Scholastic Sanskrit

TREASURY OF THE INDIC SCIENCES series
Editor-in-Chief: Robert A.F. Thurman

Jey Tsong Khapa Professor of Indo-Tibetan Buddhist Studies,
Columbia University
President, American Institute of Buddhist Studies
Series Editor: Gary A. Tubb
Dharam Hinduja Senior Lecturer in Sanskrit Teaching
and Indic Research in the Department of Religion,
Columbia University
Executive Editor: Thomas F. Yarnall
Department of Religion,
Columbia University
Editorial Board: Ryuichi Abe, Jay Garfield, David Gray, Laura Harrington,
Thubten Jinpa, Joseph Loizzo, Vesna Wallace, Christian Wedemeyer, Chun-
fang Yu
The American Institute of Buddhist Studies (ArBS), in affiliation with the
Columbia University Center for Buddhist Studies and Tibet House US, has
established the Treasury ofthe Indic Sciences series as a complement to its
Treasury ofthe Buddhist Sciences series. For over fifteen hundred years the
scholars representing various Buddhist traditions were actively engaged in
dialogue and debate with representatives of various Vedist, Hindu, Jainist,
and other religious and secular traditions. In this cosmopolitan milieu the
shared arts and sciences of all these Indic traditions were constantly refined
and developed. In full acknowledgement of this broader context, the AIBS
is widening its Buddhistic mandate with the Treasury of the Indic Sciences
series in which it is publishing translations of influential classics from the
diversity of Indic traditions. Publications from both Treasuries will bring
important Indic contributions into contemporary discourse, enriching and
contributing to our global cultural, philosophical, spiritual, and scientific
heritage.
SCHOLASTIC SANSKRIT
A HANDBOOK FOR STUDENTS
by
Gary A. Tubb
and
Emery R. Boose
Treasury of the Indic Sciences series
Published by
The American Institute of Buddhist Studies
at Columbia University in the City of New York
Co-published with
Columbia University's Center for Buddhist Studies
and Tibet House US
New York
2007 Y1J:I)}-.
l.lni\!erliliit Hamt>lKQ Asi..
Treasury ofthe Indic Sciences
A refereed series published by:
American Institute of Buddhist Studies
Columbia University
80 Claremont Avenue, Room 303
New York, NY 10027
http://www.aibs.columbia.edu
Co-published with Columbia University's Center for Buddhist Studies
and Tibet House US
Distributed by Columbia University Press
2007 by Gary A. Tubb and Emery R. Boose
All rights reserved.
No portion of this work may be reproduced in any form or by any, electronic or
mechanical, inclnding photography, recording, or by any information storage and
retrieval system or technologies now known or later developed, without written
permission from the publisher.
Printed in Canada on acid-free paper.
ISBN 978-0-9753734-7-7 (cloth)
Library of Congress Cataloging-in-Publication Data
Tubb, Gary A. (Gary Alan), 1950-
Scholastic Sanskrit: a handbook for students I by Gary A. Tubb and Emery
R. Boose.
p. em. - (Treasury of the Indic sciences series)
nCo-published with Columbia University's Center for Buddhist Studies and
Tibet House US. "
Includes bibliographical references and index.
ISBN 978-0-9753734-7-7 (hardcover: aIk. paper)
1. Sanskrit prose literature-History and criticism-Handbooks, manuals,
etc. 2. Sanskrit prose literature-Study and teaching-Handbooks, manuals,
etc. 1. Boose, Emery R. (Emery Robert), 1952- II. Title.
PK2903.T832007
891' .2808-dc22
2007000683
Dedicated to the memory of our teacher,
Daniel Henry Holmes Ingalls (1916-1999),
Wales Professor of Sanskrit, Harvard University
J
Contents
Series Editor's Preface
Authors' Preface
Abbreviations
Introduction: The Sanskrit Commentary
0.1 Importance of commentaries . . .
0.2 The five services of a commentary
0.2.1 Padaccheda.
0.2.2 Padiirthokti.
0.2.3 Vigraha...
0.2.4 Viikyayojanii
0.2.5
Part I Methods of Glossing
by Gary A. Tubb
1 The First Four Services
1.1 Introduction...
1.2 Types of glossing
2 Division of Words
1.3 Introduction............
1.4 Repetition of the words of the text
1.5 Specification of word-division . .
xxi
xxv
xxix
I
I
3
4
4
4
5
5
7
9
9
10
13
13
14
14
viii
Contents
Contents ix
1.6 Citation of individual words
1.6.1 iti .
1.6.2 -sabda, -kiira .
1.6.3 Inflected indeclinable .
1.6.4 Anubandhas
1.7 Variant readings . . . . . .
3 Giving the Meaning of Words
1.8 Introduction....
1.9 Simple synonyms
1.10 Location of glosses
1.10.1 Enclitics
1.10.2 Vigrahas
1.11 Glossing by equivalent forms .
1.11.1 Periphrasis . .
1.11.2 Participle . . .
1.12 Expansions of glosses.
1.12.1 -adi .
1.12.2 Implications
1.12.3 Implied contrasts
1.13 More removed paraphrases
1.13.1 ity r t h ~
1.13.2 iti h i i v ~ . . .
1.13.3 iti yavat ....
1.13.4 Other phrases .
1.13.5 anena, etena
1.13.6 References to speech
1.14 Meanings by extension
1.15 Glossing of puns .
1.16 Definitions. . . . . . . . . .
1.16.1 Definitions offered by the commentator
1.16.2 Definitions quoted from dictionaries
1.17 Treatment of special items
1.17.1 Proper names . . . . .
1.17.2 Names of species ...
1.17.3 Meanings of particles .
1.17.4 Other indeclinables ..
14
15
15
15
16
16
19
19
20
20
20
21
21
21
21
22
22
22
22
23
23
24
25
25
25
26
27
27
28
28
29
30
30
31
31
31
1.17.5 Pronouns .
4 Analysis of Grammatical Complexes
1.18 Introduction . . . . . . . . . .
1.19 Modes of complex formation . . .
1.20 General principles of analysis. . .
1.21 Insertion of analyses in a commentary .....
1.21.1 Glossing the words used in the analysis
1.21.2 Indicating the inflection of the original .
5 Derivative Verbs
1.22 Introduction
1.23 Desideratives
1.24 Intensives . .
1.25 Causatives . .
1.26 Denominatives
1.26.1 Desire to have.
1.26.2 Behavior toward
1.26.3 Behaviorlike ..
1.26.4 Becoming something
1.26.5 Making something so. .
1.26.6 Feeling something ...
1.26.7 Doing or making something
1.26.8 Making use of something.
6 Primary Nominal Derivation
1.27 Introduction . . . . . .
1.28 Basic bhiive meanings.
1.29 Basic kiiraka meanings
1.29.1 kartari ..
1.29.2 karrnm:zi ...
1.29.3 kara!,e .. ,
1.29.4 sampradiine.
1.29.5 apadane . ..
1.29.6 adhikara!,e
1.30 Temporal restrictions
1.30.1 bhate . . . .
