ibid:14). Had all these camels been hired for

the transport of the grain and were travelling at the

maximum speed of 25 miles a day, back and forth between

the two centers, it would have required them almost seven

years and a half to transport the available amount of

grains! Clearly, as far as circulation of grains was con-

cerned, the dominant mode of transportation worked against

the development of a fully integrated system of agricul-

tural production.

Unfortunately, there is no systematic data on the

regional distribution of transport animals over time. The

only available figures are the approximations offered by

British officers reporting on various provinces of the

country. As already mentioned, the number of camels

in the northern region was estimated at ten thousands.

For Badakhshan, the compiler of the gazetteer wrote that

"considerable attention is paid to breeding horses and

ponies ... Mules and donkeys are also fairly numerous"

(GA,Badakhshan:lO). No numerical estimate for the whole

 184

region was given but the number of pack-animals passing

yearly through the main toll station was stated as 6000

ponies and 800 camels (Ibid:67). Again, no general esti-

mates for the eastern and central regions are available.

However, 1894, British officials undertook a census of

animals belonging to the nomads who were the main carriers

between Afghanistan and India. According to this estima-

tion, the nomads possessed 32,959 camels, 1140 horses and

1241 donkeys (GA, Kabul:426ff.). There are no figures

for the southern region as such but the estimates for the

district of Kandahar were 2,000 camels 2,000 donkeys and

200 bullocks (GA,Kandahar:228). The estimates for the

western region were 50,000 camels and "a large number"

of donkeys and horses (GA,Herat:6). The overall number

of camels, therefore, may have been at the lowest be-

tween 100,000 and 200,000 in the l880s and l890s, but

guessing the number of other animals is not possible.

Terrain, distance, quantity of goods to be carried,

and availability of animals were conditioning factors

in the choice of pack-animals to be used. In the 1830s,

part of the trade route between Bokhara and China passed

through Badakhshan and was not negotiable by camels. In-

stead, ponies had to be relied on. Only after the goods

had reached the town of Khulm in the northern region were

they transferred to camels (Burnes,A.,vol.II:438). Dif-

ficulties of communication in some areas of Badakhshan,

 185

Bajour, Panjkora, and Kashmir, were such that goods could

only be carried by men (Lt. Irwin,1809/1839:1014).

When a route was equally negotiable by various types

of pack-animals, preference for a certain animal was de-

termined by the distance between centers of production

and consumption as well as the economic position of the

people who had to undertake the transport of goods. Lt.

Irwin who, as a member of the Elphinstone's mission, vis-

ited the Afghan court in Peshawar, paid special attention

in his memoir on "Climate, soil, produce, and husbandry

of Afghanistan" to the conditions of transport in various

areas of the empire. He proposed to "distinguish carri-

age as it may be, 1st, that of armies; 2nd, that of cara-

vans or of persons making distant journeys; 3rd, that of

farmers on their own farms, or for the supply of provi-

sions to towns, or the distribution of town manufactures

in the neighbourhood, or the interchange of commodities,

within small or moderate distances" (Ibid:10I2). It is

his last two categories that are, here, of interest to

me.

He affirmed that for long journeys the camel was

the most economical and that, in caravans, camels outnum-

bered all other animals. But, besides noting the situa-

tions where the terrain did not permit their employment,

he pointed out that trade between various regions depended

on the availability of an animal or animals in that region.

 186

Thus, the exchange between the northern region and the

central place of Kabul was carried from the Kabul side by

camels but from the northern side mainly by horses. Be-

bveen Kabul and Peshawar, all kinds of animals were made

use of: camels, mules, and donkeys seem to have carried

an equal quantity of goods in the trade between these two

central areas. As far as his third category was con-

cerned, Irwin acknowledged that greatest variety prevailed

yet that it was not camels but bullocks, donkeys, mules

and asses that carried the burdens on the farms and be-

tween farms and market places (Ibid:1013ff.).

The main tendency, however, was that of harnessing

the energies of every available animal for the task of

transport. Writing of the nomads, who accounted for most

of the long-distance circulation, Martin observerd that

in their caravans even the sheep carried "small packs too"

(1907:242). The inclination for making economic use of

the available animals was a function of the high cost of

transport in comparison to the price of goods in the cen-

ters of production. Although there are no available data

for systematic comparison and periodization, information

from different junctures indicates that cost of transpor-

tation constituted a significant difference in the price

of commodities at centers of production and at those of

consumption.

As cost of transport was similar throughout the area

 187

where pack-animals represented the main means of trans-

portation, I have not confined myself to examples from

Afghanistan. The first illustration concerns a Dutch pur-

chase of wheat in Jahrum, between Lar and Shiraz, in 1636.

They bought "about 17,108 man-i-Tabriz [107,040 Ibs] of

wheat, the purchase price of which was 10,264 mahmudi. The

transport costs to Gomroon [Bunder Abbas], scarcely more

than 20-25 marching days, were 50 mahmudi per man-i-Tabriz,

in all 8,554 mahmudi, i.e. nearly as much as the purchase

price of the wheat" (Steensgaard,1973:40).

In the next two examples, the cost of transportation

and its increase over a relatively short period are illu-

strated. Both examples are from 1841 and concern long-

distance trade. In that year, the rate for hire of a ca-

mel between the central places of Kandahar and Shikarpur,

a distance of 369 miles, carrying from 494 to 576 Ibs was

20 rupees (Postan,1841:16). The cost of transporting

1070 Ibs of iron from Bajour to Kunar and then to Kabul,

about 200 miles, had changed from 25 rupees in the 1830s

to 35 rupees in 1841 (Drurnmond,1841:84).

An example from 1863, directly illustrates the rela-

tive difference between price of goods at their point of

production and cost of transport to a center of consump-

tion. In that year, Arminius Vambery, an Hungarian lin-

guist who travelled in the guise of a Muslim throughout

central Asia and Afghanistan, recorded that an Indian

 188

had "purchased some loads of aniseed in Maymene for thir-

ty tenghe [15 rupees]. The carriage to Herat cost him

twenty tenghes [10 rupees] per load" (1865:303-04). The

distance between Maimana in the northern region and Herat

in the western region is about 240 miles.

Finally, Martin offered a general estimate for the

l890s. He stated that the cost of carriage over hundred

miles worked "out at about eighteen shillings [approxi-

mately 27 rupees] a hundred-weight for that distance"

(1907:242). His immediate comment was "and this in a

country where all things are cheap" (Ibid).

To convey the degree to which the cost of transport

formed an obstacle to long-distance trade in grains as

well as other commodities, I shall turn again to my ear-

lier hypothetical example and calculate the cost of trans-

port of the grain from the district of Hazhda-Nahr to the

city of Herat. According to Martin's figures, the cost

of transport of 267,245 tons of grain over the distance

of 400 miles would have amounted t0Jt19,24l,672 -- or 29

times the Afghan state's yearly revenue which was then

estimated ati670,000 (GA,1908:40).

Thus, the technical organization of circulation

severely hampered the transport of bulk goods, such as

grain or wood, and stood as an obstacle to the develop-

ment of a coordinated "home market" that would have co-

incided with the boundaries of the state. Yet, absence

 189

of a "home market" did not translate into absence 0:: cir-

culation. It is time to take account of the actual dir-

ection of movements of goods. I shall differentiate be-

tween regional patterns of movement and inter-regional

and long-distance patterns.

How did those in possession of agricultural produce

dispose of it? As far as the sale of grain is concerned,

especially wheat which formed the main item of the popu-

lation's diet, four categories of buyers can be identi-

fied: 1-, individuals and families within the locality

where the grain was produced, who for a variety of rea-

sons -- absence of males engaged in outside employment,

uneven ratio of consumers to producers and/or to the

amount of land owned, women with minor children, etc.

had to buy part or all of their supply of grain; 2-, peo-

ple in neighbouring localities whose grain production was

insufficient, either as a result of weather conditions

or because they cultivated commercial crops instead of

wheat; 3-, inhabitants of central places who did not en-

gage in agricultural production; 4-, caravaneers and

others engaged in long-distance trade and travel.

The first category has already been illustrated in

reference to the organization of production in the Kunar

valley, but fluctuations in the demands of members of

other categories through time have to be addressed more

fully.
 190

Any decrease in the amount of precipitation, by de-

termining the amount of water available for irrigation,

strongly affected the production of crops in a locality.

Yet, localities in relatively close proximity were af-

fected differently by the fluctuation in weather condi-

tions due to significant variations in altitude as well

as differences in location with respect to rivers and

main canals and, depending on the year, the same locality

might import or export grain.

Weather conditions were not the only variable affect-

ing the amount of produce. Visits by swarms of locusts

also had to be reckoned with. Such visits were not re-

gular but when they did occur the damage was substantial.

During the years 1880-1886, locusts appeared regularly

in the locality of Haibak, in the northern region, and

"consequently but little wheat was grown in those years"

(GA,Mazar-i-Sharif:237). As a result of the shortage,

localities which had a surplus of grain in normal years

had to acquire their food from the market, as was the

case in the sub-district of Pir Nakchir where "in 1886

no supplies were procurable, and the people obtained their

flour from the Tashkurghan Bazar [name of the market town

in the Haibak district]" (Ibid:565).

On the other hand, exchange between localities in

relatively close proximity was not infrequent. Produc-

tion of grain in the valley of Sighan in the central re-

 191

gion was roughly 2/5ths "wheat, 2/5ths rice, and 1/5ths

barley", and the people of the valley exchanged "most of

their rice for wheat and barley from [the valleys of]

Qalishan and Dara Yusuf" (Ibid:480). People of the lat-

ter valleys exported wheat to localities in the central

and northern region and, in years of drought, even to the

eastern region (Ibid:191).

There were also localities where ecological condi-

tions did not allow for self-sufficiency in grain but

where other types of productive activities made the con-

gregation of a large population possible. In such cases,

grain had to be regularly imported from other areas. The

three districts of Khawaja Salar, Shor Tapa, and Kerkin

on the Afghan bank of the river Oxus, in the northern re-

gion, where sericulture and sheep-raising were very pro-

fitably undertaken, represented such a situation. Each

of the districts was approximately 70-90 miles away from

three main centers of wheat production and, despite the

cost entailed, Khawaja Salar got its flour and/or wheat

from Akcha, Shor Tapa from Mazar, and Kerkin from Tashkur-

ghan (Ibid:444,447 ,449).

Whereas people in the Oxus districts were attempting

to acquire self-sufficiency in grain and not succeeding,

there were other areas where grain could be easily pro-

duced but where people specialized in commercial crops

 192

for which there was a strong demand. Sugarcane and cot-

ton from the valley of Jalalabad in the eastern region,

tobacco from Kandahar in the western region (Dr. Griffith,

1841:984-85), and fruits from various regions of the coun-

try fall within this category. When market conditions

were favorable, quite a few commercial crops were grown

in ecologically suitable environments.

In the 1840's, for instance, the following crops

were grown in the valley of Logar in the eastern region:

bean, field barley, buckwheat, clover, cabbage, blue drum-

head, cauliflower, large coriander, cumin, cucumber, egg-

plant, flax, linseed, fennel, garlic, gram, large millet,

lettuce, lucerne, leeks, maize, madder, two varieties of

water melons, two varieties of pumpkins, three varieties

of mustard, white and blue onion, poppy, radish, rice,

spot-herbs, tares, broad leaf and narrow leaf tobacco,

turnip, vines, and wheat (GA,Kabul:348ff).

The commercial crops in the valley of Logar were

primarily produced for the central place of Kabul which

lay approximately 30-60 miles from the valley. The. de-

mand in Kabul was strong enough to sustain the production

throughout our unit of long duration. As no data on the

amount of acreage devoted to the various crops over time

is available, a fuller picture of patterns of interaction

cannot be drawn.

Comparative data from other regions of the country

 193

indicate, however, that ecological conditions alone did

not determine the choice of crops but that demand from

central places was an important factor. The situation

in the valley of Arghastan, in the southern region, in

the 1870s is a casein point. In the words of a con-

temporary observer, "the principal crops are wheat, bar-

ley, and Indian corn. Cotton is also grown but nowhere

in large quantities. Lucerne, carrots, etc., are only

grown in small quantities for the use of the cultivators.

The price obtainable for these latter articles at Kandahar

in ordinary times is not sufficiently high to repay the

cost of carriage thither" (GA,Kandahar:54). Nothing in

this observation would have surprised Thunen who, as al-

ready reported, had postulated that the cost of grain

in the Town must reflect the conditions of production

and transportation faced by the most distant producer

whose grain was needed by the Town.

As the focus of this study is the process of circula-

tion over time, ideally it would have been desirable to

compare fluctuations in the demand of the central places

for agricultural produce during the whole length of the

unit of long duration, as well as information on the popu-

lation of central places, shifts in their range of in-

come, and area covered by productive zones which supplied

central places during regular and extraordinary times.

Such data are unfortunately not available at present and

 194

I have had to rely on more fragmentary information.

In a conjectural unit of time, extraordinary year(s)

can be distinguished from regular years by the occurance

of marked shifts in natural conditions affecting the us-

ual sphere of supply of a central place and/or the sudden

influx into a region of a large number of people and ani-

mals who have to be provided with food for subsistence.

Changes in price of needed articles and in distance of

centers of production from which the articles had to be

brought in can serve as an indication of the extent to

which advantage was taken of economic opportunities when

available. But if this exercise is to be of any use in

revealing long-term tendencies, it has to be combined

with the determination of the sphere of production of

central regions during regular years, the ratio of re-

gular to extraordinary years in any conjunctural unit,

and secular trends in prices over the period of long dur-

ation. Within the limits of the information at my dis-

posal, I shall attempt to draw the relevant conclusions

from an analysis of the data on wheat, the staple item

of diet in the country.

As the location of a central place from its center

of supply differed from case to case, different distances

had to be travelled in order to realize the same amount

of grain. The amount of surplus produce procurable from

the nine buluks of the district of Herat alone -- an area

 195

of 120 miles by about 14 miles that could be readily

purchased in the central place of Herat in 1885 was esti-

mated by a British officer at 6,763,416 Ibs. [3067.83 me-

tric tons] of wheat, 3,645,000 Ibs. [1653.34 metric tons]

of husk rice, and 6,763,416 Ibs. [3067.83 metric tons] of

chaff for horse feed (Stewart,C.E.:42). In 1904, the fo1-

lowing estimates of the gross and surplus produce of the

district were given by another British officer (GA, Herat:

158) :

Name of crop Gross Product Surplus Product

metric tons metric tons

Wheat 22766.16 4553.23

Barley 15301. 85 3060.37

Rice 7464.32 1492.86

Dhal 7464.32 1492.86

Total 52996.65 10599.32

The surplus was calculated at 1/5th of the gross produce.

In 1910 Russian estimates, confirmed by British observers,

put the surplus wheat of the district at 11,495 metric

tons (Ibid).

There had been no change in the technical organiza-

tion of production in the district in the meantime but,

after a series of wars from the l830s to the 1880s, a

stable political situation had emerged and the population

 196

of both the central place and the district had begun to

increase again. As such, the figures on the district's

gross and surplus produce should in no way be interpre-

ted as the limits of the district's carrying capacity in

a static sense, rather they should be seen in dynamic

interrelation with other factors.

The district of Herat was not the only source the

city of Herat could rely on for its supply of grain.

Supplies could also be procured from the districts of

Karokh, Zindajan, Ghorian, Obeh, Gu1ran, Sabzwar, and

Farah, lying respectively at a distance of 27,27 1/4,

41.5, 60, 68.5, 74, and 162 mile~ from the central place

of Herat (Ibid:each entry). Furthermore, there were

other districts located farther off. Depending on the

needs of the central place and the conditions of harvest

in the district of Herat, supplies from a number or all

of these districts could be called upon. The degree to

which supplies from these and more outlying districts

were drawn on a regular basis could be related to the

viability of the economy of the central place and the

ability of its inhabitants to underwrite the cost of

production and transportation incurred by the producers

on the outlying district(s) whose grain was needed by

the central place. Strong fluctuations in prices over

any conjectural unit and import of grains from other re-

gions on an irregular basis can be related to disruptions

 197

of the grain supply from regular centers of regional pro-

duction. Historical evidence for the region of Herat

bears out the correctness of these propositions.

At the end of the reign of Husain Baiqara -- 1470-

1506 --, when Herat had served as the capital of the

Timurid state for nearly a hundred years, a contemporary

historian claimed that the gardens, palaces, mansions,

and field of the city were connected to one another over

a distance of 60 miles, from the village of Kosoya to

that of Obeh (Khondmir,vol.IV:650). Conversion of agri-

cultural fields in the vicinity of a central place during

its periods of prosperity can be documented for other ci-

ties in Afghanistan, Therefore, the above information

on the extent of gardens and residential area is probably

correct, and it can be legitimately inferred that the

grain of the central place of Herat must have come from

its more outlying centers of supply.

While at the beginning of the sixteenth ceritury

Herat's reliance on its outlying districts was a sign of

the affluence of its residents, in 1838, the length of

the distance from which supplies had to be procured was

an indication of the total disrupture of the agricultural

production around the central place. From 23rd of Novem-

ber 1837 until 9th of September 1838 the city was under

seige by a Persian army and no agricultural production

had taken place during that year. In May of 1838, the

 198

authorities in Herat were requesting from the Persian

monarch 1866-2239 metric tons of grain as a condition for

peace (Kaye,J.W.,vol.I,1857:263). As the negotiations

broke down and the Persian army evenutally withdrew be-

cause of British pressure, the residents of Herat were

forced to buy their grain. Major James Abbot, who was

travelling from Herat to Khiva at the time, reported the

movement of a large number of caravans from places as far

as Merv, over 400 miles distant from Herat, "laden with

wheat and barley" destined for the city and its surround-

ing districts (1856,vol.I:16,28,49). Abbot, unfortunate-

ly, does not provide any data on the price of grain be-

fore and after the siege but, as he paid fourteen shill-

ings per camel for carrying his goods from Bokhara to

Khiva (Ibid:38-39), a distance of 360 miles, the price

paid for the grain in Herat must have been very high.

Even under relatively stable political conditions,

the price of grain fluctuated. The following table, based

on reports of British news-writers in Herat and repro-

duced by Hasan Kakar in an appendix (1979:237-41), show

the various shifts in the price of wheat over a short

conjectural period. In order to render the data compar-

able to information from other regions, I have converted

the local units of weight and currency to pounds and

Kabuli rupees.
 199

Table I. Wheat Prices in the Central Place of

Herat

Date Feb. July Dec.l3 Dec.29 March 8 June 6 Feb.19 May 11

1882 1882 1886 1887 1888 1889 1893 1893

Number of 14
pounds 24.5 28 28 21.21 30.8 33.2 42 1bs.of
per rupee bread

Until May 1893, grain had been very scarce in Herat and

the government had had to offer wheat for sale in the mar-

keto "After that date bread could not be found cheaper

in the bazars ll (Ibid:295). On March 10, 1893, grain had

disappeared from the markets of the city of Kandahar. By

March 17, scarcity had reached such a point in the sur-

rounding villages that IImost people subsisted on green

fodder" but, by March 31, grain had become abundant (Ibid).

Kakar does not give any information as to where the needed

grain carne from. As wheat was harvested in June in

Kandahar and July in Herat, the supply could not have

corne from areas that were close by. A remark by a British

officer who was, in May of1893, in the valley of Farah, 162

miles from Herat and 245 miles from Kandahar, may provide

the clue. He noted that during 1893, IIconsidered to be

a famine year, IIwheat was at 30 lbs. a rupee while in or-

dinary years 82 lbs.sold for a rupee (GA, Farah:74). Dur-

 200

ing ordinary years, neither Herat nor Kandahar would have

been in need of grain from Farah but, in a situation of

scarcity, one or both could have turned to the resources

of the valley. Sudden demand for the produce of the val-

ley would have resulted in the steep rise of the price.

Thus, centers of production having an unmarketable sur-

plus or being part of the sphere of one central place

would only in extraordinary years become objects of com-

petition from several centers of consumption. The situ-

ation in 1857 in Kandahar can serve as an illustration.

From February 1857 until the gathering of the grain

harvest in June, Kandahar suffered from a severe famine.

During this period, grain was imported from the districts

of Ghazni -- 87 miles from Kabul and 231 miles from

Kandahar and normally a center of supply for Kabul -- and

Sabzwar -- 80 miles from Herat 2.nd 300 miles from Kandahar

(Bellew,H.W.,1862:228). The price of 4 Ibs. of wheat

flour reached one Indian rupee or two shillings and that

of barley 8 Ibs. per two shillings (Ibid:229). Dr. Bellew,

who was a member of a mission to Afghanistan that had to

stay in Kandahar for twelve months because of the Indian

mutiny, observed that "at such prices, the poor could

get no flour at all, and for several months subsisted on

clover and lucerne, wild herbs and mulberry-leaves, which

they as often ate uncooked as cooked" (Ibid:230). It can

therefore be inferred that the quantity of grain that was

 201

imported could not have been very large. By June 9th,

the price had dropped to 24 Ibs, per rupee and it was

still going dmm (Ibid:257).

Strong fluctuation in the price of grain could also

result from an increased demand for food and supplies

generated by the sudden arrival of a large number of peo-

ple. Such was the situation when the British army ar-

rived in Kandahar on the 25th of April 1839. On the 8th

of May, a pound of flour sold for one rupee in the city

of Kandahar (Havelock,H.,vol.II,1840:21). The scarcity of

grain was so acute that the British army imported "a

large supply of grain" from Mooltan (Ibid:34) -- a dis-

tance of over 400 miles. By the 28th of June, in the

district of Kelat, 87 miles from Kandahar and one of its

supply centers, the price of wheat had fallen to 20-24

Ibs. per rupee (Ibid:49). Tendencies in a conjunctural

unit can be gathered from the data on the price of wheat

or wheat-flour in the following table:

Table II. Price of Wheat in the Central Place of Kandahar

Date Number of pounds Date Number of pounds

per·rupee per rupee
Dec.23,1881 17 Sept.23,1892 13
February 1882 17 Feb.17,1893 (flour) 11
(flour) 13
June 3,1893 17
Aug.28,1882 28
Oct.14,1893 17
January 1884 27
Dec.26,1893 17
May 6,1886 12
Feb.9,1894 26
Oct.4,1886 17
Jan.4,1895 30
Sept.14,198'J 28
March 22,1895 30
 202

Comparison of the data in tables I and II reveals

that notwithstanding yearly fluctuations, patterns of the

price of wheat in the two central places displayed dif-

ferent tendencies. Data from the central place of Kabul

for the S2.me conjunctural period reinforces the interpre-

tation that in each region the price of grain followed a

different trajectory. The information in table III, as

in tables I and II, is derived from data culled from re-

ports of British news-writers quoted in Kakar (1979:237-

41) •

Table III. Price of Wheat-Flour in the Central Place of Kabul

Date Number of pounds per rupee

July 17, 1880 114

Dec. 14, 1880 73
July 20, 1885 12
Nov. la, 1891 15
June 7, 1893 13
Oct. 11, 1893 20
Nov. 5, 1898 111
Jan. 31, 1900 147

The difference in patterns of price reflects in part

the difference in the distance between each central place

and its centers of supply and the fertility of these

centers; but the patterns also reflect differences in the

demographic histories of regions in general and central

 203

places in particular as well as vicissitudes of commerce

and politics in central places. I have already summarized

the available data on demographic changes in the regions,

and commercial and political changes in the importance of

the various regions will be taken up later. The variation

in the distance between a central place and its centers

of supply, however, has to be discussed here.

The distances of some of the districts listed as

areas exporting grain to the central place of Kandahar

were as follows: Dahla, 14 miles; Dehrawat, 60-70 miles;

Tirin, 80 miles; (all north of the city), Kushk-i-Nakhud,

40-50 miles; Maiwand, 38 miles, both west of the city

(GA, Kandahar:120,135,302,319,479). However, the dis-

tricts of Nish and Zamindawar, which were considered the

chief sources of supply of grain to the central place c£

Kandahar, were respectively 50-60 miles north and 75-140

miles -- being very extensive and divided into four sub-

districts -- south of the central place (Ibid:365j GA,

Farah:296ff). Kabul, as can be gathered from the des-

cription at the beginning of this chapter, was located

in the midst of a fairly fertile valley. Its main cen-

ters of grain supply, however, were the districts of

Ghazni, 86 miles west, and Logar and Gardez, 30-60 miles

south (GA,Kabul:118,122ff,348ff).

The length of the distance between a central place

and its center of supply, as I earlier argued, was pre-

 204

dominantly an outcome of the availability of irrigation

networks. But the need of central places for fruits and

vegetables was also a factor in the reliance of central

places on more distant centers of production for their

grain. Vegetables and fruits were cash crops which had

a ready market in the central places; the fruits found

an outlet in exports to other regions and countries as

well. The political and economic importance of a central

place was therefore reflected in an increase in the area

of the land in the vicinity of the central place that was

converted from grain to fruits and vegetables.

Throughout our unit of long duration, fruits figured

prominently in the exports of Kabul and Kandahar. Elphin-

stone observed that "there are many orchards and gardens

round Kandahar" (Vol.II:133), and of Mazar it was observed

that the town, like other towns in the northern region,

was "more of a dense mass of inhabited orchards and gar-

dens than a regular town" (GA, Mazar-i-Sharif:413).

Variations in the price of grain in the northern re-

gion seems to have been in marked contrast to that of

other regions of the country. Some time in the 1880s

probably 1885-86 a British officer visiting the region

stated that a rupee could buy 1480 lbs. or more of grain

(Ibid:115). For May 23, 1895, Kakar, without giving the

name of the locality, quotes a price of 1083 lbs. of

wheat per rupee but, for November 8, 1883 he qU0tes a

 205

price of 29 lbs of wheat per rupee (1979:237-41). With-

out having the name of the specific locality, the dis-

crepancy is too great to be credible. All British offi-

cials commented on the cheapness of provisions in the

northern region. After the famine of 1885 in Kabul, great

quantities of grain were said to have been sent there

from Khanabad -- 380-390 miles from Kabul through the

mountainous central region (GA,Badakhshan:99). This may

indicate that the price was indeed so low that even after

bearing the high cost of transportation, grain could still

be sold at a price affordable by the people of Kabul.

Cheapness of grain in the last two decades of the

nineteenth century was probably an outcome of the de-popu-

lation of the northern region after the wars and famines

of the previous fifty years. Information from the six-

teenth to the beginning of the nineteenth centuries, while

lacking in quantifiable data, does not give the impres-

sion that the situation of the 1880s and 1890s was at all

typical of the earlier period (McChesney,R.,1973:107ff.).

But whether there was a unified regional tendency in the

movement of prices even in the period of prosperity of

the region remains to be investigated. Information on

the pattern of flow of grain from the 1880s and 1890s in-

dicates at least three SUb-regions that may each have had

its own distinctive price history for grain. These sub-

regions seem to have been formed around the districts of

 206

Maimana and Shibarghan in the west, those of Akcha,

Andkovy, Mazar, and Tashkurghan, relying on the fertile

locality of Hazhda Nahr [18 canals] in the center, and

Khanabad, Imam Sahib, and Taluqan in the east (Ibid; and

GA,Mazar-i-Sharif). Since communication between sub-re-

gions was to some extent impeded by the waterless tracts

separating them from each other, it is likely that, in

ordinary years, centers of production in one sUb-region

did not supply other regions with grain.

As far as pattern of movement of grain in Badakhshan

and the central region are concerned, the available in-

formation does not allow the drawing of any firm conclu-

sions. The same difficulty faces us in assessing the

character of the secular trends for the country or any

of its regions to any degree of certainty. The dominant

obstacle in the way of generalizations here is posed by

the yearly fluctuation of prices over any conjectural

unit. Martin, for instance, asserted that the price of

bread had risen four-fold in the eighteen years between

1885 and 1903 (1907:238). Examination of the data in

table III, however, reveals that while prices between 1886

and 1893 are indeed very high in comparison to 1880, the

price of wheat in 1900 was cheaper than that of 1880. In

the first decade of the nineteenth century, when Elphin-

stone collected his information, a rupee bought 60 lbs of

wheat flour in the central place of Kabul and "in the

 207

country perhaps half as much again" (vol.I:391). Prices

in the valley of Logar, a regular center of supply of

Kabul, during the first invasion of the country by the

British army, 1839-42, had risen to 40-48 lbs per rupee

and the same price had again prevailed during the civil

war of 1867-68 (GA,Kabul,348ff) • Charles Masson, who be-

tween 1832 and 1838 lived openly as a European in Kabul,

stated that "while the quality of provisions brought into

the Kabal markets is excellent, prices are liable to much

fluctuation, especially in the various kinds of grain;

and the reason is, obviously, that the country at large

scarcely yields a sufficient quantity for the supply of

its inhabitants, and wheat becomes an article of import.

It follows hence, that not only are prices subject to

variation from extraordinary accidents, as partial or

general failure of the crops, the ravages of locusts, &c.,

but that they are affected by the ordinary and constantly

occuring changes of the season. Winter in Kabal is al-

ways distinguished by high prices, and the advance immed-

iately follows the stoppage of its communication by snow.

In the famines which, from time to time, have afflicted

Kabal, the misery has naturally been most intense within

the city during the winter; •.. The last serious famine

occured in the reign of Shah Mahmud [1809-1818]; and since

that time so great an evil has been happily averted, not-

withstanding occasional years of scarcity have, in the or-

 208

der of things, presented themselves" (vol. II, 1842/1974:

270) .

Political and ecological conditions were not alone

in affecting the fluctuations in the price of grain in a

region. Inter-regional and long distance trade have to

be taken account of as well, for the volume and direction

of trade had a direct bearing on the marketability of

agricultural surplus in areas of production as well as

on the demand for food in central places.

Whereas I have argued that the prevailing mode of

transportation formed an obstacle in the development of

long-distance trade in grain, it did not hinder long-dis-

tance trade in more expensive commodities. The cost of

carraige and the slow tempo of movement imposed by the

pace of camels and other animals forced the travellers

to interact with the people of regions through which

they travelled. Expensive articles forming the cargo of

caravans, travellers prefered to acquire their food sup-

ply from villages on or near the roads instead of carrying

it over long distances, which would have meant reducing

the volume of their precious merchandise. Consquently,

when the flow of traffic was regular, villages along the

route could count on the disposal of a portion or all of

their surplus without having to take the goods to a cen-

tral place. Furthermore, since central places were the

nodal points where those engaged in long-distance trade

 209

exchanged their wares, bought local products, rested from

their journeys, and gathered information on conditions of

trade, these centers usually had a floating sector of pop-

ulation. The actual numbers of this sector followed the

ebbs and flows of the trade but when their numbers were

regularly large, they undoubtedly exerted a strong impact

on the demand of a central place for gain and, thereby,

on the nexus between centers of production and central

place(s). I shall illustrate this assertion with some

empirical examples.

In 1884, Ashkinak, about 200 miles from Kandahar on

the Kandahar-Herat road, was a ruined fort where about

one hundred families lived. Yet, there was "considerable

cultivation, mostly wheat, [t]here and along the entire

road" (GA,Farah:23). The inhabitants disposed of their

surplus production by selling it to caravans (Ibid). All

accounts of Europeans who travelled with caravans before

and during our period of long duration confirmed this

practice. Indeed, when a caravan passes uninhabited ter-

ritory, the traveller explicitly noted the fact that pro-

visions had to be carried with them (Honingberger,M.,1834:

176) .

Closure of the port of Hormuz, on the Persian gulf,

as a result of the conflict between Portuguese and Per-

sian states from 1613 onwards and the consequent reorien-

tation of trade through Kandahar, provides a dramatic ex-

 210

ample of the impact that infusion of large-scale long-dis-

tance trade could have on a region. Richard Steel and

John Crowther, who passed through Kandahar on their way

from India to Persia, stated that, whereas before 1613

scarcely 3000 camels took this route, the number of camels

passing through after that year was between 12,000 and

14,000 (1965:269). They bought provisions from Afghans

along the route and, among changes in the city of Kandahar,

they observed that "by reason of frequent passage of Cara-

vans it is much enlarged lately, that the Suburbs are

bigger then the Citie. For within this two yeare, that

the way of Ormus is stopped up by the wars betwixt the

Persians and PortugaIs, all Caravans which passe betwixt

India and Persia, must of necessitie goe by this place.

And here they doe hire Camels to go into India, and at

their returne for Persia. They cannot returne also with-

out the Governours leave, which causeth them to stay a

moneth, and when least, sixteene or twentie days ... Pro-

visions of victuals is there in great abundance for man

and beast, yet deare by so great concourse II (Ibid:272).

But all those travellers who stayed in central places

were not forced to do so. George Forester, who, in 1783

and 1784, travelled from Kashmir and Peshawar to Herat,

-- all within the domain of the Afghan state at the time

--, wrote of Peshawar that "from its well chosen position,

which unites by a commercial chain, Persia and Afghanistan

 211

with India, Peshawar has become an important mart, and

the residence of wealthy merchants; especially of the

shaul dealers, many of whom disliking the dangerous and

tedious route of Kashmire, are here enabled to make their

purchases at a moderate advance on the first cost" (1798:

vol.II:SO).

An idea of the time this floating population actually

spent at various central places may be gathered from the

travel accounts of Burnes and Gerard in 1833. They

started from Peshawar with a caravan on March 15. After

reaching Kabul on May 1, they arrived at Balkh, in the

northern region, on June 10. Altogether they spent 61

days at three central places: Peshawar, 34 days; Kabul,

17; and Khulm in the northern region, 10 days (1833:1).

Trade, however, was in no way a constant element

throughout our period of long duration. But before as-

sessing its fluctuation, I must take account of the role

played by central places as market places and of the

structural relations of marketing.

Organizationally, the northern and southern part of

the country displayed different features. The difference

was not in the presence or absence of exchange relation-

ships but in the structuring of circulation. While all

markets in the north were periodic, those in central

places in the south were permanent.

Faizabad, the central place in the region of

 212

Badakhshan, had had two market-places -- one of 200 shops

and another of 85 shops -- which met twice a week (GA,

Badakhshan:66). In Badakhshan and the eastern part of

the northern region, which during the 1880s and 1890s

formed one administrative unit, the number of shops in

the following market-places is reported to be: Chahyab,

200; Rustak, 145; Talooqan, 250 (Ibid:52,245,277). Mar-

kets days, in each case, were confined to two a week. Ex-

cept for Talooqan, the days of the week for each market

are not given. There is also mention of bi-weekly markets

for almost every district but without any details on the

existence of a market-place or the coordination of schedul-

ing with other localities (Ibid:89,99,127,etc.)

Information from the central and western parts of

the northern region, although still fragmentary, allows

for a fuller grasp of the situation. Colonel Maitland

who, as a member of the boundary commission, travelled

throughout most of the region in 1885-86, wrote in 1889

that the people of the region "have a regular market-

place, with booths, in the center of each country district,

and here weekly or bi-weekly fairs are held, which are

largely resorted to for amusement, as well as business"

(GA,Mazar-i-Sharif:588). Nowhere did the number exceed

two days a week. The market-place of Tashkurghan, which

contained 450 to 500 shops, and was described as "the

largest and most flourishing in Afghan Turkistan" and "the

 213

emporium of the central Asian trade" (Ibid:567), was held

on Sundays and Wednesdays. At a distance of 160 miles

from Tashkurghan, the market-place of the town of Andkhui

comprised 83 shops (Ibid:65) and the market-days were

Sunday and Thursday (Yate,C.E.,1888:235). The distance

between the two centers was too great to allow for quick

flow of information on a regular basis and the fact that

Sundays were common market-days was probably a reflection

of the articulation of each center with its own sub-mar-

kets. That a regular relationship between a marketing

center and its sub-districts did in fact exist is demon-

strated by the case of Maimana -- 130 miles from Tash-

kurghan and about 30 miles from Andkui.

In 1886, Maimana's market-place contained 235 shops

and the town derived "much of its importance as being

the place of exchange for goods brought from Herat, Kanda-

har, and Mashad on one side, from Kabul and Ba1kh on the

other, and from Bokhara and Andkui on the third" (GA,

Mazar-i-Sharif:398). But Maimana also played an impor-

tant role with its own sub-districts. Although market-

days for Maimana and its sub-districts are not specified,

it is known that they were not held on the same days. Of

the sub-district of A1mar, a British officer wrote that

"in the center of the plain is a row of mud huts, con-

taining about 100 shops, which are occupied only on the

weekly ba~ar day by traders from Maimana, who sell cloth,

 214

cottons, tea, sugar, etc., and buy the produce of the

country" (Ibid:385). Weekly markets were held in at least

two other sub-districts of Maimana (Ibid:389) and if the

traders, as it was the case in Almar, actually came from

Maimana, it would seem to indicate that the periodic char-

acter of the markets allowed the traders to extend their

services to a group of population much larger than that

of the town in which they resided.

On the other hand, periodicity of markets could also

serve the interests of the peddler, part-time trader, or

producer. The town of Akcha, 90 miles from Maimana wester-

ly and 90 miles from Tashkurghan easterly, had 100 to 120

shops in its market-place. But on its bi-weekly market-

days "242 shops and stalls" were open (Ibid:47). As there

is no information on the days during which bi-weekly mar-

kets in towns such as Shiberghan, Mazar, Haibak, -- re-

spectively 20, 58, and 127 miles from Akcha -- met, it

cannot be concluded as to whether those who set up the

stalls came from these places or whether they belonged to

the categories I have mentioned. My impression from vis-

its to the area in the 1970s is that the people setting

up stalls were local peddlers and producers.

That the scheduling of market-days was not random

but involved coordination among a network of traders is

affirmed by the case of four settlements on the river

Oxus. In the sub-district of Karkin, part of Khwaja

 2 1_:J-

Salar district and 74 miles from Mazar-i-Sharif, markets

were held weekly; they were bi-weekly in the sub-district

of Dali and weekly in the Kham-i-Ab. "In Bosaga, just

across the Bokharan boundary, there [was] a fourth

bazar, said to be half-owned by the people of the Khwaja

Salar district. Fair days there once a week" (Ibid:444).

Although market-days are not mentioned, the example of

Maimana makes it highly likely that the merchants moved

from market-place to market-place.

The information at my disposal does not permit me to

draw any conclusions on the hierarchy of markets during

the regular periods of the year in the northern part of

the country. It seems likely, however, that periodicity

of the markets reflected the coordination of exchange

relationships between a market-place and its centers of

production rather than with other markets. But during

some months of the year, there seems to have been a distinc-

tive hierarchy. These revolved around a fair in Mazar-i-

Sharif and another in Hazrat-i-Iman.

Trade in the bi-weekly market-place of Mazar was

described as "chiefly local" during the regular months.

But from the 21st of March, which marked the beginning

of the Afghan year, and for two months, a fair called

Mela-i-Gul-Surkh ["fair of the red tulip", so called be-

cause of the profusion of that flower around the alleged

tomb of Caliph Ali in Mazar] was held there. The fair

 216

was attended by "buyers and sellers from the north of the

oxus, as well as from the whole of Afghan Turkistan, Kabul,

Herat, and" the central region (Ibid:251). In Hazart-i-

Iman, which derives its name from a shrine of great re-

pute there and is located just a few miles from the river

Oxus, the fair was held at the same time as that in Mazar.

However, buyers and sellers there must have corne from the

eastern part of the northern region, Badakshan, and lo-

calities north of the Oxus (GA,Badakhshan:80).

Whether market-days in towns and their surrounding

districts followed different schedules or whether centers,

like Balkh, ruined during the l880s but the dominant cen-

ter of the northern region from the sixteenth to the

eighteenth centuries, had markets that were open through-

out the week are questions that will have to await fur-

ther.research. That throughout our unit of long duration,

permanent market-places were a continuous feature of the

central places of Kabul, Kandahar and Herat, is attested

by all travellers and historians of the period.