32
33
33
35
35
39
40
41
43
43
44
45
45
46
46
47
47
47
, 47
48
. 48
48
49
49
50
52
53
54
54
54
55
55
55
56
x
Contents Contents xi
1.30.2 vanamiine ...... 56 8
Nominal Compounds 85
1.30.3 . ...... 57
A. General Principles .
1.31 Other restrictions of meaning . 57
. ........ 85
1.32 Accompanying words (upapada) 58 1.43 Introduction . . . . . . . . . . . 85
1.32.1 kartari krt with accusative 60 1.44 Basic categories of compounds . 86
1.32.2 kartari krt with non-accusative . 61
1.44.1 Avyayibhiiva 86
1.32.3 krt in the sense of another kiiraka 62
1.44.2 ... 87
1.32.4 Temporal restrictions . . . . 62
1.44.3 Karmadhiiraya 87
1.32.5 Other conditions of meaning . 63
1.44.4 Dvandva ... 87
1.33 Verbal participles ..... 64
1.44.5 Bahuvrfhi .... 88
1.34 Irregular forms (nipiitana) 65
1.44.6 Kevala-samiisa .. 88
1.45 Additional elements of meaning 89
7 Secondary Nominal Derivation
67
1.45.1 Comparison . . . . . . . 89
1.35 Introduction . . . . . . . .
67
1.45.2 Aggregation. . . . . . . 90
1.36 Dependent formulas of analysis. 68
1.45.3 Clarification of connection 90
1.36.1 Accusative 70
1.46 Optional and obligatory compounds 90
1.36.2 Instrumental 71
1.46.1 Asvapadavigraha compounds 90
1.36.3 Dative .. 71
1.46.2 Conventional meanings. 91
1.36.4 Ablative ...
71
1.46.3 Bound forms 92
1.36.5 Locative ...
72
1.47 Irregularities. . . . . . 92
1.37 Genitive formulas of analysis .
74
1.37.1 tasyapatyam 75
B. Avyayibhiiva Compounds 92
1.37.2 tasya bhiivaJ:z .... 75 1.48 Introduction . . . . . . 92
1.37.3 Others ........ 76 1.49 Optional avyayrbhiivas 93
1.37.4 General connection . 77 1.50 Obligatory avyayrbhiivas 93
1.38 Nominative formulas of analysis 77 1.51 Irregular avyayrbhiivas 95
1.39 Compound formulas of analysis 78
1.40 Possessive suffixes (matvarthlya) . 79
C. Compounds . . 96
1.40.1 Straigbtforward possession . 80 1.52 Introduction . . . . . . 96
1.40.2 Additions of meaning. 80 1.53 . . 96
1.41 Vibhaktisarrzjiiaka suffixes ..... 81 1.53.1 Straightforward analysis 97
1.42 Sviirthika suffixes . . . . . . . . . . 82 1.53.2 Additional specification of connection . 100
1.42.1 With no addition of meaning . 82 1.53.3 Partitive genitive as second member 101
1.42.2 Diminutives . . . . . . . . . 82 1.54 Karmadhiirayas of simple apposition. 102
1.42.3 Other additions of meaning. 83 1.54.1 Adjective plus substantive 102
1.42.4 cvi, etc ...... 83 1.54.2 Adjective plus adjective . . 103
1.42.5 Samasanta suffixes . . . . . 84 1.54.3 Substantive plus substantive 104
xii Contents
1.54.4 Additional specification of connection
1.55 Karmadharayas of comparison . . . . . . .
1.55.1 Upamanapurvapada karmadhilraya
1.55.2 Upamita-samasa
1.55.3 Rupaka-samasa ..
1.56 Dvigu compounds . . . . .
1.56.1 Taddhitanha-dvigu
1.56.2 Uttarapada-dvigu.
1.56.3 Samahilra-dvigu
1.57 with inseparable first member
1.57.1 ku, su, etc .
1.57.2 with fixed meaning
1.57.3 Dhiltu-lopa
1.57.4 .
1.58 Negative .
1.59 Upapada compounds . .
1.60 Irregular . . .
1.60.1 compounds .
1.60.2 Mayuravya1]1sakadi compounds
1.60.3 Suffix-formations resembling compounds
D. Dvandva Compounds
1.61 Introduction ....
1.62 ltaretara dvandvas
1.63 Samli.hilra dvandvas .
E. Bahuvnlli Compounds .
1.64 Introduction . . . . .
1.65 Appositional bahuvrfhis .
1.65.1 Simple appositional bahuvrfhis
1.65.2 Negative bahuvnnis ....
1.65.3 Pradi-bahuvrfhi compounds ..
1.65.4 Saha-bahuvrfhi compounds ..
1.65.5 Anekapada-bahuvrfhi compounds
1.66 Non-appositional bahuvrfhis . . . .
1.66. I Locative as second member
1.66.2 Genitive as first member . .
105
106
106
107
108
109
110
110
111
112
113
113
114
116
118
119
121
121
123
125
125
125
126
126
127
127
128
128
130
131
131
132
133
133
133
Contents
1.67 Special bahuvrfhis .
1.67.1 Sa1]1khya-bahuvrfhi compounds-,
1.67.2 Dig-bahuvrfhi compounds >
1.67.3 Reciprocal combat
1.68 Irregular bahuvrfhis .
F. Longer Compounds
1.69 Introduction . . . .
1.70 General principles .
1.71 Special techniques
1.71.1 Simpler formulas of analysis
1.71.2 Separate analysis
1.72 Choice of options
9 Formations
1.73 Introduction ..
1.74 ..
1.75
10 Indicating the Construction
A. Arranging the Words of the Text
1.76 Introduction . . . . . . . . .
1.77 The katha1]1bhutinf approach
1.78 The anvayamukhf approach . .
1.79 Insertion of explanatory material
1.80 Variations on the two approaches .
B. Other Techniques .
1.81 Introduction .
1.82 Direct statement of connections.
1.82.1 anvaya ..
1.82.2 sa1]1bandha .
1.82.3 . . . . . . . .
1.82.4 Connections between verses
1.82.5 Connections between statements
1.83 Supplying words needed to complete the sense
xiii
135
135
136
136
136
137
137
138
141
141
141
142
147
147
147
148
149
149
1149
149
150
151
156
161
161
161
161
162
163
163
164
165
xiv Contents
Contents xv
1.83.1 iti .
1.83.2 Words carried over .
1.84 Devices for marking features of syntax .
1.84.1 yat ... tat marking subject and predicate
1.84.2 sat marking attributive adjectives ...
1.84.3 -bhuta marking attributive substantives
1.84.4 sati marking the locative absolute
1.84.5 yathii tathii marking adverbs
1.84.6 he marking vocatives . . . . . . .
Part n The Style
by Emery R. Boose
11 The Fifth Service
2.1 Introduction
2.2 Content
2.3 Style ....
12 Abstract Nouns
2.4 Introduction.
2.5 Fonnation..
2.5.1 tii and tva
2.5.2 -bhiivalz and vrddhi derivatives.
2.6 Significance........
2.6.1 Used by necessity.
2.6.2 Used by choice ..
2.7 Common constructions . .
2.7.1 With subjective genitive
2.7.2 In compound .
2.7.3 Causal phrase within a sentence
2.7.4 iii clause without abstract suffix
2.7.5 Two genitives .
2.8 Causal ablative and instrumental . . . .
2.9 Other cases .
2.9.1 Nontinative as subject of sentence
2.9.2 Accusative as object of the sentence
165
165
166
166
167
167
168
169
169
171
173
173
174
174
177
177
177
177
178
178
178
179
179
179
180
180
181
182
182
183
183
183
2.9.3 Dative of purpose .
2.9.4 Genitive to modify another noun, .
2.9.5 Locative in a locative absolute
2.10 Predicative instrumental ....
2.11 Accusative with verb of motion.
13 Compounds
2.12 Introduction .
2.13 Analyzing long compounds.
2.13.1 Identification
2.13.2 Dissection .
2.13.3 Grouping . . . . . . . . . .
2.14 Relation to words outside the compound
2.15 ka to mark a bahuvrfhi .
2.16 tat to mark a dvandva . . . . . .
2.17 -bhuta to mark a karmadhiiraya
2.18 Pronouns in compound
2.19 Idiomatic tenns
2.19.1 -antara
2.19.2 -artha .
2.19.3 -ka/pa.
2.19.4 ogata
2.19.5 -garbha
2.19.6 -jana
2.19.7 -jata ..
2.19.8 -purva.
2.19.9 -matrii.
2.19.10
14 Word Order
2.20 Introduction . . . . . . . . . .
2.21 Subject and predicative nouns
2.21.1 Predicative nouns . . . .
2.21.2 Pronoun subjects . . . . .
2.22 Attributive and predicative nouns . . .
2.22.1 Word order with subjects . . .
2.22.2 Word order with other nouns .
183
184
184
184
185
187
187
187
187
188
188
189
190
191
191
192
193
193
193
193
194
1
194
194
195
195
195
196
197
197
197
197
198
199
199
199
xvi
Contents Contents xvii
2.22.3 Markers of attributive qualifiers
2.23 Scope ofiti .
2.24 Analyzing long sentences .
15 Causal Sentences
2.25 Introduction . . . . . . . . . . . .
2.26 Ablative and instrumental phrases
2.26.1 Ablative causal phrases .
2.26.2 Series of causes . . . . . .
2.26.3 ca in causal series. . . . .
2.26.4 Verbal nouns expressing cause
2.27 iti clauses . . . . . . . . . . .
2.28 Relative pronouns and adverbs
2.29 Locative absolute . . . . . . .
2.29.1 Expressing a cause ..
2.29.2 Expressing a condition
2.29.3 Concessive locative absolute
2.29.4 Abbreviated locative absolute
2.30 Causal particle hi . . . . . . . . . . .
16 Uses of iti
2.32 Common constructions
2.33 Quotation . . . . . . .
2.33.1 iti with source.
2.33.2 iti with ablative
2.33.3 iti with partial quotations .
2.34 Explanation . . . . . . . .
2.35 Comment . . . . . . . . .
2.35.1 Comment with iii .
2.35.2 Commonly used comments.
2.36 Constructions with the infinitive .
2.36.1 Infinitive with sak and yuj .
2.36.2 Infinitive with arh .....
2.36.3 Infinitive expressing purpose .
2.37 End of a passage .
200
200
201
205
205
206
206
206
207
208
209
210
211
212
212
213
213
213
215
215
215
216
216
217
217
218
218
218
219
220
220
221
222
222
17 References
2.39 Text and author of text
2.39.1 Verbs of taking .
2.39.2 prakrta, aprakrta .
2.39.3 Author of text commented on
2.39.4 Author as understood speaker
2.39.5 Participles of verbs of saying.
2.39.6 Instrumental of present participle
2.39.7 Auto-commentary ....
2.40 References to the commentary
2.40.1 First person plural
2.40.2 Participles. . . . . . .
2.40.3 Relative clause . . . .
2.40.4 Reference to earlier topic .
2.40.5 agre . . . . . . . . . . . .
2.41 Recognized authorities . . . . . .
2.42 Genitive in reference to particular views
18 Examples and Comparisons
2.43 Introduction . . . . . . .
2.44 Bahuvrrhis in tidi . . . .
2.44.1 Compounds in tidi
2.44.2 Series of compounds
2.44.3 ityiidi . . . . . .
2.45 Suffix vat .
2.45.1 Adverbs in -vat .
2.45.2 With standard of comparison.
2.45.3 itivat, itytidivat
2.46 iva and yathii
19 Terms of Discussion
2.47 Introduction . .
2.48 Dialectic style .
2.48.1 Oral debate
2.48.2 i i r v p k ~ and siddhtinta
2.48.3 Direct address. . . . . . .
225
225
225
225
226
226
227
228
228
229
229
229
230
230
231
231
231
232
233
>233
233
1233
234
235
236
236
236
237
238
239
239
239
239
240
241
xviii Contents
Contents
xix
Appendix: Useful Reference Works
Indices
Commentarial Passages Quoted .
Grammatical Citations
Sanskrit Index .
English Index . . . . .
2.48.4 Example of dialectic style
2.49 Objection . . . . . . .
2.49.1 nanu and atha .
2.49.2 Interrogatives
2.49.3 iti cet .....
2.49.4 ity iisankya ..
2.49.5 priipta and i i ~ i p t
2.49.6 Verbs of speaking.
2.49.7 Clauses beginning with na ca ...
2.50 Reply .
2.50.1 Verbs of speaking . . . .
2.50.2 tan na and other denials
2.50.3 prasanga and iipatti
2.50.4 d o ~ . . . . . . . . . .
2.50.5 syiit and bhavet . . . . .
2.50.6 yady api and other concessives .
2.50.7 tarhi .
2.51 Alternative interpretation . . .
2.51.1 Second interpretations
2.51.2 Citing others .....
20 Particles and Pronouns
2.52 Introduction . . . . . . . . . . .
2.53 Particles connecting sentences .
2.53.1 atka.
2.53.2 tatra .
2.53.3 tavat .
2.53.4 atro and iha .
2.53.5 kiT!' ca, etc. .
2.53.6 tu, etc.....
2.53.7 tatoo hi and tad yatOO
2.53.8 Words for "therefore" .
2.54 Particles of emphasis
2.54.1 eva .
2.54.2 api .
2.55 Pronontial adverbs nsed as prononns
2.55.1 Adverbs corresponding to pronouns
241
242
243
243
244
244
245
245
246
246
246
247
248
248
249
250
250
251
251
251
253
253
253
253
254
255
255
256
257
257
258
258
258
259
260
260
2.56
2.57
2.55.2 Adverbs as attribntives ..
Correlative pronouns and adverbs
2.56.1 Repeated correlative
2.56.2 Multiple pairs . . .
2.56.3 yatOO with tatoo .
2.56.4 yat meaning "that"
2.56.5 Repeated anya ..
Demonstrative pronouns .
2.57.1 Scale of proxintity
2.57.2 Reference to earlier and later passages.
2.57.3 tad etat and tad idam . . . . . . . . . .