The center of each city was marked by a bazar con-

sisting of covered arcades (which, at the time of con-

struction were planned to be of equal length and dimen-

sion but subsequently, because of additions or destruc-

tions, may have assumed irregular shape) as well as a num-

ber of market-places that were constructed in open air

(GA,Herat:161; Kandahar:239;Masson,C.,vol.II,1842/1974:
 217

276-68 for Kabul}. The great bazar in Kabul was destroyed

by the British army during its plunder of the city before

withdra"Ting from the country in 1842 (Kaye,J. ,1858:368-69),

yet the description by Masson of the activities of the

outdoors markets in the 1830s is closely echoed by Duke

for the 1880s (1883;192-98). Although all three central

places were in a weaker economic position at the end of

our period of long duration than at its beginning, a des-

cription of the situation in Kandahar in the l880s, the

only center aoout which statistical material of this nature

exists, will serve to underline the contrast in scale be-

tween the periodic markets of the north and the permanent

ones of the south.

At that time, there were about 1,600 shops in the

city and the number of shnps or families engaged in the

trades were as follows: armourers, 12; ata [flour], drug,

etc. sellers, 113; bakers, 30; barbers, 14; blacksmiths,

39~ .•. (GA,Kandahar:245~-461. It can be easily seen that

while most traders catered to th.e needs of residents of

the central place itself, the activities of a number of

them could have not been carried on without the patronage

of the inhaoitants of centers of agricultural production.

That people not living in the central place visited it re-

gularly on business is demonstrated by the existence of

a market-center in the open air outside the city walls,

called "the gang, where a large cattle, sheep, and grain

 218

market .IwasJ held daily, and brisk business IwasJ carried

on" (Ibid: 2 451. Those who sold their goods in the' gang

could easily buy the items needed by themselves and/or

their co-villagers from the markets of the central place.

At this juncture, it should be noted that the cost

of transport affected producers who brought their own pro-

duce to the markets of the central place quite differently

from the professional carriers or long-distance traders

who owned their own pack-animals. The investment o£ dir-

ect producers and landowners in animals was a necessity

for performing the work on the land, and the cost of their

maintenance was part of the expenses of a farm. When they

were harnessed to th.e task of carrying the agricultural

produce to the town, as far as the producer was concerned

it could be viewed in the same light as work on the farm.

It is only when the distance to the central place reached

such. an extent that the producer had to go to extra ex-

penses for his own upkeep and th.at of the animals, and

the time spent travelling prevented him from fulfilling

other tasks on the farm, that his expenses came to resem-

ble those of the professional carrier.

If inco~e of £amilies in other parts of the country

resembled that of the average family in the Kunar valley,

buying an animal was quite a significant investment. By

comparison to the average size of inheritance of 420 ru-'

pees - including land and chattels - at the time of death

 219

of the property-owner, the price in the l880s of animals

in the valley of Herat were as follows: camel, 200 rupees;

oxen, 150 rupees; cow, 100 rupees; donkey, 40 rupees. The

compiler of the Gazetteer also noted that bullocks were

"largely used for transport purposes" (GA,Herat:6). As

bullocks were the major animals used for plowing, it can

be inferred that producers and landlords, at least in lo-

calities close to the central place, took part in trans-

porting their produce to the markets. Yet, as a group

of people specialized in carrying agricultural products

from centers of production to the central place markets,

producers and landlords did not necessarily engage in that

task,

Masson, in his description of the cattle, grain,

fruits, wood, and ch.arcoal markets of Kabul, stated that

"as in other places, all traffic is transacted through the

medium cif the broker, or dalal" (l842,vol.II:269). He

also affirmed that "besides the shopkeepers, or fixed trades-

men, a vast number of itinerant traders parade the bazars"

(Ibid). While his claim that all transactions were en-

acted through brokers might be an exageration, brokers

and peddlers certainly played a very important role in the

exchange between a central place and its centers of pro-

duction, and the two functions were often carried by the

same person. Brokers visited the centers of production

regularly after each. harvest; since they were fully aware

 220

of both the demand of the central place for agricultural

produce and the conditions of production in a large area

of the country, they played an important role in the de-

termination of the price of agricultural produce. Fur~

thermore, as the data from the court archives of Kunar and

fieldwork in the 1970.s lead .me to believe, these brokers

often acted as creditors to rural producers and landlords,

a position which enahled th.em to acquire agricultural pro-

duce below the price prevailing in a locality, I will re-

turn to this group in connection with the financial struc-

ture of the state.

Besides the brokers and peddlers from ce~tral places,

the villagers had a regular exchange relationship with

the long-distance traders who regularly passed through

their areas. G.T, Vigne, who travelled with. the yearly

caravan of the Lohani clan, a group of Afghan long-dis-

tance traders who every year covered the road from Multan

in India to Bokahran and Russian fairs, and back again,

observed that, upon reaching the plain of Katawaz in the

vicinity of the town o£ Ghazni in the eastern region, on

June 19, 1836, "the caravan .might be seen dividing, each

party making for a separate village or fort, with the in-

tention of remaining there a £ew days, weeks, or even

months, according to the time required for the sale of

their .merchandis·e to the wild inhabitants of the numerous

forts on the plain, and the still wilder Hazarah tribes,

 221

who descend from the snowy range of the Narawah mountains,

for the purpose of traffic" Cl840!ll3). As these visits

were regular and social relations were cultivated over

the years, there was little need for the inhabitants of

these areas to make a journey to Kabul which was at a dis-

tance of about one hundred ~iles. Prevalence of exchange

relations, however, did not imply that the medium of ex-

change be the same in all transactions. As consideration

of this issue has become an oDject of controversy, it is

necessary to take account of the theoretical terms before

proceeding ,,,ith the empirical description,

Karl Polanyi, who called attention to the different

uses of money, offered a classification of money as equiva-

lence, means of payment, standard or measure of value,

store of wealth, and means of exchange, He argued that

"a comparative study of early money institutions must

start fram the fact, wh.ile ~odern money is ,. all-purpose'

money, i.e., the medium of exchange is also employed for

the other money uses, primitive and early moneys tend to

be 'limited purpose' moneys, i.e., different objects are

employed in different money uses" (1977:120; also 1968~

17S-203t.

Polanyi's proposal, altho~gh considered novel at the

time, was not new. Karl Marx had stated that "if we con-

sider ~oney, its existence implies a definite state in

the exchange of commodities. The particular functions of

 222

money which it performs, either as ~ere equivalence of

commodities, or as means of circulation, or means of pay-

ment, as hoard or as universal money, points, according

to the extent and relative preponderance of one function

or the other, to very di£fe.rent stages in the process of

social production {~apital,vol.I:170,emphasis in the ori-

ginal}.
Although Marx does not mention money as a measure of value

in the above quotation, earlier in the same volume CIbid~

84-142), he had devoted an entire section to the discus-

sion of the trajectory through which money embodied dif-

ferent relations, from measure of value to universal money.

The process of trans£ormation through which Marx traces

the development of universal money from barter is too ela-

borate to be summarized here. For the present purpose,

it is sufficient to note that he connected the process

to the degree to which communities possessing different

means of production are trans£ormed into one society

through reqular exchanqe (Ibid:351-52).

In my discussion in this and other chapters, I have

regularly expressed th.e costs of services or prices of

commodities in terms of money and this may give the im-

pression that, throughout the period of long duration,

money functioned as ·universal money in all the regions of

the country. In £act, the structural meaning of money

was more complex. While the notion of money as the under-

 223

lying measure of value, means of payment, and. general equiv-

alence of commodities was common, the actual means of pay-

ment mayor may not have been in monetary currency. When

payments were not in monetary terms, a series of equiva-

lences oetween different commodities was worked out. The

existence of such equivalences at times misled some vis-

itors into concluding that a certain group did not have

money. Alexander Burnes, who passed through the central

region in 1832, asserted that "these people have no money,

and are almost ignorant 'of its' value. ~-re got everything

from them by barter ••• A traveller among them can only

purchase the necesaries of life by giving a few yards of

coarse cloth, a little tobacco, pepper, or sugar, which

are here appreciated far aoove their value" <.1834/1973,

vol.I :177-78). In view of the fact that, in the same

period, taxes from the region were computed in monetary

terms (Masson, C. ,1842/1974, vol, II: 373) , it is hard to

accept Burnets assertion. Yet, as another example makes

clear, acceptance of money as means of payment did not

imply the acceptance of other equivalances of money for

cormnodities, Shah Shuja, who ruled over Afghanistan from

180.3 ..... 1809 and again 1839..... 42, relates a revealing incident

in his autobiography.

In the course of one of his failed attempts at

gaining the throne o.etween 180Q and 180.3, Sh.uj a was forced

to travel in same of the remote valleys of the eastern

 224

and southern regions. Of the Kakr clan, who inhabited the

southern region, he wrote "the simplicity of the Kakar

people showed itself to such an extent: when we needed

barley and fodder for the horses and grain for the attend-

ants and ourselves on a daily basis, they gave 3 sers

[approximately 49 lbs.] for one rupee. But at the time,

no silver currency was left with us and when for a con-

tainer of flour weighing 3 Indian sers (approximately 61

lbs.) \ve offered them a pearl valued at 100 rupees, from

ignorance and simplicity, they would not accept it and

c1enanded rupees" (1825/1954: 7) .

Preference for goods or money may have been related

to the frequency and regularity of regional and long-dis-

tance trade that passed through a locality.

As far as long-distance business transactions were

concerned, large amounts of money that would have been

needed were not actually carried around. Instead, mer-

chants and travellers had recourse to the well-developed

system of bills of exchange throughout India, Central Asia,

and Russian dominions in Asia. All European travellers

in the eighteenth and nineteenth centuries made use of it,

hut Burne's experience in 1832 may be taken as typical.

He wrote that "I had a letter of credit in my possession

for the sum of five thousand rupees, payable from the pub-

lic treasuries of Lodiana or Delhi; and the Cabool mer-

chants did not hestitate to accept it. They expressed

 225

their readiness either to discharge it on the spot with

gold, or give bills on Russia at St. MaCaire (Nijnei

Novogorodl, Astracan, or Bokhara, which I had no reason

to question: I took orders on the latter city lI834/1974,

vo 1. I ~ I 7 0) •

This banking network was intimately linked with ethnic

and political relationships in the area and as the activi-

ties of this group and that of long-distance trade in gen-

eral were strongly affected by the prevalent relations of

domination in the country I shall examine this issue in

connection with the financial structure of the state in the

following part of this study.

 ASHRAF GHANI

VOLUME II

COLUMBI..~ UNIVERSI T'l'

1982
PART I I
 226

CHAPTER FIVE

Social Relations -- An Historical Overview

Pierre Bourdieu argues that lIit is in the degree

of objectification of the accu'f.1.ulated social capital

that one find the basis of all the pertinent differ-

ences between on the one hand social universes in which

relations of domination are made, unmade, and remade in

and by the interactions between persons, and on the other

hand, social formations in which, mediated by objective,

institutionalized mechanisms ••. relations of domination

have the opaci ty and permanence of things and escape the

grasp of individual consciousness and power" (1972/1977:

184). In his view, exercise of power in situations where

a social machine for its perpetuation has not developed

constitutes "the elementary form of domination ll (Ibid:

190). He maintains that lithe reason for the pre-capital-

ist economy's great need for symbolic violence is that

the only way in which relations of domination can be set

up, maintained, or restored, is through strategies which,

being expressly oriented towards the establishment of re-

lations of personal dependence, must be disguised and

transfigured lest they destroy themselves by revealing

their true nature .•. violence is here both more present

and more hidden" (Ibid:191).

The nexus posited by Bourdieu between the development

 227

of institutions, as objectified social capital, and the

process of social reproduction of relations of domination

requires close scrutiny and I shall return to it later.

Bourdieu's conceptualization of reproduction in pre-capi-

talist societies is, however, questionable. While at

first his observations are directed towards explaining

the differences in modes of domination between societies

which "have no 'self-regulating market' (in Karl Polanyi's

sense), no educational system, no juridical apparatus,

and no State" and those where domination is "entailed in

possession of the means (economic or cultural capital) of

appropriating the mechanisms of the field of cultural

production" (Ibid:183-84), in the course of the essay the

emphasis subtly shifts to contrasting modes of domination

in pre-capitalist and capitalist societies. Bourdieu's

abrupt transition from societies characterized by the ab-

sence of institutionalized means of domination to pre-

capitalist societies as such is rather surprising for an

anthropologist. After all, the analysis of differences

in ways strategic resources have been appropriated in

such societies has constituted a main concern of the field

of political anthropology (Fried,1967).

To illustrate ethnographically his theoretical obser-

vations, Bourdieumainly refers to "ancient Kabylia." But,

as this "society" is abstracted out of historical time

and space, there is no way of knowing whether he implies

 /,28

that the end result of strategies is a simple reproduction

of the relation of domination or whether the system as

such is in flux. The former assumption would squarely

place him with the structural-functionalists who, in the

words of Meyer Fortes, declare that "tension is implicit

in the equilibirium" (1940:271). The latter assumption

would align him with Edmund Leach whose path-breaking an-

alysis of political systems of Highland Burma is the point

of departure for anthropological analysis attempting at

integrating history as a result of contradictory social

relations.

Writing of of the Kachin Hills Area, for which his-

torical records go as far back as the beginning of the

nineteenth century, Leach states that "these show clearly

that during the last 130 years the political organization

of the area has been very unstable. Small autonomous po-

litical units have often tended to aggregate into larger

systems; large-scale feudal hierarchies have fragmented

into smaller units. There have been violent and very

rapid shifts in the overall distribution of political

power. It is therefore methodologically unsound to

treat the different varieties of political system which

we now find in the area as independent types; they should

clearly be thought of as part of a larger total system in

flux" (1954/1964:6). He insists that the process of emer-

gence and breakdown of larger units be viewed not as part

 229

of the process of structural continuity but of structural

change, which he defines as "shifts in the focus of poli-

tical power within a given system" (Ibid:9).

Responding from an evolutionary perspective to Leach's

analysis of the interrelationship of different structual

forms in a total system in flux, Fried postulates a dif-

ferent trajectory for the evolution of complex society.

Starting from the assumption that stratified society must

be one "of the least stable models of organization" (1967:

225), he argues that it had either to dissolve into forms

of organization lacking differential access to strategic

resources or evolve more powerful institutions of politi-

cal control (Ibid:224-26). Thus, the stage is set for

the rise of the state, defined as "a collection of special-

ized institutions and agencies, some formal and others in-

~orrnal, that maintains an order of stratification" (Ibid:

235) .

Fried contends that the structure of the state is

essentially cellular, "that is, made up of a variety of

different kinds of components, with these components being

joined in'co subsystems that articulate wi th the whole, with

the larger structure of the government, only at higher le-

vels" (Ibid:237). Examples of such subsystems include

family and kin groups, communities and regions, offices

and bureaus, clubs and gangs, and even layers and levels

of the administrative apparatus itself. It is in the ar-

 230

ticulation of these subsystems, with each other and with

the institutions of the state, I will argue, that atten-

tion to strategies of individuals and groups will reveal

unsuspected nexuses in the mechanisms of reproduction of

a mode of domination. Fried, however, stresses another

aspect of the interelationsip. He maintains that in the

communication among these "sometimes disparate entities

there is always an understood priority of arrangement of

orders and coercive inducements to decisions, the level

of highest ultimate priority is equivalent to an internal

concept of sovereignty ••• There is also an external as-

pect of sovereignty as discrete political units recognize

or dispute each other's autonomy" (Ibid:237).

Fried's emphasis on the importance of coercion in

defining the sphere of power of a pol tical entity is an

indispensable criterion, but the notions of sovereignty

and discrete political units draw attention to the degree

of stability of these relations as well as the issue of

interpretation of ideological expressions. That political

relations in a large number of pre-capitalist systems have

been in the past characterized by flux is revealed in

Reinhard Bendix's comparative survey of kingship and

aristocracy in which he affirms that "the countries ruled

by kings may have been slow to change socially and econom-

ically, but politically they were scenes of turmoil and

bloodshed" (1978:218). He also raises the question of

 23l

the difference between formal expressions of sovereignty

and its actual content: "Vassals of the king could use

their status and resources to develop their own power,

sometime "to such an extent that kings and aristocrats

became bitter enemies despite all outward signs of majesty

and obedience" (Ibid:226). Clearly, we are in need of

criteria for determining the degree of autonomy and de-

pendence of individuals and groups within a state.

Stanley Tambiah's analysis of the Buddhist policy

in ancient India warns against the misreading of ideologi-

cal expressions of hierarchy, and attempts to explain the

survival of a highly complex pattern of stratification

over a wide area in the absence of a centralized state.

Refuting the notion that the Buddhist state under emperor

Ashoka was bureacratized centralized monarchy, he affirms

that it was "a kind of galaxy-type structure with lesser

political replicas revolving around the central entity

and in perpetual motion of fission or incorporation. In-

deed .•• a king as a wheel-rolling world ruler by defini-

tion required lesser kings under him who in turn encom-

passed still lesser rulers" (1976:70). As far as the

consequent fragmentation of the large empire is concerned,

he claims that the contours of the caste-based social or-

der had no compulsive need for an expression in kingship

and that the existing cultural and linguistic diversity

allowed it to resist political unification. "In the In-

 232

dian context the social order was prior to its political

expression ... fragmenting did not damage the economic

prosperity of the time, for the new commercial segments

functioned and flourished despite the political decline

of the empire" (Ibid:71-72). This observation is methodo-

logically very helpful and the presence or absence of cul-

tural mechanisms that define a social order transcending

the boundaries of a state or states has to be investiga-

ted from a comparative perspective.

Although at first Tambiah's orientation might seem

to contradict Fried's stipulation on the necessity of

coercion for the reproduction of stratification, atten-

tion to the unit(s) of analysis, in my opinion, makes the

two approaches complementary. Whereas the use or threat

of use of force may place a preponderant role in the re-

production of the mode of domination in each of the con-

stituent political units of a larger social order, cul-

tural mechanisms may play a dominant role in the inter-

connections within the larger system. Furthermore, Fried's

distinction between "pristine" and "secondary" states is

relevant here. Pristine states are defined as those that

"emerged from stratified societies and experienced the

slow, autochthonous growth of the specialized formal in-

struments of social control out of their own needs for

these institutions" (1967:231). But states which form

under the pressure of a pristine state "do not repeat the

 233

steps which the original state experiences. These are

secondary states" (Ibid:240). Although departing frcm

Fried's original categorical use in the context of this

work, I shall use the term secondary state analytically

whenever a new state emerges in the context of an already

existing social and political order. The essential theo-

retical point to be remembered is Fried's insistence that

"secondary states emerge through processes quite different

from those that give rise to the pristine states" (Ibid:

242). I will argue that it is at such junctures that

Bourdien's conceptualization of institutions as objecti-

fied cultural capital can be most judiciously applied.

Once in existence, not only can these institutions be put

to use by different masters but, despite the flux of the

larger system, they might significantly condition the

reproduction, restructuring or de structuring of modes

of domination at the level of subsystems.

As the trajectories of fragmentation and centraliza-

tion differ significantly in pre-capitalist formations,

focus on the articulation of different institutions both

in pre-capitalist and capitalist societies over relevant

periods of long duration may allow a better grasp of re-

production as an historical process. This procedure,

however, requires that reductionist tendencies be resisted.

"For there is," warns Braudel, "no single conjuncture: we

must visualize a series of overlapping histories, develop-

 234

ing simultaneously. It would be too simple, too perfect,

if this complex truth could be reduced to the rhythms of

one dominant pattern ... It is impossible to define even

the economic conjuncture as a single movement once and

for all, complete with laws and consequences" (1966/1973,

vol.II:892). Tensions and contradictions in the articula-

tion of relations of production and domination in histor-

ically specific circumstances will provide the locus of

analysis.

As far as the economic structure of society over any

unit of time is concerned, Marx's definition provides a

useful departure. Society is viewed as equivalent to

the "social process of production in general. The latter

is as much a production process of material conditions of

human life as a process taking place under specific his-

torical and economic relations, producing and reproducing

these production relations themselves, and thereby also

the bearers of this process, their material conditions of

existence and their mutual relations, i.e., their particu-

lar socio-economic form. For the aggregate of these re-

lations, in which the agents of this production stand with

respect to Nature and to one another, and in which they

produce, is precisely society, considered from the stand-

point of its economic structure" (Capital,III:818).

This posited link between the economic structure of

society and its political structure is, however, problema-

 235

tic. Marx claimed that "the specific economic form, in

which unpaid surplus-labour is pumped out of the direct

producers, determines the relationship of rulers and

ruled, as it grows directly out of production itself and,

in turn, reacts upon it as a determining element. Upon

this, however, is founded the entire formation of the eco-

nomic community which grows up out of the production rela-

tions themselves, thereby simultaneously its political

form. It is always the direct relationship of the owners

of the conditions of production to the "direct producers

a relation always naturally corresponding to a definite

stage in the development of the methods of labour and

thereby its social productivity -- which reveals the in-

nermost secret, the hidden basis of the entire social

structure, and with it the political form of the relation

of sovereignty and dependence, in short, the corresponding

form of the state" (Ibid:79l). Significantly, he immedi-

ately added a caveat which almost amounts to the negation

of the grand pronouncement: "This does not prevent the

same economic basis the same from the standpoint of

its main conditions due to innumerable different

empirical circumstances, natural environment, racial rela-

tions, external historical influences, etc., from showing

infinite variations and gradations in appearance, which

can be ascertained only by analysis of the empirically

given circumstances" (Ibid:792).

 236

Marx seemed to assume an organic unity between the

boundaries of the economy, as defined by the organization

of the labor process, and the state. Application of his

definition of economic structure of society to concrete

circumstances, however, reveals there is no empirical or

theoretical justification for the validity of his assump-

tion. Furthermore, the character of property relations

and the uses to which the extracted resources are put can-

not be reduced to the organization of labor. Abundant

historical examples demonstrate the compatability of the

same forms of labor with very different total systems of

domination.

Marx's implicit identification of the state with the

class of property owners and his view of wider social re-

lations as mere appearances are equally problematic.

Since these remarks are contained in the discussion of

pre-capitalist rent, they were presumably intended to shed

light on the structure of production and domination in

these societies. Marx seemed, however, to assert by theo-

retical fiat what he was in need of demonstrating analyti-

cally and empirically.

Analysis of the complexity of social relations in

concrete societies in time and place requires that atten-

tion be paid to the articulation of the control both of

the means of production, domination -- administration and

persuasion (Max Weber) -- and the means of destruction

(Goody,J.,1971:39-57). Attention to the unity and multi-

plicity of loci of control and patterns of extraction as

well as patterns of redistribution in determinable con-

junctural units of time will allow for a more dynamic

understanding of the production and reproduction of these

relations. A concrete examination of these issues will

be permitted by investigating how political bodies in a

system finance their activities and control their re-

sources.

Max Weber defined financing as the "provision of cor-

porate activity with economically scarce means" and as-

serted that "the most direct connection between the eco-

nomic system and primarily non-economic organizations lies

in the way in which they secure the means of their corpor-

ate activities as such; that is, the activity of the ad-

ministrative staff itself and that which is directed by

it" (1956/1978:194). The advantage of this broader forrnu-

lation is that it leaves the actual character of the po-

litical bodies in question to be determined by empirical

analysis. The state is undoubtedly the most widely known

of such bodies but it is, by no means, the only one. His-

torically, there has been wide variation in the degree to

which a state has been able to carry out its financial

activities autonomously or has had to depend on other

organizations for the realization of its resources. Sim-

ilarly, the degree to which all subjects of the state

 238

have been submitted to equal treatment, as far as realiza-

tion of resources was concerned, varied from period to

period.

Unitl, now, I have deliberately kept to a minimum

the discussion of categories through which the people of

Afghanistan identified themselves. It is time to take

account of these relations and examine the degree to which

they corresponded with regional boundaries, the most pre-

valent unit of reproduction of the economic structure dur-

ing our period of long duration. In the remaining part

of this chapter, I shall discuss patterns of socio-politi-

cal relations prior to the rise of the Afghan state in

1747. In the next chapter, I shall then discuss the pro-

cess of emergence and reproduction of that state through

time and the role of Islam as ideology and law in the pro-

cess of de structuring and restructuring of relations of

domination.

* * * *

From the beginning of the sixteenth century upto the

eighteenth century, the present territory of Afghanistan

was part of the Uzbeck, Safavid, and Mughal empires, as

shown in maps 8, 9, and 10. All these empires had been

formed in the first quarter of the sixteenth century and

demarcation of their spheres of influence was the out-

 239

come of direct military confrontation among them. In the

sixteenth century, shifts in the balance of forces re-

flected itself in the frequency with which various regions

changed hands. During the seventeenth century, however,

the boundaries remained relatively stable: Badakhshan and

the northern region formed part of the Uzbeck empire, the

western and southern regions part of the Safavid empire,

and the eastern region part of the Mughal empire. Various

localities in the central region fell within the zones of

one or other of the three empires or remained relatively

autonomous. The degree of exercise of power by an empire

within a region was not uniform and, as we shall see later

differed considerably in central and peripheral locations.

The complexity of the process of consolidation of a

secondary state is best revealed through the example of

the western region, centered on the city of Herat. Hav-

ing served as the capital of the Timurid empire for a

century, Herat had all the institutions of a stratified

society when Muhammad Khan, the Uzbeck ruler, defeated

the last Timurid rulers of the city in 1606. Upon the

departure of the Timurid princes, in the words of a con-

temporary historian, the "saaat [those whose claim of

descent from the Prophet of Islam is accepted by the

community], quzat [judges], akaber [great man], and

ayan [notables]" of Herat gathered for deliberation and

decided to pledge obedience to Muhammad Khan (Khawandmir,

 240

vol.IV:376). As a result of negotiations between repre-

sentatives of this group and the Khan, it was agreed that

a payment of 100,000 tanga be made by "the common people"

to the Khan. "The great people" were to pay 20,000 tanga

to the Khan and 15,000 to Maulana Abdur Rahim, the reli-

gious dignitary attached to the service of the Khan, pre-

viously a student in Herat, who had arranged the negotia-

tions. The city was then taken over by Maulana Abdur

Rahim and other Uzbeck officials and, within a week, the

agreed sums were realized from its inhabitants. During

the Friday prayer, the khutba, the public acknowledgement

which legitimated the authority of the new ruler, was

read in the name of the Uzbeck Khan (Ibid:378-79).

Herat remained under Uzbeck rule until 1510 when

Muhammad Khan was killed in a battle against the Safavid

Shah. The ashraf and ayan (nobles and notables) of Herat

then received the representative of the Shah who was

bearer of an edict especially issued in their name (Ibid:

514). The amount of tribute imposed on the people of

Herat by the Shah was not specified in the edict but when

the Safavid ruler entered the city a thousand men were

needed to carry the gifts presented to him by the newly

appointed governor of the city (Ibid:518). There is lit-

tle doubt that the governor had realized most of these

"gifts" from its citizenry.

Safavid conquest did not confine itself to this well-

 241

established form of extraction. Having declared the

Shi'a form of Islam as the official religion of the state,

Safavid rulers put intense symbolic pressure, backed by

the actual use of forces, to convert to Shi'ism a popula-

tion that hitherto had adhered to the Sunni faith. The

people of Herat were given a demonstration of the reli-

gious commitment of their new rulers upon the arrival of

the Shah's representative to the city. At a gathering

of both the "select and common" people in the major mosque

of the city was read to the population the edict issued

by the Shah, by an eminent religious scholar. Yet, des-

pite urgings by Safavid soldiers present in the mosque,

he refused to engage in the ritual cursing of the first

three caliphs of Islam and of a wife of the Prophet as

was sanctioned by the Shi'a establishment. In order to

impress the gathering, the Shah's representative immed-

iately killed with his own hands the dignitary there in

the mosque. Thereafter, the ritual was followed (Ibid:

515-516) .

Prior to the emergence of the Safavid empire in 1503,

Shi'ism had relied on persuasion rather than force for

gaining adherents. Although the shi'a doctrine had been

shaped by its ulema, the main avenue for seizure of po-

litical power by the Safavid family had been a Sufi order.

The role of Sufi orders in the reproduction and spread

of the Islamic social order has been one of the most ne-
 242

glected topics in the social history and anthropology of

the Islamic world. Since these activities occur before

and during our period of long duration, it is necessary

to analyze briefly the organizational patterns of one

such order which emerged in Mughal India and has continued

to play an important role in the history of Afghanistan

uptil this day.

The activities of Sufi orders have been eminently

connected with the mobility of their members, bypassing

the boundaries of anyone state or region. Movements of

people have played an extremely important part in the so-

cial and economic orientations of the formative period

of the Islamic world, i.e. between 700 and 1100. Andrew

Watson, who has analyzed the "Arab agricultural revolu-

tion and its diffusion" during that period, affirms that

all classes of people travelled a great deal -- lithe

rich and poor, the scholar and the illiterate, the holy

and the not so holy. Poverty was no obstacle: one could

move by foot, begging along the way; relatives could be

imposed upon endlessly; patrons were readily found for

scholars or holy men, or those who posed as such; a place

to bunk, and perhaps to eat, was available outside the

main mosque in most cities. Lured on in search of money,

adventure or truth, Muslims from every region left home

and roamed to and fro over the continents II (1974:21).

That the social order had managed to reproduce itself suc-

 243

cessfully after the fall of the Abbasid state which had

played a crucial role in shaping it (Shaban,M.A.,1970:138-

69), is indicated by a revealing comment of Marshal

Hodgson. He states that: "in 1500, Islamdom was expanding

over the hemisphere as a relatively integral cultural and

political order. Despite considerable diversity of langu-

age, custom, artisitic tradition, and even religious prac-

tice, the unit of Dar aI-Islam was a more significant fact

politically than the existence of any of the states with-

in it, which were of a local and in many cases a transi-

ent character" (1974,vol.III:4). That, despite the fra-

gile character of the states, the Islamic social order was

being successfully reproduced reflects the strength of

its institutions. Sufi orders were one of these institu-

tions.

Not only did Sufi orders transcend the bounds of the

locally based political and economic social relations,

they also gathered in their ranks members from different

strata of society. This can best be illustrated from

reference to the re-organization of an order by Shaykh

Ahmad Sirhandi (1564-1624), honored by his followers with

the title Mujadid-i Alf-i-Thani (the Renovator of the

Second Millenium). Since studies of the development of

his intellectual views as well as the reception of these

views by his contemporaries and subsequent generations

are readily available (Friedmann,Y.,197IiRar~an,F.,1968),

 244

I have confined myself to a description of the organiza-

tional features of the movement.

A biographical account of Sirhandi, his children and

twenty of his chief disciplines, written by a follower

shortly after the death of the Shaykh (Hazart Al-Quds, re-

printed in 1971) as well as a volume of Sirhandi's cor-

respondence which was compiled in his lifetime (Maktubat,

reprinted in 1968), contain valuable information on the

structural aspects of the Naqshbandi Sufi order.

The data indicates that mobility in search of learn-

ing the Sufi vision from the right master as well as the

goal of spreading that vision formed an integral part of

the life of the Sufi disciples. Coming from areas as

far apart as Central Asia and Bengal, most of them had

travelled through a considerable part of India before

being assigned as representatives of the order either to

their place of origin or to other localities. But even

with the attainment of the vision their wandering did not

cease and some continued to Arabia and other localities

in the Middle East. Both rich and poor were found in the

ranks. Some were men who had taken a vow of poverty and

earned their living from such pursuits as copying books,

others were landlords, connected to the court or possess-

ing considerable wealth from previous careers.

The master determined the state of knowledge at

which a disciple was j udqed competent of initiating others
 245

into the calling. Consequently, the length of time spent

by the disciple with the master varied considerably from

case to case. A closer reading of the biographies, how-

ever, suggests that factors other than the spirituality

of the disciple may have been instrumental in the deci-

sion. Whereas some spent as many as seven or eight years

before they were given permission to leave, an Afghan

devotee was allowed to depart after only a week. At the

time, a rival religious movement had mobilized a number of

Afghan clans in armed uprisings against the Mughal empire

and Sirhandi might have wanted to spread his ideas and

influence among the Afghans.

The ijazat (permission) granted to a khalifa (lit.

follower) could also be restricted in character. Thus,

an Afghan landlord who, upon seeking Sirhandi, had been

immediately initiated, was first allowed to lead ten fol-

lowers, then seventy; only at the end of his third visit

was the limitation on the number of his followers re-

moved (Hazart al-Quds:358). Upon his appointment to a

locality, the khalifa received a specially written docu-

ment delineating the territory in which he could exercise

specified spiritual functions. The extent of the area in

question varied from small districts on the one hand to

such large cities as Lahore on the other. Bonds between

the shaikh and his khulafa were maintained through regu-

lar correspondence and visits. In addition, the khulafa

 246

sent their students to the shaikh for further instruction.

Before joining the Naqshbandi order, which was to be-

come the main instrument for the propagation of his ideas,

Sirhandi had been granted an ijaza by his father to ini-

tiate the devout into the Qadirya and Chishtia order (Ib-

id:28-33). In 1598, feeling the need for further illumin-

ation, he decided to take a trip to Arabia but, in Delhi,

he met a leading shaikh of the Nazshbandi order who ini-

tiated him into that order and persuaded him to change

his travel plans. The shaikh, according to the standard

biography, was so impressed with the rapidity with which

Sirhandi passed through the stages of illumination that,

after a while, he turned over the leadership of the order

to him. Having become the undisputed master of the

Naqshbandi order, Sirhandi bestowed ijazat mainly for

this order. Yet, in at least one case, he granted one

of his khulafa the right to initiate followers into all

three orders (Ibid:320).

Regardless of the historical truth of the actual de-

tails through which the leadership of the order was passed

to Sirhandi, the fact remains that, historically, the pas-

sing of leadership to a person outside the genealogy of

the founder -- the more usual way of institutionalization

-- has allowed for the smooth transition of Sufi orders per-

mitting them to nect the challenge of novel historical cir-

cumstances. In the case of Sirhandi, there can be little
 247

doubt that his ability to dispense ijazat in three or-

ders simultaneously allowed him to reach a larger number

of people than would otherwise have been the case. More-

over, the extent of his influence was not confined to

Sufi orders. Having accomplished the standard education

of a scholar of religion in Islamic law and sciences (Ib-

id:32), Sirhandi taught these subjects to students who

joined the ranks of ulama, from among whom judges of the

Islamic courts were appointed. While these scholars did

not necessarily join Sufi orders, their ties to their mas-

ter remained strong. Sirhandi was acutely aware of the

importance of this group as he remarked that lion the day

of judgement, we will be asked about shari'a and not

tasawuf" that is about law and not mysticism. Students,

carriers of shari'a and vehicles of salvation of the pop-

ulation, are to be preferred over Sufis who had freed them-

selves from the world (Maktubat:185-86).

It is therefore necessary to avoid positing an a

priori opposition between ulama and Sufi orders and, in-

stead, analyzing the concrete historical process of in-

teraction between the two groups when, indeed, there were

two groups rather than one. During the sixteenth and

seventeenth centuries, ulama experienced considerable

geographical mobility and the extent of their social ties

with scholars in other regions, especially when reinforced

through nexuses of Sufi orders, had considerable implica-

 248

tions for structural continuity or change in the repro-

duction of the Islamic social order.

The political role played by the orders, however, de-

pended on the character of the state or states with which

the orders had to deal. At issue, to put it simply, was

the extent to which a state allowed orders to function as

a parallel structure of power dictating its own policies

or maintaining the total control of its own machinery.

The relative balance of power between the two systems can

be best seen in the financial sphere. A weak state usual-

ly exempted the properties of Sufi orders and ulama, whe-

ther held as awqaf or private property, from taxation and,

in order to placate them, it even granted them new large

endowments. A strong state, or one that felt threatened

by the order and/or ulama, attempted to impose taxation

and confiscate properties, resorting at time even to their

physical elimination. But since orders and a number of

ulama were supported by direct contributions from their

followers, wealthy individuals and even members of the

court, the success of such state pOlicies was uncertain.

Furthermore, an active policy of recruitment carried out

by an order among soldiers of a state could effectively

hamper its coercive power. Fear of the growing popular-

ity of an order within the army or among rebellious so-

cial groups might clearly prompt preemptive moves against

it.
 249

Thus, two of Sirhandi's most prominent khulafa, be-

lieved to pose a serious threat to the state because of

their popularity with Afghan and Uzbeck soldiers, were

put under strict surveillance. One of them was finally ex-

iled (Hazrat al Quds:305,385). It may have been because of

their activities that Sirhandi was first imprisoned for

one year and then placed in military detention. The

Mughal emperor Jahangir (1605-1627) noted in his auto-

biography for the year 1619: "At this time it was reported

to me that a Shayyad [a loud talker, a cheat] of the name

of Shaikh Ahmad had spread the net of hypocrisy and deceit

in Sirhind, and caught in it many of the apparent wor-

shippers without spirituality, and had sent into every

city and country one of his disciples, whom he called his

deputy (khalifa), and whom he considered more skilled than

others in the adorning of shops (of deceit) and selling

of religious knowledge, and in deceiving men. He had

also written a number of idle tales to his disciples and

his believers, and had made them into a book which he

called Maktubat (letters). In that alabum (Jung) of ab-

surdities many unprofitable things had been written that

drag (people) into infidelity and irrpiety. Amongst these

he had written in a letter as follows: 'In the course of

my travels I had come to the dwelling of the Two Lights

(the Sun and Moon), and saw a very lofty and very splen-

did building. From there I passed to the abode of Dis-

 250

crimination (Faruq), and from there I passed to the abode

of Truth (Siddiq), and to each I wrote a suitable explan-

ation (or perhaps, of each I wrote a suitable description).

From there I reached the abode of Love, and I beheld a

brilliant dwelling. It had divers colours and lights and

reflected glories. That is to say (God forgive us! -- an

exclamation of Jahagir's), I passed from the abode of the

Vicegerents (khulafa) and attained to the highest rank."

There were other presumptuous expressions which it would

be too long to write, and would be contrary to good man-

ners. I accordingly gave an order that they should bring

him to the Court that is based on justice. According to

order he came to pay his respects. To all that I asked

him he could give no reasonable answer, and appeared to

me to be extremely proud and self-satisfied, with all

his ignorance. I considered the best thing for him would

be that he should remain some time in the prison of cor-

rection until the heat of his temperament and confusion

of his brain were somewhat quenched, and the excitement

of the people should also subside. He was accordingly

handed over to AnIra'I Singh-dalan to be imprisoned in

Gwalior fort" (Tuzuk-i-JahangIrI Or Memoirs of Jahangir,

vol.II, trans. by A. Rogers,2nd ed.1968:91-93).

Thus, within twenty years from his initiation into

the Naqshbandi order, Sirhandi had managed to become a

power in the realm. In his case, the strength of the

 251

state and the religious diversity of the population in

India allowed for the effective checking of the threat

he was posing to the state.

The Safaviyya order in Iran faced a different set of

political circumstances and was consequently able to be-

come an empire. Although there is no satisfactory account

of the history of the order between the death of its foun-

der in 1334 and the founding of the Safavid empire by his

fifth lineal descendant, Shah Isma'il I, in 1501, descrip-

tions of some of the order's activities (Parizi,B.,1978:

9-17;Savory,R.,1980:1-27) clearly indicate that, struc-

turally, it shared the pattern of organization of other

Sufi orders. From its base in the city of Ardabil in

northern Iran, it rapidly extended its network through

its khulafa in eastern Anatolia and Syria. At first, it

attracted large endowments from local states. Soon, its

growing powers brought it into direct conflict with these

polities and both the father and grandfather of Shah

Isama'il (1501-1524) lost their lives on the battlefields.

Because military functions of the Turkic followers

of the order from the highlands between Anatolia and

Iran were crucial in the seizure of power, historians,

such as Savory, have concentrated almost exclusively on

the relationship between the order and these clans.

Marshal Hodgson, however, calls attention to the fact that

although the majority of the population between the Nile

 252

and the Oxus seem, by 1500, to follow the Sunni interpre-

tation of Islam, Shia orders had steadily gained adherents

in the urban centers of what today constitutes Iran and

Iraq. He argues that "Isma'il had then two sorts of Shi'i

support: the Shi'ism of the tribes, not only Sufi in feel-

ing but often deviating widely from the norms of the

Shar'i ulamai and the Shi'ism of the great city families,

no doubt also of a Sufi cast, but relatively bourgeois and

Shari'ah minded in many cases, though still willing to

look to a new social order within these limits" (1961,vol.

111:30). When the head of the order, upon becoming ruler

of the state, declared Shi'ism the religion of the state,

this combination made the conversion an enduring success.

Although the Safavid Shahs continued to claim legit-

imacy in their capacity of "Murshid-i-Kamil" (perfect

guide), the structural pose of the order after the es-

tablishment of the empire is not clear. Savory argues

that Shi'i ulama gradually managed to consolidate their

authority at the expense of the order (1980: ). Al-

though the hostile attitude of some of the leading ulama

against the order is well documented, the fact remains

that, even at the end of the Safavid period, it was offi-

cially stated that "Khalifas are also appointed through-

out the God-protected kingdom for the maintenance of law-

ful practices and for the prohibition of unlawful (munkar)

ones" (Tadhkirat AI-Maluk,1943:55). Clearly, there is

 253

need for further investigation of transformations in the

structure and changes, if any, in the social membership

of the order after the formation of the Safavid empire.

In Hodgson's view, the consolidation of the Shi'i

state that emerged from the activities of the Safaviyya order

radically changed the character of interrelationships

within the Islamic world. He claims that the newly em-

bittered quarrel between Shi'is and Sunis and the rise of

regional empires, forging their own separate worlds, had,

by 1550, dealt a major blow to the "cosmopolitan compre-

hensiveness of Islam" (1974,vol.III:4). Whereas the

Sunni-Shi'i division provided the ideological justifica-

tion for some of the most intense military conflicts be-

tween the Iranian empire on the one hand, and the Ottoman

and Uzbeck empires on the other, the fact that equally

intense conflicts had characterized the relationship be-

tween the Uzbeck empire and the small Timurid states in

Central Asia, Afghanistan and Iran, and the Mughal empire

in India, where all rulers adhered to the Sunni interpre-

tation of Islam, raises the question as to whether these

conflicts were not the products of the process of struc-

tural transformation of smaller states within larger

territorial units. Although available evidence is not

definitive, it seems to me that Hodgson greatly exager-

ates the degree to which flows of information, goods, and

people between different parts of the Islamic world were

 254

hindered by the advent of a Shi'i state in Iran.

As far as the rise of territorial states within the

Islamic world is concerned, Hodgson -- as the subtitle

of his study "Gunpowder Empires" suggests -- puts causal

emphasis on the emergence of artillery. He asserts that

"the command of siege and field artillery quickly came

to be of fateful significance politically throughout the

Dar aI-Islam ... The relative expensiveness of the ar-

tillery and the relative untenability of stone for-

tresses gave an increased advantage over local military

garrisons to a well-organized central power which could

afford artillery -- not always a decisive advantage, to

be sure. Perhaps at aleast as important was that gunpow-

der weapons seemed to imply, evidently from the start,

a continuous development of new techniques ... Such ad-

vantages for rulers with resources brought with them ex-

tensive possibilities for general political change" (Ibid:

18) •

Although there are new elements in Hodgson's formu-

lation, the argument itself is not novel. W. Barthold

(n.d.1930?:142-43) had already argued the case for Iran

and Irfan Habib (1963:317) had shown that, as far as

Mughal India was concerned, it was not the artillery but

the cavalry that formed the most important branch of the

army. This seems to have been true for most of the

Safavid era in Iran (Lambton,A.,1953:106-107) as well as

 255

for the Uzbeck empire. Indeed, once the difficulties

of moving and feeding large bodies of troops over vast

distances, in the absences of modern means of communica-

tion, is taken into account, the superiority accorded

by these states to the cavalry -- armed with guns --

becomes self-evident. Sieges of walled urban centers

lasting close to a year were common in this period and

Hodgson's emphasis on the untenability of fortresses a-

gainst the artillery is not convincing. The rapidly

changing technology of the means of destruction is impor-

tant but instead of merely relating it to cost it could

be more profitably related to availability and the degree

of political influence exerted by states controlling the

production of superior armaments over consumer states.

In short, mere acts of conquest, facilitated by

sophisticated means of destruction, should be analytically

distinguished from the institutionalization of power. We

should be able to differentiate between ideological claims

of absolutism and the degree to which centralization of

power is an actual reality. The link posited by Hodgson

between the cost of the means of destruction, the finan-

cial base of a state, and the degree of centralization of

power allows for empirical testing of the issue. The

question is not that of the realization of resources from

direct producers but whether the central government exer-

cises effective power in the recruitment, financial main-

 256

tenance, and control of the armed forces alone or whether

it shares these functions with other political bodies. In

Mughal India, where the cash nexus was probably more high-

ly developed than in Iran and the Uzbeck empire, the rele-

vant information is available in monetary figures and al-

lows for succint presentation.

Different systems of revenue prevailed in Mughal In-

dia, extracting in the name of the empire one third to

three quar.ters of the actual produce from direct producers.

The right of collection of land revenue and other taxes

was, over large portions of the empire, assigned by the

emperor to some of this subjects. The areas whose re-

venues were thus assigned were known as jagirs (less often,

tuyuls); and the assignees were termed jagirdars (Habib,

I.,1965:58-59). The jagirdars were officials of the state

and their jagirs were normally assigned to them in lieu

of salary. The payment made to these officials was deter-

mined by a dual system of ranking. The zat (body,person,

self) rank determined the personal pay and the sawar

(horse) the size of the cavalry contingent that the

mansabdar (holder of rank) was supposed to maintain for

the emperor's service. The basic objective of the system

being that of settling precedence and gradation of pay,

ranks might not, in fact, imply the command of any body

of troops (Irvine,W.,1903:4). As far as recruitment was

concerned, it is essential to realize that "few soldiers

 257

were entertained directly by the emperor himself; and

for the most part the men entered first the service of

some chief or leader" (Ibid:3). The system of patronage

tied all ranks to their immediate superiors rather than

to the whole army (Ibid:58).

While the income of officials carne mainly from their

jagir, the emperor drew his cash income mainly out of the

revenue of the Khalisa (Habib,1965:61). Compared to the

total revenue of the empire, the income derived from

khalisa varied from about 5 to 25 percent (Ibid). Irfan

Habib has tabulated, for the year 1647, the salary bill of

chief officials, compared in percentage to the empire's

revenue, and amounts supposedly devoted to the upkeep

of cav81ry (Ibid:62-64). Instead of reproducing his fig-

ure in darns, which was a lower unit of accounting, I have

converted them to rupees, the more common silver currency

which was the equivalent of 40 darns.

Table I
Percentage of
Rank Total Salary Bill Estimated Revenue-
for both zat and sawar Income (Jama) of
ranks (million of rupees) the Empire
4 princes of imperial family
family with ranks above
7,000 zat 18.1 8.2
21 mansabdar of 5,000 to 7,000 zat 35.44 16.1
43 mansabdar of 3,000 to 4,000 zat 27.0 12.3
151 mansabdar of 1,000 to 2,500 zat 36.35 16.5
226 mansabdar of 500 to 900 zat 18.38 8.1

Total:
445 mansabdar of 500 and above 135.27 61.5
 258

Out of a total of 8,000 mansabdar, 7,555 were hold-

ing ranks below 500 zat -- the lowest rank being command

over ten or twenty men -- and they probably accounted for

barely 25%, or at most 30% of the total revenue of the

Empire (Ibid:63). As for the 445 top mansabdar, they

must have "held over half of the Empire in jagir" (Ibid);

table II indicates the extent they were supposed to spend

from their income on the upkeep of their cavalry.

Table II

Total Pay Against % of Pay Against % of

Salary Zat Ranks Total Sawar Ranks Total
Class of mansabdar Bill (million Salary (million Salary
(million of rupees) Bill of rupees) Bill
of rupees)

Princes of above
7,000 zat 18.1 3.1 17.1 15.0 82.9

Mansabdar of 5,000
to 7,000 zat 35.44 5.74 16.2 29.7 83.8

Mansabdar of 3,000
to 5,000 zat 27.0 6.67 24.7 20.34 75.3

Mansabdar of 1,000
to 2,500 zat 36.35 9.97 27.4 26.38 72.6

Mansabdar of 500
to 900 zat 18.38 5.24 28.5 13.14 71.5

For the same year 1647, the total revenue of the em-

pire was estimated at 220 million rupees, and the direct

income of the emperor from khalisa was started as 30 mil-

lion rupees, or 13.6% of the total (Lahouri,M.A.,1868,

 259

vol.II:7l0-l3). At the time, the empire was divided in

22 suba (provinces) but there were significant differences

in the revenue of each province. The revenue from suba

in the present territory of Afghanistan which, except

for the western region of Herat, were, in that year, all

subordinate to the Mughal state, were as follows:

Badkashan, 1 million; Balkh, 2 million; Kandahar, 1.5

million; and Kabul, 4 million rupees. These sub a were

among the lowest eight in yielding revenue. By contrast,

figures for suba with the highest yield were as follows:

Shahjahanabad, 25 million; Akbarabad, 22.5 million;

Lahore, 22.5 million; Ajrnir, 15 million (Ibid). The sal-

ary of the subadar (governor) of Kabul, who was, at the

time, one of the highest ranking officials, amounted to

3 million rupees (Ibid:75).

Royal princes and high-ranking officials were as-

signed to duties in the provinces and were in a position

to make use of their salaries, as well as the larger re-

sources of the territory under their command, to support

their own armies. These were indeed private armies and

were used as such, as indicated by the fact that succes-

sion to the imperial throne was, in almost evey instance,

decided on the field of battle. Princes and officials

derived from their jagir a revenue fairly equal to the

emperor's income from khalisa, which served for maintain-

ing military forces owing direct allegiance to his own

 260

person. Emperors could, and indeed did, frequently ro-

tate, dismiss, imprison, or even eliminate princes and

officials from their posts. Yet, the reproduction of the

system revolved around the maintenance of close personal

relations between an emperor and some five hundred top

officials. The personal qualities of each emperor had

strong implications for the whole system and the elaborate

exchange of gifts between an emperor and his officials,

as well as the frequent use of violence against individ-

ual officials, were also aspects of the same process for

maintaining the state.

The officials' financial ability and political re-

sponsibility for retaining a large military force did

not, however, mean that, in practice, such a force was

actually kept. A dramatic illustration of the gap be-

tween theory and reality is provided by an incident dur-

ing the reign of Jahangir (1605-1627). In February 1626,

Mahabat Khan, one of the highest officials of the realm,

became aware of an imperial plan against him and was a-

fraid for his life. He made use of his personal force of

four to five thousand horsemen to take the emperor hos-

tagE and carried him off to Kabul (Tuzuk-i-Jahangiri,

Persian ed.:4l2ff.). Theoretically, the emperor's army

should have numbered hundreds of thousands (Irvine,W.,

1903:61). Yet, by May of the same year, the heir appar-

ent had managed to gather only a mere thousand horsemen,

 261

five hundred of whom dispersed after their leader died

(Ibid:421) .

Even when a large army could be assembled, the state

was dependent for its supplies on the cooperation of one

specific group, the banjaras. An observation of Jahangir,

when preparing to fight the Shah of Iran after the latter's

annexation of Kandahar in 1622, forcefully undelines this

dependence. He stated that "as there was little cultiva-

tion between Multan and Qandahar, the despatch of a large

army without provisions was not to be thought of. It was

therefore decided to encourage the grain-sellers, who in

the language of India are called banjara, and, providing

them with money, to take them along with the victorious

army, so that there might be no difficulty about supplies.

The Banjaras are a tribe [in Persian taifa, social group,

guild, people, band, etc. A.G.]. Some of them have 1,000

bullocks, and some more or less. They take grain from

different districts (bulukat) into the towns and sell it.

They go along with the armies, and with such an army there

would be 100,000 bullocks or more" (Rogers,A.,Tuzuk-i-

Jahangiri,vol.II,1968:223-34) .

Banjaras seem to have had close relations with vil-

lage merchants and money-lenders and might themselves

even have performed these functions for the direct pro-

ducers. Direct producers were compelled to maintain a

relation with the market in order to raise cash since it

 262

was expected that state taxes would be mainly paid in

cash. A highly developed monetary system existed and

Irfan Habib has shown that "the cash nexus was firmly

established in almost evey part of the empire. Its pre-

valence meant simply that the peasant was normally com-

pelled to sell a very large -- in not a few cases, the

larger portion of his produce in order to meet the re-

venue demand" (1963:239). The cash nexus did not, how-

ever, necessarily imply that the producers relationship

to the market was unmediated. Habib has argued that "it

is possible that a very large number of peasants were

not able to reach the open market at all, being compelled

to sellon contracted terms, to their creditors. Whether

the creditors were merchants or the village money-lenders,

the result was always to depress the prices received by

the peasants" (Ibid:78). Since the ratio of urban popu-

lation to the total population was probably very high (Ib-

id:76), it is s~fe to assume that banjaras must have had

a rather extensive network of social relations in every

region of the country to be able to deliver the needed

grain to urban centers and armies.

Functions performed by banjaras during periods of

conflict imply that their leaders who, according to a

European traveller, put as much pomp as princes

(Travernier,J.B,,1676/l905,vol.I:3l), must also have main-

tained close relations with the jagirdars who commanded

 263

the troops. But the character of the relationship, if

any, between banjaras, rural money-lenders and merchants,

troop commanders and their agents is not clear. Although

"jagirs were divorsed, as far as possible, from any rights

to the land, and were essentially assignments of revenue,

assessed and expressed in cash" (Habib,1963:318-19),

jagirdars had to send their own agents to collect the

revenue on their behalf (Ibid:283). In order to prevent

jagirdars from developing a territorial base of power,

emperors, as part of an elaborate bureacratic series of

checks, changed their jagir frequently (Ibid:269). Habib

asserts that the local element, except for officials

in charge of local records and revenue collection, was

"almost entirely excluded from the assignees' administra-

tion" (Ibid:287). Nonetheless, since his conclusion is

based on an order of Jahangir prohibiting the agents of

khalisa and jagirdars from "forming family-ties with the

local gentry" (Ibid:287); the degree to which practice

corresponded to the wishes of the authorities remains

open to question.

The existence of a financial machinery for linking

villages with the institutions of the larger society was

observed by Jean Baptiste Tavernier. He stated that "in

the Indies, the village must be very small where there

does not reside a banker, whom they call cheraff [from

the Arabic sarraf]; whose business it is to remit money

 264

and Bills of Exchange" (1676/l905,vol.I:2l). Unfortunate-

ly, there is little information on the social relations

between these bankers in rural areas and those in urban

centers who played a crucial role in the long-distance

trade. On the other hand, it is known that sarrafs "en-

joyed no caste-monopoly over money-lending which was car-

ried on by various classes of persons in different forms.

Merchants advanced loans to cultivators and artisans to

be repaid in the form of produce or manufactured material.

Village headmen, as well as the administration, gave

taqavi (modern 'taccavi') loans to peasants to enable them

to engage in cultivation. Usurious money-lending appears

to have been carried on in the countryside by grain-mer-

chants in the Punjab and sanyasis or mendicants in Bengal.

Finally, there were the professional money-lenders, known

as mahajans and sahukars, who must have commanded consid-

erable resources of money" (Habib,1960:l8-l9). Given

the existence of such well-organized groups with ready

access to cash, it seems hard to conceive that jagirdars,

who were always in need of cash and often in places dis-

tant from their jagir, would have dispensed with the ser-

vices of these groups. Whether such a connection did in

fact exist remains to be investigated.

Individual officials and the imperial administration

clearly made use of sarrafs for transferring large amounts

of money from one part of the state to another (Ibid:lO-

 265

11). The degree of political dependence of the state

on these financial networks, and the shift of support,

at the beginning of the eigtheenth century by the "great

firms" from the central government to regional powers

are, however, subjects on which there is no scholarly

consensus (e.g., Leonard,K.,1979:151-167;Richards,J.F.,

1981:285-313).

Peasants met with increasing oppression from the

state (Habib,1963:319ff.) and, with the passage of time,

this oppression became so severe that the choice left to

them was that of starvation, slavery, or armed resistance

(Ibid:329). Peasants were not the only elements, with-

in the differentiated structure of the rural community,

whose interests were pitted against the central govern-

ment and the jagirdars. Intermediate, if not mediating

between producers and the state, was the class of

zamindars (landowners). The word has many connotations

but Habib contends that the system of zamindari may in

essence be described as "a right superior to that of the

peasants, and originating, in the main, in the existing

imperial power. It implied a claim to a share in the pro-

duce of the soil which was completely distinct from, al-

though it might be laid side by side with, the land-

revenue demand" (1965:68). As the right was "fully

salable with almost no restraints" (Ibid:70), it contri-

buted to the movement of new individuals and groups into

 266

the villages. The point of conflict between zamindars

and the state revolved around their respective shares of

the produce (Habib,1963:334). What made zamindars a force

to be reckoned with was that, as a class, "command over

armed retainers was usually a necessary complement of

their right, and they were frequently leaders of caste-

groups" (Ibid:334). Increasing demands of the state for

revenue drove producers and zamindars into a series of

alliances, usually expressed through new religious sects,

giving rise to armed revolts in different parts of the

country and ultimately bringing about the collapse of the

empire (Ibid:330-35l).

Attempts at building centralized structures in the

sixteenth and seventeenth centuries and tendencies towards

the breakdown of these structures towards the end of the

seventeenth and beginning of the eighteenth centuries also

characterized the history of both the Safavid and Uzbeck

empires. The two empires differed in their organization

and in the strategies adopted for restructuring relations

among the main political bodies. I shall briefly consid-

er these developments; focus on the empire's financial

basis will allow for analytical isolation of the main ac-

tors.

The Safavid empire, like the Mughal empire, distin-

guished between the general revenue of the empire that was

administered by the divan (central department) and revenue

 267

directly under the control of the ruler and administered

by the khasa (private department). The only relatively

reliable figures on the amount of these revenues comes

from a manual of administration that reflects the situa-

tion at the beginning of the eighteenth century. v.

Minorsky, who prepared the text of the manual, Tadkhirat

AI-Maluk, for publication, attempted to make sense of the

figures in his extensive commentary. Therefore, instead

of computing the figures from the original Persian text,

I have made use of his calculations (Minorsky,1943:110-

86) .

The total revenue of the empire was calculated in

monetary terms at 785,623.88 tomans, 608,652.34 of which

from the divan and 176,971.24 tomans (22.5% of the 'total)

from the khasa (Ibid:174). For the same period, the

total expenditure of the empire was stated at 625,274.56

tomans, of which 507,400.63 were listed under the divan

and 117,873.93 (18.85% of the total) under the khasa

(Ibid:183) .

The main items of expenditure are summarized in

Table III. Total sums listed by the author of the manual

for each division and additions of the items as enumer-

ated in sub-heaf .1gS do not usually concur. Minorsky

is of the opinion that "the author's totals must be more

reliable than the totals calculated on the basis of the

scribe's copy" (Ibid:183). Therefore, he states both the

indicated and calculated sums.

 268

Table III. Expenditure of the Safavid State

A. Yearly Salaries in tomans

(a) Civil Staff

1. Some Princesses 2191.63
2. Arnirs and Governors 396792.00
3. Farashes (carpet and tent spreaders) 4956.5
4. Yasavulan-i-Suhbat (aides de camp) 4721. 742
5. Doctors 4998.13
6. Workmen 6542.1
7. Yassvuls (aides of lower rank than 4) 1587.17

(S) Military Staff

8. Artillery men 1942.50
9. Qurchis (praetorians) 25572.67
10. Ghu1ams (slaves) 18261.52
11. Tufanchis, etc. (equipped with hand guns) 21960.33
12. Ghazis (original Sufi corps) 2124.87

(y) Others 7892.17

As calculated 503,320.917

As indicated 491,896.570

B. Tahvil, sums credited to various officials for producing

or acquiring certain commodities.

As calculated 36,483.9800

As indicated 40,391. 4569

C. Various Categories of Grants

As calculated 67,235.1100

As indicated 93,032.0500

Total As indicated 625,273.60

Sum of A,B,C, on basis of

indicated figures 625,320.0769

Sum of A,B,C, on basis of

calculated figures 607,040.007
 269

These figures do not refer to the whole area under the

jurisdiction of the empire, as was the case with figures

for the Mughal empire, and reference to territorial divi-

sions reveals a complex set of political and administra-

tive arrangements between the central and provincial

governments. Although seventeen territorial divisions are

mentioned in the text, figures on income and expenditure

encompass only thirteen divisions. The four territories

excluded from the statistics -- Arabistan, Luristan,

Georgia, and Kurdistan -- were areas that "belonged to

ancient families or hereditary rulers and, in spite of

their incorporation in the Safavid state, enjoyed inde-

pendence. Their revenue was not included in the budget,

and apart from military assistance, they owed their

suzerain nothing but tribute disguised as gifts" (Ibid:

112). The remaining provinces were divided into two cate-

gories: mamalik or "state" provinces, administered by

governors with wide powers over the disposal of revenue,

and khasa or "crown" provinces where governors were dir-

ectly accountable to the Shah's private department. The

distinction originated during the reign of Shah Abbas I

(1588-1629) and, as it was an outcome of strong internal

tensions and external conflicts during the earlier period,

especially from 1524-1588 (Savory,1980:40ff.), and had

strong implications for the subsequent history of the

empire, it needs to be briefly considered.

 270

The Turkish clans, whose victories on the battlefield

brought about the transformation of the Safavid house from

leaders of a Sufi order into Shahs of Iran, were known as

qizilbash, after their distinctive red headgear signify-

ing their loyalty to the order. Prior to the reign of

Shah Abbas I, "the government of the provinces was alloted

to the qizilbash chiefs in the form of assignments known

as Tiyul. The governor of the province was allowed to

consume the greater part of the revenue of the province

on the condition that he maintained, and mustered when

required to do so by the shah, a stated number of troops"

(Ibid:79). There is no statistical analysis of the length

of tenure of these governors or the frequency with which

they were succeeded by their own descendants. At least

in one case, the governorship of Shiraz remained within

the same family during the first half of the sixteenth

century (Savory,1964:115).

The shahs seem to have had a direct presence in the

provinces through provincial wazirs (comptrollers) and

sadrs (heads of religious establishment) whom they ap-

pointed directly and transferred freely from one province

to another (Ibid:119,123). Labels of office should not

be interpreted as corresponding to clear functional di-

visions; the shah could, and did, invest administrative

and religious officials, most of whom were non-Turkish

and Persian-speaking, with the insignia of military of-

 271

fice, and some of them even discharged actual military

duties (Ibid:115,126;Khwandmir,vol.IV;491,507,527,534, and

603-85 for examples). By manipulating personal rivalries

among these dignitaries as well as among leading qizilbash

amirs belonging to different clans, the shah could exert

direct influence on the conduct of affairs of the state.

Nonetheless, governors seem to have been able to put an

end violently to the ambitions of religious and admin-

istrative functionaries who posed a threat to the consoli-

dation of their power (Savory,1964:126 for one prominent

example; and 1980:33ff. for others). Despite the ideology

of absolutism and religious sanctity surrounding the shahs,

relations between shahs and amirs were full of violent

episodes during which some amirs reduced shahs to mere

figureheads (Ibid). The fact that governors "were able

to SUb-assign the area under them" (Lambton,A.,1953:107)

must have considerably enhanced their ability to build

and maintain their own network of support.

The continuing control of governors and amirs over

a major portion of the empire's economic resources through-

out the Safavid era is borne out by the data in Table III.

The sum of 396,792 tomans alloted to them constituted

63.46% of the total expenditure of the empire. That num-

ber of men belonging to this class, as in the case of the

Mughal empire, was actually very small is reported by the

court historian who stated that the figure was 114 amirs
 272

in 1576 and 92 in 1629 (Eskandar Monshi, tr.by Savory,

1978,vol.I:222;vol.II:1309) .

The social composition of this class and their con-

trol over the means of destruction had, however, greatly

changed as a result of policies pursued by Abbas I (1588-

1629). In order to counterbalance the power of the

qizilbash arnirs, Shah Abbas attempted to create a military

force that owed loyalty only to him. Thus there carne into

being the ghulam (so-called slave) regiments. "He en-

rolled in the armed forces large numbers of Georgian,

Circassian, and other golarns, ... Several thousand men

were drafted into regiments of musketeers from the

Cagatay tribe, and from various Arab and Persian tribes

in Khorasan, Azerbaidjan, and Tabarestan. Into the re-

giments of musketeers, too, were drafted all the riff-

raff from every province -- sturdy, serviceable men who

were unemployed and preyed on the lower classes of soci-

ety ... All these men were placed on the golam muster

rolls" (Monshi/Savory,1978,vol.I:527).

As the military vehicle of Abbas's policies, the

ghulam regiments and other forces were used to achieve

the forceful relocation of groups of people according

to John Perry's calculation, some fourteen instances in-

volving at least 100,000 families (1975:202f.). Shah

Tahrnasp (1524-1576) had already had some 30,000 Caucasians

forcefully brought over to the territory of the Safavid

 273

empire after their military defeat in 1554. Among the

populations affected by Shah Abbas's schemes were the

qizilbash clans. Roger Savory affirms that Abbas "em-

barked on a systematic policy of transferring groups of

qizilbash belonging to one tribe, to an ulka, or tribal

district, held in fief [sic.; Persian original tiyul] by

another tribe ... in some cases, he would deliberately

place an officer in charge of a tribe who was not himself

member of the tribe; in other cases, he would allege that

the tribe did not possess an officer capable of leading

it, and would appoint a ghulam to act as chief of the

tribe" (1980:81).

Although the creation of the new military force im-

plied a reduction in the number of qizilbash fighters

and the restructuring of relations between the clans and

the state, it did not mean the end of these clans as a

military force. The ghulam forces, once formed, numbered

20,000 horsemen and 12,000 infantry; at the same time,

the qizilbash force still numbered 30,000 horsemen, though

it had been reduced by half since the beginning of Abbas's

reign (Bayani,K.,1974:70).

To maintain the loyalty of the new force, the ruler

provided its payroll from the financial resources direct-

ly at this disposal. Hence the need to expand these re-

sources and the conversion of semi-autonomous provinces

into khasa (Savory,1980:80) whereby agents, directly ap-

 274

pointed by the crown, administered the territories and

were moved from one post to another at will. The effect

of these policies on relations between the state and

qizilbash clans is reflected in the changes in the social

composition of the amirs.

In 1576, there were 114 amirs in the Safavid state;

72 were quoted in name by the court historian, out of

which 59 were of lITurkish origin". The majority, if not

all, of the amirs were affiliated to a clan (Minor sky ,

1943:15). In 1629, the year of shah Abbas's death, out

of 93 amirs 21 were recruited from among ghulams (Monshi,

1978,vol.II:1309-l7). An impressive reflection of the

power of ghulam amirs that year is provided by the fact

that they supplied eight out of the fourteen governors

of large provinces (Bayani,K.,1974:91).

There is no satisfactory analysis of the social

relations between the two categories of amirs during the

subsequent history of the Safavid empire. Khanbaba

Bayani claims that, at the end of shah Safi's reign (1629-

1642), only three large provinces out of fifteen were

governed by ghulams but that, out of 37 governors ap-

pointed by shah Abbas II (1642-1666), 23 were ghulams

and 2 more might have been (Ibid). Savory, who does not

provide any statistical information, claims that, at times

of peace, the general tendency was to appoint ghulams to

administer provinces while during periods of danger

 275

qizilbash governors were preferred (1980: ).

Regardless of their composition at any particular

time, the amirs, as a class, continued to exert a large

degree of power throughout the Safavid empire. A strong

shah could check their power by playing on their faction-

alism but the amirs seem to have played an important role

in the running of affairs of the empire. Data in Table

III has already shown that, financially, the amirs re-

mained in a strong position until the end of the Safavid

period.

The fact that the Saravid forces did not constitute

a regular standing army may have been an important factor

in the ability of the shahs to check the development of

social ties between amirs and soldiers. Bayani quotes

the official history of the reign of shah Abbas I I to the

effect that, except for the royal guard who were always

in attendance, soldiers were stationed on their tiyuls

throughout the realm. Even the infantry, who were e-

quipped with hand-guns, only gathered at times of war

(Ibid:92-93). During periods of intense and lasting

military conflict, such as the reign of Abbas I, men who

drew salaries must have spent considerable ·U.me in actual

combat. But a treaty of peace with the Otto~an empire

in 1639 stopped the armed conflict between the two rival

powers for 74 years (Ibid:77). During that period, the

gap between the number of men on the payroll and the ac-
 276

tual force was considerable. Thus, in a review held in

1660, Abbas II "discovered that the same arms, horses, and

men passed before him 10-12 times" (Minorsky,1943:35).

Lack of trained men was not the only problem. Bayani

marshals considerable evidence to show that, after the

reign of Abbas I, procurement of equipment for the army

was almost completely neglected (1974:78ff.). In his

capital city of Isfahan, Abbas I had made use of his com-

mercial contacts with Europe to initiate a policy of

building factories for the manufacture of arms and cannons

(Ibid:7l) but little attention was subsequently paid to

maintain these factories (Ibid:82). Indeed, from the fig-

ures in table III, it can be seen that sums paid to car-

pet and tent spreaders were 2.5 times as much as those

paid to the artillery men. It was, therefore, not very

surprising that a force of Ghilzai Afghans who, in 1709,

had gained independence from Safavid rule in Kandahar,

conquered ln 1722, after a siege of six months, the cap-

ital city of Isfahan and put an end to the dynasty.

Unlike the Mughal and Safavid empires which relied

on administrative means to contain regional forces, the

Uzbeck empire came to terms with these powers through a

very different institutional mechanism. ROlJert McChesney

argues that, theoretically, the organizational principles

of the Uzbeck states were two: first, that only an agnatic

descendant of the Mongol conqueror Chhingiz Khan (ca.1155-

 277

1227) was eligible to rule; second, that the right to

govern resided not in an individual but in the entire

Chingizid ruling lineage -- "among those individuals who

had reached the 'age of discretion'" (1973:ix). The ter-

ritory ruled by a member of the lineage was called an

1
appanage. Within each appanage, its ruler called

sultan -- was "the first authority, possessing complete

control over the finances and foreign policies of his

territory" (Ibid:xvii). To underline the importance of

appanages, the office of Khan, the nominal ruler of the

empire, was invested in the senior member of the lineage

whose actual authority in the empire was moral rather

than real (Ibid:xiv-xv). The institutional mechanism

through which territorial distributions were made, new

Khans chosen and decisions of the empire taken was an

assembly -- called a qurultay -- of the members of the

ruling lineage (Ibid:xvi).

Leaders of the clans, providing the military back-

bone of the empire but not being part of the Chingizid

g~ilealogy, were another group of power-holders, Their

title was that of amir, and each sultan assigned his

amirs territories for their support, called iqta (pl.

lThe use of this term has been based on the analysis by

the Russian historians of the organizational parallels
between the political systems of Kiev, in the 11th and
12th, and Muscovy in the 13th and 14th centuries
(McChesney,1973:xvi) .
 278

iqtaat). Although amirs exercised within their respec-

tive iqta powers similar to that of a ruler, they could

be either dismissed or transferred by their sultan (Ibid:

xvii) .

In practice, tensions within the empire, as well as

in the larger system of interrelations with which the

Uzbeck empire was locked in conflict, made the successful

realization of the Uzbeck political norms dependent on the

coordinated effort of members of the ruling lineage for

acquisition of new territory. When efforts at expansion

were blocked by other powers, the result usually was in-

tra-appanage warfare.

From the formation of the empire in Dasht-i-Qipchaq

to the conquest of Herat and subsequent routing by the

Safavid forces in 1510, there had been considerable fluc-

tuations in the degree of cooperation and conflict among

members of the lineage (Dickson,M.,1960:208-216). But

in 1512, the Uzbecks decisively defeated a joint Mughal-

Safavid force that had aimed at dislodging them complete-

ly from Central Asia and, by 1526, they had conquered

Balkh, which, with brief exceptions, remained under their

control until 1711. Badakhshan, on the other hand, was

only conquered during the seventeenth century.

The ruling group was not homogeneous. Once the dis-

tribution of appanages had taken place, the significant

geneaological link was no longer to Chingiz Khan or even

 279

the founder of the state but to the closest ancestor who

had first consolidated his authority in the appanage.

Thus, in the two assemblies held in 1512/1513, the four

main appanages of Samarqand, Tashkrand, Bukhara, and

Miyankal-Soghd later augmented by Balkh -- were dis-

tributed to four named lineages (McChesney,R.,1973:27).

Each appanage was, in turn, divided among the members

of the named lineage and the amirs (Ibid:32). There is

no information, however, on whether any descendant of

the founder of the dynasty was excluded from a share dur-

ing the distribution.

The history of the Uzbeck empire after 1540 is, to

a large degree, the chronicle of struggle among these

four named groups for the dominance of the whole empire.

The background for this internecine warfare was, once a-

gain, provided by the checking of Uzbeck expansion by the

Safavid empire. For seventeen years, inter-appanage wars

were so intense that the four units became virtually

separate states (McChesney,R.:42).

This period was followed by an attempt at completely

reshaping the structure of uzbeck polity. Abd Allah Khan,

who first emerged into prominence with his conquest of

Bukhara in 1557, forged an alliance with the amirs and

proceeded to destroy the power of members of other named

lineages as well as his own (Ibid:43-44). Abd Allah Khan's

internal victories were followed by his conquest of Herat

 280

and other Safavid centers in 1588/89. That his appointee

as governor of Herat was a scholar from a "noble family

of Samarkand" , who had first been appointed by him as

sadr (chief religious dignitary) then raised to the rank

of arnir (Monshi,vol.II:56l-562), shows the degree of de-

parture from Uzbeck constitutional theory. With the

death of Abd Allah Khan in 1589 and the subsequent assas-

sination of his son and successor shortly after (Ibid:728-

729), the old political theory was reaffirmed by a new

ruling lineage.

In pursuit of their policies of centralization of

power, Abd Allah Khan and his son had systematically e-

liminated other members of the ruling lineage who could

have served as rallying points of opposition. They had

been so successful in this endeavor that, after the

assassination of Abd Allah's son the only member of the

lineage who was alive was an opium addict -- considered

too insignificant to pose a threat (Ibid:738). Elected

Khan by a number of amirs, he was, after a brief per-

iod, deposed by the forces of a new named lineage whose

head was the son of Abd Allah's sister. Thus, power was

transferred to another line of the Chingizid lineage.

During the first assembly of the new ruling lineage,

held in 1599, in Samarkand, the decentralized principles

of government were reaffirmed and appanages distributed

(McChesney,1973:53-56). The period of expansion of the

 281

new state was only from 1599 to 1603, in the course of

which it defeated a Safavid force and annexed the central

area of Balkh and the northern region of present day

Afghanistan. Once the period of expansion carne to an

end, a series of internal tensions emerged.

From 1610 onwards, the two appanages of Balkh and

Bukhara, assigned to two brothers of the ruling lineage,

emerged as the dual centers of the empire and, despite

frequent attempts at consolidation, remained autonomous

until the end of the dynasty in 1711 (Ibid:53-70). The

most significant event of the period was, however, the

increasingly important political role of the amirs.

McChesney divides the history of relations between

sultans and amirs into two distinct periods. He argues

that, from 1599 to 1641, iqtaat "were neither immutable

nor heritary" (Ibid:65) and transfer of amirs from one

iqta to another was quite common (Ibid:67). But from

1641 to 1711, the identification of amirs with their

iqta became more firmly established and ultimately led to

the emergence of a number of independent principalities

in the eighteenth century.

This process of restructuring was facilitated by two

sets of factors. First, even in the earlier period, amirs

had retained fiscal authority from the sultans, except

for passing a portion of the spoils of war. They had re-

lied on the same administrative mechanisms for the extrac-

 282

tion of resources from direct producers as those prevail-

ing in an appanage (Ibid:79). Second, a process of seden-

tarization of the clans seems to have been occuring in

the second half of the seventeenth century (Ibid:84f.).

Also, since the beginning of the century, the amirs had

been conducting their own raiding expeditions against non-

Uzbeck elements (Ibid:65) and were consequently in command

of their own forces; they could easily redirect their for-

ces against other Uzbeck groups. Indeed, the second half

of the seventeenth century witnessed considerable inter-

iqta warfare (Ibid:83). When the last Chingizid khan,

whose policies had alienated the amirs, was assassinated

in 1711, a number of new principalities came into being.

In Bukhara, one clan had gained dominance and was able to

seize power, but in the northern region of present day

Afghanistan, separate states were formed in Qunduz, Balkh,

Andkhuy, Maymanah, and Khulm (Ibid:102). The story of the

incorporation of these statelets within the Afghan state

during the eighteenth and nineteenth centuries will be

one of the themes addressed in the next chapter.

Our review of the Mughal, Safavid, and Uzbeck em-

pires reveals that, depite differences in size and com-

plexity of administrative and financial institutions,

structural tensions in the relations between centripetal

and centrifugal forces remained unresolved in all three

empires, and periods of centralization were followed by

 283

periods of reassertion of regional powers.

The ability of regional powers to reassert themselves,

I will argue, was closely related to the fact that politi-

cal conquest of territories had not been followed by the

incorporation of these areas within an economic system

that corresponded to the political boundaries of the state.

As none of these empires had an integrated domestic mar-

ket and methods of production were similar in all the cen-

tral areas of the whole region, distinctive regional eco-

nomies persisted throughout the duration of these empires.

The very ease with which various regions were transferred

from the sphere of influence of one empire to another, as

illustrated by the example of Herat, was a reflection of

the economic autonomy of these regions. These regionally

enduring economic units were, in turn, the foundations

upon which the very wide political and financial powers

of the provincial authorities were built. While rulers,

through frequent transfer of individual officials, could

and did exert influence in the provinces, they were not

in a position to reduce these officials into mere func-

tionaries. Thus, fluctuation between exchange of gifts

as formalized expressions of dependence and frequent re-

sort to actual violence was a mechanism of underwriting

that dependence.

The absence of integrated domestic markets did not

imply the absence of economic interrelations among re-

 284

gions, regardless of state boundaries. Indeed, long-dis-

tance trade played an important role in the history of

the area.

The extent and direction of flow of commodities, at

the time of formation of centralization and de structuring

of these empires, was closely linked to the emrging cos-

mopolitical economy of Europe. This nexus was not only

important in the relationship of each of the empires with

Europe, but also in the economic articulation of relations

among these empires. The European connection, however,

seems to have manifested itself in very different forms

at different periods.

What has to be clearly grasped is the fact that,

despite the discovery of the sea route to India by the

Portuguese, its disruptive impact on the intercontinental

trade was not felt until after "the elapse of an entire

century. After a set-back at the beginning of the 16th

century the trade routes through the Middle East regained

their former importance, and at the end of the 16th cen-

tury the transcontinental caravan trade reached dimen-

sions which must presumably be regarded as its histori-

cal culmination" (Steensgaard,1973:9).

The nature of co~~odities moved in the course of

long-distance trade and the costs of protection incurred

along the land and sea routes between India and Europe

were both factors in the competition between caravans and

 285

carracks. Commodities in demand were luxury and not bulk

goods. As such, the overall quantity transferred was

rather small by comparison to heavy articles. Niels

Steensgaard, using Lane's concepts, has carefully compared

the available evidence on costs of transportation and pro-

tection between the routes Isfahan-Gombroon-Surat and

Isfahan-Kandahar-Lahore. His conclusion is that, for

goods in demand in the 16th century Europe, the cost of

transport by caravan was cheaper than that by ship (Ibid:

40). However, the cost of protection along the land

route was both higher and unpredictable, a factor that

made sharp fluctuation in the market unavoidable (Ibid:

40,58,67). He argues that the secret of the survival of

the caravans lied in the high protection costs imposed

by the Portugese on the sea route (Ibid:lOl).

The amount of goods conveyed along the two routes

is estimated by Steensgaard in the following table (Ibid:

168) .