261
261
261
262
262
263
263
264
264
265
265
267
271
271
278
281
290
It was only after this book was accepted for publication in the Treasury
of the Indic Sciences series that I was caIled upon to become editor of the
same series, and thus eventually found myself in the rather odd position of
welcoming my own work into the series. I am grateful for this opporturtity
for two reasons. First, it affords me a place, separate from the preface and
introduction that Emery Boose and I wrote together, to describe subsequent
developments in the handbook that have involved only my portion of the
work and not his. Second, it allows me to make a few comments, at a point
near the beginrting of this new series, on my hopes for the series and their
connection with the subject matter of the present volume.
This is the first printed edition of a handbook for students on the read-
ing of Sanskrit commentaries that has been circulated privately in five pre-
vious versions over a period reaching back twenty-five years. The original
version was a typewritten draft contairting two sections corresponding to
the two parts of the present edition. The first section was mine, and dealt
with commentarial techniques for explaining the meanings of words in a
text, and thus provided a survey of methods of glossing. The second sec-
tion, written by Emery Boose, covered the distinctive features of the type
of expository prose used in discussing the contents of the text, and was thus
concerned with what Renou and others have referred to as the h i i ~ y style
of Sanskrit.
Subsequent versions of the handbook have largely been the result of
my own attempts to find the most useful way to deal with a complex topic
treated only sporadically in the original version-the use in commentaries
of the technical devices of traditional PaI)irtian grammar, a topic .that is
difficult to introduce concisely and simply, and impossible to treat com-
prehensively in any portion of a single volume. Throughout the history
xxi
xxii
Series Editor's Preface xxiii
of these attempts, Emery Boose's portion of the handbook has remained
essentially the same, with the exception of progressive changes in visual
formatting and the correction of minor typographical errors. Neither I nor
any of the many readers of the handbook I have heard from has discovered
any way in which his contribution could be improved. I have also avoided
making any changes to the preface and introduction that we wrote together,
other than those required by the alterations in my own material that I shall
now briefly describe.
In the second version of the handbook, the material on PliJ;linian gram-
mar was extracted to form the core of a separate part of the work, and at
the same time the other parts of work were typed into computer files and
reformatted. My treatment of this PiiJ:tinian material became progressively
more unwieldy, and once it had been circulated among colleagues in the
third version of the handbook, I realized that this material was too difficult
and extensive to be handled with the same comprehensiveness as in the
part of the handbook dealing with glosses in non-technical language, and
that the same range of technical information was beginning to be covered
in recent work by George Cardona and others. In the fourth version, the
other portions of the handbook were reformatted separately from the ma-
terial on Panini which was then withdrawn. In the fifth version, the first
part was in view of the new structure, both parts were completely
reformatted using the I5I'EX macro system with automated cross references
and indexing, and a shortened overview of PiiJ:tinian technical terminology
was added to the first part to compensate for the withdrawal of a separate
section on this topic. The present edition is an improved reformatting of
this fifth version, reshaped to fit the series in which it appears.
Along the way this handbook has been read by many students and col-
leagues, and in recent years here at Columbia University I have used it
occasionally in courses with a focus on the study of commentarial tech-
niques, in which we have read from samples of commentaries on poetic
and philosophical works that I have annotated with cross-references to the
sections of the handbook. I am grateful to all those students who have
offered comments on the handbook, and I would especially like to thank
Rupa Viswanath for her detailed comments and suggestions.
It was only after coming to Columbia that I was exposed to the ex-
tensive role played by Sanskrit commentarial techniques in Buddhist texts.
For this I have Professor Robert A. F. Thurman to thank. He encouraged
me and others to be attentive to the interplay between Brahmanical and
Buddhist intellectual activities, and it was in the setting of joint meetings
of classes of advanced Sanskrit and advanced Tibetan, in which we read the
same texts in both languages, that the interconnections became most clear
to me. In reading with our students a commentary by CandrakIrti, I real-
ized that some of CandrakIrti's points were often lost on Tibetan-oriented
readers who were not versed in the technicalities of Sanskrit grammatical
explication. and at the same time that CandrakIrti had expounded detailed
priniciples of exegesis that had remained largely unknown to readers who
specialized in non-Buddhist texts.
I was therefore pleased to learn of the initiation of the Treasury of the
Indic Sciences series, which is designed to publish works of non-Buddhist
origin that are relevant to the history of Buddhism in India and beyond,
and I was more particularly delighted by Professor Thurman's invitation to
include this handbook on Sanskrit commentaries in the series. I am also
grateful for the concurrent inclusion in the same series of the catalogue of
Sanskrit manuscripts at Columbia University compiled by David Pingree.
Both works are perhaps atypical of the sort of text that will constitute the
bnlk of the series, but both are tools for pursuing its aim of exposing the
intellectual interplay between Buddhist and non-Buddhist activity within a
shared intellectual world.
The opening volume in the series presented a Brahrnanical work on
logic. translated by Stephen H. Phillips and N.S. Ramanuja Tatach,Jrya.
The vitality of this sort of text as a participant in cross-cultural conver-
sations was made clear once again last year in a conference on Mind and
Reality held here at Columbia, in which Professor Phillips engaged in stim-
ulating dialogue with experts on Buddhist logic. I am confident that sub-
sequent volumes in this series will make available texts of similar interest
and usefulness.
In addition to my gratitude to Professor Thurman for all of his many
acts of inspiration and support, I feel a deep indebtedness to the technical
editor of the series, Thomas Yarnall. His knowledge -of the interacting
systems of thought I have mentioned is matched by his expertise in all the
departments of publishing technology, and enriched by his generosity and
patience.
My longest-standing debt of gratitude is to myoid friend Emery Boose.
His contribution to this handbook speaks for itself in its insight and rigor,
xxiv
Gary A. Tubb
Series Editor, Treasury ofthe Indic Sciences
and without him none of this would ever have been written, I hope that the
publication, at long last, of this handbook might serve in some small way
to begin to repay his many kindnesses.
In this edition, chapters are numbered sequentially throughout the book,
but the numbering of sections is separate for each part.
are by section or subsection, and the numbers referring to these with
the number of the part in which they appear, so that it will be Immediately
apparent whether the topic in question is being treated in terms of glossmg
or of argument. .
I have typeset this book using open-source computer typesettmg
software running on open-source Debian GNUlLinux operating systems.
New York
November 2006
Authors' Preface
This handbook is intended to give students in the early years of their study
of Sanskrit some help in using Sanskrit commentaries and in reading the
style of Sanskrit prose, which is employed not only in commen-
taries but in many other works as well. We try to bring together simple
explanations of the things that often make it difficult for students to begin
to deal with scholastic works in Sanskrit, and of the things that are most
readily useful once they are understood. We hope that the handbook will
serve both as a reference work-by using the index or table of contents it
should be possible to find explanations of many puzzling problems as they
arise-and as a general introduction to the techniques used in commen-
taries and in the style in general.
In identifying the difficulties that scholastic styles of Sanskrit present
we have tried to follow our own observations of the obstacles most of-
ten encountered by students-observations that include our own memo-
ries of the problems we grappled with as students. We were fortunate in
having teachers whose explanations made accessible to us much that had
seemed impenetrable. What has prompted us to attempt the writing of this
handbook is the fact that so many things that turned out to be so simple
and useful did have to be explained to us, unless we had managed to fig-
ure them out on our own in the process of struggling with the texts: in
printed sources much of the information we needed was buried, scattered,
or nonexistent.