Table IV. Estimate of European Consumption of Asian Goods around 1600

the Cape route the caravan route in all

Pepper 1-2,000,000 lb. 3-4,000,000 lb. 5,000,000 lb.

Cloves
Indigo 350,000 - 700,000 -
1,350,000 lb •
Drugs . 650,000 lb. 1,000,000 lb.
Mace
Nutmeg
Textiles ? ? ?
Raw Silk 500,000 lb. 500,000 lb.
°
 286

Assuming that most spices and about half the silk

came from India and adopting Steensgaard's estimate of

440 lbs. per camel load, it appears that 9,000 to 12,000

camels must have been needed to carry the above amounts.

Steensgaard does not discuss the time spent on each route

but Tavernier, who between 1636 and 1662 made six jour-

neys in Asia and knew both routes well, provided the rele-

vant data. The Isfahan-Kandahar-Labore journey took 106

days (1676/l905,vol.I:72) while sailing between Surat

and Hormus could take from fourteen to thirty five days.

But the sailing was seasonal and could only be done from

November to March from Surat and From November to Febru-

ary from Hormus (Ibid:2). With technical improvements

in sailing and changing political circumstances in Eur-

ope, reducing the cost of protection at sea, the ships

were clearly destined to gain a decisive edge over the

caravans.

The Portuguese monopoly at sea was broken with the

advent of new European companies at the beginning of the

seventeenth century. Unlike the Portuguese who were

after tribute, these companies operated as capitalist

enterprises. They undertook to internalize costs of pro-

tection thus putting the caravans, which had no control

over their own costs of protection, on the defensive. By

1625, the sea route had triumphed -- a victory vividly

marked by the spectacle of Levantine towns which now im-

 287

ported their needed Indian spices from Europe (Ibid:171-

173) .

The long-distance trade was not confined to goods

bound for Europe. There was an extensive trade among

Asian states, Babur, the founder of the Mughal empire

who conquered Kabul in 1504, provided a vivid description

of the inland commerce of the area in his memoires. He

stated that "just as Arabs call every place outside Arab

(Arabia), Ajam, so Hindustanis call every place outside

Hindustan, Khurasan. There are two trade-marts on the

land-route between Hindustan and Khurasan; one is Kabul,

the other Qandahar. To Kabul caravans came from Kashgar,

Farghana, Turkistan, Samarkand, Bukhara, Balkh, Hisar

and Badakhshan. To Qandahar they come from Khurasan.

Kabul is an excellent trading-center; if merchants went

to Khita [China] or to Rum [ottoman Empire], they might

make no higher profit. Down to Kabul every year come 7,8

or 10,000 horses and up to it, from Hindustan, come every

year caravans of 10, 15, or 20,000 heads-of-houses, bring-

ing slaves (barda), white cloth, sugar-candy, refined and

common sugar, and aromatic roots. Many a trader is not

content with a profit of 30 to 40 on 10 [300% to 400%].

In Kabul can be had the products of Khurasan, Rum, Iraq,

and Chin(China); while it is Hindustan's own market

(Beveridge,tr., 1969:202)

Kabul had retained its position as an inland trading

 288

center towards the second half of the seventeenth century

as indicated by Tavernier. He described the place as a

"large city, very well fortified; and is the place where

those Usbek corne every year to sell their horses. They

reckon, that there are bought and sold, every year, above

sixty thousands. They bring also out of Persia, great

numbers of sheep, and other cattle; it being the general

concourse of Persians, Tartarians, and Indians" (1676/

1905,vol.I:73) .

Animals were not the only items of trade. Tavernier

claimed that out of 22,000 bales (2,200,000 lbs.) of silk

exported every year from Bengal, the "Hollanders usually

carry away six or seven-thousand bales, and would carry

more, did not the merchants of Tartary and the Moghul's

empire oppose them; for they buy up as much as the Hollan-

ders; the rest the natives keep to make their stuffs"

(Ibid,vol.II:299-300). Tavernier, unfortunately, does

not provide any information on the relative share of Mid-

dle Eastern and Central Asian merchants on the one hand,

and European merchants on the other, in their demand for

the numerous Indian commodities.

By 1676, the time of Tavernier's writing, the impact

of the emergence of the Indian port of Surat as the main

outlet of export was already being felt in the manufac-

turing centers of the north. He stated that "vast quan-

tities of linen Calicuts" made in Multan and Lahore, had

 289

to be carried to Agra and from Agra to Surat. "But in

regard carriage is so dear, very few merchants traffic

either to Multan or at Lahore; and many of the workmen

have also deserted those places, so that the King's re-

venues are very much diminished in those provinces" (Ibid,

vol.I:72) .

Holden Furber demonstrates that the European trading

companies were the beneficiaries of this shift of trade

routes: "for the European East India companies, the Per-

sian Gulf was a profitable trading area in the seventeenth

century primarily because of the demand for Persian silk

and the marketability of East India goods in Persia" (1976:

9). Until the beginning of the eighteenth century, the

function played by the Europeans as importers and trans-

porters of Asian goods was still more important than

their role as promoters of their own manufactures. Sim-

ilarily, they were still vulnerable to political events

in the area and the overthrow of the Safavid dynasty in

1722 profoundly upset their Indian trade (Ibid:143).

Tavernier's remark on the emergence of the sea route

and its impact on the revenues of the state has important

implications for the changing relations between central

and provincial authorities and needs to be explored fur-

ther. Babur (1504-1530), who stated the revenues of

Kabul at 320,000 rupees, claimed that the bulk of it came

from transit duties (1969:221,250). Jahangir (1605-1627),

 290

at the beginning of his reign, affirmed that transit du-

ties were "in fact the chief revenue" of the provinces

of Kabul and Kandahar (Tuzuk-i-Jahangiri,tr.by Beveridge,

1968,vol.I;47). Jahangir claimed that he had forbidden

the imposition of all tolls and other burdens which "the

jagirdars of every province and district had imposed for

their own profit" (Ibid:7); but the testimony of contem-

porary English travellers reveals that the measure was

ineffective (Steel and Crowther,1616/1965,vol.IV:266-277).

with the re-emergence of regional units costs of protec-

tion imposed on caravans, most probably, must have signi-

ficantly increased (Habib,I.,1963:68).

The available information on the finances of the

three empires is not detailed enough to allow a breakdown

of the respective shares of taxes on land and trade in

the total income of ~he empires across time. Without im-

plying any causal connection; it is intriguing to notice

that the reigns of the most outstanding Uzbeck, Mughal,

and Safavid rulers -- Abd Allah Khan (1557-1598); Akbar

(1556-1605); Shah Abbas I (1588-1629) -- corresponded

with the period of flourishing intercontinental trade and

that all three took vigorous measures for the promotion

of trade, illustrated by the building of caravan sarais

and ensuring the security of the roads.

Indeed, in both the Mughal and Safavid empires, it

became established practice that the officer in whose

 291

jurisdiction a robbery or theft of merchants or residents

took place, had to recover it or pay compensation to the

victims out of his own funds (Habib,I.,1963:68; Monshi,

1978,vol.I:527). This required the greater involvement

of officials with populations located along the trade

routes. My analysis, in the next chapter, of the social

structure of the Afghan clans within the context of the

larger political and economic forces will serve as an

illustration of the intricacies of this interrelationship.

 292

CHAPTER SIX

The Afghan State: Formation and Reproduction

In an essay following the publication of his clas-

sic analysis of the inner Asian frontiers of China (1940),

Owen Lattimore examined, from a comparative perspective,

the question of the interrelationship between agricultur-

al and steppe societies. "In India," he argued, "the

picture is not so stylized as i t is in China. In North-

west India, which is not bordered by a steppe plateau

but encircled by mountains, the irrigated and unirrigated

lands are not set off from each other in large blocks

but interpenetrate each other in a rather complicated way.

Moreover the adjacent pastoral economy, as in Afghanistan,

is more often a semipastoral economy, associated with

fixed villages and village institutions, than true no-

madism of the open steppe. In Iran and Iraq the contrast

between the steppe and the sown land is more stylized,

as in China; but the irrigated agiculture is partly of

an oasis type and partly strung out along rivers that

have steppe or desert on both sides and consequently lacks

the scale and mass that are so impressive in China. In

Turkey the agriculture is more massive, but there is more

of the semipastoral economy that is linked with villages,

and less open steppe nomadism" (1947/1962:148).

Unlike those approaches which interpret nomad-seden-

 293

tary relations as containing "modes of production," inher-

ent opposition, or symbiosis, Lattimore's emphasis on the

engulfing ecological and political processes alows us to

investigate the historically changing patterns of inter-

relationships. To elucidate the structures of regions

and frontiers of states, he advocates that the respective

geographical range of unification by military action, cen-

tralization under uniform civil administration, and eco-

nomic integration be analytically distinguished and con-

cretely investigated. He argues that, prior to the

"Great Navigations that opened up the oceans at the end

of the fifteenth century," the following generalization

was applicable to the old world as a whole:

"a) The radius of military action was greater than

that of civil administration. An inner radius reached

over territories that could, after conquest, be added to

the state, and an outer radius reached into territories

that could be invaded, with profit in plunder or tribute,

or for the purpose of breaking up barbarian concentra-

tions dangerous to the state, but that could not be

permanently annexed.

lib) Civil administration tended to be stronger re-

gionally than nationally. The solidity of the state was

closely associated with the fact that similar administra-

tions duplicated each other in region after region .•.

"c) Economic integration had the shortest range. It

 294

was a function of the ability to transport bulk goods

(especially food) at a profit. For many centuries, and

not only in China, this meant that where transportation

by water was available within an empire, linking several

regions to each other, the empire that could be built was

larger and more stable than where transportation depended

in the main on carts, pack animals, or human porters"

(1956/1962:480-81) .

My analysis in the preceding pages of the economic

and political articulation of regional and central struc-

tures serves as a confirmation of Lattimore's generaliza-

tion. It should be noted, however, that he carefully

limits the applicability of his observations to the sit-

uation prevailing prior to the emergence of the cosmopoli-

cal economy of Europe in the sixteenth century. During

the following four centuries, there was hardly any part of

the world that was spared the destructive and possibly,

transformative impact of that expanding system. At the

time of the foundation of the Afghan state, 1747, the ma-

jor impact of the cosmopolitica1 economy on the area man-

ifested itself through the shift of trade routes from

land to sea. Understanding the genesis of the Afghan

state, therefore, requires grasping the complex interplay

of forces between changing state structures and the no-

madic and sedentary populations in the area.

The categories of clan and lineage will be frequently

 295

used in the next pages; since their usage in the anthro-

pological literature is not uniform, I follow the standard

social anthropological distinction. This is exemplified

by Morton Fried who singles out demonstrated descent as

a characteristic feature of lineages and stipulated des-

cent as that of clans. In the case of lineage "the gen-

ealogical basis of all relationships within the group is

explicit, i.e., connecting links between tertiary or more

distant kin can be identified," but in the case of a clan

"the genealogical basis for relationships is implicit,

i.e., connecting links between tertiary or more distant

kin are often unknown and cannot be identified, but are

assumed to have existed. Even in cases where the con-

necting links are known and can be given, their presence

as links is sociologically unimportant, relationship be-

ing a categorical and not a geneaological function" (1957:

23). Fried notes that "both kinds of groups can exist

at the same time in the same society. However, there are

societies in which one or the other predominates or is

the sole kind of organization at the appropriate social

level. This is said to be related to the divergent func-

tions of stipulated and demonstrated descent. The former

accords with easier access to corporate holdings of the

group; the latter tends to be associated with increasing

use of the principle of economic scarcity and the narrow-

ing of rights of access to basic resources" (1968:371).

 296

Following Paul Kirchoff's lead in his analysis of the

"conical clan" (1935/1968:371-81), Fried links the emer-

gence of corporate unilineal descent groups to the de-

velopment of social stratification. His notion of cor-

poration, which he defines as "continuity of possession

to an estate which consists of things, persons, or both"

(1957:23), is narrower than the usage of jurists such as

Saviqny who stressed sanction by the state as well as con-

tinuity as the essential elements of corporation (1884:

177-201) .

Since Fried is concerned with the evolution of stra-

tification and "pristine" states, his discardi~g of the

criterion of sanction is understandable. In the context

of secondary states where the formation and reproduction

of clans and lineages is intimately linked to the mili-

tary, political, and economic politices pursued by the

state, analysis of conditions which affect the corpor-

ate continuity of lineages and clans cannot be avoided.

Indeed, Fried persuasively argues that the political and

economic actions of states on populations living in their

periphery can lead to the formation, or even con~c:ous

creation, of secondary tribes (1975:101-102). Once these

tribes are in existence, the development of lineages, re-

alignments among clans, and the degree of corporate con-

tinuity in their social organization, can be drastically

affected by their interaction with states.

 297

The record of relations between Afghan clans and the

states ruling the area between the sixteenth and eigh-

teenth centuries offers many empirical examples for the

theoretical postulates. It is time to take brief account

of some instances. Limitation of space has forced me to

confine my remarks to the Pashtun clans.

The ethnogenesis of Pashtun clans, as is the case for

most other clans in the area, is obscure. Professor

Georg Morgenstierne, who was an authority on the linguis-

tic history of the area, believed that the Pashtu langu-

age was derived from Saka, a language spoken by a central

Asian people who conquered the present territory of

Afghanistan in the second millenium B.C. He pointed to-

wards the Sulaiman range, in the eastern region, as the

place where the hammering out of common identity by the

Pashtuns took place (1975:4). Pashtun folklore, on the

other hand, claims a Jewish origin and the mountains of

Ghur, in the central region, as their original location.

There is no historical evidence to support this assertion

and, as argued by Sir Olaf Caroe, its advocacy might be

related very strongly to the desire for a monotheistic

pedigree (1958:10). The first historical references to

Pashtuns occur in the tenth century, as documented by

the researches of Ghulam Muhammad Ghubar, an Afghan his-

torian. At the time they seem to have been inhabiting

the Sulaiman mountains and their emigration from these

 298

mountains began in the eleventh and twelfth centuries

(1968 :308-10) .

It is not the divergence of these views but their

convergence on two elements that I wish to emphasize. In

all views, origin of the Pashtuns is postulated at a time

when states already existed in the area. Furthermore,

the posited place of origin is always in a peripheral

area, movements to and from which are related to acts of

states.

Prior to the sixteenth century, a main avenue of

contact of Pashtuns with various states had been their

recruitment in the armies of rulers bent on conquering

India. The gradual rise of some men to positions of in-

fluence in the state culminated in the establishment of

a Pashtun dynasty that ruled northern India from 1451 to

1526.· Bybverthrowing this dynasty, Babur, a Mughal

prince who has been ousted from central Asia by the Uz-

becks, founded the Mughal empire. But before the con-

solidation of the Mughal empire under Akbar (1556-1605),

there was an interlude between 1540 and 1555 during which

another Afghan dynasty, again rising through the army,

ruled India. Until the reign of Shah Jehan (1627-1658),

Pashtuns were prominently represented in the ranks of

Mughal nobles; their influence seems, however, to have

steadily declined from then onwards (Rahim,196l:284).

The character of relations maintained by these nobles

 299

with their clans of origin is not clear but their impact

on the structure of clans in the present territory of

Afghanistan was probably minimal.

The same cannot be said of the impact on Pashtun

clans of the three empires -- Safavid, Mughal, and Uzbeck

-- that were formed during the sixteenth century. Since

reliable ma.terial from this time orL\lards has become avail--

able, it is possible to analyze the processes of inter-

action between these empires and the Pashtun clans. I

shall pay particular attention to changes in the internal

structure of clans and lineages that resulted from these

interactions.

Before proceeding any further, it is essential to

describe the terms through which the Pashtun themselves

refer to their social organization. Of the four terms,

kahul, zay, khel, and qaum, which can be translated re-

spectively as household, lineage, clan, and tribe, only

the term for household is unambiguous. Zay and khel are

suffixes to proper names but can mean either lineage or

clan. Qaum is a more abstract term and its reference to

lineage, clan, or group of clans, is situational, and

an inquiry may elicit anyone of them. The reason for

this seeming confusion between lineage and clan is, I

think, fairly simple. As Pashtuns theoretically accord

the same importance to stipulated and demonstrated des-

cent, they have no cause to stress differences which

 300

they have glossed so carefully through the manipulation of

genealogy. Furthermore, despite the situational meaning

of the terms, Pashtuns do have in fact a way of separating

lineages from clans. An individual has to know the names

of his seven ascending ancestors and there is a specific

term designating each one of the seven descending gener-

ations. That each of these lines is not formally con-

verted into a named lineage is a reflection of the fact

that the process of naming, as mentioned earlier in the

case of the Uzbeck ruling lineage, may be strongly linked

to control over strategic resources.

This practice is not limited to the Pashtuns. Abd

Allah Hammoudi, who offers a cogent critique of Ernest

Gellner's application of "segmentary theory" to Morocco,

affirms that "tradiation demands for certain groups of

Ait-Atta that, to establish 'good birth' seven ancestors

must be mentioned. Others are less demanding •.• this

criterion immediately excludes from the race for prestige

and power all imigrants and strangers" (1980:286). He

calls attention to the analysis of conditions that deter-

mine the choice of those ancestors who identify social

segments as well as discerning the reasons why common re-

ference for a group is not always a person's name (Ibid:

282,299). Hammoudi's emphasis on the details of history

is based on his argument that Gellner "attempts a histor-

ical reconstruction by brushing aside all history" (Ibid:

 301

298) and his insistence that "a careful appreciation of

history can avoid overly systematic interpretations which

force reality" (Ibid:300).

Hammoudi neglects, however, to draw another important

conclusion regarding segmentary theory, namely the un-

founded character of the assertion that structural equiva-

lence exists between segments in stratified societies. Indeed,

Maurice Freedman demonstrated in the context of south-

eastern China that "if one sub-lineage was richer than

other sub-lineages, the degree of segmentation within was

likely to be greater ... In such a system segments could

not always stand balanced against other segments of like

order within a more comprehensive segment H (1958/1970:49).

The difference between Freedman's analysis of segmenta-

tion and that presented for example, by Paul Bohannan

can be best gathered from the following diagrams.

Paul Bohannan's Diagram

~1
 302

"The lineage whose apical ancestor is some three to six

generations removed from living elders and who are asso-
ciated with the smallest discrete territory (tar) I call
the minimal segment . . . it can vary in population from
200 people to well over a thousand. . . . The territory
of a minimal segment adjoins the territory of its sibling
minimal segment. Thus, the lineage comprising two minimal
segments also has a discrete territory, and is in turn a
segment of a more inclusive lineage, and of its more in-
clusive territory. In Fig. 1, the whole system can be
seen: the father or founder of segment a was a brother
of the founder of segment b. Each is a-minimal segment
today, and each has its own territory. The two segments
taken together are all descended from 1, and are known
by his name -- the children of 1. In the same way, the
territory of lineage 1, made up as it is of the combined
minimal territories a and b, combines with the territory
of lineage 2, made up of the "combined minimal territories
of c and d,-to form territory A, occupied by lineage seg-
ment A, all descended from a single ancestor "A." This
process is extended indefinitely right up to the apex of
the genealogy, back in time to the founder who begot the
entire people, and outwards in space to the edges of
Tivland. The entire 800,000 Tiv form a single "lineage"
(nongo) and a single land called Tar Tiv. The geograph-
ical position of territories follOWS the genealogical
division into lineages" (Bohannan,1954:3).

Maurice Freedman's Diagram

(1958/1970:49)

A __________________________________ ~B

Bl
A A2
Bla

The organizational principles of structures portrayed

in the two diagrams are quite different. One of these

principles for societies like Tiv is that of complementary

opposition between segments. E.E. Evans-Pritchard, wri-

ting about the Nuer, described the idea through the fol-
 303

lowing diagram.

Evans-Pritchard's Diagram

A B
x Y

Xl Yl

Zl
X2 ----------
Z2
Y2

"When Zl fights Z2 no other section is involved. When

Zl fights Yl, Zl and Z2 unite as Y2. When Yl fights Xl,
Yl and Y2 units, and so do Xl and X2. When Xl fights A,
Xl, X2, Yl, and Y2 all unite as B. When A raids the
Dinka A and B may unite" (1940/1977:143-44).

Marshall Sahlins, who saw segmentary lineage organi-

zation as a means for the predatory expansion of one

tribe "moving against other tribes, in a tribal inter-

cultural environment" (1961/1967:91, emphasis in the

original), argued that opposition created the structure.

"Tiv-Nuer lineages do not corne into existence except

through the massing effect, in opposition to equivalent

groups. They are not permanent, absolute social entities,

but relative ones" (Ibid:105).

As claims regarding the segmentation of Tiv and Nuer

peoples were made on the basis of informants' ideal model

of social organization and without detailed analysis of

historical information, the degree to which they repre-

 304

sent a given reality remains in doubt.

In south-eastern China, clans were not functioning

in a "tribal intercultural environment" but in the con-

text of a relatively centralized state. Segments were

not structurally equivalent, as Freedman demonstrated,

and Sahlins' claims of complementary opposition and struc-

tural relativity leading to automatic collective action

(Ibid:103) cannot even be formally made.

My reasons for considering segmentation theory are

related to the fact that Pashtun clans have described

their own social organization in such terms. Moreover,

during my fieldwork between 1973 and 1977, informants

repeatedly asserted that in making alliances Pashtun

clans adhered to a segmentary model. Yet, ample histori-

cal examples show that the model has, in fact, been fre-

quently disregarded and even that levels of segmentation

have been the outcome of state policy.

The contact of state(s) with clans did not need to

be continuous for the impact on clan organization to be

significant. Before his conquest of India in 1526, Babur

who, from 1507 onwards, ruled in the eastern region and

parts of the southern region of present day Afghanistan,

led frequent expeditions against Pashtun clans in these

regions and in the area that forms today the north-western

frontier province of Pakistan. Most of these expeditions

were in search of loot and did not aim at subjugating

 305

people to administrative or economic integration. Pashtun

clans seem to have recovered from the successive blows

though, in most instances, the number of casualties suf-

fered by clans -- which Babur records as number of skulls

used to erect pillars to victory -- were relatively large

(Beveridge,trans.1969:232-33,234,etc.). These losses

must have affected the structure of clans as well as the

relative position of various segments to each other.

There is, however, no information from the time of Babur

apart from an account of an encounter between the army of

the Persian ruler Nadir Shah Afshar (1736-1747) and a

Pashtun clan. Since its impact on the internal organiza-

tion of that clan have been recorded, I will discuss this

encounter.

In 1879, when the British officer who gathered the

information visited the Shorawak district, the Barechi

clan was divided into four named sections: Mandozai (1500

fighting men; 5 sub-sections); Zakozai (1500 fighting men;

4 sub-sections); Badalzai (400 fighting men; 3 sub-sec-

tions); and Shirani (300 fighting men; 3 sUb-sections).

It can be readily seen that even in sections with the

same number of fighting men, the number of segments was

not equivalent. Moreover, all sections did not claim the

same origin. Three claimed descent from Barech but the

fourth "section" [was] a fragment of another tribe, but

now completely assimilated to the other sections" (GAt

 306

Kandahar: 89) . In 1883, the khan (chief) of all four sec-

tions was the leading man of the Mandozai section which,

in the light of the strength of that section, is not sur-

prising. This had not always been so and can be traced

back to encounter between the Barechis and the army of

Nadir.

In 1737, when Nadir, after defeating the Ghilzai

Pashtuns who had conquered Iran in 1722 marched on their

seat of government in Kandahar, the Barechis opposed him.

"To punish them, Nadir Shah despatched his General, Kamran,

with a large force. The bulk of the Barechis retreated

at once to the mountains, but the Badalzai .•. remained in

their homes ..• They were surrounded and surprised by

Kamran, ... and the greater portion of the people put to

the sword. Only a few hundred escaped to join the other

sections. The ruins round Badalzai [village] still tes-

tify to the extent of the catastrophe. Up to this time

the Badalzais were not only the eldest, but the strongest

section of the Barechisi and their village the largest in

Shorawak. Since then they are the weakest, excepting the

semi-alien Shiranis, and the village, though large, is

not a third of the size of the Mandozai" (Ibid:45l-52).

A contemporary account of the event estimated the number

of the inhabitants of Shorawak who were killed in the bat-

tle at six hundred (Mirza Mehdi,n.d.:174).

Despite the change in balance within the Barechi

 307

clan, its association with the territory of Shorawak was

not altered. This was not always the case and migration

of clans often occurred as a result of confrontation with

state power.

Some time between 1698 and 1709, the Safavid gover-

nor of Kandahar forced the Abdali clans out of their lands

in the vicinity of the city (Tarikh-i-Sulatni,1880:69).

The clans first moved to the districts of Shorawak and

Farah, northwest of Kandahar. In 1709, Safavid rule in

Kandahar was overthrown by the Ghilzai clans. The Abdalis

were then persuaded by the Safavid governor to join

Safavid troops sent against the Ghilzais. The failure of

the expedition made the Abadlis aware of the weakness of

the Safavid empire; they in turn conquered Herat in 1717,

proclaiming it an independent state, and maintained their

control until they were defeated by Nadir Shah Afshar in

1732. Some 60,000 families of the Abdali clans were then

forcefully made to move to the province of Persian

Khorasan (Koudusi,1960:548, reproducing Nadir's original

edict). A good number of Abdali men were, however, re-

cruited into Nadir's army and in return for their loyal

services, when Nadir vanquished the Ghilzai in 1738, he

qranted them the lands of the Hotak Ghilzai clan in the

district of Kandahar. The ruling house of the Hotak clan

was banished to the province of Mazendaran and the clans-

men to the province of Persian Khorasan (MirzaMehdi,n.d.:181).

 308

Relocation of population was not confined to Pashtun

clans. After conquering Herat in 1738, Nadir forcefully

moved some of its Persian-speaking population as well as

the Jamshedi clan to his capital city of Mashad and other

districts (Ibid:55). In their place, he installed "sixty-

thousand families" from his own clan of Afshar (Koudusi,

1960:111). Nadir Shah seems to have undertaken forced

migration on a massive scale, as illustrated by map 11.

John Perry, who drew the map, argued that Nadir's prin-

ciple aims were the recruitment of manpower for his army,

the fragmentation of recalcitrant clans, and increasing

the available manpower for production in the region of

Khorasan where his capital city of Mashad was located

(1975:208-209).

If Nadir was able both to force populations out and

to decide the place of their relocation, other rulers were

not always as successful in determining the destination

of expulsions. The example of the Yusufzai clans is a

case in point.

In his narrative of events for the year 1586, Abul

Fazl Allami, an outstanding scholar of the reign of Akbar

(1556-1605), provided some valuable insights into the pat-

tern of movement of the Yusufzai clans. According to his

account, the Yusufzais used to live in Kandahar and

Karabagh. In the fifteenth century, they were forced to

move to Kabul where, shortly after, they were attacked by

 309

Mirza Ulugh Begh (d.1501), the ruler of Kabul. Their re-

maining members moved again to the district of Lagr~an in

the eastern region, and from there to Hashtneghar in the

vicinity of Peshawar, then again to neighbouring Swat and

Bajaur. "It is now close to one hundred years that they

have spent their time in Swat and Bajaur in robbery and

rebellion," wrote Abul Fazl who estimated their numbers

at 30,000 in Bajaur and 40,000 in Swat (Akbarnarneh,vol.III:

475,481-482) .

Prior to the conquest of these territories by the

Yusufzais, the area was ruled by a group of people who

claimed descent from a daughter of Alexander the Great and

gave their ruler the title of Sultan (Ibid:475). Babur's

account of his conquest of Bajaur in 1519 -- during which

3,000 male inhabitants were executed confirms the ear-

lier existence of Sultans of Bajaur and Swat. He also

mentions that when Yusufzais took over Swat, which up till

then was considered a dependency of Kabul, they stopped

paying revenue to the ruler of Kabul (Beveridge trans.,

1969:370-375;207-221) .

The process through which the Yusufzais gained con-

trol over Swat and Bajaur is not clear; events in 1593

may provide a hint. In that year, under strong pressure

from Mughal armies, some Yusufzais were forced out of

Swat and Bajaur. They, in turn, pushed on to a new dis-

trict which, according to Abul Fazl, was inhabited by

 310

kafirs (non-Muslims, lit. infidels) and managed to consoli-

date their hold there (Akbarnameh,vol.III:625).

Expansion into the land of so-called kafirs was not

an isolated phenomenon. An account of a series of wars

between Muslims and kafirs in the valley of Laghman, in

eastern Afghanistan, originally written in 1582, documents

the establishment of Muslim settlements in the area and

the forceful conversion of kafir people to Islam (Sifat-

Nama-i Darvis Muhammad Khan-Gazi,Persian text,Rome ed.,

1965:1-75). Since the sixteenth century and until 1896 when

the area of Nuristan, as shown on map 12, was conquered

by th8 Afghan state and its inhabitants converted to Islam,

there seems to have been continuous expansion of Pashtun

clans and other Muslim groups into kafir territories, con-

version of kafirs to Islam, and probable incorporation

within the clan structures. As illustrated by the inci-

dent of 1593, the impetus for such movements may have

come from the policies of pacification and incorporation

pursued by the surrounding states.

In his perceptive remark on the encapsulated charac-

ter of nomadism in Afghanistan, Lattimore noted that

large concentrations of settled population and areas of

habitation of nomadic and semi-nomadic groups interpene-

trated each other in a complex fashion. The spatial lo-

cation of irrigation agriculture and pastures was such

that the seasonal movements of nomadic populations between

 311

their summer and winter pastures inevitably brought them

in close proximity to the areas where they were vulner-

able to attacks from state armies. Babur gives frequent

examples of his forays against Pashtun clans. The scale

of losses suffered by nomadic clans can be gathered from

the results of a raid carried out against the Ghilzai

clans in 1507. Counting the spoils, Babur stated his

own share, which was a fifth of what was taken, "came out

at 16,000, that is to say, this 16,000 was the fifth of

80.000 sheep; no question, however, that with those lost

and those not asked for, a lakh (100,000) of sheep had

been taken" (Beveridge trans.,1969:325). Such plundering

expeditions seem to have forced most nomadic clans to

agree quickly to paying some kind of tribute (Ibid:413).

From the records of the revenue of the state provided by

Abul Fazl for the central places of Kabul and Kandahar

and their dependent districts, it seems certain that, by

the reign of Akbar, most nomadic clans in these areas

were paying taxes in kind and/or money (Ain-i-Akbari,

Gladwin trans.,vol.II,1800:304-307).

Similarly, the large-scale raidings undertaken by

the Uzbeck state may have been a strong factor in forcing

Pashtun clans into seeking protection from the Mughal and

Safavid states. Furthermore, the extensive economic in-

terconnections between nomads and the inhabitants of "cen-

tral places" gave some nomadic clans a common interest

 312

in those policies of the state that aimed at the esta-

blishment of order and security for long-distance trade.

Most central places had a special marketplace where

animals were sold and, as data for subsequent periods in-

dicates, nomads must have been the major patrons of such

markets. Nomadic women, who combined herding and domes-

tic manufacture, must have sought there an outlet for

their products -- mainly carpets and rugs. The extensive

development of manufacture of luxury carpets in urban

workshops in Persia and India in the sixteenth centuries

would not have been possible without the supply of wool

by nomads.

The main economic functions performed by nomads, how-

ever were those of breeders of transport animals and

carriers in long-distance trade. Data from the end of

the eighteenth century clearly indicates that, although

merchants kept their own camels and organized their own

caravans, nomadic clans played the main role in long-dis-

tance trade by hiring out their camels to other trading

clans as well as engaging in direct trade (Elphinstone,

1832,vol.I:378-382). This picture, as I will show later,

is confirmed for the nineteenth century as well. Indeed,

the Lohani clan, from the time of Babur on (Beveridge

trans.1969:235), is continuously mentioned as a major

trading group between Afghanistan, India and central Asia.

Passing in slow stages through the districts lying be-

 313

tween the major central places, these trading clans ful-

filled the functions of a regular seasonal market for the

settled population (e.g., Vigne,1840:113).

Relations between the trading nomadic clans and the

groups through whose territory they passed were not always

peaceful, even though at times they belonged to the same

ethno-linguistic group and maintained regular contact.

Indeed, during the eighteenth and nineteenth centuries,

some of the most intense fighting took place between the

Lohanis and another Pashtun clan, the Wazirs, through

whose area the Lohanis had to pass on their way to and

from India. The reason for the conflict was the Lohani's

refusal to pay the Wazirs for cost of protection (Ibid:

78). As they did pay the powerful Sulimankhel clan for

passing through the plain of Katawaz (Ibid:lll) and every

other regional power from India to the borders of Russia

(Ibid:78-79) , the Lohanis must have considered the Wazirs

not too serious a threat at the beginning. But to effect

their passage through the Wazir mountains, they were "com-

pelled to move in large bodies of from 5,000 to 10,000

and regular marches and encampments [were] observed, under

an elected Khan or leader, exactly like an army moving

in an enemy's country" (GA,Kabul:426).

Excessive demands for costly protection on the part

of empires and other powers were met with the same re-

sistance. Around the beginning of the eighteenth century,

 314

the Nawab of Dera Ismail Khan -- now part of the north-

west frontier province of Pakistan -- engaged the Lohanis

in battle but lost 400 men and some pieces of canon a-

gainst 250 men for the Lohanis. He was forced to retreat

and, as a result, duties paid by the Lohanis "were re-

duced from nine to six annas a maund for their cloth goods,

from eight to six rupees for every sale camel, and from

five to three for every sale horse; and the impost levied

at Derabund was reduced from one thousand four hundred,

to one thousand one hundred, rupees a-year" (Vigne,1840:

56-57). Thus, control over means of violence was always

an important variable in the relations between states and

nomadic clans as well as between these clans and their

more sedentary kinsmen.

As for sedentary clans, their relations with states

were also conditioned by their location, that is whether

they inhabited an econologically central or marginal zone,

and how distant they were from the main arteries of long-

distance trade. Clans whose lands were within the vicin-

ity of a central area or within zones on which a central

area depended for its supply of food, were bound to ex-

perience strong pressure from wielders of political power

-- whether provincial governors of a larger state or re-

gionally autonomous rulers -- for payment of tribute.

Thus, when Babur took over Kabul in 1507, in order to

provide for his central Asian followers, he imposed a

 315

grain levy of 30,000 ass-loads on the people of Kabul,

Ghazni and "their dependencies ... the impost was exces-

sive, and under it the country suffered very grievously"

(Beveridge,trans.,1969:228). As these were areas where

production for the market was regularly undertaken and

circulation of land as commodity widely practiced, the

economic and political fluctuations of the state must

have been closely felt by clans settled there.

The policies of states towards areas with an inter-

mediate scale of production were dictated by the charac-

ter of the state, the degree of ease with which military

operations could be conducted, and the location of these

areas in relation to major routes of communication link-

ing the central area of various regions. For a regional

state, such as that of Babur in Ka~Hl, it was profitable

to keep its forces continually busy and find new append-

ages for the displaced central Asian aristocrats, a state

with a huge territory under its control, such as the

Mughal state under Akbar, found it easier to come to

terms with the local population. Imperial policy towards

security of travel and communication within the state, as

well as with areas within other states, took a very dif-

ferent form from that of regionally confined power-hold-

ers. Raiding traders and imposing a special tribute on

them were normal practices with Babur (Beveridge trans.,

1969:235,331,etc.). By contrast, Akbar and his succes-

 316

sors spent enormous efforts to provide traders and other

travellers with security of travel. In the context of

these efforts, a most important aspect of the relations

between the Mughal state and Pashtun clans was worked out.

From the sixteenth century up to the end of the nine-

teenth century, our sources indicated that various Pash-

tun clans -- though not always the same in the same places

-- were located on the main routes between India and Iran

and central Asia. More specifically, the number of moun-

tain passages through which an army or a caravan could

move to and from India were limited to the passes on which

Pashtun clans lived. Although most of these areas were

in ecologically marginal zones, successive states had to

adopt coercive policies towards the clans inhabiting them

and cultivate among them a relation of dependence, in or-

der that routes remain open to travellers.

Contol of the Khyber pass, one of the easiest pas-

sageways between Afghanistan and India, has had, since

the time of Akbar, a high priority on the part of all

states ruling the area. Before the construction of a

paved road in the twentieth century, the length of the

pass was estimated at 28 miles. Its description in 1878

closely echoes earlier accounts. Major C.J. East wrote

that lithe road is chiefly the bed of a torrent, subject

at certain seasons to be suddenly flooded; at first it

is flat, covered with loose stones and gravel, and en-

 317

closed between perpendicular hills; in the middle portion

it is narrow, winding and steep, being in parts much cut

by protruding pieces of rock, &c.; lastly it consists of

a made road down the side of a steep hill ... With the

exception of the valley of Lalabeg, which is six miles

long, the remainder of the pass, or 22 miles, is commanded

by common musket shot, the hills on either side varying

from 200 yards to 50 feet in distance apart" (L/P&S/18/

C71:10). Clearly, closing the pass to traffic was an

easy venture.

Abul Fazl, while mentioning the fact that Akbar ex-

tended the route (Akbarnarneh,vol.III:487), recorded a

number of instances in which it was closed by Pashtun

clans -- to whom he referred by the more common name

of Afghans (Ibid:473,5l2,702). Elsewhere, too, Pashtun

clans interfered with the flow of traffic. Abul Fazl

stated that "from Kabul to the river Indus, various

groups of Afghans live and every now and then because of

ignorance and self-interest, they harm the travellers and

oppress the hapless" (Ibid:5l9). To control the activi-

ties of Pashtun clans, Akbar ordered the construction of

eleven forts along the road from Kabul to Attock on the

Indus. This was done within a short time and the forts

were manned by soldiers (Ibid). Mughal armies were also

sent on several plundering expeditions against the clans

of the Khyber until those clans, in sign of submission,

 318

gave hostages to the state and accepted collective respon-

sibility for the safety of the Khyber (Ibid:513,521,531).

Similarly, Abul Fazl related the campaigns of Mughal armies

against the Yusufzai clans in Swat and Bajaur, as well as

the execution of a number of them and the sale of others

as slaves in Iran and central Asia, to the interference of

these clans with the safety of travel (Ibid:486,708).

With the decline of the Mughal state at the beginning

of the eighteenth century, Pashtun clans on the Khyber

regained their power and managed to impose fees for going

through the pass on subjects of all states, including the

British government of India.

There seems to be a fairly close relation between

proximity of location to Padshah-i-Lara (king's road) and

the rise of aristocratic lineages among clans. Indeed,

the rise of such lineages among the Khattak, Mohmand,

eastern and western Ghilzais, and the Abdali clans can

be linked to the policies of the Mughal and Safavid

states.

In 1586 and 1589, a Khattak by the name of Akory and

an Abdali by the name of Sado were appointed, respective-

ly by Akbar and Shah Abbas I, to watch the safety of the

highroads between Attock and Peshawar and between Kanda-

har and Herat (Caroe,1958:212,223). No reliable account

of Sado's appointment can be found and Caroe, who gives

1589 as the date of his appointment by Shah Abbas I, does

 319

not indicate the source of his assertion. As for Akory,

Afzal Khan, chief of the Khattaks at the beginning of the

eighteenth century and sixth patrilineal descendant of

Akrory, provides the following account of his selection.

"It is said [of Akory] that he was both very coura-

geous in the use of his sword and extremely generous"

(Tarikh ~urassa, 1974:265). Brought to the notice of

Akbar by some of his officials, that monarch "offered him

a mansab [rank] but malik [title used by headmen; original

meaning in Arabic is ruler] Akory was a wise man and re-

fused the mansab. He brought to the attention of the

king that if he, alone, received a mansab, the rest of

his clan would become jealous and the service of the king

would not be fulfilled" (Ibid). Instead, he asked that

the whole clan be made responsible for the safekeeping

of the highway and share in the tolls collected from

travellers. The king accepted this arrangement and Akory

was able to secure the services of the Khattak clan for

the king (Ibid). The distribution of shares between

Akory and the rest of his clansmen is not known but, in

addition to his income from tolls, he was given a jagir

by Akbar along the bank of the Kabul river until its

confluence with the Indus (Ibid; Caroe,1958:2l2).