Most of what we present here is not our own discovery: In particular we
have made frequent and grateful use of information gleaned from remarks
made in class by Professor Daniel H. H. Ingalls and from the works of
Speijer and Coulson mentioned in the Appendix.
We are grateful to Professor Ingalls for reading the first drafts of Part
xxv
xxvi Authors' Preface
xxvii
One and for making many detailed and helpful suggestions.
Part One of the handbook, written by Gary A. Tubb, deals with the de-
vices used in Sanskrit commentaries to explain the words of the text being
commented on and their meanings; it covers fully the more important tech-
niques of glossing in non-technical language used by the commentators,
and gives a brief introduction to the rudiments of analysis in the techni-
cal terminology of PiiJ).inian grammar as employed in commentaries. Part
Two, written by Emery R. Boose, deals with the devices used in commen-
taries to discuss the ideas contained in the text being commented on, and
provides an introduction to the bhiieYa style and to the teChniques of San-
skrit expository prose in general, whether employed in a commentary or in
an independent work. In each part there are occasional references to sec-
tions in the other parts, and both parts are covered simultaneously by the
indices at the end of the handbook, but neither of the parts presupposes a
knowledge of the material covered in the other.
The examples used in the handbook are drawn as far as possible from
works that are easily accessible and that students of Sanskrit are likely to
read. In Part One they come chiefly from well-known commentaries on the
great mahiikiivya poems, especially from the commentaries of Mallinatha,
and occasionally an example is borrowed from the works of the grammar-
ians. In Part Two the examples come from texts of the Sfuikara school
of Vedanta, especially from Sailkara's own and in a
few instances constructed examples are given for the sake of simplicity.
Although Part One thus relies primarily on literary works for its examples,
and Part Two on philosophical works for its examples, the points illustrated
in each part are applicable to most commentarial or expository works On
any subject.
The members of Sanskrit compounds are separated by hyphens as far
as possible, and a circumflex accent is used to mark a long vowel or diph-
thong produced by sandhi between words or members of a compound.
Square brackets indicate words not present in the Sanskrit text that must
be supplied from context. In Part One, words quoted from the text being
commented on in an example are in bold type. Also in Part One, indica-
tory phonemes (anubandhas) in technical elements of PiiI)inian grammar
are printed as uppercase letters.
Ellipsis in a passage cited in Sanskrit is marked by three dots, and is
indicated only when it occurs within the words actually quoted. Words
preceding or following the quoted words in the same original sentence are
often omitted without notice when the quoted words yield the same mean-
ing that they have in the full sentence. ,-
In the translations of the examples, parentheses indicate words that are
necessary to give a clear translation in normal English and that are im-
plied in the Sanskrit text but do not correspond directly to words actually
present in the Sanskrit text. Square brackets indicate words corresponding
to terms that in the Sanskrit text are not expressly present but that must
be understood from the context; the English words in square brackets are
often translations of Sanskrit words supplied in square brackets in the text.
In Part Two a very literal translation is sometimes followed by a colon and
a more natural translation, and alternative translations are separated by a
semicolon.
Reference is often made to the works of Coulson, Speijer, and Whit-
ney where these authors have additional discussion or examples for the
grammatical point in question. References to Coulson are by page num-
ber; references to Speijer and Whitney are by paragraph number. See the
Appendix for bibliographic information on these and other works that may
be useful in learning to read scholastic styles of Sanskrit.
Abbreviations
Coulson Michael Coulson, Sanskrit: An Introduction to the Classical
Language (Teach Yourself Books)
GBh of SaJikara (SlivilI)ivilasa)
KiiSikii Kiisikii Vrtti on PilI)ini's
Kir. Kiriitiirjunrya of Bharavi (NirJ.1aya-Sagara Press)
Kir.M. Mallinatha's commentary on the Kiriitiirjunrya (Nin;mya-Sagara
Press)
Kum. Kumiirasambhava of Kalidasa (NirJ.1aya-Sagara Press)
Kum.C. Caritravardhana's commentary on the Kumiirasambhava
(Gujarati Printing Press)
Kum.M. Mallinatha's commentary on the Kumiirasambhava
(NirJ.1aya-Sagara Press)
Kum.V. Vallabhadeva's commentary on the Kumiirasambhava (ed. M. S.
Narayana Murti)
Megh. Meghaduta of Kiilidasa (ed. M. R. Kale).
Megh.M. Mallinatha's commentary on the Meghaduta (ed. M. R. Kale).
Megh.V. Vallabhadeva's commentary on the Meghaduta (ed. E.
Hultzsch).
Pii. PilI)ini's
xxix
xxx
Abbreviations
PP Paiicaplidikii of Padmapiida (Madras Government Oriental Series
155)
Ragh. RaghuvarJ1sa of Kalidasa (NirJ:taya-Sagara Press)
Ragh.H. Hemiidri's commentary on the RaghuvarJ1sa (ed. Rewa Prasada
DwivedI)
Ragh.M. Mallinatha's commentary on the RaghuvarJ1Sa (NirJ:taya-Sagara
Press)
SBh Brahmasiitrabhiisya of Satikara (Sviimi Satyananda SarasvatI)
SDS SarvadarsanasOligraha of SayaJ.lamiidhava (Bhandarkar Oriental
Research Institute)
Sisu. Sisupiilavadha of Magha (Kashi Sanskrit Series 69)
SiSu.M. Mallinatha's commentary on the Sisupiilavadha (Kashi Sanskrit
Series 69)
SiSu.V. Vallabhadeva's commentary on the Sisupiilavadha of Magha
(Kashi Sanskrit Series 69)
Speijer J. S. Speijer, Sanskrit Syntax (Rpt. Motilal Banarsidass)
Whitney William Dwight Whitney, Sanskrit Grammar (Harvard
University Press)
Introduction
The Sanskrit Commentary
0.1 Importance of commentaries
The Sanskrit tradition offers to each student of its texts a rich source both
of potential assistance and of potential difficulties in the form of an ex-
traordinary abundance of works of commentary. The importance that these
commentaries have for students is due in general to two characteristics-
the simple fact of their abundance, and the specific ways in which they
offer assistance in the reading of a Sanskrit text.
Works of commentary pervade the history of Sanskrit thought to a de-
gree that is unparalleled in the writings of most other traditions: it is no
exaggeration to say that of all the expository works available in San;krit
most are, at least in external form, commentaries. There are several rea-
sons for this prevalence of commentaries in Sanskrit, and some of them are
tied to features that are peculiar to the Sanskrit tradition.