Beyond his concern for the safety of the road,

Akbar's intention in grating privileges to Akory was to

build a local basis to counterbalance general Pashtun re-

 320

sistance to the consolidation of Mughal rule in the area.

His policy was eminently successful; until the reign of

Aurangzeb (1658-1707), the Khattak clan always sided with

the Mughal state against other Pashtun clans (Afzal Khan,

1974:253,268-74).

It is only with Akory's son that the title of khan

was first mentioned in Afzal Khan's narrative of the

Khattak clan. Major R. Leech, who spent from 1839 to

1842 as political agent in various parts of the southern

region of Afghanistan, states that the title of sultan --

an official rank in the Safavid state -- was first men-

tioned in reference to Sado's grandson in a manuscript

account of the Abdalis prior to their emergence as the

founders of the Afghan state in 1747 (1845c:458). A

Ghilzai contemporary to the Abdali sultan, who bore the

same title and belonged to the Tokhi clan, was also men-

tioned in the manuscript (Ibid). Indeed, of all amirs

alive in 1629, two Pashtun notables are listed as "(72)

Sir Khan Afghan, governor of Pusang (Fusanj), a depend-

ency of Kandahar [and] (73) Maqdud Sultan Afghan, sub-

ordinate to the governor of Kandahar" (Eskandar Monshi,

1978:1315). Maqdud Sultan cannot be easily related to

a clan, yet he might be the same man as the one referred

to in Pashtun as Sultan Malakhy Ghilzai. Leech reported

seeing an edict of a Mughal ruler, dated 1022 A.H. (1613),

which he incorrectly identified as Aurangzeb, appointing

 32l

Malakhy Sultan "to the charge of the highroad from Kalat

to Karatoo, (the former in the Tarnak valley, and the lat-

ter in the Arghandab,) to protect it from Hazarah robbers.

[Jahangir] no doubt found Malakhe the most powerful of

the Khilzye chiefs at enmity with Hazarasi as patronizing

an officer of his own creation at court, he no doubt

found very different from supporting a newly created

chief over his tribe" (1845a:3l0). During the same per-

iod, an individual by the name of Jabar from the Ahrnadzai

clan gained the position of khan of the eastern Ghilzais

in return for the protection of the road between Kabul

and Jalalabad (Ibid:3l0-l2). In his account of the early

Ghilzais, Leech makes an observation that is of wider

significance. He stated that "the Hotakees I suppose from

being removed from the high road were not required by

[Jahangir], and therefore remained unnoticed; that

monarch's sole object being to secure his communication

with Ghuznee, Cabool and Hindustan, and not being coveting

revenue from their Karazees [underground channels], and

almond orchards" (Ibid:310).

Mir Wais, a Hotaki clansman, after being appointed

khan of the western Ghilzais by the Safavid governor of

Kandahar, founded an independent state in Kandahar in

1709; under his son Mahrnud, the Ghilzai Pashtuns con-

quered the Safavid capital of Isfahan in 1722 thereby

putting an end to that dynasty.

 322

It is with the advent of individuals who emerged as

khans or sultans that aristocratic lineages, expressed in

Pashtu through the notion of Khankhel (lit. the lineage

of khans) make their appearance. At this point the dif-

ference between historically demonstrated and stipulated

descent becomes clear. Afzal Khan, for instance, traces

his descent through six individuals, covering a period

of about one hundred years, to Akory. But he only men-

tions seventeen individuals, about whom he provides no

information and whose existence cannot even be verified

from other sources, as his links to Qais, the putative

ancestor of all Pashtuns who allegedly lived at the time

of the Prophet of Islam. Each of the male lineal des-

cendants of an ancestor who first founded a Khankhel lin-

eage had, on the other hand, the potential for being con-

sidered as the beginning of a distinct sub-lineage. Whe-

ther he actually did become the founding ancestor of a

new segment was linked to the degree his name could serve

as justification for strategic resources.

Despite the patrilineal ideology, the matrilineal

link could also become significant in the determination

of segmentation. This fact is especially clear among the

Morchakhels, who formed the Khankhel of the Lalpura clan

of Mohmands which controlled the road between Jalalabad

and Peshawar. The lineage was named after Malik Morcha

"who lived during the reign of Akbar Shah and served at

 323

Moradabad returning afterwards to take charge of the

Moghul fort near Dakka, of which he and his sons appear

to have acquired hereditary command" (Merk,1881:11). His

son's grandson had two wives called Jahana and Araba. The

descendants of the former, called Jahana Kor (lit. house

of), never held the Khanship, while the Araba Kor was the

Khankhel (Ibid).

The financial base of khans was provided by tolls

on roads, jagirs in more centrally located areas, and,

possibly, stipends in cash from the state. Moreover, as

they furthered the cause of the state in their wars with

autonomous clans and participated as well in joint actions

with other forces of the state, they must have received

a regular supply of arms and the means of destruction from

the states. They made full use of these resources for the

expansion of their own territories and the increase of

their revenue as demonstrated by reports of constant war-

fare against their less well-endowed neighbours, whether

Pashtun oronon-Pashtun.

The financial security of a khan could be radically

affected, however, by actions of the ruler or the agents

of the state. These took the form of change of jagirs,

assigning profitable jagirs for higher revenue to other

individuals, and even interfering with the collection of

tolls (Afzal Khan:275-76,348,368,etc.). Indeed, it was

the resistance of Khushal Khan (the grandfather of Afzal

 324

Khan and the most renowned Pashtun poet and warrior) to an

edict of Aurangzeb forbidding the collection of tolls,

which brought about his imprisonment and subsequent strug-

gle against the Mughal state (Ibid:275ff).

Policies of the states aimed at checking the ambi-

tions of khans were helped by the inherent oppositions

within a Khankhel. There was no fixed order of succes-

sion to the khanship and, as power and financial rewards

of the office were chiefly concentrated in one person and

the direct group of his adherents, competition for the

office was intense. The main contestants were the des-

cendants of a khan as well as descendants of the colla-

teral branches. As the khan's appointment had to be con-

firmed by the state which could also remove him from his

functions at will, various factions in the Khankhel had

to cultivate relations with state dignitaries in their

provinces. The conditions imposed on an aspirant could

be stiff. When Khushal Khan, who was the father of fifty-

six sons, revolted against Aurangzeb, the terms for the

appointment of one of his sons as khan included the cap-

turing of his own father. This he agreed to undertake

but failed to achieve (Ibid:30l).

State support alone was not sufficient to underwrite

the authority of a khan. He had to command the loyalty

of a sufficiently large number of his clansmen to prevent

his father, brothers, and cousins from disturbing the

 ~­
3 -~

peace of the highway. Indeed, an obvious avenue for the

demonstration of the power of a deposed or aspirant khan

was his ability to disrupt the flow of traffic and this

measure was frequently resorted to; when the show was im-

impressive enough, it brought the desired end (Ibid:304,

316,etc.). The outcome of conflicts between khans and

states was not the reassertion of some sort of structural

equilibrium. It often brought about fundamental altera-

tions in the balances of power between the two parties.

And, depending on the location of his clan, a khan could

engage in relations with more than one state and thereby

help in determining the rapidity with which territory

changed hands between states. The example of Sher Khan

Abdali will serve to illustrate the point.

Abdul Hamid Lahouri, contemporary historian of the

reign of Shahjahan (1627-1658), reported that Sher Khan's

father, under pressure from the Mughal governor of

Kandahar, was forced to move to Iran where his son was

raised. When Shah Abbas I conquered Kandahar in 1621,

he appointedSherKhan as khan of the Pashtun clans of

Kushan (1867,vol.I:419). As Sher Khan is reputed to be

the grandson of Sado, founder of the Sadozai lineage and

an ancestor of Ahmadshah, the founder of the Afghan state

(Fofalzai,1967,vol.I,21), it can be presumed that he al-

ready led a faction among the clans. After the death

of Shah Abbas, relations between Sher Khan and the Safavid

 326

governor of Kandahar deteriorated, and Sher Khan started

to interfere with the flow of long-distance trade; "fre-

quenters of the road between Iraq and India could not tra-

vel to and from easily" (Ibid). The governor of Kandahar,

therefore, with 3,000 horsemen of his own and 1,000 horse-

men of the zamindars (landowners) of Kandahar, attacked

Sher Khan's fort while the latter was on a raid in

Baluchistan, and subsequently forced him to India. In

1603, Sher Khan presented himself to Shahjahan and was

appointed to the rank of 2,000 zat, 1,000 horses, and

granted a reward of 20,000 rupees in cash as well as a

jagir in Punjab (Ibid:420-42l). Shahjahan's generous

attitude towards Sher Khan was due to the fact that the

Mughal ruler had long planned the conquest of Kandahar,

which he eventually achieved in 1637. A local account

of Sher Khan's tribulations, which neglected to mention

his stay in either Iran or India, reported that Sher

Khan was forced to seek aid from the Mughal state when

the Safavids extended patronnage to a rival cousin. Sher

Khan seem to have remained in control of the khanship and

was succeeded by his son (Leech,R.,184c:458-462; Tarikh-

i-Sultani,188l:6l-64) .

In general, Safavid policy was less tolerant towards

the consolidation of hereditary leadership within a lin-

eage. As I have already indicated, the Safavid governor

of Kandahar managed to force the Abdalis out of Kandahar;

 327

he also replaced the old Khankhel of the western Ghilzais

by an appointee of his ow~ from the Hotak clan. The weak-

ness of the Safavid state, however, allowed the new

Pashtun leadership to put an end to it. Yet, the Ghilzais

failed to consolidate their power either in Afghanistan

or Iran, and Nadir Afshar, after ousting them from Iran,

conquered their stronghold of Kandahar. Nadir's attitude

towards aristocratic Pashtun lineages is best summarized

in the words of the court historian of Ahmad Shah, the

founder of the Afghan state: "In whichever clan where

there was a respected and powerful lineage or group he

would attempt to reduce their dignity in public. In their

place he would promote people without fame and distinction

so that considering their importance as derived from him

they would not think of disobeying him" (Mahmoud Monshi,

1974,vol.I:34). Thus, when Nadir subdued the Abdalis

in Herat, and demoted Ahmad Shah's father and brother,

he appointed one of their former chamberlains as khan of

the Abdalis and governor of Kandahar (Ibid:35). Nadir

did, however, select the young Ahmad as one of the com-

manders of the Abdali contingent of his army. This con-

tingent, numbering five to six thousands men was often

preferred to the qizil.bash soldiers. Upon the assassina-

tion of Nadir in 1747, his empire crumbled overnight and

the various contingents started to carve out states for

themselves. The Abdali contingent fought its way back to

 328

Kandahar and, after some deliberations, the Pashtun khans

agreed to the choice of Ahrnand Khan as their king.

What were the structural contours of this newly

formed state and how was it reproduced over our unit of

long duration? I shall attempt to provide an answer in

the remaining part of this study.

Within months of his election, Ahmad Shah (1747-1772)

led his army to the successful conquest of the towns of

Ghazni, Kabul and Peshawar, where the "Ayan and Izza"

(nobles and notables) carne to present themselves and "un-

dertook the payment of the rights of divan and the ob-

ligatory dues of the ruler" (Mahmud Monshi,1974:8). As

shown by map 13., he had, by the time of his death and

through the power of arms, imposed tribute over a large

expanse of territory.

He fought his memorable battle in 1761 on the fields

of Paniput near Delhi, where he decisively defeated the

newly risen Maratha power in India. Yet, when after his

victory he returned to his capital of Kandahar and at-

tempted to build a new city there, he was twice refused

the 12 qulbas, roughly 300 acres, that were needed for the

site, by two Durani clans that owned the land. The con-

struction could only start when the Popalzais, his own

clansmen, offered him the needed land (Faiz Muharnrnad,vol.

I,1913:25-26) .

These two events bring to sharp relief the external

 329

and internal aspects in the operation of the Durani state

during Ahmad Shah's reign. Both the internal and external

aspects of the relations of domination underwent radical

changes after Ahmad Shah's death. By concentrating on

four conjunctural periods, I shall examine changes in re-

lations between rulers and other political forces in the

state. Before doing so, I shall first take account of

the ethno-linguistic diversity that has been character-

istic of the people living in the present territory of

Afghanistan.

In contemporary Afghanistan, linguists have recorded

the existence of some thirty-two languages (A.G. Rawan

Farhadi,1977,personal communication). These can be traced

back to three main families, Iranian, 86 percent; Turkic,

13 percent and Dardic, 0.5 percent; others make up 0.5

percent (Humlum,1959:95). But people who have spoken kin-

dred languages or even the same language have identified

themselves through other attributes. Pashtuns, Tajiks,

and Hazaras, who are all speakers of Iranian languages --

the former speaking Pashtu and the latter two Persian --

have had considerably different im~ges of themselves and

others from that of members of a common language family.

The same holds true for speakers of Turkic languages, the

Uzbecks, Turkomans, and Qizilbashes. Even religion has

not been a factor of unity. While Pashtuns, Tajiks,

Uzbecks, and Turkomans have, in most cases, adhered to the

 330

Sunni interpretation of Islam, Hazaras and Qizilbashes

have generally followed the Shi'i interpretation. Non-

Muslim groups such as Hindus, Jews, Armenians and, until

their conversion in 1896, Nuristans -- commonly called

kafirs (infidels) -- have also coexisted at various per-

iods. There is no reliable data on the relative number

of these groups for our period of long duration. The

following two sets of figures J offered by Elphinstone and

British intelligence officials for the beginning and end

of the nineteenth century, can be taken as educated

guesses. Elphinstone's figures refer to the territory

of the Durani state while figures from the military re-

port refer to the present territory of Afghanistan.

Table I. population Breakdown at the Beginning of the Nineteenth

Century (Elphinstone,vol.I,1839:114)

Pashtuns 4,300,000
Beloches 1,000,000
Tartars [Turkic populations] 1,200,000
Persians (including Tajiks) 1,500,000
Indians (Casmeerees,Juts,etc) 5,700,000
Miscellaneous Tribes 300,000

Total 14,000,000
 331

Table II. population Breakdown at the End of the Nineteenth Century

(Military Report on Afghanistan, 1906:194-9)

Pashtuns 2,684,000 Qizil-Bashes 100,000

Tajiks 1,500,000 Kafirs 70,000
Uzbecks 750,000 Baluchis 40,000
Hazaras 516,000 Hindus 30,000
Char Aimaqs 245,000 Arabs 25,000

Total: five to six millions.

The political and economic role played by various

groups was not a reflection of their numerical strength.

The differences can be traced to the position of a group

in relation to the organization of production, circula-

tion, and domination. Since the mid-eighteenth century,

and despite the mobility of groups, the ethnic distribu-

tion was as follows. Hazaras were the main group in the

central region and Uzbecks and other Turkic-speaking

groups -- with the exception of the Qizilbashes -- the

major group in the north. Badakshan was for the most

part inhabited by the Persian-speaking Tajiks as well as

speakers of a number of other languages. "Kafirs" in-

habited the mountainous terrain of Nuristan. Pashtuns,

Tajiks, Hazaras, and various minor groups were inter-

mingled in the western, southern, and eastern regions.

At the beginning of the Durani state, the Persian-

speaking population, identified by the name of Tajik,

Parsiwan, and so on, was occupying the areas in or around

 332

central places in the western, southern, and eastern re-

gions while Pashtuns, in general, lived in more marginal

places. In intermediate places, there seems to have been

more of an intermixture.

Besides general remarks on the part of contemporary

historians and travellers, the predominance of the Per-

sian-speaking population in central areas can be quanti-

tatively demonstrated in one case. In a census of Kabul

-- capital of the state from 1775 onwards -- taken in

1876, the ethnic breakdown was given as follows (GA,Kabul:

221ff.):

Pashtuns: Duranis 3,000

Ghilzais 2,000
Safis 4,000

Persian-
Speakers: Kabulis 103,050
Tajiks 12,000

Others: Armenians 100

Hindus 4,000
Jews 50
Kashmiris 3,000
Parachas 3,000
Qizilbashes 6,500

Total 140,700

As people of Kabul usually identify themselves as Tajiks,

it is not entirely clear as to why the Afghan official

in charge of the census classed Tajiks and Kabulis in two

different categories. It is possible that through the

 333

category of Tajik he might have wanted to differentiate

the newly migrant from the older established inhabitants

of Kabul. By contrast to other major groups of popula-

tion -- Hazaras, Pashtuns, and Uzbecks -- , Tajiks were

not organized in clans. The continual presence of

state power in central places and the consequent sub-

ordination of the people living in these areas might ac-

count for the absences of clanties among Tajiks.

Unlike the Tajiks, who were mainly engaged in agri-

cultural production, the Hindus performed a major func-

tion in the sphere of circulation. In their capacity as

bankers, along with Armenians and Jews, they cashed bills

of exchange in all the central places of Afghanistan,

Iran, India, and central Asia. These bills of exchange,

which obviated the problem of carrying money in bulk as

well as the danger of robbery in unsettled times, were

extremely important in the functioning of long-distance

trade. Inside Afghanistan, their operations were not

confined to long-distance trade. In the western, south-

ern, and eastern regions, they formed the bulk of shop-

keepers in villages. Besides being the conduits in the

rural areas for products from central and foreign places,

they played, as grain merchants, an important role in the

circulation of agricultural products. Their function as

middlemen between producers and consumers of rural and

urban products was considerably facilitated by their ac-

 334

tivity as money-lenders.

The existence of Hindus as a well organized group in

Afghan society pre-dated the rise of the Abdali clan to

power. Ahmad Shah and the subsequent Abdali powerholders

made use of the banking and other services they provided

so that their position may not have undergone radical

change. Some Hindu bankers may have benefitted from their

connection with Abdali khans.

The same cannot be said, however, of the Tajik and

Hazara populations in the districts of the central place

of Kandahar. As this is one of the situations where

relationship to mode of domination directly affected the

relation of ethnic groups to the means of production and

to each other, I shall briefly examine the situation. The

word "Afghan," until the beginning of the twentieth cen-

tury, was exclusively applied to Pashtuns, and whenever it

appears in a quotation it should be understood in that re-

stricted sense.

Sir Henry Rawlinson, who served as political agent

in Kandahar during the first Anglo-Afghan war (1839-1842)

and, in that capacity, had access to state papers that

have since been lost, stated, in his report on the Durani

(names assumed by the Abdalis after Ahmad Shah assumed

power) clans, that prior to the conquest of Kandahar by

Nadir in 1738, "the land had been cultivated by a mixed

peasantry, composed of Parsiwans, Hazaras, Kakars,

 335

Baluchis, etc., with a small proportion of Afghan colon-

ists, all of whom considered themselves from long posses-

sion to have a right of proprietorship to the soil, and

who during the short period of Ghilzai sovereignty paid

their land tax and other duties to that Government at the

same rates to which they had previously been liable under

the Saffaveen monarchs" (1841/1980,GA,Kandahar:509).

Nadir, as a consequence of his conquest, claimed the

rights to the land had been transferred to the state and

was therefore entitled to bestow it to his Abdali soldiers.

In response to appeals from the peasantry, he divided the

lands into three categories. He reserved about 500 kulba

of the best land in the vicinity of the city of Kandahar

as khalisa (state land) and assigned it to Pariswan

(Persian-speaking) cultivators, according to a system

where half of the produce went to producers and the other

half to the state. On the basis of the provision of one

horseman per kulba, he assigned 6000 kulba to the Abdalis

in tiyul. Contiguous to every village, he assigned a

section of land, called ryoti (lit. that of subjects) for

the support of the former proprietors. These lands were

subject to taxation but supposed to have been free of

Abdali interference. As the officials who determined the

location of these lands were themselves Parsiwans, these

lands were among the most valuable. Rawlinson estimated

their extent at about one third of the lands alloted to

 336

the Abdalis (Ibid:509-5ll).

At first, the Abdalis employed the services of the

original owners for the cultivation of their lands, but

they gradually took over the management, and, being free

of taxation, they planted "gardens and vineyards, proper-

ty which always gives large return" (Ibid:5l0). In addi-

tion to the irrigated land, the Abdalis had been granted

considerable land in the valleys of the Kadanai, Do:i,

Afghastan, and Tarnak rivers which, due to insufficiency

of water, were called khushkaba (waterless). For the

Abdalis who maintained a total or partial nomadic style

of life, these valleys must have provided an important

source of pastures. Prior to their explusion from

Kandahar by the Safavid state, the Abdalis had moved be-

tween Kandahar and the valley of Toba and Zamindawar in

the northwest and Karabagh, Ghazni, and Kabul in eastern

region (Mahmud Monshi,vol.I,1974:3l). Unfortunately,

neither Rawlinson not any other source provides any in-

formation as to the location of the Abdali summer and

winter pastures and the routes followed by them.

Rawlinson does, however, mention the fact that the val-

leys of Kandahar provided an important outlet for the

expansion of Abdali lands through their investments in

karezes [underground irrigation channels] and, because

of their privileged position under Ahmad Shah, the Abdali

clans "spread themselves over these valleys almost to the

 337

exclusion of the native peasantry" (Rawlinson,184l/l980,

GA,Kandahar:5ll). The special relation of the Abdalis to

the state was due to the circumstances under which the

state was founded.

In October of 1747, when the Pashtun khans gathered

to elect one of their number as king, Ahmad Khan had been

neither governor of the province of Kandahar and, thus,

khan of the clans there, nor commander of the Abdali con-

tingent in Nadir Shah's army. Since both of these posi-

tion, according to Ahmad Shah's biographer (Mahmud Monshi,

vol.I,1974:43,50), were filled by members of the Alizai

clan, it is clear the Nadir had not promoted a rival

member of Ahmad Khan's Sadozai lineage or even of the

larger Popalzai clan, to a position of major authority

among the Pashtuns. Ahmad Khan's father and brother had

both ruled Herat during Abdali hegemony there, but even

in Herat their power had not been unchallenged. Members

of another Sadozai sub-lineage, which traced its descent

to Sher Khan's brother and which with the Mughal with-

drawal from Kandahar had moved to Multan, had provided

more khans for the clan than Ahmad Khan's branch (Tarikh-

i-Sultani,188l:65,97-l04) .

In a deadlocked assembly of khans who could not a-

gree to the election of a ruler from a military strong

clan, a Sufi proclaimed Ahmad Khan king. The historian

of a rival clan asserted that the more powerful khans

 338

consented to the choice "because the Sadozai lineage was

small in number and they knew that in ruling he would pay

attention to the wishes of the Durani nobles and if he

attempted to become authoritarian or oppressive, his over-

throw would be easily within the power of the Durani clans"

(Ibid:123). Indeed, during the first months of his reign,

Ahmad Shah was challenged by a group consisting of the

former governor of Kandahar, khans of the Barech and Tarin

clans, and a number of Abdali clans, but he was able to

defeat them and either execute or imprison the main ring-

leaders (Mahmud Monshi,1974:59-77). He owed his victory

to the fact that most of the Abdali khans that he had

appointed to the major offices of the state closely

modeled on the Safavid and Nadir's state in formal struc-

ture -- as well as the majority of the Abdali clansmen

had remained loyal to him (Ibid:71).

Ahmad Shah had gained their loyalty through two mea-

sures. First, after his proclamation as king, a convoy

was seized which was carrying the taxes of the Kabul and

Peshawar provinces to Nadir -- variously estimated from

one million three hundred and sixty thousand rupees to

twenty million rupees (Ghubar,1967:360i Tarikh-i-Sultani,

1881:124). "Depending on their rank and importance

[Ahmad Shah] bestowed on every Durani elder (lit. white-

beard) and leader an appropriate gift and stipend" (Mahmud

Monshi,vol.I,1974:56). His second set of measures in-

 339

corpora ted the clans directly to the structure of the

state and needs to be considered more fully.

After his assumption of kingship, Ahmad Shah took the

title of Dur-i-Duran" (the pearl of pearls or the pearl

of ages) and renamed the Abdali clans as the "Durani"

clans. Having conquered the central places of Kabul and

Peshawar, he gave the Pashtun clans residing in the coun-

try "enclosed between the range of Hindoo Coosh, the

Indus, the Salt Range, and the Range of Solaiman" (Elphin-

stone,vol.II,1839/1972:2) the name of Ber-Duranis (the

upper Duranis). They consisted of the Yusufzais, Othman

Khails, Tarklanris, Khattaks, the clans of the Khyber and

the plain of Peshawar as well as those of Bangush (Ibid).

Through their re-naming, Ahmad Shah was clearly aiming at

forging a closer relationship with these powerful clans.

According to traditional Pashtun genealogies, as shown

in the following charts, some of the clans were reputedly

descendants of the same son of Qais, the alleged ancestor

of all Pashtuns, yet other clans, such as the Khattaks,

belonged to different branches. Closer at home, in

Kandahar, the renaming of clans might, however, have been

intended to emphasize a more restricted rather than a

broader level of segmentation.

 340

Diagram 1. Sarbanri I -- \\Testern Afghanistan

Sharkbun (or Shakarbun or Sharjyun)

I
Sherani Tarin Barech
(from a Kakar woman, so
went to join Ghurghusht I
branch to which Kakars
belong) Spin Tor Abdal
(white) (black)

Sheranis
I I
Spin Tarins Tor Tarins
(Duki in (Pishin Abdalis
Baluchistan) B I tan) or
Duranis
(Afghanistan)

I
I
Zirak

?=
Popalzais Alikozais Barakzais Achakzais Nurzais Alizais Ishaqza
(Afghanistan) (Pishin \. J
and
Toba)
V
(Afghanistan, also
"Multan Pathans")
Sadozais Muhammadzais
(One branch) (one branch)
Ruled Afghanistan Ruled Afghanistan
1747-1818 1826-1978

Source: adapted from Caroe, 1958:12.

 341

Diagram 2. Sarbani II -- Eastern Afghans

Kharshbun (or Krishyun)

I
I
Kand Zamad Kasi
I
Khweshgis Muhammadzais
I
Kasis
I
... <' - - - Shinwaris Ketrans
(Quetta) (Ningrahar (Baluch
-:v"
(Hashtnagar) and border)
Khaibar)
Ghoriah Khakhay

Ii
Ghoriah Khel Khakhay Khel

I
I I I
Khalils Daudzais Mohmands Chamkanis
adopted
(Peshawar and adjacent (Kurram &
border Peshawar)

Mand Muk Tarklanris

Bajaur

Gigianis
I
,----'--..,. (Doaba)
Yusuf Umar

Yusufzais
I I
l'1andar
I
(Dir, Swat,
etc. )

Mandar Yusufzais
(Marden and Swabi)

Source: Caroe,1958:13.
 342

Diagram 3. Bitan (or Batni or Bait) [Qais' second son]

Bitan

warSh~in Bibi :>:lato Shah Hussain of Ghor

Bhitannis
(Bannu and Tank)
Mati tribes
I
I I
Ghalzoe Ibrahim Lodi
(conceived out of
wedlock) I
Lodis
I
all the Ghalji (Ghilzai)
tribes of Afghanistan

Niaz Siani Dotan

Niazis
I
Dotanis
(Isa Khel and (South
Mianwali) Wariristan)

I
praray Ismcdl
I
I
Lodis and Biluts Sur Lohanis
---(paniala)

I
Lodi dynasty Sur dynasty Marwats
of Delhi of Delhi Daulat Khel
Barbars etc.
A.D. 1451-1526 A.D. 1539-55
(Bannu,Tank,etc.)

Source: Caroe,1958:15
 343

Diagram 4. The Kar1anris

Kar1anri (father unknown)

Kakay

~
KOday
1st
if
I
Utman Di1azak Orak Su1aiman Shitak

utmal Khe1 Di1Jzaks oraJzais

(Peshawar (extinct) (South
border) Tiran) Daurs
Bannuchis
(Tochi
&
(Bannu)

Mani Lugman Manga1 Khugi

I I I I
F ar:.. dun Khataks Manga1s Khugianis
(Kohat Muqbils Turis
Peshawar Zadrans Zazis(Jajis)
and (Khost) (Kurram
Afridis Mardan) border)
(North
Tirah,
Bazar
and Khaibar, Malikmir
Wazir
Kohat Pass)
I I
I I Bangash
Musa Darwesh Mubarik l>lahmud
(Kohart
I I
I Gurbuz Mahsud
and
Kurram)
I I (Khost)
Ahmad Utman
I I
Ahmadzai Utmanzai Mahsuds
Wazirs Wazirs (Central
(South Waziristan) (North ~'Iaziristan)
Waziristan)

Source: Caroe, 1958:21

 344

From diagram I., it can be readily seen that the

Duranis or Abdalis are believed to be descendants of Tarin

who, beside Abdal, is shown to have had two other sons.

Sher Khan, father of Ahmad Shah's paternal grandfather,

was regularly referred to as Tarin in the Mughal contem-

porary sources. This may have been due to the fact that

the Safavid and Mughal states appointed only one khan for

all the western Sarbanri Pashtuns. If this were the case,

it follows that the state, through appointing khans at

certain levels of segmentation, can strongly influence

the tendencies towards the production of smaller or lar-

ger number of segments. The empirical example for this

assertion is provided by Ahmad Shah's example of the re-

cognition of the Achakzais as a full-fledged Durani clan.

In diagram 1., they are represented as a segment struc-

turally equal to other segments. In fact, they were a

sub-clan of the Barakzais "from whom they were separated

by Ahmad Shah, to reduce the formidable numbers of the

latter" (GA,Kandahar:16). The exclusion of the Tarins

and Barechis from the framework of the Durani clans

might have been related both to the armed opposition of

these two clans to Ahmad Shah at the beginning of his

reign and to a desire not to increase their strength

which closer association with the state would have brought

about.

The Durani power being the mainstay of the state in

 345

its foundation, Ahmad Shah had to seek ways to create a

firm nexus between clans, their khans, and the state. To

this end, the category of lands called khushkaba (water-

less), which were located in the outer valleys, were

granted to the Durani clans as mawrusi (hereditary) and

from then on fully treated as private property. This

was justified on the assumption that these lands had been

the original settlements of the Duranis before their ex-

pUlsion from Kandahar. Prior to Ahmad Shah's reign, these

lands, according to Islamic laws of taxation, had been

subject to the payment of one-tenth of the produce to

the state. Ahmad Shah reduced the obligation to a "small

supply of wheat, barley, or chaff on the occasion of the

passage of the army in their vicinity; and under this

light obligation, the valley of Tarnak was soon occupied

throughout by the Alikozai; the Arghastan fell to the lot

of Barakzai and Popalzai; and the Kadanai and Dori to the

Achakzai and Nurzai, a very small proportion of any of

these lands being left to the cultivation of the native

peasantry" (Rawlinson,GA,Kandahar:512).

Lands held as tiyul were not formally converted to

any other form of tenure and Ahmad Shah required two

horsemen from each kulba, on the understanding that only

one would actually serve at any time. But instead of

considering the land as the soldier's pay, Ahmad Shah al-

so paid him for the duration of his service as well as

 346

for a quarter of a year when his services were not called

upon. Therefore, the Durani clansmen looked upon the land

as fully belonging to them.

The concessions made to the khans were even more

significant. The military pay of the leaders "varied ac-

cording to the rank of the individual from 100 to 1,000

tomans [2,000 to 20,000 rupees] annually" (Ibid:511). But

instead of being paid in cash, the Durani khans were given

assignments on the produce of khalisa lands which fell un-

der their mana9"ement. Furthermore, they were 9"i ven the right

to farm revenues of the districts as payor on behalf of

the state treasury. These revenues were calculated from

taxes on ryoti (subjects) lands and other taxes extracted

from non-Durani populations. The income from these

sources was usually "debited at an aggregate equivalent

in money to the Chief of the Durani Ulus [clan], who re-

sided upon the spot, and to whom this general superinten-

dence of the local revenues gave the most favorable oppor-

tunities for improving the condition of his own followers"

(Ibid:5l2). In addition to these financial powers, Durani

khans -- the highest ranking of whom were called sardar

-- commanded their followers in battle and acted as "the

civil magistrates of the country allotted to maintain

them" (Elphinstone,1839/l972:10l).

~lilitary divisions were made to correspond to those

of the clans. Thus, the Sardar of each clan commanded

 347

the troops of that clan, and his subordinate officers

were the khans, maliks (headmen) and mashirs (elders) of

the sub-clans and lineages (Ibid:102). Significantly,

while the king selected the sardars governing each of the

major Durani clans "out of the head family," khans of the

subdivisions were "appointed out of their head families

by the Sirdars; and the Maulliks and Mushirs of the still

smaller divisions [were] in most cases, elected from the

proper families by the people, subject to the Sirdar's

approval; but in others, appointed by the Sirdar, with

some regard to the wishes of the people" (Ibid:104).

Khans were bound to make use of their extensive po-

wers for their own enrichment as well as that of the

clansmen on whose loyalty their position depended. Not

surprisingly, they turned towards acquiring the lands of

the Parsiwan and other local landowners and, through a

combination of financial and political pressures, com-

pelled them to dispose of their lands to Duranis "at a

rate very far beneath their value, exchanging their for-

mer independence to work as hired labourers on the Durani

lands" (Rawlinson , GA,Kandahar: 515) .

Access of all Duranis to strategic resources was, of

course, by no means equal. The most powerful Durani khans,

who controlled large tracts of land, did not directly man-

age their lands but entrusted them to subordinate clans-

men. During the reign of Ahmad Shah (1747-1772) I through-

 348

out Kandahar the lucrative positions of the management

of lands [entrusted to kishtgars] fell into the hands of

the Duranis, while the actual manual labor of cultivating

the land [undertaken by buzgurs] "resolved upon Parsiwans

and other ryots, who received from their employers a daily

pittance, just sufficient to subsist them" (Ibid:5l6).

Thus, it was through the agency of the state that class

positions were seen to be derived and coincide with eth-

nic identities.

The expression of these tensions was not confined

to the sphere of production. It also conditioned the

relation between the Durani khans and the kings. Rawlin-

son, from whose report I have been quoting extensively,

claimed that "Ahmad Shah appears to have hardly been a-

ware of the danger to which he was subjecting the State,

in thus laying the foundation of formidable and almost

independent Durani power. He considered the Durani tribes

to constitute the true and intrinsic strength of his king-

dom, and he believed that the more their power was devel-

oped the stronger would be his means for achieving for-

eign conquest, and the safer would be his bulwark against

foreign aggression" (Ibid:5l7). As far as the situation

in the southern region was concerned, this description is

largely to the point. The situation was, however, far

more complex as far as the overall organization of the

state and ethnic relations in other regions of the coun-

 349

try were concerned.

The seizure of the treasure intended for Nadir Shah,

which laid the financial basis of Ahmad Shah's state,

brought him an additional asset. The official in charge

of the convoy was a Qizilbash by the name of Taqi Khan

Shirazi. After being informed of Nadir's assassination,

"Taqi Khan and a number of the Qizil Bash were persuaded

to accept employment with Ahmad Shah and on the orders of

Ahmad Shah he gathered the Qizil Bash who were dispersed

in Kabul, Lahore, and other places around himself and

convinced them to take service with Ahmad Shah" (Mujmal

Ul Tawarikh, contemporary account, 1965 ed.:60-6l). This

force became the nucleus of the Ghulaman Khasa (slaves of

the state) (Mahmud Monshi,vol.II,1974:990) more generally

known as Ghulam Khana (household slaves). Owing loyalty

only to the person of the Shah, this force, as we have

seen, was used from the beginning of Ahmad Shah's reign

to balance the influence of the Durani forces and played

a very important role in the reign of his successors.

In recognition of this service, Taqi Khan was later as-

signed to farm the revenue of the provinces of Derajat

and Multan. When he died in 1756, despite outstanding

financial debts to the state, his descendants were not

held accountable and were given permission to take re-

sidence anywhere they pleased (Ibid,vol.I:596).

The recruitment of troops from among minor ethnic

 350

groups was not confined to the Qizilbash. After the con-

quest of the northern region in 1756, the commander of

Ahmad Shah's army brought him a force of 5,000 Uzbecks

taken hostages (Ibid:609). In 1767, another Durani khan,

after conquering the northern region and Badkhshan, uhired

servants from all Uzbeck clans" (Ibid,vol.II;12ll). Fur-

thermore, throughout Mahmud Monshi's narrative of the var-

ious battles, there are references to the Topchian-i-

Farhangi Nizad (artillerymen of European descent) (Ibid:

236,367,519,782,1007), which although not supported by

other sources, implies the hiring of Europeans on the part

of Ahmad Shah. Nonetheless, the commander of the artil-

lery was a Durani khan.

An important factor in the consolidation of the

Durani power in Kandahar was probably the decision of the

Ghilzai contingent of Nadir's army to remain in Iran.

This force, under one co~~ander by the name of Azad Khan,

first found a theater of action in Azerbaijan. Azad Khan

later conquered the Safavid capital of Isfahan and, until

his defeat by more powerful rivals in 1760, was one of the

contenders for the mastery of Iran (Perry,J.R.,1979:48-

79) .

The armed forces \'lere not the only elements whose

careers, in periods of rapid political change, were

characterized by a high degree of geographical mobility.

The specialized staff needed for the management of the

 351

financial and administrative organizations of the state

were equally ready to offer their services to those capa-

ble of seizing and maintaining political power. The a-

bility of the bureaucratic staff to adjust to new loca-

tions was considerably facilitated by the fact that, for

several centuries Persian had been the main language of

administration in Iran, central Asia, Afghanistan, and

India. Any newly-formed state, if it intended to realize

regular revenues from the subjugated populations, had to

allot a place in its structure for Persian-speaking func-

tionaries. A striking illustration of this principle was

provided by the "British policy of continuing Persian as

the language of administration [in India] until 1835 and

of continuing the pre-British pattern of drawing civil

servants from families of Persian origin who made their

livin'3" as administrators and scribes" (Cohn,B. ,1969:79).

Pashtun rulers who, out of long association with the

Persian and Mughal states, were fully conversant in the

Persian language, did not have any hestitation in adopting

the administrative structure of the Safavid state and hir-

ing a large number of Qizilbash and other Persian-speak-

ers as accountants and scribes (Mahmud Monshi,1974,vol.I:

533-534; Ghubar,1967:359). Their services were not lim-

ited to the central government. Throughout our period

of long duration, tax-farming remained the main form

through which the state collected its revenues. Pashtun

 352

khans, who became the major though by no means the only

group of tax-farmers, needed their own bureaucratic es-

tablishment. Thus, despite the reduction of the majority

of the Persiwan landowners of Kandhar to the status of

dependent cultivators on Durani lands, the specialized

knowledge of some members of the group made the function-

ing of the whole system dependen~ on them.

Even more significantly, the existence of these

groups freed the ruler from an exclusive reliance on the

power of Durani khans and clans. Yet, unlike states where

ethnic or religious groups other than that of the ruler

were totally relied upon for the armed forces and the

civilian administration and ultimately gained control of

the machinery of the state, the access of the Durani and

other Pashtun clans to the means of destruction, as well

as to the highest offices of the state, allowed the ru-

ler to maintain a degree of control over the situation

by carefully balancing and manipulating the power of the

various ethnic groups within the bureaucracy and the army.

Relations among ethnic groups were characterized by a

high degree of tension. Whether these tensions broke into

actual armed confrontations leading to the de structuring

of the total system depended on the financial resources

of the state, the wayan individual ruler had gained the

throne, and the capacity of the ruler in catering to the

interest of the various groups. Before describing the

 353

the actual unfolding of these complexities, I must take

account of two elements in the structure of domination.