One striking feature is the frequency with which we find works in San-
skrit that seem to require, or even to presuppose, the eventual services of
a commentator. In works of the stistra type, this situation is an ancient
one. The old siitras standing at the head of several of the philosophical
schools are essentially signposts in a line of oral argument (which may in
itself be a form of commentary on an older text, as in the exegesis of the
p n i ~ d s underlying the composition of the Vediintasiitrds), and in the ab-
sence of that oral corpus the siitras often have meaning for us only as they
are expounded in a full scholastic commentary. Throughout the history of
Sanskrit thought a similar situation is produced by authors who give a bare
outline of their material in short siitras or kiirikiis that are then fleshed out
I
2 Introduction
0.2. The five services ofa commentary
3
by the author's own commentary; the Tarkasarigraha of Annambhaga, for
example, is little more than a string of lists without the explanations sup-
plied in Annambhaga's own Tarkadfpikii or in the similar works by other
scholars.
In these laconic works the underlying motive is clearly a desire to fa-
cilitate memorization of the chief points of a system. Another frequent
source of works requiring commentary is the deliberate practice of com-
posing works of great difficulty, a practice that is explicitly and proudly
acknowledged in the poet Bhagi's famous boast, that his mahakiivya could
be understood only with the aid of a commentary (vyiikhyii-gamyam idarrr
kiivyam, etc., Bhat!ikiivya 22.34). Bhagi's verse was taken up by Bhiimaha
in his work on poetics and twisted into the petulant observation that such
poems are more siistra than kiivya (Kiivyiilarikiira 2.20), but we should
also remember that Bhiimaha himself, when describing the characteris-
tics of an ideal mahakiivya, does not say that the poem should require no
commentary, but rather that it should not require too much commentary
(nfitivyiikhyeyam, Kiivyiilarikiira 1.20).
Yet poems far less recondite than Bhagi's have attracted many com-
mentators (wituess the enormous number of available commentaries on
KiUidiisa's mahakiivyas), and for this the reasons are obviously pedagogi-
cal. Indeed, the fact that we have old commentaries on these mahakiivyas,
and only much more recent ones on great plays of equivalent age and dif-
ficulty, may be presumed to indicate that the mahiikiivyas made up the
standard cutriculum for Indian students of Sanskrit kiivya.
Another striking feature of the Sanskrit tradition is the frequency with
which works that may as well have been independent treatises are cast into
the external form of a commentary on an earlier text In this way many
treatises of great originality have been made to depend, at least nominally,
on earlier works that they leave far behind---one thinks of Prasastapiida's
Padiirthasarigraha, which is ostensibly a commentary on the
sutra, and which in turn provides the occasion for Udayana's equally origi-
nal Kiraniivalf. And the Kiraniivalfstands at the head of a very long string
of upon with varying degrees of dependence
on the work being commented upon in each instance. In many such works
the motives for writing in the form of a commentary go beyond the aim of
providing exegesis, and include the desire to associate oneself with an es-
tablished authority-to present one's views as a worthy unfolding of time-
honored tradition.
In many other works in which the commentarial format is freely cho-
sen the underlying concern is, once again, pedagogical. We often find a
scholar breaking a single body of material into several parts of increas-
ing difficulty, which are then presented as a string of commentaries, one
upon another. An example of this is Hemacandra's work On poetics, the
Kiivyiinuiiisana, which appears in the form of a number of brief sa/ras
supporting his own commentary, the Alarikiiracu(iiimarri, which in turn is
expanded upon in another commentary by the same author, the Viveka.
The material in the Viveka is of an advanced nature and could easily have
been presented in the form of an independent treatise, but by shaping it
into the form of a commentary on his more elementary work Hemacandra
allows the students in his advanced class to continue their studies using a
framework with which they are already familiar.
In some ways a work like Hemacandra's is equivalent to one of our
modern books in which extensive footnotes and appendices are provided,
but there is an important practical difference between the two formats; in
a modern work the additional material is presented in a number of sepa-
rate notes whose connection with the main text is merely signaled by ty-
pographical symbols or references, but in a Sanskrit work the additional
material is usually presented in a more continuous composition that relies
on a large body of conventional terminology and special syntactical devices
in fulfilling its role as a commentary on the original text. As a result; an
understanding of the special techniques of the Sanskrit commentarial style
is necessary in the reading of such works.
0.2 The five services of a commentary
In identifying the useful services offered by Sanskrit commentaries, we
have been guided by the list of the five characteristics of commentaries
given in a verse from the PariiSarapuriirra that is quoted in the NYiiyakosa:
pada-cchedal,z padflrthOktir vigraho viikya-yojanii
samiidhiinarrr vytikhyiinarrr pafica-Iaksanam
These five services correspond, in this same order, to the most
often faced by students attempting to decipher a Sanskrit text:
4 Introduction
0.2. The five services ofa commentary
5
0.2.1 Padaccheda
Word-division, separation of the words of the text the rules of sandhi and
the peculiarities of the Devanligar'i script make this the first problem that
confronts a student attempting to tnmslate a text. Before one can begin to
look up the words one must know where one word ends and the next be-
gins. Any commentary that is not too long-winded can immediately solve
this problem for even a beginning student, yet many students do not know
enough about the basic nature of a commentary to take advantage of this
simple service.
0.2.2 Padiirthokti
Stating the meaning of the words, paraphrasing: once it has been deter-
mined what the words of the text are, the next problem is to find the ap-
propriate meaning in the dictionary. Because most Sanskrit words have so
many possible meanings, the most frustrating part of the study of Sanskrit
for many students in their early years is the necessity of laboriously writing
down or holding in mind several definitions for each word until it can be
determined which possible meaning is most appropriate for each word in
the present context. While it is true that the glosses offered by commen-
tators may often be as unfamiliar to a student as the word being glossed,
even a beginning student who is confronted, for example, with the word
bhiiskaral;! in a Sanskrit text may find a simple gloss such as bhiiskaral;!
saryal;! accessible enough to save some of the time spent flipping through
the dictionaries. On a higher level, of course, these glosses often provide
information that is difficult or impossible to find in the standard reference
works, or that at least is more precise in its appropriateness to the text being
commented on.
0.2.3 Vigraha
Analysis of granunatical complexes (i.e., of nominal compounds and of
derived stems): here the dictionaries are often of no use, especially where
nominal compounds are involved. Time spent in learning to understand the
basic formulas of analysis used by commentators is useful not only because
it makes it possible to take advantage of this service of commentaries, but
also because it is a good way to acquire a grasp of the different types of
complex formations used in Sanskrit. c,:
0.2.4 Viikyayojanii
Construing the sentences, indicating the construction of the text: once
again, a student can avoid the tedious procedure of trying to figure out the
text in several stages, if only the basic principles by which commentators
repeat the words of the text being commented on are understood. These
principles are really fairly simple, but until they have been explained to a
student most commentaries appear to be masses of unconnected explana-
tions.
0;2.5
The answering of objections: this part of a commentary is concerned with
the ideas contained in the original text rather than with its words. Accord-
ingly, it differs from the four preceding services in two ways. First, at this
level a commentary goes beyond straightforward exegesis and becomes an
argumentative treatise in its own right. Second, this portion of a commen-
tary is based on the style of oral debate rather than On the style of oral
instruction, so that it uses a different arsenal of vocabulary, syntax, and' or-
ganization. An understanding of this style can be applied to the reading of
any text written in expository Sanskrit, whether that text is a commentary
or not.
The first four of these services form the subject of Part One of this
handbook in their achievement through methods of glossing using non-
technical language. Their achievement through the use of the technical
elements of PliJ.1inian granunar is touched on here in Sections 1.18-1.20,
but a full treatment of the topic lies outside the scope of this handbook in
its present form. The fifth service forms the subject of Part 1\\'0.
Chapter 1
The First Four Services
1.1 Introduction
Of the five services that a full commentary is expected to provide, the first
four are services of glossing and deal directly with the words of a text,
while the fifth is a service of discussion and is concerned more with the
ideas expressed in that text. Generally the first four services-the four
principal elements in the glossing of the words of a text----{;onsitute the bulk
of a commentary on a literary text, but may appear only intennittently in a
commentary on a philosophical or scientific text. In commentaries of lmy
type, however, the basic techniques of glossing are the same. In Sans)crit
commentaries these four services are achieved partly through the use of
the technical elements of ~ i n i n grammar, and partly through methods
of glossing using non-technical language. Part One of this handbook deals
with the use of non-technical language to fulfil the first four services:
I. padaccheda-division of words, i.e., the resolution ofexternal sandhi
(Sections 1.3-1.7).
2. padiinhokti-statement of the meaning of the words, Le., the giving
of synonyms, paraphrases, definitions, and examples (Sections 1.8-
1.17).
3. vigraha-analysis of complex formations, i.e., breaking them down
into their constituent elements (Sections 1.18-1.75).
4. viikyayojanii----{;onstruingthe sentences, i.e., indicating how the words
fit together syntactically (Sections 1.76-1.84).
9
10 Chapter 1. The First Four Services
1.2. Types of glossing
11
A full treatment of the achievement of these same services through the use
of the technical elements of P3J)inian grammar lies outside the scope of
this handbook.
The fifth of the services provided by commentaries, that involving dis-
cussion of the ideas contained in the text, will fonn the subject of Pan
Two.
1.2 Types of glossing
In the following sections, methods of glossing will be taken up one by
one, but in an actual commentary they often work together in explaining a
particular word. As preliminary examples of the types of glossing used by
commentators and of how the explanation of those types is distributed in
this handbook, here are the full remarks of the commentator Mallinatha on
three words from the Raghuvatpia of Kalidasa.
In the first example, Mallinatha is glossing the word saifave, "in in-
fancy":
Ragh.M. 1.8 fifor bluivaiJ faifav",!, balyam.
vacanodgatr-iidibhyo 'fl" ity aii-pratyayaJ:z. "sisutva1Jl saisava1Jl bdlyam"
ity amaraJ:z. tasmin vayasi [JaiSave}. Infancy is the condition of being
an infant, childhood. The suffix aN (i.e., -a, with vrddhi strengthening of
the first syllable of the stem to which it is added), by (satra 5.1.129 in
grammar) "aN is added to words for species of animals and for
the ages of life, and to udgatf, etc. (in the senses of the condition of being
that or the activity of that)." "Infanthood, infancy, childhood"-Amara. In
that age of life (viz., in infancy).
Here fifor bhiiva/.! is an analysis in ordinary language of the fonnation
faifavam, an example of secondary nominal derivation (Section 1.37.2).
This is followed by a simple synonym, biilyam (Section 1.9). The next
sentence gives the taddhita suffix by which the word is fanned according
to the PiiJ)inian system of grammar, and quotes the spe9ific rule by which
its fonnation is justified. The next sentence cites the treatment of the word
in the dictionary of Amara (Section 1.16.2). The final sentence, after these
explanations in which the word has been treated in the nominative case,
in effect gives it the locative case ending that it has in the original text
(Section 1.21). This last step is necessary so that the word will fit properly
in the syntactical arrangement of the entire verse (Section 1.78).
In the examples given in this handbook, the ';;ords repeated from a
text being commented on are emphasized in bold type. In this example,
the word being explained never actually appears in the commentary in the
same fonn that it has in the original text, and we have therefore supplied it
in square brackets at the point at which the words of the commentary serve
to reconstitute it.
In the next example, Mallinatha explains the word prajanam, "of the
people":
Ragh.M. 1.24 prajiiyanta iti praja janaf:r. "upasarge ca sarrzjiiayiim" iti
a-pratyayatr "prajii sylit santataujane" ity ttlstim (prajiiniim).
They are generated, thus progeny, people. The suffix J?a (i.e., -a, with loss
of the final vowel and final consonant of the root to which it is added), by
(satra 3.2.99 ofP3J)ini) "And (J;Ja is added to the root jan) when a verbal
prefix is used, if a name is being formed." ''praja in the senses of offspring
and people"-Amara. Of them (viz., of the people).
Here the complex fonnation being analyzed is an example of primary nom-
inal derivation (Section 1.29.1), and thus involves a krt suffix, but otherwise
the elements of Mallinatha's remarks are the same as in the preceding ex-
ample. '
In the final example Mallinatha chooses to end his remarks by repeating
in its original case the word being explained, rather than simply indicating
that case by the use of a pronoun:
Ragh.M. 1.54 dhura,!, vahantfti dhurya yugya/.!. "dhuro Ya4-flhakau" iti
"dhar-vahe dhurya-dhaureya-dhUrl(lii/.!
sa-dhuralJ'ldharab" ity amarab. dhuryiin rathiisvan viSramaya
vinfta-sraman kuru. They bear a yoke, thus YOke-bearers, draft animals.
The suffix yaT (i.e., -ya, with acute pitch accent on the. preceding
syllable), by (satra 4.4.77 in Piini) "yaT and flhaK are added to the root
dhur (in the sense of 'bearing that')." "In the meaning of a yoke-bearer:
dhurya, dhaureya, dhurin, and dhurQ1!uihara"-Amara. Rest make quiet
and rested the the chariot horses. .
12 Chapter 1. The First Four Services
Of the four services of commentaries listed in the preceding section, the
first service, division of the words of the original text, is accomplished in
each of these examples simply by isolating the word in order to explain it.
The second service, that of explaining the meaning of the individual word,
is accomplished partly by giving synonyms, partly by quoting dictionaries,
and partly by performing the third service, that of analyzing complex for-
mations, since each of the three words being explained in the comments
quoted here happens to be an example of such a formation. The fourth
service, that of indicating the syntactical structure of the original text, is
generally achieved through the use of methods that can be glimpsed in the
ways in which the commentator has returned each word to the required
grammatical case at the end of each of the examples.
It is in the analysis of complex formations that the role of PaI,tinian
grammar becomes most visible. The formulas used to give the meanings
of these items-sisar saisavam, prajiiyanta iti etc.-are
based on the treatment of such formations in grammar, but the
formulas themselves are in ordinary language and can understood without
a detailed knowledge of the technical elements of PaI,tinian grammar. The
same is true of much of the terminology used by the grammarians to name
important classes of complex formations. The artificial names for elements
such as suffixes, however, together with the meanings of the indicatory
letters added to them, reqnire special explanation; a brief overview of this
topic is given in Sections 1.18-1.20.