Throughout the rule of the Durani Sadozai clan (1747-

1818), the religious stratum "was in control of the judi-

cial courts and the mosques" (Ghubar:1967:357). The for-

mal list of religious functionaries as recorded by

Elphinstone (vol.II,1839/1972:262-65,277-78), resembles

that of the Safavid state but, unlike the situation in

Iran, the dominant form of Islam in Afghanistan was the

Sunni interpretation. Courts of sharia were present in

all central places and had deputies "over the whole coun-

try" (Ibid:263). Qudat (judges), upon the advice of the

Imam (prayer-leader) of the king's household, were

appointed by the king but their pay was provided in most

cases from a "small tax imposed on every family in the

district" (Ibid). They could nowhere initiate a case

unless an application was made to them and, although

they based their judgements on sharia, it was in practice

modified by Pashtunwali (way of the Pashtuns) or "custom-

ary law" (Ibid:262). In this period, some new awqaf,

such as that around the alleged cloak of the Prophet of

Islam in Kandahar, were founded by the state and, as ln

Safavid Iran, there was an office holder, bearing the

title of sadr, who kept a registry of all awqa£ in a

central place, whether granted by private persons or by

the state (Ibid:276). In general, there seems to have

 354

been little interference with members of the religious

stratum and, as the policy of territorial expansion was

usually justified in terms of restoration of "true ortho-

doxy," rulers regularly met with groups of the leading

ulama. The actual influence of ulama over the course

of events seems to have differed greatly from the reign

of one ruler to that of another. In rural areas, prea-

chers in charge of mosques, called imams or mullahs, had

"grants of land from the headmen of the tribe, or from

the tribe itself" (Ibid:278) and many also received grants from
the state as well as voluntary contributions from the

people (Ibid).

Historical precedents which were drawn upon to

shape the Durani state policies in regard to the religous,

the bureaucratic, and the military strata, also guided

the relations of the state with non-Durani Pashtuns and

other ethnic groups. As far as Pashtun clans were con-

cerned, the most important aspect of this relationship

was the extension of the recognition and/or creation of

khans among all clans. It is indeed from this period on-

wards that the title of khan gained general currency and

ended up being used as a qeneral form of address in

the twentieth century.

Under Ahmad Shah, the major thrust of the state was

that of territorial expansion. To achieve this goal as

well as to prevent dissension at horne, he needed to se-

 355

cure the cooperation of other Pashtun clans. The recog-

nition or creation of khans served both to give other

clans a stake in the future of the state and to provide

the state with the necessary manpower for its geographical

expansion. As no other Pashtun clan held all its lands

as a grant from the state, Ahmad Shah required, instead

of taxes which would have been resisted, an agreed upon

number of horsemen. The difference in the relation of

Durani and other Pashtun clans to the state can be gath-

ered from the following table which portrays the distri-

bution of the Kandahar horse (Rawlinson,GA,Kandahar:5l3).

Distribution of Kandahar Horse, with their allotment of lands under

Ahmad Shah

Name of Tribe Number of Kulbas Quota of Horses

Durani
Popa1zai 965 1/4 806
Alikozai 1,050 851
Barakzai 1,018 1/2 907
Alizai 661 3/4 819
Nurzai 868 1/2 1,169
Ishakzai 357 1/2 635
Khogani 163 423
Maki 121 1/2 100
Total 5,206 5,710

Tribes not Durani

Tokhi 14 1,061
Hotak 10 507
Kakar 56 30
Dawi 5 45
Tirin 25 729
Barechi 518
---
Total llO 2,890

Royal attendants 3,995

Total of ku1bas 5,316; Total of horse 12,559.
 356

The notion of supplying horse in lieu of taxation

was, however, fictional. In the case of the Tokhi and

Hotak clans of the western Ghilzais, Leech reported that

whenever their services were called upon by the state,

rulers "invariably found it necessary to give them pay

from the Royal Treasury" (L/P&S/3/ll,184l:6). Since

these contingents were raised by and fought under their

own khans, they became the main link between the state

and non-Durani Pashtun clans. Yet, as the state kept

as its prerogatives the appointment of khans, their

scale of pay, and the recognition of a clan as a segment

of another clan or as an independent clan, the state's

impact on the structure of segmentation and on relations

between different named groups and lineages within a clan,

was not negligible. Elphinstone's diagram portraying

the various levels of segmentation among the Ghilzai

clans and the names of khans appointed by the state, to-

gether with the positions left vacant, illustrates the

general relation between the state and these clans. As

can be seen, no cornmon chief was appointed for the

Boorann branch of Ghilzais.
 357

Daigram 5. The Ghilzais

Ghiljie (chief) Abdooreheem

Boorann Toraun
(no common chief) (chief) Abdooreheem
I I
Tokhee Hotukee
Irjub Moosa (chief) (chief)
(no common (no common Sahaubodeen Abdoorehem
chief) chief)

Sukhank
I
Turruk Undar
(chief) (chief) (chief)
OtInaun Hyder Lauleh

Soliman Khail
I
Ali
I
pitch
I
Ismaelzye
(no common chief)
I
Kyser

Dosso
I
Abdoorehem Sooltaun Ahmedzye
(chief)
Khaunan
I
Baubukr Maroof Alladeen Bauraun Yehya Esau Moosa
Khail Khail Khail Khail Khail Khail Khail
(chief) (chief) (chief) (chief) (chief) (chief) (chief)
Afzal Aluned Khudraun unknown Soliman Abdoola Khaunan

Source: Elphinstone,vol.I:2l2.
 358

The importance of the connection with the state for

the finances of a khan can be gathered from the fact that

direct allowances received by Abdur Rahim Khan Hotak, re-

presented in diagram 5. as khan of all Ghilzai clans, as

well as khan of the Turan section and the Hotak clan, was

150,000 rupees during the reign of Timur Shah (1772-1793).

In addition, he was given the right to tax-farm several

districts in what is today the north-west frontier pro-

vince of Pakistan, "by the collection of the revenues of

which districts he derived much profit" (Leech,184l:6).

At that time, he was supposed to supply the ruler with

7,000 horsemen, the choice of whom must have provided

him with strong means for maintaining his influ~nce with-

in his clan. The relationship to the state of other

khans who, according to the diagram, were formally sub-

ordinate to the Hotak khan, was not mediated ti1rough him,

as shown by the fact that the khan of Tokhis received a

pay of 160,000 rupees. In addition, Tokhis were allowed

to collect taxes from all caravans passing between Kabul

and Kandahar; they were also in control of the revenue

of some land (Ibid:7-ll).

Exemption of all Pashtun clans from taxation was not

automatic but reflected the balance of power between the

state and the various clans. According to the revenue

settlement of an official of Ahmad Shah, all Ghilzai clans

living in the Logar valley, immediately south of Kabul,

 359

were paying taxes. But even here, different ethnic groups

were treated differently by the state. Parsiwans (Per-

sian-speakers, consisting of Tajiks, Qizilbash, Hazara,

and others) were paying twelve time as much in taxation

as the Pashtuns. However, when Pashtuns bought land from

Parsiwans, they were expected to pay the same rate of

revenue as the former owners. Unlike the situation in

Kandahar, in Logar, the correspondence of class and ethni-

city was more complicated. Pashtuns were described as

small crofters, farm-labourers, merchants, camel-breeders,

shepherds and drovers; Tajiks as the chief carriers; and

Hazaras mostly as farm servants. Qizilbash, on the

other hand, were said to be landowners who "mostly re-

tain Hazaras, Ahmadzais or Abdrahimzais [Pashtun Ghilzais]

as servants. They live on their estates, or serve the

Crown as Cavalry soldiers" (GA,Kabul:348-370).

Here again, the differential access of different

ethnic groups to the state had a considerable impact on

their access to means of production. At the same time,

it can be readily seen that the position of various

groups in relation to means of production and domination

took on very different configurations in the different

regions of the country and, consequently, found very dif-

ferent forms of expression in day to day as well as long-

term patterns of interaction. Furthermore, considerable

differences in the financial basis of the state at differ-

 360

ent conjunctural periods during our unit of long duration

gave expression to different forms of contradictions

among the state, the different classes, ethnic groups,

and strata, in the different regions of the country. It

is towards an examination of this issue that I will now

turn.

The following are estimates of the revenue of the

state during the reign of the Durani rulers.

Ahamd Shah (1747-1772), 31 million rupees (Ghubar,

1967: 358) i

Timur Shah (1772-1793),10 million rupees (Fofalzai,

vo 1. I I , 1967 : 35) i

Zaman Shah (1793-1800),10 million rupees (Husaini,

1797/1967:32) ;

Shah Mahmud(1800-1803), ?i

Shah Shuja (1803-1809), 20 million r~pees (Elphin-

stone,vol.II:258)i

Shah Mahmud(1809-1818), ?i

Elphinstone, who first offered a figure of 30 million

rupees as the revenue of the total territory claimed by

the Duranis, considered that 10 million of it went direct-

ly to "half-subdued princes" and could, therefore, not be

counted as part of the income of the state. That figure

corresponds, however, to the estimate by Ghubar of the

income during Ahmad Shah's reign. Of the remaining 20

million rupees, Elphinstone wrote that "of this a great

part (about halfj is assigned in Teeool (Jageer), most

 361

of it granted on condition of military service ... ; the

rest is alloted to maintain the Moolahs or religious

officers, or given in charity to dervises and Syuds".

He estimated the sum directly received by the ruler at

"upwards of nine millions of rupees" (Ibid). This later

figure comes very close to the estimates of Husaini and

Fofalzai and it can be assumed that these historians were

referring to this figure as the income of the Crown.

Husaini's territorial breakdown of the revenue, al-

though incomplete, reveals some major differences in the

relation of the state with different regions. He gave

no information about the revenues from the provinces of

Kabul, Herat, and Badakhshan. Of Kandahar, he wrote that

the whole province is "allole:d as jaqir to the Duranis"

and, of the province of Balkh, that "its income is not

even sufficient to cover the expenses of the governor and

the army that is stationed there and these expenses are

covered by His Majesty" (1797/1967:31-32). On the other

hand, he reported that the amount paid by tax-farmers

or hereditary rulers of the provinces of Kashmir, Multan,

Dera Ghazi Khan, Dera Ismael Khan, and Peshawar were

respectively 2.4 million, 2-300,000; 700,000; 250,000; and

700,000 rupees (Ibid:30-32). He also claimed that the

tax-farmer of Kashmir realized 1.6 million rupees above

what he paid to the state and tax-farmer of Dera Ismael

Khan realized 150,000 rupees (Ibid:31).

 362

Thus, the fiancial basis of the Durani state and its

sanctioning of privleges for special classes and ethnic

groups at home rested on income from provinces which be-

came part of "British India" in the nineteenth century.

It can also be seen that access to privileged positions

of influence in the state -- such as an assignment of

tax-farming, governorship, or khanship -- could almost

immediately turn into immense financial advantage. Ten-

sions arising out of access to these positions and the

protection of already established privileges permeated

the history of the entire period bringing about the de-

structuring of the Durani state, the loss of the Indian

provinces and the re-structuring of relations of domina-

tion in the present territory of Afghanistan. I will now,

briefly, consider the highlights of these complexities

between 1747 and 1818.

When Ahmad Shah was elected king in 1747, the inter-

nal crisis of the neighbouring states was so advanced

that, despite their larger resources, they were not able

to check the territorial expansion of the Durani state.

Consequently, the flow of cash to coffers of the Afghan

state and khans was constant. Ahmad Shah's conquest of

Punjab in 1751, resulted in the payment by the defeated

governor of the province to Ahmad Shah and his officials

of a huge sum of IT".'l1ey as we 11 as the acceptance of an

annual tribute (;..:. ten million rupees. In the same year 1

 363

the Emperor of India, in return for an undertaking by

Ahmad Shah not to invade his capital, ceded the province

of Multan permanently and offered ten million rupees in

cash (Mahmud Monshi,vol.I,1974:326,332). Even the means

of destruction with which Ahmad Shah effected his con-

quests were largely collected from his opponents (Ibid:

44,88,114). After his historic victory over the Mara-

thas in 1760, for instance, he seized 5,000 artillery

guns, 100,000 handguns and 5,000 elephants (Ibid,vol.II:

1025). The impact of these victories on the finances of

the state was such that, in 1758, Ahmad Shah decreed that

all forms of taxation in the realm were to be remitted

for two years (Ibid,vol.I:588).

The Durani khans, who occupied most positions in the

state, were undoubtedly the major beneficiaries of this

expansion. This has led most Afghan and other historians

to conclude that khans of the major Durani clans had an

hereditary claim to the main offices of the state. A

careful reading of the evidence presented in Ahmad Shah's

biography, which prior to 1974 was not available to re-

searchers, shows that, from the beginning of his reign

onwards, he exercised the right of dismissing, appointing,

creating new officers, choosing the tax-farmers, and pro-

moting his own dependents to the rank of khan. These

were not mere theoretical prerogatives. Indeed, he ap-

pointed a number of non-Durani and non-Pashtuns as gover-

 364

nors and tax-farmers of some of the major provinces. He

also bestowed the rank of khan on some of the eunchs who

had accompanied him from the court of Nadir and appointed

one of them as the governor of the predominantly Pashtun

province of Peshawar (Ibid:546,l2l6).

He exercised power in the choice of khans of Durani

and other clans. Even in those cases where a khankhel

already existed, limiting his selection to members of

certain families, as the order of succession was not

fixed, his decision confirming one of a number of rival

candidates could materially affect differential access

to strategic resources and contribute to the formation

of distinctive sub-lineages within a lineage. The in-

dividuals passed over could look to improve their for-

tunes by seeking an office in the state or hope to create

a faction in the clan and, by offering their services to

claimants to the throne, hope to gain khanship or other

high offices in the state. Khans dismissed from lucra-

tive posts or demoted to less influential positions could

also opt for armed confrontation with a ruler. The net

result of these tensions was the creation of long-lasting

feuds among various individuals, lineages, and clans,

which affected the whole system of domination.

The trend started in the reign of Ahmad Shah reached

crisis proportions under his successors. In 1751, Ahmad

Shah's brother's son, who, in the absence of his uncle

 365

was acting as his deputy in Kandahar, with the help of a

number of prominent Durani khans and his own Arab body-

guards, declared himself king. His rebellion was crushed

and he and his adherents executed (Ibid:348-350).

An even more important event, as an early expression

of future tendencies, was, in 1764, the dismissal by Ahmad

Shah of Shah Wali Khan, chief-minister and the most power-

ful Durani khan of his reign. Shah Wali Khan had made

use of his control of the financial machinery of the state

to build an extensive network of followers throughout

the realm and seemed directly to threaten Ahmad Shah's

control of the state. He was consequently dismissed from

the control of finances without being asked questions

about his accounts and given the rank of cornrnander-in-

chief of the army. As Mahmud Monshi remarked, "from the

requisi tes of the office he had nothing but the name" (Ib-

id:1148). Even more significantly, he was replaced by

a non-Pashtun mulla from the town of Daulatabad, in the

northern region (Ibid). Yet, after the death of his suc-

cessor in 1769, he was reappointed as chief-minister

(Ibid: 1219) .

When Ahmad Shah died, his second son and heir appar-

ent, Timur, was ruling in Herat with the title of Shah

and full powers of appointment and dismissal of his of-

ficials. But Shah Wali Khan declared another brother of

Timur, who was also his son-in-law, as king in Kandahar.

 366

On the arrival of Timur's army in the vicinity of Kanda-

har, Shah Wali attempted to seek an interview with him

but was seized and executed. Sensing the alienation

caused by this measure among the Durani khans, Timur Shah

appointed his own son as governor of Kandahar and moved

to Kabul, which he soon declared as his capital in 1775.

Even before Timur's accession to the throne, the po-

litical situation of the neighbouring states had consider-

ably changed. From a persecuted religious sect, the Sikhs

were transforming themselves into an organized state and

Ahmad Shah, in his last campaign in 1769, had failed to

destroy their hold over the Punjab (Ganda Singh,1959:320).

The British had become a strong territorial power in

India, and in Bukhara and Iran new dynasties, bent on con-

solidating their power at horne and seeking external con-

quests, were established. The expansi0n of the territory

of the Afghan state was no longer a viable alternative

to consolidation of its power.

To increase his own power at the expense of Pashtun

khans, Timur relied heavily on the Qizilbash and other

members of the Persian-speaking ethnic groups. He even

placed his twelve-thousand strong bodyguards under the

command of a Qizilbash. "From the powerful khans, he

only accepted the services of those individuals who un-

conditionally obeyed the monarch" (Ghubar,1967:374). It

was not long before Pashtun khans reacted to Timur's

 367

policies. In 1774, Durani and Ghilzai khans declared a

Durani khan, previously tax-farmer of the province of

Sind, king of the country and, at the head of an army of

twenty five thousands, marched on Kabul. However, a num-

ber of Durani khans defected to Timur who was able to

crush the rebellion (Ibid:374). One of these khans,

called Painda Khan, was awarded by Timur the khanship of

the Barakzai clan, replacing his brother. His descendants

ruled Afghanistan from 1826 to 1978.

After the rebellion of 1774, Timur tried to keep

Pashtun khans in check by granting or withdrawing the tax

privileges of influential khans as well as regulating the

access of various individuals to positions of authority

in the state. He did not dare, however, confront the

financial privileges of the khans as a class or those of

the Durani clans as a group. Despite a number of minor

rebellions at home and a few foreign wars, his reign was

the calmest period of Durani power in Afhghanistan. With

his death in 1793, all the tensions within Afghan society

broke to the surface.

Timur left behind thirty three sons all but two of

whom (governors of the provinces of Kandahar and Herat)

resided in Kabul. In return for the dismissal and execu-

tion of some of Timur's most influential Parsiwan offi-

cials, the Pashtun khans recognized his fifth son, Zaman,

as king, and imprisoned all his other brothers who were

 368

in Kabul. The two brothers in Kandahar and Herat refused

to accept Zaman's authority and confronted him militarily.

War followed war. After accepting the authority of one

brother, disgruntled khans would, as the autobiography

of Shah Shuja (1954:1-62) shows, flock to the standard of

another brother. A brief battle or usually the defection

of a powerful khan would decide the issue of the contest

one way or the other.

A major reason for these shifts of allegiance was

the attempt of Timur's sons to imitate the example of

their father in checking the power of Pashtun khans. By

the time of Shah Mahmud's second reign (1800-1803;1809-

1818), most of the powerful Durani khans had been killed

in battle, executed, or had died. Mahmud's son, gover-

nor of Kandahar, was also initiating a systematic policy

of eliminating them through a series of individual plots

without challenging the financial privileges of the group.

His last victim was Shah Mahmud's chief-minister whom he

had blinded in 1817. But the chief-minister, a son of

Painda, khan of the Barakzais, had managed to place his

numerous brothers in such major posts as the governor-

ships of Kashmir, Sind, Peshawar, and elsewhere. His

blinding led to a confrontation between his brothers and

Shah Mahmud as a result of which the Sadozai dynasty in

Afghanistan came to an end.

The Barakzai brothers had managed to overthrow a

 369

structure but were unable to agree among themselves as to

how to replace it. Between 1818 and 1826, they plunged

the country into a series of civil wars. By 1826, a num-

ber of regional states had emerged: the western region of

Herat was ruled by the former king Shah Mahmud and his

son Kamran; the southern region of Kandahar was ruled by

five Barakzai brothers who were from the same mother; the

eastern region of Kabul was ruled by another Barakzai

brother by the name of Dost Muhammed; the central region,

northern region, and Badakshan were ruled a number of

autonomous chiefs. Until 1834, when the area was com-

pletely annexed by the Sikh state, a number of Barakzai

brothers also ruled in Peshawar and other localities in

the present territory of the north-western frontier pro-

vince of Pakistan.

During the years 1826 to 1839, which form my second

conjunctural unit of analysis, the history of the nor-

thern part of the country was almost totally separate from

the southern part and the Pashtun influence on the shape

of politics of the northern region, Badakhshan, and a

considerable part of the central region, was either negli-

gible or totally absent.

During the Sadozai period, these regions had contri-

buted little to the financial revenues of the state and

even the existing revenues were assigned either to khans

or to clans. with the constellation of forces that

 370

emerged after 1826, the financial needs of states had to

be met from local resources. Confrontation between rulers

and hitherto privileged elements was therefore inevitable.

Given the complex pattern of interpenetration of central

and peripheral zones of production in Afghanistan, each

of the states faced a different set of tasks and their

degree of success differed considerably. The trans-

formation of the Durani clans from a formidable military

and political power into a subjugated group can serve as

a general example of the restructuring of relations of

domination during this conjuncture, and I shall now out-

line the salient features of this change.

The Barakzai brothers who took over Kandahar, re-

ferred to as the Sardars of Kandahar, relied on two groups

to reduce the power of Durani clans. In return for the

support extended to them by their own Barakzai clansmen,

the tax-exempt status of the clan was not interfered with

(Rawlinson,GA,Kandahar:529). But the main instrument of

their power was a force 6f 3,000 mercenary horsemen "from

whose ranks the Duranis were jealously excluded ... These

were all picked men under the command of a few noted des-

perados, whose only guides to action were their own per-

sonal advantage and the will of the ruler whom they

served" (Ibid:533). It was with this body that the

Sardars goaded segments of the clans into rebellions and

speedily crushed them.

 371

The Sardars' main object was the commutation of the

Durani horse into revenue, and the reduction of the

Duranis to the same status of tax-payers as other groups.

Indeed, since the time of Ahmand Shah, the Kandahar horse

had rarely been called upon, yet, in return for supplying

it, the Duranis held their financial privileges (Ibid:

526). The direct suppression of these privileges would

have confronted the Saradars with the united opposition

of all Durani clans. Instead, they attempted to achieve

their goal through indirect means.

The neasures resorted to were mainly financial in

character and their imposition was entrusted to descen-

dants of the Parsiwan ministers of Timur Shah who had been

executed by Zaman Shah upon the insistance of Durani khans.

Backed by the armed might of the non-Durani troops, these

agents started imposing taxes on the direct producers who

cultivated Durani lands. This was gradually extended to

taxes on lands converted to gardens, bought from Parsiwans,

newly brought under cultivation, or converted from khalisa

into private property. These financial maneuvers were so

successful that the Duranis were finally compelled to

"come forward and declare their readiness to submit to a

reasonable assessment of their tiyul lands, on the under-

standing that the liability was to be clearly sp~~ified,

and was to secure for them the guarantee against any of

the grievous exertions to which they had been lately sub-

 372

jected" (Ibid:528). Their taxes, however, were still not

calculated at the same rate as those of other subjects

of the state. At the time of the British invasion of

Afghanistan in 1839, despite the fact that tiyul lands

had long been treated as private property and "about two-

thirds of the dependent tiyul-kulbas had passed from the

possession of the parties on whom they had been original-

ly conferred by Nadir Shah" (Ibid:532) , the Sardars were

still planning on revoking the grant. By this time, the

financial obligations of the Duranis in the areas border-

ing the central place of Kandahar were such that they were

"gradually led to dispose of their horses and arms as use-

less and expensive encumberances" (Ibid:534). Having been

thus dispossessed of their means of destruction, the

Duranis responded to the Sardars' new measures by selling

their rights of cultivating tiyul lands to their Parsiwan

producers or even abandoning them (Ibid:534-35).

In the outlying Durani districts and the lands in-

habited by Ghilzais, the situation was entirely different.

Leech stated that the Sardars of Kandahar "completely

failed in establishing an influence in the Guljie

[Ghilzai] country, and their troops ... experienced a total

defeat by a small party of the Hotuk headed by Gool

Muhamad Khan" (1841:4). The Tokhi clan, which controlled

the road between Kabul and Kandahar, also managed to main-

tain its independence from the rulers of Kabul and

 373

Kandahar alike, and imposed a toll on caravans passing

through its territory.

In other parts of the country, the situation was much

the same and every region witnessed a series of wars

through which an enterprising ruler or rulers extended

their Rr.en of influence at the expense of competitors.

In contrast to our first conjunctural unit (1747-1818),

during this period of the breakdown of central government

and the re-emergence of regional powers, a greater subju-

gation of the population of the country to the power of

the state took place. Tension between states and the

forces in society had reached such an intensity that when

Shah Shuja, son of Timur, marched at the head of a British

army to Afghanistan, both the ruler of Kabul and the

Sardars of Kandahar fled from their seats of government.

The British invasion of Afghanistan, as J.A. Norris's

re-examination of the evidence shows, was rooted on the

one hand in the fear posed by Russia to India, and on the

other in the belief in the capacity of central Asian mar-

kets to absorb British manufactured goods (1967:59). In-

deed, in 1841, these markets were flooded with English

cotton goods (Lurnley,1862:298). The immediate political

cause of the invasion was the dominance of Russian in-

fluence in the Persian court and the invasion of the re-

gion of Herat by the Shah of Iran in November of 1837

(Kaye,J.W.,1857:211-300). The Sardars of Kandahar had

 374

just entered into an alliance with the Shah of Iran when

a British mission, sent to win over Dost Muhammad, the

ruler of Kabul, failed to achieve its mission. The gover-

nor-general of India opted then for an invasion of

Afghanistan. The British government also sent a naval

expedition to bombard the Iranian island of Kharg bring-

ing about the withdrawal of the Iranian army from the

gates of the beseiged city of Herat.

In the British Foreign Secretary's words, the object

of invading Afghanistan was "to reorganize the country

under one Chief ... A good Afghan state in connection with

British India would make a better Barrier than Persia has

been, because it would be more under our control. We

should have the same kind of geographical Pull upon such

a state that Russia has upon Persia" (Palmerston, Aug .

25,1838,quoted in Norris,1967:209). To implement these

...
. ~~,--~ . ..,. ~

policies, Shah Shuja who, after his overthrow in 1809,

had lived as pensioner of the British government in India,

was put at the head of a British army and, finding no re-

sistance, entered the city of Kandahar on April 25, 1839.

On the 6th of August, he was reinstated in his palace in

Kabul. From then on, the policies of European powers,

either directly or indirectly, determined the territor-

ial limits of Afghanistan as well as the character of re-

lations of domination in the country.

The ease with which the British gained possession of

 375

the southern, eastern, and parts of the central region,

did not translate into an ability to restructure state

and society as initially intended. By 1842, the British

garrison In Kabul was destroyed by a popular uprising and,

although an "army of retribution" marched through the

country, destroyed the great bazar in Kabul and looted

the city, the British government was forced to evacuate

Afghanistan. By 1843, the Sardars of Kandahar had re-

turned from their Iranian exile to regain the southern

region and Dost Muhammad, after two years of exile in

India, had resumed his position in the central place of

Kabul and the eastern region.

From Malcolm Yapp's detailed description of the

resistance (1962:499-523;1963:288-313;1964:333-381), it

is clear that the revolts first started in peripheral

areas then gained momentum in central areas. The Afghans

managed to eliminate entirely British garrisons in the

towns of Ghazni and Kabul in the eastern region, but they

failed to dislodge the British forces from the central

place of Kandahar in the southern region, and the town

of Jalalabad in the eastern region. In claiming that the

movement was "very much a revolt of Sunni Pashtu speakers"

and that, because the total idiom of revolt was pervaded

by Islamic symbolism, it could not in any sense be con-

sidered nationalist, Yapp commits a factual error and mis-

ses the essential part played by Islam in the social re-

 376

lations of the society. Kabul being in the midst of a

mainly Persian-speaking area, some of the prominent lead-

ers of the revolt, as well as a major section of the rank

and file, were Tajiks. The military leadership of the

revolt was provided by Durani, Ghilzai, and Persian-

speaking khans, whose financial privileges had been se-

verely curtailed by the British, but the main ideologi-

cal leadership of the groups came from heads of Sufi or-

ders, whose organizational networks, as I have argued

earlier, crosscut all forms of ethnic and class cleavages.

Yapp is, however, correct in pointing out that one of the

important long-term consequences of the invasion was the

education "provided in the creation of stronger systems

of government, and particularly by the example of the use

of disciplined forces and the training of Afghan troops

which paved the way for the subsequent creations of a

powerful standing army by Dost Muhammad with which he

could extend his power to the rest of Afghanistan" (Ibid:

381) .

Indeed, my third conjunctural unit, encompassing the

period from 1843-1878, was dominated by the changes in-

·troduced in the mode of domination during Dost Muhammad's

second reign (1843-1863). Through military or political

maneuvers, Dost Muhammad brought all the regions of the

present territory of Afghanistan under his control (Faiz

Muhammad,vol.II,1913:198-251). The effective use of means

 377

of destruction was certainly an important part of the pro-

cess through which he consolidated his power. There also

was an equally important, if subtle, shift in the legiti-

mation of relations of domination.

Pashtun clans, as we have seen, regulated their so-

cial relations through a code they referred to as "Pash-

tunwali" in which notions of honor, revenge, and hospi-

tality were important elements. The code, as an astute

foreign observer noted, did not make any reference to

"paying taxes and following kings" (Broadfoot,1842/1886:

361). Nevertheless, Dost Muhammad's motto was that the

notion of Pashtun honor should not stand in the way of the

state interests. This was illustrated by his dealings

with the Sardars of Kandahar.

Following a conflict with the forces of Kandahar over

the district of Kalat, Dost was visited by one of the

Sardars of Kandahar who, in return for the gesture and on

the basis of Pashtun notions of honor, expected him to

leave the district. Dost retorted that "if a person dis-

obeys the monarch, rebels, and then requests a province

of the country, and if the king agrees to the demand and

turns over the province, within a short time, the order

of the state would be destroyed, bloodshed committed, and

every destitute will turn into a ruler or a minister"

(Faiz Muharnrnad,vol.II,19l3:2l4). Dost was able to main-

tain firm but peaceful relations with Kandhar. In 1854,

 378

two of the Kandahar Sardars died and a struggle for suc-

cession between the remaining Sardar and the sons of the

deceased brothers broke into violent confrontation. The

parties agreed to invite Dost Muhammad to mediate the dis-

pute. After entering the central place of Kandahar with

his army, Dost assigned the disputants large grants in

cash from the taxes of Kandahar but annexed the southern

region to his domain.

Dost justified his notion of monarchy by reverting

to Islam and the danger posed to the country by "infidel ll

powers. Born of a low-ranking mother, his education had

been neglected and, during his military adventures, he

had become addicted to wine. After conquering Kabul in

1826, he learnt how to read and write, quit drinking, and

strictly enforced the Islamic injunction against alcohol-

ic beverages. In 1838, when the Sikhs conquered Peshawar

and were threatening to move on to Kabul, he prepared a

force to combat them. He assumed the title of Arnir jus-

tifying it with the following declaration: "Upon the re-

quest of the ulama of the nation ..• that jihad [holy war]

is dependent on the existence of a ruler and the raising

of the flag of authority, and that such a person should

issue the coin and khutba [declaration of legitimacy read

every Friday in the mosques], the ruler declares jihad so

that whoever disobeys him would have disobeyed the in-

junctions of God and His Prophet" (Ibid,vol.I:127).

 379

Islamic theory of government was not limited to de-

claration of jihad. Payment of taxes to a legitimate

government was another element and, from the time of Dost

onwards, demands of the state for a share of the produce

kept increasing. Collection of the revenue was still car-

ried on through tax-farmers, but Dost Muhammad entrusted

the tax-farming of the provinces and major districts and

sub-districts only to his sons who, in turn, sub-farmed

the districts and sub-districts to the previous rulers

and notables, or to individuals of their choice. In 1843,

he also raised five divisions of infantry of 800 men

each -- which were placed under the command of his sons

who had to provide for their maintenance from revenues of

of the territories assigned to them (Ibid:200).

Dost had twenty seven sons, some of whom died before

him. As his domain increased, more sons acquired access

to the financial and military means necessary to support

their own basis of power. Upon his death, civil war raged

in the country until one son imposed his authority over

all the regions of the country. Sher Ali (1863-66;1868-

1879) embarked on a gradual policy of building centralized

institutions but his efforts were cut short by the second

invasion of the country by the British army.

During this conjuncture, the impact of British poli-

cy on the course of events in Afghanistan underwent sever-

al changes. Until 1855, there had been no contact between

 380

the two governments as such. The implications of the

Crimean war (1853-56) for British interests in Asia, com-

bined with the fear of an Iranian invasion of Herat --

which actually occurred in 1856 -- brought the British

government to offer Dost Muhammad a treaty of friendship,

which he accepted. During the years 1856 to 1858, the

material results of the treaty were the acquisition by

the Afghan state of 4,000 flint muskets, 4,000 percussion

muskets, together with 2,000 rounds of ammunition for

each musket, and 2.6 million British rupees in cash (Par-

liamentary Papers,48,1882). Dost Muhammad made use of

these resources for his conquest of Herat in 1863. Be-

fore that could take place though, the British had to en-

gage in a three-months war to bring the withdrawal of

Iranian forces from Herat in 1857 (Alder,1947a:186-209i

1974b:287-311). Alder, who has provided a detailed ac-

count of the reasons for British concern over Herat,

states that during the twenty five years following 1838,

Herat "had directly and indirectly cost Britain two wars,

over twenty million pounds, at least as many thousand

lives" (Ibid:186). In the northern region, the British

government reached, in 1872-1873, an understanding with

the Russian government recognizing the Oxus river as the

northern boundary of Afghanistan and acknowledging the

region of Badakhshan as Afghan territory. Without these

agreements, which in 1885-1887 led to the physical demar-

 381

cation of the northern boundaries of the state, the cur-

rent map of Afghanistan could have looked very different.

The British government was bent on preventing other

powers from annexing various regions of Afghanistan. Yet,

at the same time, it had not devised a clear and consis-

tent policy regarding the Afghan state. until 1872, non-

intervention and the strengthening of the Afghan state

were i ts ~:1.liding principles. Thus, during the civil war

of 1863-1868, the British government remained neutral,

but as soon as Sher Ali gained the throne, he was active-

ly helped. Between 1868 and 1873, the Afghan ruler re-

ceived 33,600 muskets and rifles -- with the necessary

ammunition --, 1,000 pistols, 50 friction tubes, one

million percussion cups, 10 pieces of artillery -- moun-

tain, siege, and howitzers --, and 2.4 million Indian

rupees in cash from the British government. The money

and the means of destruction were instrumental in the

creation of a large standing army and the reconquest of

the country by the state.

From 1872 on, relations between the two governments

symbolised by Sher Ali's refusal to draw a cash sub-

sidy of 1.2 million Indian rupees granted to him in 1873

-- started to deteriorate. Two sets of factors brought

about the clash of interests. On the one hand, the

Afghan ruler was requesting a guarantee of protection a-

gainst Russian aggression, the recognition of his des-

 382

cendants as the only legitimate rulers of the country, and

the formal acknowledgement of his youngest son as heir

apparent. On the other hand, there had been a consider-

able debate regarding the northern boundary of British

India. By 1872, a new doctrine known as the "forward

policy," which proclaimed that the Hindu Kush range formed

the "scientific frontier" of British India (Fraser-Tytler,

1967:70f£) had gained acceptance among British policy-ma-

kers.

It was in this context of events that, when Sher Ali

reacted positively to a Russian proposal to enter into an

alliance, the British invaded Afghanistan yet gain in

November 1878. Sher Ali left his capital of Kabul for

the northern region to seek the promised Russian help.

Little was he aware that the Congress of Berlin of July

1878 had resolved the outstanding differences between the

British and Russian governmentsi he was not to receive

any assistance. As he lay on his deathbed, the British

army entered Kabul unopposed and recognized Sher Ali's

eldest son -- just released from a confinement of some

six years on account of his refusal to accept the selec-

tion of his brother as heir apparent -- ruler of the

country. He promptly accepted the terms of a treaty dic-

tated by Brit~in.

The colonial policy-makers, having forgotten the des-

truction of the entire British army, the expenditure of

 383

thirteen to twenty million pounds and a permanent increase

of charges on India of 5,500,000 pounds (Dupree,1973/1980:

400), were overjoyed by the ease of their entry into the

capital. By April 1881, as a result on the one hand of

successful revolts led by a combination of religious lead-

ers (Pashtun and Tajik khans) with military leaders (of

all ethnic groups) and, on the other hand, of the elec-

toral victory of Gladstone's Liberal party, the British

forces again withdrew from Afghanistan.

The single major defeat of the British army was ef-

fected at the hands of a force mostly drawn from the

Afghan army, on July 27, 1880. Prior to that date, the

British were planning to sever Kandahar and the southern

region from Afghanistan. The army, on which Sher Ali had

spent the bulk of the state resources, had ensured that

it remained part of the country. During our last conjunc-

tural unit of analysis, 1880-1901, except for a brief

encounter with Russian forces in 1885, the army did not

engage against any major foreign power. It was, however,

constantly employed in the reconquest of every part of

the country, in some instances several times over.

Before the withdrawal of their forces, and in return

for the control of the foreign relations of Afghanistan,

the British had recognized the claims of the son of a

brother of Sher Ali as the ruler of Aghanistan. His name

was Abdur Rahman and he was to rule till 1901. During

 384

the civil war of 1863-1868, he had twice acted as king-

maker, first placing his own father, then his father's

brother from the same mother, on the throne of Kabul.

Defeated by Sher Ali, he had gone into exile in central

Asia where he lived as a pensioner of the Russian govern-

ment. In February 1880, he crossed the Oxus river with

some one hundred followers. By July of the same year, he

had entered the eastern region and declared himself Amir.

Hasan Kakr has provided a detailed descriptive account

of Abdur Rahman's military campaigns, from his first

battles in 1880 to the conquest of Kafiristan and the

forceful conversion of its people to Islam in 1896. He

has also recorded the major features of his administra-

tion in two books written in English (1971;1979). I

shall, therefore, concentrate only on those aspects of

Abdur Rahman's policies which had major implications for

the de structuring and restructuring of social relations

in the country. This requires that I take account of the

main features of the relations between the state, classes,

strata and ethnic groups at the beginning of his reign.

The available evidence strongly suggests that

throughout the years 1747-1880, most local power-holding

families in most regions had managed to reproduce them-

selves socially and politically. But the encounters be-

tween the centralizing government and local forces since

1826 had considerably modified the situation prevailing

 385

in the conjunctiral period of 1747-1818. By 1878, Afghans

in most regions had been compelled to pay some form of

taxation to the central government. This can be seen by

a comparison of the revenues of the central government

at different periods. Before 1880, the value of the

Anglo-Indian rupee relative to the Afghan rupee was rough-

ly equivalent; after 1880, there was great fluctuation in

favor of the Indian rupee. The revenue figures from 1857

onwards were estimates given by Afghan rulers to British

officials. The figures are:

late 1820s 1.4 ~illion rupees (Masson,vol.I,1842/1974:250)

late 1830s 2,509,238 rupees (Mohan Lal,contemporary observer,

quoted in Habibi,1970,vol.II:118)

1857 3.5 million Indian (Afghanistan Imperial Gazetteer,

rupees 1908:40)

1869 7 million Kabuli (Ibid)

rupees

1874 (exclusive of the (Ibid)

northern region)
10 million Kabuli
rupees

1885 10 million Indian (Ibid)

rupees

The regional breakdown is only available for the

year 1885: the eastern region and the revenue-paying

parts of the central region, about 5 million Indian ru-

pees; the northern region, 1.4 million; the southern

region, 1.35 million; the western region, 1.15 million;

 386

and Badkhshan, nearly 500,000 Indian rupees (Ibid). All

regions of the country were not equally subjugated to the

state and, within a region, taxes differed from one 10-

cality to another. Furthermore, the state was not in

control of all the revenue from the regions and had to re-

distribute a considerable part of its assumed income to

other forces in society.

The situation involved a broad range of classes,

strata, and ethnic groups, as the data on the expendi-

ture of the very large district of Jalalabad, in the

eastern region, indicates. In 1878, British officials

who took over the administration of the district provided

the following information (GA,Kabul:205): the total re-

venue of the district amounted to 697,038 Kabuli rupees.

Of this amount, 75.53% was derived from revenue, 19.3%

from trade related taxes, and the rest from a range of

other taxes. Of t:le fixed expenditure of the district

of Jalalabad, 37.31% was devoted to the following five

items:

"1. Wazifas -- These are allowances in cash

and grain made to men of priestly and religious
classes .•• The total amount given away under
this head is Rs.56,621 per annum. Every learned
and religious person in the district appears to
be in receipt of a wazifa ...
"2. Malikana -- These are small allowances
in cash or kind granted to the headmen of the vil-
lages. They amount to Rs.16,521 per annum ...
"3. Takhfif -- Is the name applied locally
to remission of revenue, or payments on account
of diluvion, etc. In 1879, the amount so paid
was Rs.3,614.
"4. Jagirs -- These are remission or assign-
 ,~ 387

ments of revenue in favour of certain Khans,

Sardars, and Maliks residing in the district.
They amount to Rs.15,585.
"5. Tankhwah-i-\~ilayati Under this pe-
culiarly named heading [lit. provincial allow-
ances] are included all the allowances made to
independent and semi-independent tribes, e.g.
Afridis, Shiwaris, Mohmands,etc; also the sal-
aries granted to the officials and chief men
of the district. They amount to Rs.I,67,7l5."

The khans of Mohmand clans in addition to receiving sub-

sidies had retained, until 1879-1880, the right to levy

their own taxes on goods passing through their territory

(Merk,188l:3) .

Throughout the period from 1747 to 1878, local for-

ces in the intermediate and marginal zones had engaged

in conflict with neighbouring groups without reference

to the state. On occasions, a combination of local for-

ces would manage to inflict heavy defeats on armies of

the state sent to conquer them. It was the ability of

the Pashtun clans, today living in the "tribal agencies"

of the north-west frontier province of Pakistan, to re-

sist the armed encroachment of the Afghan state (Faiz

Muhammad,vol.II,1913:198), that forced Dost Muharrmad and

the subsequent rulers of the country to expand into the

northern region and Badakhshan while maintaining an in-

direct influence among these clans. During the reign of

Abdur Rahman, the clans were formally severed from

Afghanistan through the demarcation of the boundaries of

"British India," an event which has, to a great degree,

 388

determined the conflict of interest between the govern-

ments of Afghanistan and Pakistan in the twentieth cen-

tury.

Although the territorial limitations of Abdur

Rahman's sphere of operations were thus defined by out-

side powers, he showed every sign of exerting state sov-

ereignty within that territory. To this end, he fought

four major civil wars and some one hundred major and mi-

nor revolts -- including a ten-year war with the

Shinwaris and other Pasthun clans in the district of

Jalalabad whose financial privileges and tax-exempt sta-

tus he discontinued. Towards the end of his reign, he

told a British engineer who worked for him that he had

"ordered over a hundred thousand to be executed" (Martin,

1907:157). His policies, however, were not based on

violence alone but involved an attempt fundamentally to

restructure the relations between state and society.

Abdur Rahman's image of the state is revealed in a

number of religious handbo'oks, issued under his instruc-

tions, and drawn on for preaching by religious dignitar-

ies across the country. It was asserted that "without

the monarchy, there would be encroachment of everyone on

everyone, and this would bring about certain chaos"

(Resala Mawaza,1893:61). Kings are not accountable for

the oppressiveness of their rule since injustice "is a

result of our sins, not of faults on their part" (Abu

 389

Bakr,Taqwim Din,1888:116). Similarly, it was claimed that

as all Islamic activities would come to a standstill

without a king, even the rule of an evil king was better

than perpetual mischief" (Ibid:121-122)

Abdur Rahman's conception of the monarchy was not

shared by other leading contenders for the throne, two

of his father's brothers I sons. Ayub Khan, a son of

Sher Ali, commander of the forces that defeated the Bri-

tish in 1881 and in control of both the western and sou-

thern regions, wrote to Abdur Rahman:

"We were both in exile, but now God has given

us Afghanistan between us ... What I propose
is this: do go to Turkistan [the northern re-
gion], which was your father's share. Let the
son of Azim Khan take Kurram. Let Kabul go to
Yakub; Kandahar to the sons of Amin Khan;
Girishik to Hashim; Herat to mell (quoted in
Kakar,H.1971:77).

At the same time, Ishaq Khan, referred to by Ayub as the

son of Azim Khan, who controlled the northern region,

dre\" on the same set of precedents, declaring that

"Turkistan was acquired at the cost of my bro-

ther's life. Your father during the reign of
Amir Dost Muhammad Khan -- on whose throne you
are now seated -- was acknowledged the semi-
independent ruler of Turkistan, and only under
exceptional and pressing circumstances did he
assist his father, the Amir. Under these cir-
cumstances, it is proper and right that you
should acknowledge me as the exclusive owner
of Turkistan, liable to render you help in case
of extreme emergency. My father was Amir as
well as yours, and my claims must not be over
looked II (quoted in Adamec,1975:168).
 390

Abdur Rahman managed to defeat Ayub Khan in battle in

1881, establishing his rule over the southern and western

regions, but the northern region came under his authority

only after he defeated Ishaq in 1888.

Opposition to Abdur Rahman's policies was not confined

to members of the royal lineage but included every politi-

cally influential sector of Afghan society. To consolidate

the power of the state, Abdur Rahman undertook both to re-

press this general opposition and to reintegrate these

groups within the structure of the state. In order to

analyze this complex process, I shall examine the inter-

related changes in the control of the means of destruction,

administration, persuasion, adjudication, and finance.

When Abdur Rahman acceded to the throne, the destruc-

tive impact of the British armed intervention was clear.

In the words of the Viceroy of India, "the military opera-

tions in the country [Afghanistan] during the years 1378-

79-80, and its prolonged occupation by British troops, have

left the Civil Government and the military resources of the

Afghans in a state of dilapidation which will require a

long time to repair" (FO 539/19,May 8,1881:258). By con-

trolling the supply to the ruler of the Afghan state of

the means of destruction as well as money, British policy-

makers were able significantly to condition the pace of

restructuring the relations of domination in the country.

They were aware of their weight, as is shown in the fol-

 391

lowing passage of a letter, sent to Abdur Rahman in 1889,

in answer to his insistence on conducting internal affairs

independently and his refusal to open up the country to

British influence. The letter read:

"I may, perhaps, remind your Highness that since

the time when, nine years ago, you were placed
on the Throne of Afghanistan by the Government
of India, you have received from successive Vice-
roys a steady support, both moral and material,
which has contributed in no small degree to the
permanence and stability of your rule. It has
been closely shown to the world that the Govern-
ment of India was on your side; those who re-
belled or intrigued against you have been dis-
couraged; and you have received, in money and
military stores, assistance to the following
extent: 144 lakhs [one lakh = 100,000 rupees];
74 guns with ammunition in proportion; 25,000
breach-loading and 11,500 muzzle-loading ri-
fles, with several million rounds of cartridges.
You are receiving from my Government a lakh of
rupees every month" (Viceroy of India to Abdur
Rahman,FO 539/44.Sept.lO,1889:15).

The timing of the delivery of these armaments was

often crucial to the success of Abdur Rahman's policies.

In the midst of a revolt of Ghilzai Pashtuns (1886-1887)

and just before the outbreak of Ishaq Khan's revolt of

1888, the British speedily arranged for the delivery of

one million "Martini-Henry and the same number of Snider

Cartridges" (FO 539/36.Inc.4-43,Aug.28,1887:31). Abdur

Rahman had requested that the cost of cartridges be de-

ducted from his subsidy, but the Viceroy informed him that

he had "much pleasure in presenting the ammunition to you

as a gift, and as an indication of the unfailing good-will

of the British government" (Ibid,5-43,Sept.15,1887:32).

 392

On the other hand, the British also made use of their

control over the timing and delivery of means of destruc-

tion to extract concessions from the Amir. In 1893, when

Abdur Rahman finally agreed to a demarcation of the boun-

dary with British India, which removed the Pashtun clans

in the present territory of Pakistan from under his in-

fluence, his yearly subsidy was increased by 600,000 ru-

pees. Even more important to the Amir was the agreement

that "restrictions on purchase of arms and ammunition

will be removed. Ameer will also receive presents of

arms and ammunition from British Government" (Viceroy of

India to Secretary of State for India,FO 539/64,Nov.8,1893:

89). Indeed, the Amir had embarked on a full program for

the production of means of destruction at home but, given

the landlocked nature of Afghanistan and the costs of

transportation involved, he needed the cooperation of Bri-

tish officials in the procurement of the needed supplies.

To fight his wars, the Amir needed soldiers. The

size of the army had, throughout his reign, been estimated

at 79,000 men (IO,Military Report on Afghanistan,1906:ll2).

Besides regular troops, Abdur Rahman made use of quasi-

regular and irregular troops. In 1885, the pay of all

these forces amounted to 7,262,670 rupees. This consti-

tuted 58.6% of the total expenditure of the state which,

including the subsidy from the British government, was

at 12,389,492 rupees (Amir to Viceroy, FO 539/27,July 18,

 393

1885:216-222) .

As the reign of Abdur Rahman witnessed some of the

most extreme fluctuations in the availability and price

of food, there might have been a very strong incentive for

the hard-hit inhabitants of central and peripheral areas

to volunteer for relatively lucrative service in the army.

Enf0rcing the principle of collective responsiblity -- as

demonstrated in the material in the court records of the

valley of Kunar -- the Amir tried to discipline his sold-

iers. Throughout his reign, he also called on the ser-

vices of the cavalry, supported and commanded by local

landowners; he even made use of them in times of peace

and brought them under more strict discipline (Faiz

Muhammad,vol.III,hereafter ST,1915:289,1153). Abdur

Rahman's basic goal, inspired by tha example of the Otto-

man Empire (Abdur Rahman,1886:2-8), was that of general

conscription. In 1885, he ordered that people be classi-

fied in groups of 20; each year, one of them would serve

at the borders while the others provided for him and his

family. In 1895, he was demanding that one in every

eight men, between the ages of 15 and 20, should report

for duty (ST:541,777,872,and 1217).

Adbur Rahman also made extensive use of the hostility

among different clans and ethnic groups to suppress the

people who revolted. The army and other institutions of

the state were effectively operated to regulate the access

 294

of all social groups to means of destruction. In cases of

populations where revolts had been constant, a continuous

policy of confiscating arms was put into effect.

Ruthless utilization of means of destruction insured

the de-structuring of existing forms of relations of dom-

ination. To restructure these relations, a new set of re-

lationships had to be institutionalized. Abdur Rahman's

enduring success lay in the fact that he made use of exist-

ing elements to bring about a structural transformation of

relations between the state and the social order. The

new role of the religious strata, the khans, and the royal

lineage can illustrate the extent of the changes.

The revolt of 1886-1888, known as the Ghilzai upris-

ing but in fact consisting of some Ghilzai Pashtuns,

Hazaras and Tajiks, was led by a religious leader whose

father had emerged as the single most prominent leader of

the struggle against the second British occupation of

Afghanistan. The son actually assumed the title of Arnir

and set up a rival administration to that of the state

(ST:5l8ff). In this revolt, as in many others, the cen-

tral issue was that of centralization of power and focused

on the obligations demanded of the population in the name

of Islam.

In 1885, Abdur Rahman appointed Maulawi (religious

title) Ahmad Jan Khan Kandahari as the head of the depart-

ment of estimations and accounting. Ahmad Jan Khan immedi-

 395

ately "undertook the measurement of lands and properties

of the subjects and, on the basis of acreage, increased

the taxes of the state" (ST:475). In accordance with the

Islamic theory of taxation, he demanded one fourth of the

produce from irrigated lands and one-tenth from lands that

were rain-fed. In addition, he imposed taxes on flocks,

gardens, marriages, inheritance (0.5%), and on all other

transactions in the country. He decided "the confiscation

of awaqaf and the assignment of their revenue for the up-

keep of imams [religious dignitaries in charge of prayer]

and muizins [persons who give the call to prayer] of the

mosques as well as the expenses incurred in providing

covering, candles, repairs, etc., of the mosques. He also

investigated all edicts by which, in the name of knowledge

and descent from the Prophet [of Islam], former rulers

had assigned stipends to all sorts of people. Since most

recipients had no inkling of knowledge and, with time,

edicts had passed hands enabling people to draw undeserved

sums from Bait ul-Mal [the Islamic treasury], he cancelled

these privileges. Only those whose claims were valid were

given new orders, stamped with the seal of His Majesty ...

Finding these measures in accordance with political law

and sharia, His Majesty approved them" (Ibid).

Ahmad Jan Khan also compiled two handbooks, known as

Asas-ul-Quzat (The Guide to Judges) and Dastur ul-Amal-i-

Hukam (Instructions to Governors) . "as he was an extreme-

 396

ly able writer [general form implying knowledge of law,

finances, and composition], he was able to employ many peo-

ple who had previously had no access to the diwan [admini-

stration] of the state. Many new departments came into

being" (Ibid:63). Furthermore, the Amir uniformally im-

posed taxation on all hitherto tax-exempt lands of reli-

gious scholars and officials (Ibid:878).

In most cases, alliances formed among members of the

religious stratum, khans, landowners, and the rest of the

population, actually challenging the authority of the

state through armed uprising, can be directly traced to

the imposition of these policies. Abdur Rahman success-

fully suppressed all the revolts and his measures were im-

plemented. Since his power over the whole country was

uniformly established only in 1896, resistance to his po-

licies, though widespread, assumed crisis proportions at

different times in different regions or districts. No

simultaneous uprising from all regions of the country

ever faced the Amir, and he was able to deal with the in-

dividual uprisings as they occurred.

Abdur Rahman, as well as his opponents, sought legiti-

mation in Islamic symbols. The leaders of the revolt jus-

tified their action on the basis of resistance to an un-

lawful ruler while the Amir claimed enforcement of the de-

mands of sharia. I have elsewhere, described the version

of Islam that was officially propagated as well as the

 397

other versions totally challenging its validity (1978:269-

284). The elements stressed in official handbooks in-

cluded the search for true knowledge and practice based

upon it, the absolute duty of working for livelihood, ob-

servance of contracts, and payment of zakat (taxes; parti-

cipation in jihad, service at the borders of Dar-ul-Islam,

and obedience to the ruler.

Acceptance and dissemination of these doctrines be-

came the prerequisites for the assignment of stipends to

members of the religious stratum. The members of the

committee who, under the direct supervision of the Arnir,

prepared the handbooks, were also appointed to a committee

entrusted with examining the religious knowledge of all

claimants to stipends from the state. Even members of the

most influential families of ulama were made to come to

Kabul and be examined by the committee (ST:475,516,517).

Examinations did not merely involve matters of doctrine,

but also aimed at systematically suppressing forms of or-

ganization, such as Sufi brotherhood, that had shown an

ability to challenge the state. The Arnir, himself, took

an active part interviewing these individuals. In a let-

ter dated Shawal 15, 1305 A.H. (May 26, 1888 A.D.), the

Arnir informed the governor of the southern district of

the easterm region that

"Abdul Latif Sahibzada, a dignitary of the Khost

region and a partisan of the state, carne to us.
We advised him never to participate in mystic or-
 398

ders or try to gather adherents, because all

these acts give rise to evil. He confessed
to the truth of the statement and gave a writ-
ten guarantee signed and sealed in the court
of sharia to this effect ... Therefore, we
allowed him to leave, and write that his
salary in cash and kind as registered in the
books, is to be reinstated to him. You
should pay attention that he does not act in
violation of his pledge as the condition for
his stay in that region is the observance of
the said items" (Afghan Archives, Correspond-
ence between Amir and governor,item 53).

Financial privileges, as records of the courts of

sharia of Kunar show, were renewable only on a yearly basis

and depended on the strict observance of conduct laid down

by the Arnir. A large number of spies, kept the Amir (Ibid,

Correspondence between Amir and Prince, throughout the

four volumes) his governors and qudat informed of the ac-

tivities of the individual members. As for sufi orders,

they had shown in other situations a remarkable ability

to turn themselves into secret societies and, although

the state under Abdur Rahman was able to drive them under-

ground, it was not able to destroy their organizations.

The most successful aspect of Abdur Rahman's religious

policy was his resorting to ulama to staff a centrally

directed system of courts of sharia. I have elsewhere

(1982a) provided a detailed analysis of the pervasiveness

of the theme of centralization in Asas ul-Quzat, the hand-

book compiled by Maulawi Ahmad Jan Khan, and its operation

in practice. By contrast to the situation prevailing prior

to Abdur Rahman's reign, where in some extreme cases qudat

 399

were appointed by the local khan (Merk,188l:5), during his

reign, qudat became full-time functionaries of the state

-- not even allowed to engage in teaching -- who, as mem-

bers of a hierarchial organization, had to submit regular

reports of the proceedings of their courts, through higher

echelons, to Qadi-ul-Qudat (the chief-qadi). They were

forbidden to render judgements in their homes or in mosques

and had to attend to business in the courts. The state

became an automatic party to a case of homocide and, in

addition to a blood-price of 3,500 rupees to be paid to

the family of the victim, the murderer had to pay a fine

of 3,500 to the state as well. The governor of the south-

ern district of Kabul, where feuds among individuals, lin-

eages and clans, had been very common prior to Abdur

Rahman's reign, wrote to the Arnir that the imposition of

the fine was leading to the ruin of many people and should,

therefore, be reduced. The Arnir, refusing to lower the

fine, replied that the object of the directive was the

elimination of conflict, not the collection of revenues

(ST:I065) •

Controlling feuds was not the only measure through

which Abdur Rahman led his attacks on the institutions of

the Pashtun and other clans. Sharia became the law of the

land and a vigorous attempt to suppress customary forms

of adjudication was undertaken; all the available data in-

dicates that it was successful. A detailed breakdown of

 400

the case-material from the archives of the court of Kunar

(Ghani,1982b) leads me to conclude that the fact the state

upheld the rights of women to a share of the property of

a deceased relative -- unlike custom which disinherited

them -- was a very significant factor in the ability of

courts of sharia to supercede the institutions of custom.

Decisions reached in courts of sharia, unlike those made

through the channels of custom, were sanctioned by the

power of the state and, as such, were binding on all par-

ties to a dispute. Since there were records of decisions

and disputes could be speedily terminated, recourse to

the courts soon became desirable. Thus, the existence

of conflicts of interests among members of families, other

kin groups, and localities, made the recourse to the in-

stitution of the court, and thereby further consolidation

of the state, possible.

The negative implications for local notables were

severe and there is direct evidence that, through economic

means at their disposal, they tried to prevent people from

seeking access to the courts (Afghan Archives, Correspond-

ence between the Arnir and the governor of the southern

district of the eastern region). Abdur Rahman could not,

however, tolerate the existence of autonomous centers of

power and took energetic measures completely to subordin-

ate to the state khans and other men of local influence.

He regularly inteviewed influential men from all regions

 401

of the country and personally decided their fate. His let-

ter to the governor of the southern district of the eastern

region, in 1887, provides an insight into his method of

operation. He wrote

"Regarding the influential men of the Mangal clan

sent to us, you have acted in a stupid and care-
less manner. No information with regard to in-
fluence, reasons for detention, or for return,
was provided. You should have given the name
and background of every man, stated whether two
hundred or five hundred men in the clan followed
him, and what kind of service he would be able to
render the state" (Afghan Archives,Correspondence:
11-12) .

The Arnir then informed the governor that, as a result of

the interviews, he had classified the khans into three

categories: the least influential, whom he had allowed

to return to their localities; some of the influential

ones, who were given employment by the state; and the re-

maining, who were kept as hostages (Ibid).

Those chosen for appointment were usually given com-

missions in the army and sent to areas outside their own

localities, becoming in effect hostages. A large number

of the leading khans were liquidated, forced into exile,

or imprisoned in the capital of Kabul. Colonel West

Ridgeway, British commissioner appointed to the demarca-

tion of the northern boundary of Afghanistan, interviewed

the Arnir in 1886 and reported that "he inflicts savage

punishments, but so long as the people are submissive, par-

ticularly as regards to taxation, he leaves them alone.

 402

He wages war on classes not on the masses" (Great Britain,

Biographical Accounts of Chiefs, Sardars and Others of

Afghanistan, 1887:21).

Yet, Abdur Rahman did not intend to eliminate the

khans as a class. His goal was to deprive them of autono-

mous political power and completely subdue them to the

central government. The state sill needed the khans to

perform a number of functions in the intermediate and mar-

ginal zones of the country. To this end, throughout Abdur

Rahman's rule, khans were appointed within a number of

clans and, in some cases where, due to the remoteness of

the area, khanship had not been instituted, it was pro-

moted by the Amir. In these cases, the khans were pro-

vided with funds to hire a number of retainers from their

own clan that would be ready for serving the state under

their command (ST:923). The differences in allowances to

khans varied from 2 ,000 to in some cases, even more than

50,000 rupees (Ibid:923,954).

Controlling the financial basis of power of the khans,

Abdur Rahman was able to compel them to act according ~o

his wishes. In 1897, the Amir, who had earlier confiscated

the property of the khan of the Mohammad clan -- a family

that had maintained its position since the reign of the

Mughal ruler Akbar (1556-1605) -- rest.ored him to his for-

mer position and pay. Of the 47,704 rupees in cash and

20.825 metric tons of grain alloted to the khan as his

 403

overall allowance, he was to consider 12,000 rupees as the

stipends of his mother and wife, 17,704 rupees and all the

grain as his own pay, and 18,000 rupees as the salary of

175 horsemen and infantry under his command (Ibid:954).

When regular forces of the army were not sufficient to

suppress a revolt, these levies were called upon. Loyal

khans were also appointed governors of districts outside

their own areas.

The degree of success of khankhel to maintain some

influence during the reign of Abdur Rahman was directly

related to the speed with which it managed to accept the

supremacy of the state. Ghilzai and Hazara khans, who

led their clans in some of the bloodiest battles of the

period, came for the heaviest punishment. Thousands of

Ghilzai Pashtuns were forced to leave their place of re-

sidence in the southern and eastern regions and settle in

the nort~ern region. The fate of the Hazaras was even

worst.

Prior to the reign of Abdur Rahman, Hazaras, in a

large part of the central region, had been little sub-

jected to the central government and their khans exercised

extensive power over their people. At first, Abdur Rahman

exploited local rivalries among Hazara khans to acquire

rates of taxation two or three times those paid to earlier

governments (Ibid:399,514-515,654). But khans in the in-

terior of the central region, known as Yaghistan (the

 404

land of rebellion), were not paying any revenue to the

state. In 1891, Abdur Rahman conquered all of the central

region. His repressive measures soon led to a general

uprising of the Hazara clans and, between 1891 and 1893,

a savage war raged in the central region (Ibid:583-l238).

In his war against the Hazaras, most of whom followed

the Shi'i persuation of Islam, the Arnir gained a fetwa

from Sunni ulama declaring the Hazaras "infidels" and fit

to be sold in slavery. Faiz Muhammad, the official his-

torian of Abdur Rahman's era who had full access to all

state papers and was himself a Hazara, provided a full

account of the sale of Hazaras into slavery, the execu-

tion of their khans and religious leaders, and the in-

human punishments inflicted on all the population. The

intensity of the suppression of Hazaras in the central

region was undoubtedly the most severe case of the per-

secution of a single religious-regional group by the

state. A careful reading of the account of other revolts

makes it clear that the dominant motive of the Arnir was

the complete pacification of all the population regardless

of regional, religious, or linguistic identities. However,

military forces recruited from one or several regional

groups were widely used in suppressing other regional

groups, and actors perceived their relations to the state

in even more pronounced regional terms. Assignments of

lands of rebellious groups to people from other groups

 405

(Ibid:788,924,934,982,1183,etc.) fueled the tensions even

further.

Abdur Rahman's policy of subordinating the khans to

the state, rather than liquidating them as a class, can

be documented, even in the case of Hazaras. After the

complete pacification of the central region, the privi-

leges of some khans were restored and the groups that re-

turned to their localities were treated leniently (Ibid:

l182,1237,1238,etc.). From a long-term perspective, two

of Abdur Rahman's measures were crucial in the social re-

production of khanly families within the structure of the

state in twentieth century Afghanistan. Young boys were

brought to Kabul as hostages ~s a guarantee for the loyal

conduct of their families. They were given the title of

ghulam-bachas (slave boys) attending the Amir in the court,

and were provided with an education. Many of them became

prominent officials of the state subsequently. Recruiting

the royal guard from sons of prominent families was equal-

ly significant. Originally formed from four hundred

Durani khanzadas (descendants of khans), the guard was

later expanded to include members of all other regional

groups. Yet, after the Hazara revolt, Shi'is were ex-

cluded from it (Ibid:440,979iAfghan Archives, Correspond-

ence between the Amir and prince,vol.IV,ite,7:6). Service

in the guard was considered a high mark of distinction

and candidates had to present a document certifying they

 406

were the descendants of a khan or headmen in their locali-

ties. The Arnir personally interviewed all candidates to

make sure of their credentials (Ibid:items 8 and 13:7,10).

Even members of the royal lineage volunteered to serve in

the guard (Ibid:items 6,11,12:5,9-10;ST:980).

Relations between the Arnir and members of his own

lineage, namely the Mohamrnadzais, underwent a radical

transformation during the second decade of his rule. The

implementation of his policy of centralization had brought

him into direct conflict with most of the powerful ~embers

of the lineage and they had been forced to go into exile

in India and elsewhere. In the 1880s, after the Amir had

acquired a firm grip over the country, most were allowed

to return and some were appointed to influential positions

in the bureaucracy (ST:708,842,868,902,etc.). The changed

character of the relations between the Arnir and the

Muhamrnadzai lineage was formally marked in 1892 when the

Amir sanctioned the special status of the lineage in the

state. He issued the following edict:

"Because of their kinship and solidarity with the

royal family, His Majesty has chosen the members
of the Muhamrnadzai lineage to be superior to the
Ghilzai and Durani clans and that they should be
more prosperous. Therefore, it is decided that,
in order that their lives should be more comfort-
able than that of other people, each man should
receive a yearly salary of 400 rupees and each
woman 200, so that the foundation of the state
and the dynasty be stable" (Ibid: 914) .

Muhamrnadzai women (who married outside the lineage) were

 407

also to benefit from this measure (Ibid). The Amir person-

ally checked the genealogy of the claimants before the

treasury assigned them their salaries (Archives,Correspon-

dence between Amir and prince,vol.III,2l Rabi' al Awal

1311/1893:3). Shortly afterwards, all major sub-lineages

inside Afghanistan signed, in 1896, a covenant pledging

themselves "to recognize [Abdur Rahman's] sons as inheri-

tors to his crown and kingdom; not to deviate from his will,

to keep on obeying them whether he be alive or dead ..• it

is for his sons, neither for us nor for our descendants,

to choose the inheritors" (Archives of the Afghan Ministry

of Foregin Affairs, quoted and translated in Kakar, 1971,

appendices XII and XIII:293).

Muha~~~zais were appointed to positions of influence

within the civil administration but were treated like other

members of the bureaucracy and frequently dismissed by the

Amir when their actions displeased him. The degree of

subordination of the lineage to Abdur Rahman is best re-

vealed in a request made by a large number of its members

to appoint them as "inspectors and spies" for the deter-

mination of the honesty of officials (ST:933). The Amir

did appoint some of them to inspect financial irregulari-

ties in the southern region but, when they began to make

use of their position to enrich themselves, he warned them

that "they would be executed" (Ibid:969,1058).

Abdur Rahman's policy of pacification and his system-

 408

atic demands for increasing taxes brought about an expan-

sion in the size and role of the administration. Entrusted

with the task of breaking down popular resistance to con-

solidate the state, officials made use of their wide powers

to extort the maximum of resources from the population for

their own quick enrichment (ST:475,653,700,763,788,833,94l,

l158,etc.). Tax-farmers who, as in previous periods, con-

tinued to be the main collectors of revenue, also attempted

to abuse their power (Ibid:1186,1226027,etc.).

To check the power of officials and tax-farmers,

Abdur Rahman appointed them to short terms of office and

dismissed them frequently. In addition, he attempted to

restructure the financial administration. He tried to set

up a budget; the appointed officials having proven unequal

to the task, he expanded the functions of the aUditing de-

partment. Books of the officials were regularly sent

for inspection and members of the department frequently

travelled to the provinces (Ibid:638,668-69,700,842,86l).

Furthermore, contracts of tax-farmers had to be recorded

in the courts of shari a of the locality leased to them and

qudat attended to the complaints of the subjects.

In general, the second decade of the rule of the

Amir was devoted to the institutionalization of gains

achieved through the ruthless use of the means of destruc-

tion at his disposal. The degreee of his success in re-

structuring tne relations of domination in the country was

 409

revealed on the day of his death, October 1,1901. For the

first time since the foundation of the Afghan state in 1747,

the eldest son of a ruler, though born of a concubine, suc-

ceeded his father to the throne without encountering any

armed opposition. Abdur Rahman had successully replaced

the old mode of domination with a new one where political

power resided in the institutions of a centralized state.

The ease of transition came as a surprise to the Bri-

tish officials, who had watched the scene closely and ex-

pected to gain a number of concesssions vital to their

interests from the successor of the Amir. One of these

officials, forecasting chaos in the country following the

death of Abdur Rahman, had stated that "we shall have an

opportunity of making our political and strategic fron-

tiers identical, and of completing our defences by bringing

Afghanistan proper under our effectual control, making the

continuation of railroads and telegraphs, and the opening

out of the country the conditions on which its new Ruler

is allowed to exist" (FO 539/33, Dec.20,1886:20).

Despite his dependence on arms and money, Abdur

Rahman had successfully resisted the imposition of these

measures. Similarly, he had accepted only Indian Muslims

as the British representatives to his court; they were

kept in isolation from the people and treated as virtual

prisoners (Martin,1907:301). In 1896, having completed

the consolidation of his regime, Abdur Rahman asked to

 410

send an ambassador to London and to have free hand in the

conduct of his own foreign policy (Military Report of

Afghanistan,1906:l83). His request was flatly refused and

British demands for concessions ~.vere maintained (L/P&S/18A

l896,8l:l). In that year, a British official wrote "If

only an exceptional man can hold the whole of Afghanistan,

and if the only means by which the integrity of Afghanistan

can be preserved are a cruel depotism, are the results

worth purchasing at the cost of an unfriendly Amir who

will have his own way and break his agreements?" (Ibid,l09:

6) •

The British were not to win any of their demands dur-

ing the reign of Abdur Rahman. When his son succeeded him,

they were still in no position to enforce them on the

Afghans. ~nstead of being dismembered and further en-

croached upon by imperial powers, Afghanistan won its

political independence from a British government that had

just recovered from World War I, under Abdur Rahman's

grandson, in 1919. The changes leading to that event and

the reproduction of relations of production and domination

during the twentieth century will have to await future

analysis.

Before concluding, however, and in order to complete

the picture of the Afghan state at the turn of the century,

I must briefly take account of the economic articulation

of the country's regions under Abdur Rahman.

 411

The distinctive regional patterns of agrj.cultural pro-

duction were not altered during this last unit of our con-

juntural analysis. Despite the Amir's refusal to con-

struct railways in the country and although the imposition

of high custom duties had resulted in the decline of trade

with India (Kakar,1979:213f) I changes in the larger economy

of the area and in the international system of communica-

tions were undoubtedly felt, at least in certain sections

of Afghan society and in certain parts of the country.

Nomads were most receptive to these changes. From

1840 onwards, unaffected by the cost of transportation,

nomadic sheep-owners had become involved in the sale of

their wool to firms which eventually exported it, via

Karachi, to London. The impact on the local scene, as

was the case in the Mizan district of the southern region,

translated into flock-owners maintaining the sheep "for

their wool and unwilling to part with them" (GA,Kandahar:

344). Nomads also served as carriers for agricultural

products that were either cultivated for export or gathered

especially for the Indian market. Of a total amount of

2,197,000 rupees worth of goods taken from Afghanistan to

India during this period, the main items were: fruits,

31.86%; madders, 27.31%; raw silk, 22.76%; wool, 6.8%;

and chars (an intoxicating drug extracted from hemp) ,

4.64%. The goods they transported back from India into

Afghanistan were valued at 1,907,000 rupees, with the fol-

 412

lowing breakdown: Indigo, 31.46%; English manufactured cot-

ton goods, 26.22%; Indian manufactured cotton goods, 31.46%:

tea, 5.24% (GA,Kabul:426ff). With the availability of rail-

roads, those groups of nomads who, besides carrying goods

engaged in money-lending and peddling of cloth, extended

their sphere of operations to Indian provinces as far as

Bengal, and some even went to Australia (Roninson,1935:2-

4) •

Abdur Rahman's attempts at turning the political bor-

ders into effective frontiers, where the flov1 of goods

would be controlled by the state, were not successful and

the smuggling of goods was widespread. Lord George Curzon

who, prior to his tenure as viceroy of India (1899-1905),

travelled extensively in Afghanistan and Russian dominated

central Asia, made a revealing analysis of the impact of

British and Russian railroad systems on the structure of

marketing in Afghanistan. He wrote that, from the two

termini of the railroad on the frontiers of the Afghan

southern and eastern regions

"long strings of camels convey British and Indian

merchandise into the interior: the caravans in
correspondence with the sind-Pishin line serving
the Kandahar region, and ultimately Herat; those
that start from Peshawur serving Kabul, and ulti-
mately Afghan Turkestan [the northern region].
On the north-west, the Russian railway runs par-
allel with the English, several hundred miles a-
part, but at a rather greater distance from the
Afghan frontier on that side than is the Indian
railway on the south-east. From Merv, however,
from Tcharjui on the Oxus, and from Bokhara com-
munication is made with the Afghan interior; and
 413

caravans, made up for the most part in

Bokhara, but charge with Russian merchandise,
serve the frontier markets of Maimena, Andkhui,
Shiberghan, Akcha, and Siripul, where the goods,
are redistributed onto the inland villages and
towns. The situation which it is my object to
indicate is this. The Russian, or mosre strict-
ly Bokharan, caravans, in correspondence with
the Russian railway on the north, are not only
seriously competing with, but are even beating
the Afghan or Indian caravans in correspondence
with the Anglo-Indian lines on the south. In
other words Afghanistan, which has hitherto
been regarded as a peculiarly sacred preserve
of the British or Indian trader, is fast becom-
ing a battle-ground of international rivalry,
and is little by little yielding to Russia
that which it steals from Great Britain"
(Curzon,1889:447) .

Curzon, who illustrated his observations with data

from official British and Russian sources, provided a

description of a fundamental tension in the articulation

of poliltical and economic relations in the country.

Abdur Rahman had managed to achieve an unprecedented de-

gree of political centralization. Yet the economic in-

tegration of the country had not been reached. Tensions

in the articulation of relations of domination within the

system of production, despite many changed circumstances,

remained as high as they had been at the be inning of our

unit of long duration. Conflicts and alliances had altered

relations of domination but had not transformed relations

in the social and technical organization of production.

At a time when internal political forces had been made to

converge under the sole leadership of the ruler, external

economic forces were pulling the various regions of the

 414

country away from each other. Realigning economics with

politics remained to be the strongest challenge of

Afghanistan in the twentieth century.

PART I I I
 415

Chapter VII

Conclusion

In the preceding pages I have analyzed the historical

process through which structures of production and domina-

tion were reproduced as a totality from 1747 to 1901 in

Afghanistan. In this endeavor I have paid equal attention

to structure and history. Systematic relationship among

a set of elements in such a way that changes in one effect

changes in others has been the defining characteristic of

concept of structure to me. But a combination of elements

can be called structural if, and only if, their interrela-

tionship endures the test of duration, i.e. history.

To bring the changing articulation of structures in

the total system into sharper relief I opted for a com-

bination of temporal and spatial frameworks. Within the

basic unit of long-duration, 1747-1901, I delineated appro-

priate conjunctura1 units of 20 to 60 years to document

the degree of continuity or change. Spatially, instead

of taking the whole territory of the state as the unit of

analysis I focused on six regions. within each of these

regions, I distinguished between central and marginal eco-

logical zones and then explored the interrelations among

these elements within regional, inter-regional and inter-

territorial units.

This procedure al1o~led me to chart the actual tempo

 416

of the reproduction of differeLt st~uctures and to deter-

mine the ways in which the combination of these structures

gave rise to systemic transformations. The pace of move-

ment in the structures of production being relatively

slow a unit of long-duration, 1747-1901, was the most

suitable means of presentation of elements in a structure

such as the technical and social organization of labor

as well as the relations among the entire sets of

structures in the system of production.

This slow pace of movement is not surprising. Karl

Marx provided a graphic portrayal of the differences in

the conditions of reproduction under conditions of capi-

talist and pre-capitalist production. He affirmed that

capitalist production is "not only rapid, so that the

commodity quickly acquires the form in which it is suit-

able for circulation, but it is continuous. Production

here appears only as constant reproduction and at the same

time it takes place on a mass scale" (Theories of Surplus

Value,III:286). Under these conditions "the stock of

commodities which a shopkeeper "accumulates and keeps in

store, will be small. It is different in the less devel-

oped stages of production where reproduction proceeds

slowly -- where, therefore, more commodities must remain

in the circulation reservoirs -- the means of transport

are slow, the communications difficult and, as a conse-

quence, the renewal of stock can be interrupted and a

 417

great deal ef time elapses as a result between the empty-

ing and refilling ef reserveir -- that is, the renewal ef

steck in hand. The pesitien is then similar to. that ef

preducts whose repreductien takes place yearly er half-

yearly, in shert in mere er less prelenged perieds ef

time, ewing to. the nature ef their use-values"(Ibid:287).

If these differences in centinuity, scale and tempe

ef repreductien ef ecenemic structures in capitalist and

pre-capitalist fermatiens are net clearly grasped, it

can give rise to. false abstractiens and analytic cenfu-

sien. Jenathan Friedman who., fer instance, has rightly

insisted that enly threugh repreductien dees a mede ef

preductien beceme a system (1976:3-16), articificially

impeses in his ethnegraphic analysis ef the Kachin (1972)

categeries that are specific to. capitalist preductien.

In anether piece (1975: 201-276), wi theut effering histerical

or theeretical justificatiens fer his assumption, Friedman

assumes lineages to. be the selid reck areund which his

epi-genetic medel ef evelutien unfelds itself. The ser-

ies ef magical transfermatiens is, hewever, perfermed by

deliberately breaking-dewn elements in an existing tetal

structure and re-arranging them in an alleged evelutien-

ary sequence ef transfermatiens. There is a deuble

ireny in Friedman's methed ef eperatien. In attempting

to. impreve ever Leach's werk en the Kachin, he returns

to. treating the varieus lineages as iselated systems

 418

while one of Leach's major innovations was his analysis

that these were not isolated societies but elements in a

total system in flux. Furthermore, Friedman achieves the

elegance of his model by sacrificing the very el~ment on

which he insists theoretically, namely systemic reproduc-

tion. Thus, by refusing to confront his model with tests

of concrete durations, Friedman, instead of enriching our

understanding of historical processes, impoverishes it.

The imposition of organic characteristics on social

processes is not confined to Friedman. Marx, in his dis-

cussion of the Physiocrats, asserted that "for them the

bourgeois forms of production necessarily appeared as

natural forms. It was their great merit that they con-

ceived these forms as physiological forms of society: as

forms arising from the natural necessity of production

itself, forms that are independent of anyone's will or

of politics, etc"(Theories of Surplus-Value,I:44). Sub-

sequently, Marx justified his concept of accumulation

by directly relying on Charles Darwin. He wrote that

accumulation "means assimilation, continual preservation

and at the same time transformation of what has already

been handed over and realized. In this way Darwin makes

'accumulation' through inheritance the driving principle

in the formation of all organic things, of plants and

animals" (T3V,III:294-95). Marx's reliance on Darwin was

not confined to seeking support in an isolated insistance.

 419

In a letter to Lassale he asserted that "Darwin's book is

very important and serves me as a natural-scientific

basis for class struggle in history" (January 16, 1861,

Selected Correspondence n.d.:151).

The theoretical area where Marx applied the concept

of inherited social combination was on the issue of domin-

ation. He asserted that the "political economy in its

classical period, adopted a severe attitude towards the

machinery of the State, etc. At a later stage it rea-

lized and as shown in practice too -- learnt from ex-

perience that the necessity for the inherited social com-

bination of all classes, which in part were totally un-

productive, arose from its own organization" (TSV,I:175).

In accounting for the expanding scale of primitive accumu-

lation, he put heavy emphasis on the role of the state,

stating that "the different momenta of primitive accumu-

lation distribute themselves now, more or less in chrono-

logical order, particularly over Spain, Portugal, Holland,

France, and England. In England at the end of the 17th

century, they arrived at a systematical combination, em-

bracing the colonies, the national debt, the modern mode

of taxation, and the protectionist system. These methods

depend in part on brute force, e.g., the colonial system.

But they all employ the power of the State, the concen-

trated and organized force of society, to hasten, hothouse

fashion, the process of transformation of the feudal

 417

great deal of time elapses as a result between the empty-

ing and refilling of reservoir -- that is, the renewal of

stock in hand. The position is then similar to that of

products whose reproduction takes place yearly or half-

yearly, in short in more or less prolonged periods of

time, owing to the nature of their use-values"(Ibid:287).

I t these differences in continuity, scale and tempo

of reproduction of economic structures in capitalist and

pre-capitalist formations are not clearly grasped, it

can give rise to false abstractions and analytic confu-

sion. Jonathan Fr~edman who, for instance, has rightly

insisted that only through reproduction does a mode of

production become a system (1976:3-16), articificially

imposes in his ethnographic analysis of the Kachin (1972)

categories that are specific to capitalist production.

~\{1(e.
In another (1975:201-276), without offering historical

or theoretical justifications for his assumption, Friedman

assumes lineages to be the solid rock around which his

epi-genetic model of eVQlution unfolds itself. The ser-

ies of magical transformations is, however, performed by

deliberately breaking-down elements in an existing total

structure and re-arranging them in an alleged evolution-

ary sequence of transformations. There is a double

irony in Friedman's method of operation. In attempting

to improve over Leach's work on the Kachin, he returns

to treating the various lineages as isolated systems

 418

while one of Leach's major innovations was his analysis

that these were not isolated societies but elements in a

total system in flux. Furthermore, Friedman achieves the

elegance of his model by sacrificing the very element 'on

which he insists theoretically, namely systemic reproduc-

tion. Thus, by refusing to confront his model with tests

of concrete durations, Friedman, instead of enriching our

understanding of historical processes, impoverishes it.

The imposition of organic characteristics on social

processes is not confined to Friedman. Marx, in his dis-

cussion of the Physiocrats, asserted that "for them the

bourgeois forms of production necessarily appeared as

natural forms. It was their great merit that they con-

ceived these forms as physiological forms of society: as

forms arising from the natural necessity of production

itself, forms that are independent of anyone's will or

of politics, etc"(Theories of Surplus-Value,I:44). Sub-

sequently, Marx justified his concept of accumulation

by directly relying on Charles Darwin. He wrote that

accumulation "means assimilation, continual preservation

and at the same time transformation of what has already

been handed over and realized. In this wa'5J.' Darwin makes

'accumulation' through inheritance the driving principle

in the formation of all organic things, of plants and

animals" (TSV,III:294-95). Marx's reliance on Darwin was

not confined to seeking support in an isolated insistance.

 419

In a letter to Lassale he asserted that "Darwin's book is

very important and serves me as a natural-scientific

basis for class struggle in history" (January 16, 1861,

Selected Correspondence n.d.:lSl).

The theoretical area where Marx applied the concept

of inherited social combination was on the issue of domin-

ation. He asserted that the "political economy in its

classical period, adopted a severe attitud8 towards the

machinery of the State, etc. At a later stage it rea-

lized and as shown in practice too -- learnt from ex-

perience that the necessity for the inherited social com-

bination of all classes, which in part were totally un-

productive, arose from its own organization" (TSV,I:17S).

In accounting for the expanding scale of primitive accumu-

lation, he put heavy emphasis on the role of the state,

stating that lIthe different momenta of primitive accumu-

lation distribute themselves now, more or less in chrono-

logical order, particularly over Spain, Portugal, Holland,

France, and England. In "England at the end of the 17th

century, they arrived at a systematical combination, em-

bracing the colonies, the national debt, the modern mode

of taxation, and the protectionist system. These methods

depend in part on brute force, e.g., the colonial system.

But they all employ the power of the State, the concen-

trated and organized force of society, to hasten, hothouse

fashion, the process of transformation of the feudal

 420

mode of production into the capitalist mode, and to shcrt-

en the transition. Force is the midwife of every old so-

ciety pregnant with a new one. It is itself an economic

power" (Capital,I:175).

This passage is in sharp contrast to his remarks on

the achievements of physiocrats. Instead of explaining

how the state arose out of the "physiological forms of

society" he brings in the state, i.e. politics, to ex-

plain the emergence of the capitalist mode of production.

The contradiction between the two remarks is due to Marx'

systematic theoretical neglect of the analysis of rela-

tions of domination as structural relations that constant-

ly affected the relations of production. If we are to

grasp the totality of social reproduction of systems in

history we can no longer limit our attention to the struc-

tures of production and ignore structures of domination.

Understanding the genesis and reproduction of the absolu-

tist state, in conjunction with changes in the economic

structures of societies, becomes an important task.

In his contribution to the understanding of the char-

acter of the absolutionist state, Perry Anderson wr5:tes

that "in nature and structure, the Absolutist monarchies

of Europe were still feudal states: the machinery of

rule of the same aristrocratic class that had dominated

the Middle Ages. But in Western Europe where they were

born, the social formations which they governed were a

 421

complex combination of feudal and capitalist modes of pro-

duction, with a gradually rising urban bourgeosie and a

growing primitive accumulation of capital, on an interna-

tional scale ..• The increase in the political sway of

the royal state was accompanied, not by a decrease in the

economic security of noble landownership, but by a corre-

sponding increase in the general rights of private pro-

perty. The age in which 'Absolutist' public authority

was imposed was also simultaneously the age in which

'absolute' private property was progressively consolidated"

(1974:428-29) •

Anderson's observations offer an interesting contrast

to the situation in Afghanistan. As I have shown in chap-

ters five and six, the structure of domination underwent

rapid changes while, the structure of production regis-

tered very slow changes. Furthermore, Anderson's juxta-

position of the absolutist state and modes of production

is important in focusing attention on their interrelation,

yet he has not demonstrated the precise nature of this

"complex combination".

Niels steensgaard has noted that "Marxist scholars

have so far have been unable to determine the nature of

the absolutist state in relation to modes of production"

(1981:253). He points out that "relagating the profits

of the large trading companies and the early colonial

ventures to the category of plunder and windfall gains

 422

is highly unsatisfactory, considering the place they occupy

in the preindustrial economy. Ignoring the enormous con-

centration of demand resulting from the consolidation of

the early modern state is even more unsatisfactory" (Ibid:

254). Using Fredrick Lane's concept of cost of protection,

Steensgaard argues that, "in the centuries before the fi-

nal introduction of a capitalist economy, the dynamics

of the European economy were guaranteed by the competition

among the entrepreneurs in the protection business -- the

only large-scale business that always could find a market.

Parasitic empires have often been the victims of their

own greed when they killed the goose that laid the golden

egg. The Western European governments of the Ancien

Regime were the first parasites to become the victims of

their farsightedness when they fed the goose so well that

it became able to take over. By protecting the productive

investments and innovations the seigniorial sector had

forged the weapon of its O'Yffi undoing" (Ibid:272).

My analysis of the changing balance between the cen-

tral government and the class of khans over the control of

means of destruction and imposition of costs of protection

reveals how different the situation was from that in Wes-

tern Europe. It underlines the crucial nature of the link-

age between the central government and the European power

as suppliers of means of destruction, either directly or

by allowing the purchase of arms in the international mar-

 423

keto

The character of the relationship between the absolu-

tist state, the nobility and the peasantry, however, was

not uniform throughout west Europe. Robert Brenner char-

acterizes the contrast between the French and the English

experiences in the following words: IIBy the early modern

period the consolidation of peasant property in relation-

ship to the development of the French state had created a

very different sort of class structure in the French coun-

tryside from that which emerged in England. And there

is no better index of these contrasting structures than

the dramatically different sorts of peasant revolts which

marked the early modern era in both countries. In England,

of course, peasant revolt was directed against the land-

lords, in a vain last-ditch struggle to defend disinte-

grating peasant proprietorship against advancing capita-

list encroachment. In France the target of peasant revolt

typically, the crushing taxation of the absolutist

state, which ironically had been instrumental in securing

and protecting peasant proprietorship (and thus impending

capitalist development). Thus in France strong peasant

property and the absolutist state developed in mutual

dependence upon one another. The state increased its

own power by virtue of its ability to get between the land-

lords and the peasants, to ensure peasant freedom, hered-

itability and fixed rents, and thuse use peasant produc-

 424

tion, via non-parliamentary taxation, as the direct source

of reVeii.'-l6 for royal strength and autonomy... In England,

by constrast, monarchial centralization developed, especi-

ally from the latter fifteenth century, in relationship to

and with ultimate dependence upon the landlord class, as

was most dramatically evidenced in the contemporaneous

growth of parliamentary institutions (while they decayed

in France)" (1976:70-71).

In my analysis of conjunctures of domination I showed

that the relationship of the khans with the central govern-

ment underwent fundamental changes during my unit of long-

duration. There is marked contrast between the two per-

iods 1747-1818 and 1880-1901. During the first period

the khans enjoyed maximum control over the means of de-

struction, administration, finance, adjudication and per-

suation. During the second conjuncture their powers were

derived from the state. Through institutions such as

the standing army, the centrally directed bureacracy,

the judiciary, the religious establishment, a~d the con-

trolled - if not directed - system of finance, the state

was able to get between landlord and direct producers.

Through its redistributive network, the state was able

to directly influence the direction of mobility of in-

dividual landlords as well as members of other sectors

of society.

However, as long as centralization of power was

founded on an economic structure with a slow tempo of
changes, the reproduction of political centralization re-
mained precarious. For as John Austi~, some one-hundred
and thirty years ago pointed out the tendency "to streng-
then the central authority is aided or impeded by causes
extrinsic to political institutions. Such for example
is the facility or difficulty of communication between the
parts of the country; the diposition of their respective
populations to union or separation; the attachment or
aversion of the bulk of the community to the actual hold-
ers of the sovereign power; and the docility or indocility

inherent in the national character" (1847: 226) . Austin also dis-

tinguished between formal and practical centralization, assert-
ing that" in imperfectly civilized countries whose governme!lts
are apparently centralized (as, for example Russia,) the cen-
tralization is really formal. In such a country, the adrnini-
strative authority, however perfect their arrangement, are
not practically dependent on the sovereign power" (Ibid: 234) .
Austin's characterization of centralization is con-
siderably more sophisticated than Karl wittfogel's attempt
to derive the structure of domination in "Asiatic" soci-
eties from their mode of irrigation. By avoiding to con-
cretely investigate his theory in definite historical
periods, Wittfogel often reduces his effort to a random
catalogue of confirmations of his thesis. Furthermore,
for someone who advocates the comparative nature of
 426

the social sciences, his abstraction of the notion of

private property derived from L~e capitalist period of

European history and his silence on the character of

European absolutist states is amazing.

If this work has revealed the inadequacy of such

misplaced concreteness and shown that the anthropological

mode of analysis is capable of posing -- if not success-

fully answering -- the issue of articulation of complex

structures through time, it would have served its purpose.

 427

BIBLIOGRAPHY

Theoretical Literature

Arensberg, Conrad
1957 "Anthropology As History"
in Karl Polanyi, Conrad Arensberg & Harry
Pearson, eds. Trade and Market in the Early
Empires
Glencoe, Illinois: The Free Press pp.97-114

Anderson, Perry
1974 Lineages of the Absolutionist State
London: New Left Books

Baran, Paul
1957/1968 The Political Economy of Growth
New York: Modern Reader

Bendix, Reinhard
1962 Max Weber, An Intellectual Portrait
New York: Doubleday and Company

1978 Kings Or People Power and the Mandate to Rule

Berkeley: University of California Perss

Bloch, Marc
1971 Feudal Society
trans. by L.A. Manyon
2nd edition London: Routledge and Kegan Paul

Bohanan, Paul
1954 "The Migration and Expansion of the Tiv"
in Africa XXIV:2-16

Bourdieu, Pierre
1972/1977 Outline Of A Theory of Practice
London: Cambridge University Press

Braudel, Fernand
1958/1972 "History and the Social Sciences"
in Peter Burke, ed. Economy and Society in Early
Modern Europe
New York: Harper & Row

1966/1973 The Mediterranean and the Mediterranean World

in the Age of Philip II
2 vols. New York: Harper & Row
 428

Chi, ChIao Ting

1936/1970 Key Economic Areas of China
New York: Augustus M. Kelly Publishers

Cohen, Myron
1976 House United House Divided The Chinese Family
in Tai\van
New York: Columbia University Press

Dalton, George
1960 "A Note of Clarification On Economic Surpluses"
American Anthropologist 62:483-490

1974 "How Exactly Are Peasants Exploited?"

American Anthropologist 76:553-561

Derman, William and Levine, Michael

1977 "Peasants, Propaganda, Economics, and Exploita-
tion: A Reply to Dalton"
American Anthropologist 79:119-125

Evans-Pritchard, E.E.
1940/1977 The Nuer
New York: Oxford University Press

Fortes, Meyer
1949 "Time and Social Structure: An Ashanti Case
Study" in Meyer Fortes, ed. Social Structure
Studies Presented to A.R. Radcliffe-Brown
Oxford: Oxford University Press

Fried, Morton
1952 "Land Tenure, Geography and Ecology in the CC71-
tact of Cultures"
American Journal of Economics and Sociology
11:331-412

1957 "The Classification of Corporate Unilineal

Descent Groups"
Journal of the Royal Anthropological Institute
87:1-29

1967 The Evolution of Political Society

New York: Random House

1968 Introduction to Paul Kirchoff's "The Principles

of Clanship in Human Society"
in Morton Fried, ed. Readings in Anthropology
vol.II
New York: Thomas Y. Crowell Co.
 429

Fried, Morton (continued)

1975 The Notion of Tribe
Menlo Park, California: Cummings Publishing Co.

Friedman, Jonathan
1975 "Religion As Economy and Economy as Feligion"
Ethnos 40 (I-IV): 46-63

Friedmann, Maurice
1958/1970 Lineage Organization in South-Eastern China
New York: Humanities Press

Godelier, Maurice
1966/1972 Rationality and Irrationality in Economics
New York: ~-1onthly Review Press

1973/1977 Perspectives in Marxist Anthropology

London: Cambridge University Press

1975a "Economy and Religion: An Evolutionary Optical

Illusion" in Johnathan Friedman and M.J.
Rowlands, eds. The Evolution of Social Systems
Liverpool: Duckworth pp.3-13

1975b "Politics As 'Infrastructure': An Anthropolo-

gist's Thoughts on the Example of Classical
Greece and the Notion of Relations of Produc-
tion and Economic Determination"
in Jonathan Friedman and M.J. Rowlands, eds.
The Evolution of Social Systems

Goody, Jack
1971 Technology, Tradition, and the State in Africa
London: Oxford University Press

1976 Production and Reproduction A Comparative

Study of the Domestic Domain
London: Cambridge University Press

Hammoudi, Abdallah
1980 "Segmentarity, Social Stratification, Political
Power and Sainthood: Reflections on Gellner's
Theses" Economy and Society 9:280-303

Harris, Marvin
1959 "The Economy Has No Surplus?"
American Anthropologist 61:189-199

Jones, Richard
1831 Distribution of Wealth
London: John Murray
 430

Jones, Richard (continued)

1833 An Introductory Lecture on Political Economy
London: John Murray

Kirchoff, Paul
1935/1968 "The Principles of Calnship in Human Society"
in Morton Fried, ed. Reudings in Anthropology
vol. II
New York: Thomas Y. Crowell Co.

Kula, witold
1962/1976 An Economic Theory of the Feudal System
trans. by Lawrence Garner
London: New Left Books

Lane, Fredrick
1942/1966 Venice and History
Baltimore: John Hopkins University Press

Lattimore, Owen
1940 Inner Frontiers of China
New York: American Geographical Society

1947/1962 "Inner Assian Frontiers: Chinese and Russian

Margins of Expansion"
in Owen Lattimore Studies in Frontier History
London: Oxford University Press pp.134-l64

1956/1962 "The Frontier in History"

in Owen Lattimore Studies in Frontier History
London: Oxford University Press pp.469-500

Leach, Edmund
i954/l964 Political Systems of Highland Burma
Boston: Beacon Press

Levi-Strauss, Claude
1955/1977 Tristes Tropiques
trans. by John and Doreen Weightman
New York: Pocket Books

1962/1966 The Savage Mind

Chicago: The University of Chicago Press

1963 Structural Anthropology

trans. by Claire Jacobson and Brooke Grundfest
Schoepf
New York: Basic Books
 431

Marx, Karl
1983/1971 Theories of Surplus-Value Part Three
Moscow: Progress Publishers

Selected Correspondence
Moscow: Progress Publishers

1867/1975 Capital vol.I

New York: International Press

1894/1977 Capital vol. III

New York: International Press

1859/1971 A Contribution To the Critique of Political

Economy
London: Lawrence & Wishart

Mill, John Stuart

1864 Principles of Political Economy With Some of
Their Applications to Social Philosophy
2 vols. London:

Moore, Jr., Barrington

1966 Social Origins of Dictatorship and Democracy
Boston: Beacon Press

Murphy, Robert F.
1971 The Dialectics of Social Life
New York: Basic Books

Newcomer, Peter
1977 "Toward a Scientific Treatment of 'Exploitation':
A Critique of Dalton"
American Anthropologist 79:115-119

Orlove, Benjamin
1980 "Ecological Anthropology"
in Annual Review of Anthropology IX:235-273

Pearson, Harry
1957 "The Economy Has No Surplus: Critique of a
Theory of Development"
in Karl Polanyi, Harry Pearson, and Conrad
Arensberg, eds. Trade and Market in the Early
Empires
Glencoe, Ill.: The Free Press pp. 320-341

Polanyi, Karl
1968 Primitive, Archaic and Modern Economics Essay
of Karl Po1anyi
George Dalton ed.
Boston: Beacon Press
Polanyi, Karl (continued)
1977 The Livelihood of Man
Harry Pearson, ed.
New York: Academic Press

Radcliffe-Brown, A.R.
1935/1952 IIPatrilineal and Matrilineal Succession"
in A.R. Radcliffe-Brown Structure and Function
in Primitive Society
New York: Free Press pp.32-48

1950 "Introduction ll
in A.R. Radcliffe-Brown and Daryll Ford, eds.,
African Systems of Kinship and Marriage
London: Oxford University Press pp.1-85

Ricardo, David
1821 On the Principles of Political Economy and
Taxation
London:

Sahlins, Marshall
1961/1967 "The Segmentary Lineage: An Organization of
Predatory Expansion:
in Ronald Cohen and John Middleton, eds.
Comparative Political Systems
Austin: University of Texas Press pp.89-ll9

Smith, Adam
1776/ An Inquiry Into the Nature and Causes of the
Wealth of Nations
New York: Modern Library

Smith, Carol
1974 "Economics of Marketing Systems: Models From
Economic Geography"
in Annual Review of Anthropology 1I:167-20l

1976 "Regional Economic Systems: Linking Geographi-

cal Models and Socioeconomic Problems"
in Carol Smith, ed. Regional Analysis vol.1
New York: Academic Press pp.3-63

Tambiah, Stanley
1976 World Conqueror and World Renouncer. A Study
of Buddhism and Polity in Thailand Against a
Historical Background
New York: Cambridge University Press

Thompson, E.P.
1978 The Povertv of Theory and Other Essays
London: Me~lin Press
 433

Vincent, Joan
1977 "Agrarian Society as Organized Flow: Process of
Development Past and Present"
Peasant Studies IV:

Von Savigny
1884 Jural Relations or the Roman Law of Persons
As Subjects of Jural Relations
trans. by W.H. Rattigan
London: Wildy and Sons

Von Thunen, Johann Heinrich

1826/1966 Isolated State
trans. Peter Hall
London: Pergamon Press

Weber, Max
1956/1978 Economy and Society
2 vols. Berkeley: University of California Press

Wittfogel, Karl
1938/1969 "The Theory of Oriental Society"
in Morton Fried, ed. Readings in Anthropology
vol.II
New York: Thomas T. Crowell Co. pp. 179-198

1957/1976 Oriental Despotism

New Haven: Yale University Press

Wolf, Eric
1969 Peasant Wars of the Twentieth Century
New York: Harper and Row
 434

Biographies, Autobiographies, and Contemporary Accounts in

Persian and Pashtun

Afghanistan National Archives

1887-1900 Correspondence Between Arnir Abdur Rahman and
Prince Habibul1ah vo1s. I-IV

1887-1891 Correspondence Between Arnir Abdur Rahman and

the Governor of the Southern District of the
Eastern Region

1885-1890 Diwan-e-Qaza-i-Kunar ("Court Recrods of Kunar")

Published Sources

Abdur Rahman, Arnir

1886 Sar Reshta-i-Is1amia-i-Rurn ("The Organization
of the Islamic Ottoman State")
Kabul: Government Press

Abru, Hafiz
1970 Geographia-i-Hafiz Abru ("Geography of Hafiz
Abru") e.d Mayil Hirawi
Tehran: Bunyad-i-Farhang

Abu Bakr, et al.

1888 Taqwim Din ("The Essence of Religion")
2nd editlon
Kabul: Goverment Press

Abul Fazl
1886 Akbar Namah ("History of Akbar")
3 vols. ed. Mau1awi Abdur Rahim
Calcutta: Baptist Mission Press

1880 Ayeen Akbery or the Institutes of the Emperor

Akbar
trans. by F. Gladwin 2 vo1s.
London: J. Swan and Co.

Afghanistan
1977 Statistical Information of Afghanistan 1976-
1977
Kabul: Government Press
 435

Afghanistan (continued)
1927 Nizam Nama-:i.-Wakil ul-Tujar (IIRegulations De-
fining the Duties of Trade Agentsll)
Kabul: Government Press

Afzal Khan
1974 Tarikh-i-Murassa (liThe Jewel Studed Historyll)
Peshawar: university Book Agency

Al Karaji
1966 Istikhraj-i-Abha~i Pinhan (IIExtraction of
Hidden Waters Ii)
translated from Arabic by Husain K. Jam
Tehran: Bunyad Farhang

Babur
1969 Babur Nama ("Memoires II)
trans. by A.S. Beveridge
London: Luzac & Co.

Badakhshi, Sang Muhammad

1959 Tarikh-i-Badakhshan
A.N. Boldyrev, ed.
Leningrad:

Chekovich, O.D.
1965 Bukharski Dokumentary XIV Veka ("Waqf Documents
of Bukhara Ii)
Tashkent (text in Persian)

1974 ("Documents From Central Asia and Afghanistan")

Moscmi (text in Persian)

Faiz Muhammad
1913 Saraj ul-Tawarikh (liThe Lantern of History")
Kabul: Government Press vols.I and II

1915
Kabul: Government Press vol. III

Fofalzai, A.W.
1958 Durrat Al-Zaman ("History of Shah Zaman")
Kabul: Goverment Press

1967a Timur Shah-i-Durani

2 vols.
Kabul: Historical Society

1967b Tarikh-i-Khirqa-i-Sharifa-i-Kandahar (IIHistory

OI the Holy Cloak of the Prophet at Kandahar ll )
Kabul: Historical Society
 436

Gulistana, Abul Hasan

1965 Mujrnal ul-Tawarikh ("History of the Afshar and
Zand Periods")
Tehran: Ibn Sina Publishers

Hafiz r Nur Muhammad

1946 Tarikh-i-Mazar-i-Sharif ("History of the Holy
Shrine" )
Kabul: Gcvernment Press

Hirawi, Qasim A.
1967 Irsjad ul-Ziraat ("Guide to Agriculture")
Tehran: Tehran University Press

1968 Tariq Qismat-i-Ab-i-Qulb ("Division of Waters")

Tehran: Bunyard Farhang

Husaini, Imam ul-Din

1797/1967 "Fasl-i-Az Tarikh-i-Husain Shahi" (A Chapter
From Husain Shahi's History"
Aryana IV: 28-35

Ibn Huqal
1966 Surat ul'-Ard ("Description of the Earth")
translated from Arabic by G. Shuar
Tehran: Bunyard Farhang

Kandahari, Ahmad Jan Khan, ed.

1885 Asas-ul-Quzat ("Handbook of Judges")
Kabul: Government Press

1885/1891 Qanun Kar Guzari Dar Mamelat Hukumati Wa Tain

Gerayim Wa Siyasat ("Law Defining the Proce-
dure of Actions in Gubernatorial Affairs and
the Determination of Offenses and Punishments"
Kabul: Government Press

Khazini, Abul Fatah

1967 Mizan ul-Hikmat ("Scales of Wisdom")
ed. M. Razawi
Tehran: Bunyard Farhang

Khawandamir, Ghayyath ul-Din

n.d. Tarikh-i Habib al-Siyar Fi Akhbar Afrad al-
Bashar ("Universal History")
Tehran

Lahouri, Mullah Abdul Hamid

1868 Padshah Namah ("History of the King")
eds. Maula\vi Kabir ul-Din and Maulawi Abdur
Rahim 3 vols. '
Calcutta: College Press
 437

Minorsky, V. ed.
1943 Tadhkirat Al-Muluk ("Manual of Administration")
London: Luzac & Co.

Mirza Mehdi
n.d. Tarikh-i-Ayam-i-Nadir ("History of the Times
of Nadir"
Publisher Not Listed

Monshi, Eskander
1978 History of Shah Abbas the Great
2 vols. trans. by R.M. Savory
Boulder, Colorado: Westview Press

Monshi Mahmud
1974 Tarikh Ahmad Shahi ("History of the Reign of
Afuriad Shah")
Moscow: Danesh Publishers

Muhamrnadzai, G.M.
1893 Resala l--1awaza ("Handbook of Preaching")
Kabul: Government Press

Shah Shuja
1954 Waqa-i-Shah Shuja ("Events of Shah Shuja's
Life" )
Kabul: Historical Society

Sifat Narna-Darvis Muhammad Zehan Ghazi

1582/1965 Persian text ed. G. Scarcia
Roma: Instituto Italiano Per II Medico Ed
Estremo Oriente

Sirhindim Ahmad
19G8 Maktubat ("Selected Correspondence")
Karachi: Iqbal Society

Sirhindi, Badr ud-Din

1971 Hazarat AI-Quds ("The Saint" A Biography of
Sirhinddi)
Lahore: Punjab Court of Awqaf

Sultan Muhammad Khan

1880 Tarikh-i-Sultani ("Royal History")
Bombay: Muhammad Press

Tarikh-i-Sistan ("History of Sistan")

1976 trans. by Milton Gold
Roma: Instituto Italiano Per II Medico Ed
Estremo Oriente
 438

Sultan Muhammad Khan (continued)

Tuzuk-i-Jahangir ("Memoirs of Jahangir")
--~978 trans. by A. Rogers
Delhi: Munshiram ManorIal Publishers

n.d. Persian text

Luknahow, India: Monshi Nawkishtwar
 439

British Official Sources

British Government of India

1886-1887 Imperial Gazetteer of India
vols. 1-12
London: Trubner & Co.

1888 Biographical Accounts of Chiefs, Sardars and

Others of Afghanistan
Calcutta: Government Press

1906 Military Report On Afghanistan

Calcutta: Government Press

1907-1908 Imperial Gazetteer of India

vols. 1-22
Oxford: Clarendon Press

1908-1914 Imperial Gazetteer of Afghanistan

six Parts
Calcutta: Government Press

1972-1980 Historical and Political Gazetteer of

Afghanistan
L. Adamec, ed. 5 vols.
Graz, Austria: Akademische Druck-u-Verlagsaltalt

Lumley, J. Savile
1862 "Manufacture and Commerce and Trade of Foreign
Countries -- Russia"
Accounts and Papers 58:270-317

Leech, R.
1841 "Report On the Tooran Guljies"
L/P&S/3/1l (February 10)

Merk, W.R.H.
1881 Report on the Mohmands
Calcutta

Rawlinson, H.
1841/1980 "Report on the Duranis"
in L. Adamec, ed. Historical and Political
Gazetteer of Afghanistan, Kandahar vol.V:
509-577

Robinson, J.A.
1935 Notes on the Nomad Tribes of Eastern
Afghanistan
London:
 440

From the files of the Public Records Office, India Office

Records, the following have been quoted in the text:

L/P&S/18, FO 539, and Parliamentary Papers 48.

 441

Contemporary Accounts in English

Abbot, James
1856 Narrative Of a Journey from Heraut to Khiva,
MOscOwand St. Petersburgh
2 vols. London: James Madden

Bellew, H.W.
1862 Journal Of a Political Mission to Afghanistan
in 1857
London: Smith, Elder and Co.

Burnes, Alexander
1834/1973 Travels Into Bokhara
3 vols. Karachi: Oxford University Press

Curzon, Lord George

1889 "British and Russian Commercial Competition
in Central Asia" in Asiatic Quarterly VIII:
438-451

Drummond, Capt.
1841 "On the Mineral Resources of Northern
Afghanistan" in Journal of the Asiatic society
of Bengal X:74-93
Duke, Joshua
1883 Recollections of the Kabul Campaign, 1879&
1880
London: W.H. Allen & Co.

Elphinstone, Monstuart
An Account Of the Kingdom of Caubul
1839 vol.l London: Richard Bently
1972 vol.2 Karachi: Oxford University Pres
For~ster, George
1798 A Journey From England to Bengal
2 vols. London: R. Faulder

Griffiths, Dr.
1841 "Extracts From a Report On Subjects Connected
With Afghanistan"
in Journal of the Asiatic Society of Bengal
X:797-816, 9177-1038

Havelock, Henry
1840 Narrative Of the War In Afghanistan In 1838-39
2 vols. London: Henry Colburn Publishers
 442

Honisberger, Martin
1834 "Journal Of a Route From Dera Ghazi-Khan
Through the Veziri Country To Kabul"
in Journal and Procaed~ngs of the Asiatic
Society of Benga~ III: 175-178

Irwin, Lt.
1809/1839 "ClimatE'!, Soil and Products Of Afghanistan"
in Journal of the Asiatic Society of Bengal
VIII: 475-477; 779-805; 869-890; 1005-1016

Kaye, John W.
1857/1858 History Of the War In Afghanistan
3 vols. London: Richard Bentley

Lal, Mohan
1846 The Life Of Amir Dost Muhammad Khan Of Kabul
London: Longman, Brown, Green, and Longmans

Leech, R.
1844a "Route From Dera Ghazeen Khan To Candahar,
Through the Sakhee Sarwar Pass and Buzdar,
with Other Routes"
~n Journal of the Asiatic Society of Bengal
XIII: 519-527

1844b "Itinerary From Yazd To Herat and From Herat

to Cabool, Via Candahar"
in Journal of the Asiatic Society of Bengal
XIII: 827-852

1845a "An Account of Early Ghiljees"

in Journal of the Asiatic Society of Behgal
XIV: 306-328

1845b "Supplementary Account of Hazaras"

in Journal of the Asiatic Society of Bengal
XIV: 313-340

1845c "An Account of Early Abdalees"

in Journal of the Asiatic Soci~ty of B~riga1
XIV: 445-470

Martin, Frank
1907 Under the Absolute Amir
New York: Harper and Brothers
Masson, Charles
1842/1974 Narratives Of Various Journeys In Balochistan,
Afghanistan and the Panjab
Karachi: Oxford Unlverslty Press
 443

Poshtan, J.
1841 "Memorandwl1 On the Trade Between the Towns Of
Shikarpure and Candahar"
in Journal of the Asiatic Society of Bengal X:
12-17

Steel, Richard and Crowther, John

161/1965 "A Journal Of the Journey Of Richard Steel and
John Crowther From Azmene In India, the Place
Of the Great Mongols Residence, To Spahan the
Royal Seat Of King Of Persia, In the Affaires
Of the East-India Society. Ann. 1615-1616"
in Hak1uytus Posthumus Or Purchas His Pi1grimes
New York: AMS Press IV:266-280

Tavernier, Jean B.
1676/1905 Travels in India
trans. by John Philips
Calcutta: "Banga~asi" Office

Vambery, Armenius
1865 Travels in Central Asia
New York: Harper and Brothers

Vigne, G.T.
1840 A Personal Narrative Of a Visit To Ghuzni,
Kabul, and Afghanistan
London: whittaker & Co.

Yate, C.E.
1888 Northern Afghanisl.an
London: Wi11iar.1 Blackwood and Sons

Broadfoot, J.S.
1842/1886 "Report on Parts of the Ghi1zai Country and
On Some Tribes in the Neighbourhood of Ghzani
and On the Route From Ghazni To Dera Ismae1
Khan by the Ghwa1ari Pass" Royal Geographical
Society Supplementary Paper
 444:

Afghanistan, Middle East, South Asia -- Studies and Surveys

Adamec, Ludwig
1975 Who's Who of Afghanistan
Graz: Akademische Druck-u-Verlagsantalt

Alder, G.J.
1974a "Key to India?: Britain and the Herat Problem
1830-1863"
Part I Middle Eastern Studies X(2) :186-209

1974b " "

Part II Middle Eastern Studies X (3) :287-311

Amin, H. and Schilz, G.B

1974 A Geography of Afghanistan
Nebraska: Center For Afghanistan Studies

BailIe, Neil B.E.

1832 The Moohummudan Law of Inheritance According
to Abu Hanifa and His Followers
Calcutta: BAptist Mission Press

Bartold, V.
n.d. Iran
trans. by G.K. Nariman

Bayani, Khanbaba
1974 Janqha-i Dawra-i Safawya ("The Wars of the
Safavid Period")
Tehran: Stad Buzurg Urtishstaran

Bulliet, Richard
1975 The Camel and the Wheel
Cambridge: Harvard University Press

Caroe, Sir Olaf

1958 The Pathans 550 B.C.-1957 A.D.
London: Macmillan & Co.

Cohn, Bernard S.
1969 "Structural Change in Indian Rural Society"
in Robert Frykenberg, ed. Land Control and
Social Structure in Indian History
Madison and London: The University of Wisconsin
Press

Dickson, Martin n.
1960 "Uzbeck Dynastic Theory in the Sixteenth Centt:ry"
Trudy Mezhdunardnaia Kongresso Vosto-Kovedov
III:208-216
 445

Dupree, Louis
1973/1980 Afghanistan
Princeton, N.J.: Princeton University Press

Fraser-Tytler, W.K.
1967 Afghanistan
3rd edItion London: Oxford University Press

Friedmann, Yohanan
1971 Shaykh Ahmad Sirhindi
Montreal: McGill-Queen's University Press

Furber, Holden
1976 Rival Empires of Trade in the Orient, 1600-1800
Minneapolis: University of Minnesota Press

Ferrier, R.W.
1973 "The Armenians and the East India Company in
Persia in the Seventeenth and Early Eighteenth
Centuries"
Economic History Review 2nd Series XXVI:38-62

1976 "An English View of Persian Trade in 1618"

Journal of the Economic and Social History of
the Orient XIX (part two) :182-214

Ghani, Ashraf
1978 "Islam and State-Building in a Tribal Society:
Afghanistan 1880-1901"
Modern Asian Studies XII (2): 269-284

1982a "Disputes in a Court of Sharia; Kunar Valley,

Afghanistan 1885-1890"
International Journal of Middle East Studies
(in press)

1982b "Sharia in the Process of State-Building;

Afghanistan 1880-1901"
under consideration

Ghubar, M. Ghu1am Muhammad

1967 Afghanistan Dar Masir-i-Tarikh ("Afghanistan
in the Course of History")
Kabul: Government Press

Gregorian, Vartan
1969 The Emergence of Modern Afghanistan Politics
of Reform and Modernization 1880-1946
Stanford: Stanford University Press
 446

Habibm Irfan
1960 "Banking in Mughal India"
in T. Raychaudhuri, ed. Contributions to Indian
Economic History
Calcutta: Firma K.L. Mukhopadhyay No.1 pp.1-20

1963 The Agrarian System of Mughal India

Bombay: Asia Publishing House

1965 "Distribution of Landed Property in Pre-British

India"
Enquiry (Delhi) n.s., 11(3) :21-75

Hodqson, Marshal
1974 Venture of Islam
Chicago: The University of Chicago Press

Humlum, Johanes
1959 La Geographie de l'Afshanistan
Copenhagen: Gyldenal

Inalick, Halil
1969 "Capital Accumulation in the Ottoman Empire"
Journal of Economic History XXIX (1): 97-140

Irvine, William
1903 The Army of the Indian Moghuls
London: Luzac & Co.

Issawi, Charles, ed.

1971 The Economic History of Iran 1800-1914
Chicago: The University of Chicago Press

Kakar, Hasan
1971 Afghanistan: A Study in Internal Political
Development 1880-1896
Lahore: Punjab Educational Press

1979 Government and Society in Afghanistan The

Reign of Arnir Abd al-Rahman Khan
Austin: University of Texas Press

Koudusi, M. Husain
1960 Nadir Nama ("History of Nadir")
Khurasan, Iran: Asar-i-iAnjuman-i-Milli

Lambton, Ann
1953 Landlord and Peasant in Persia
London: Oxford University Press
 447

Leonard, Karen
1979 "The Great 'Firm' Theory of the Decline of
the Mughal Empire"
Comparative Studies in Society and History XXI
(2) : 151-167

McChesney, Robert
1973 Waqf At Balk: A Study of the Endowments At the
Shrine of Ali Ibn Abi Talib
unpublished Ph.D. dissertation, department of
Near Eastern Studies, Princeton University

Morgenstierne, Georg
1975 "On Pashtu and the Ethnogenesis of Pasht'.:ms"
Paper presented to the International Seminar
on the Pashtu International Center, Kabul,
Afghanistan

Norris, J.A.
1967 The First Afghan War l~38~~842
Cambridge: CAmbridge University Press

Parizi, Bastini
1978 Siasat Wa Iqtisad Asar-i-Safawi ("Politics
and Economics of the Safavid Period")
Tehran: Safi Ali Shah Publisher

Perry, J.R.
1975 "Forces Migration In Iran During the Seventeenth
and Eighteenth Centuries"
Iranian Studies VIII(4): 199-215

1979 Karim Khan Zand A History of IranlJ47-1779

Chicago: The University of Chicago Press

Rahim, M.A.
1961 History of the Afghans in India A.D. 1545-1631
Karachi: Pakistan Publishing Company

Rahman, Fazlur, ed.

1968 Intikhab Maktubat Shaikh Ahmad Sirhindi ("Se-
lected Letters of Shaikh Ahmad Sirhindi")
Karachi: Iqbal Academy
Richards, J.F.
1981 "Mughal State Finance and the Premodern World
Economy"
Comparative Studies in Society and History
XXI I (2) : 2 6 5 - 313
 448

Savory, R.M.
1964 "S ome Notes on the Provincial Administration of
the Early Safawid Empire" XXVII:114-128
Bulletin of the School of Oriental and
African Studies

1980 Iran Under the Safavids

Cambridge: Cambridge University Press

Schran, Peter
1978 "A Reassessment of Inland Communications in
Late Ch'ing China"
Ch'ing Shih Wen-To 111(10) :28-48

Shaban, M.A.
1970 The 'Abbasid Revolution
Cambridge: Cambridge University Press

Singh, Ganda
1959 Ahmad Shah Durrani Father of Modern Afghanistan
London: Asia Publishing House

Steensgaard, Niels
1973 Carracks, Caravans, and Companies
Kobenhavn: Student-litteratur

watson, Andrew
1974 "The Arab Agricultural Revolution and Its Dif-
fusion, 700-1100"
The Journal of Economic History XXXIV(l) :8-35

Yapp, Malcolm
1962 "Diaturbances in Eastern Afghanistan, 1839-42"
Bulletin of the School of Oriental and African
Studies XXV(3): 499-523

1963 "Disturbances in Western Afghanistan, 1839-41"

Bulletin of the School of Oriental and African
Studies XXVI (2) : 288-313

1964 liThe Revolution of 1841-42 in Afghanis~3n"

Bulletin of the School of Oriental and African
Studies XXVII (2) : 333-381

1960 liThe India Office Records As a Source for the

Economic History of the Middle East"
in M. Cook,ed. Studies in the Economic History
of the Middle East From the Rise of Islam To
the Present Day
London: Oxford University Press pp.501-513