Chapter 2
Division of Words
1.3 Introduction
In Sanskrit sentences the constant application of rules for the euphonic
combination of words often makes it difficult to determine where one word
ends and the next begins. The resolution of sandhi between words is
the chief purpose of the oldest surviving exegetical works in India! to
each of the collections of Vedic hymns there was added in ancient times
a padapii!ha ("word-text") version, in which each word-isolate is exhib-
ited separately, with terminal sandhi. This straightforward approach to the
separation of words can still be found in the padaccheda feature of many of
the modern editions of Sanskrit texts printed as ponies for Indian students.
In most Sanskrit commentaries, however, the indication of boundaries be-
tween words is not given a distinct section reserved for that purpose, but
rather is achieved naturally through three general practices; the repetition
of the words with intervening explanatory material (Section 1.4), the oc-
casional specification of word-division in doubtfnl instances (Section 1.5),
and the citation of individual words for the purpose of making special re-
marks (Section 1.6). In the course of displaying the words of the text.being
commented on, a commentator may also mention variant readings (Sec-
tion 1.7).
13
14 Chapter 2. Division of Words
1.6. Citation ofindividuai words
15
1.4 Repetition of the words of the text
Ordinarily a continuous commentary repeats each word found in the text.
For examples of how this is done and for a description of common ex-
ceptions to this general rule, see Sections 1.76-1.80. Because each word
repeated from the text is usually followed by a gloss or by other inserted
material, separation of the words of the text is automatically provided in the
course of presenting the words for explication. The value of the repeated
words in providing this service is enhanced if they are printed in bold type;
if they are not, they may be sought out and underlined (Section 1.79).
1.5 Specification of word-division
For purposes of word-division, the repetition of words from the text in-
volves ambiguity only when two adjacent items in the text are repeated
next to each other in the commentary without the insertion of intervening
explanatory material. Because it may not be clear whether or not such
items form a compound word, the commentator will sometimes explicitly
state whether they are two words or one, or will specify where the division
should be made.
Ragh.M. 1.17 ii mana!" manum iirabhyety abhividhih. pada-dvaya,!,
cliitat, samii.sasya Since Manu-beginning with Manu;
this is an inclusive limit. And these are two words, because the compound
is optional.
Kum.M. 5.55 tadeti cchedah. radii prabhrti. tata iirabhyeti artha!,.
Divide after tadii. From then on, i.e., beginning then.
Kum. V. 5.63 tvayii yathii srutalf' tathiliva tat. ... yathii srutam iti dve
pade. It is just as you have heard. yatha srutam-these are two words.
1.6 Citation of individual words
In the course of giving a running gloss, the words repeated from the text
are generally presented in the same form in which they appear in the text
(Section 1.9). Often a commentator will also single out a word for further
remarks, and when this is done there are four methods of citing
the word (Le., of referring to it as a word, rather than simply using it in the
course of a continuous gloss):
1.6.1 iti
The word may be folJowed by iti (corresponding to our use of quotation
marks):
Kum. v: 1.58 vikn'yanta iti karma'}i. (The verb) vikriyante is passive.
Kum. V. 1.25 um iti praraya-kope. (The particle) um is used in
feigned anger.
1.6.2 -sabda, -"kiira
The word may be placed in compound with -sabda or (especially if it con-
sists of a single sylJable) with -kiira:
Ragh.M. 1.2 dvau kva-sabdau mahad antarOf!! sucayata/L The two
words 'where' suggest a great difference.
Ragh.M. 1.1 piirvalfparamiSvarau. parama-sabda!,
sarvottamatva-dyotaniirtha!l. Piirvati and the Supreme Lord. The
word 'supreme' has the purpose of revealing his preeminence above all
others.
Kum.M. 1.4 ki,!,ceti cakiiriirtha!,. The meaning of (the particle) ca is
"furthermore."
1.6.3 Inflected indeclinable
If the word is indeclinable, it may be given an inflectiorrai ending and be
declined like a masculine noun. The presence of the case-ending will then
be enough to indicate that the word is being referred to as a word rather
than being used in its own meaning. Ambiguity exists only when the"inde-
clinable ends in a long vowel and is presented in the nominative, but here
16 Chapter 2. Division of Words
1.7. Variant readings
17
the context will usually make it clear that the word is being referred to as a
word.
Ragh.M. 1.24 samuccaye. api (is used) in (the sense of)
conjunction.
Kum. V. 2.26 ksatdham kala s'a . '"
. . - - 1J1sma ltl Jarat-piitho 'tra ramyat h
Here the old reading "!qataha1J1.-kiira-SQTflSinal/' ij' more ..
Kum. V. 2.28 priirthayadhvam ity apapathah.
"prtirthayadhvam" is a
bad reading. . .
Kum. V. 3.44 iYildfni1lp chandasi vidhiinild bhiIsdyam triyambakam ..
prayogo durlabha iti maheivaram iti pathaniya . B' ,ttl
b . . m. ecause (sandhi
su sUtutes) such as "iy" are prescribed (only) for the" d' I
.. . e IC anguage the
usage tnyambakam" is difficult to arrive at in the classical Ian '
that one ought to read "maheivaram." guage, so
Kum. v: 4.36 marmlirpitam iti piithlintaram.
"mamdrpitam." . There is another reading,
Ragh.M. 1.4 athava palqdntare. athava (is used) in (the sense of) an
alternative.
Kum. V. 2.42 parvdpelqaya samuccaye. hfti palhe cdrthe
(The particle) ca (is used) in (the sense of) conjunction with repect to what
precedes. In the reading "hi," (the particle) hi (is used) in the meaning of
(the particle) ca.
As the word earthe in the last example shows, this device can also be used
in compound, where the inflectional ending will, of course, be absent. This
should not result in ambiguity as long as it is recognized that a compound
is being used.
1.6.4 Anubandhas
The word may be displayed along with the indicatory letters (anubandhas)
used in the Pii\linian system of grammar:
Ragh.M. 1.90 Ii tapo-vaniit. an maryadayam. As far as the
penance-grove. (The preposition) a [called an in Plilrini's grammar] (is
used) in the sense of an exclusive limit.
Kum.M.1.35 nano ca-sabda-prayogiid evajneyam.
And that the word na [called naliin PliJ;1ini's grarumar] is a modifier can
be known simply from the use of the word ca.
For an overview of the use of anubandhas and the technical citation of
grammatical elements such as roots, seeSections 1.76-1.80.
1.7 Variant readings
Variant readings are referred to by using the noun palha or any of the forms
of the verb palhati.
KumM. 3.44 keeit sahasiklilz tn'-loeanam iti pethuh.
(commentators are said to have) read "tn' I . .. .
- oeanam.
Certain reckless
Chapter 3
Giving the Meaning of Words
1.8 Introduction
The basic task of giving the meaning of a word in the text being commented
is carried out by placing an explanation of that meaning, at the simplest
level often merely a synonym for it (Section 1.9) alongside the wor4 in
question. Ordinary this simple gloss follows the word being explained,
but the location of the gloss may vary (Section 1.10). Commentators Ijlay
also gloss words with equivalent forms other than simple synonyms (Sec-
tion 1.11), and may expand the gloss in various ways (Section 1.12) or ap-
pend more removed paraphrases (Section 1.13). Techniques are available
for dealing with words that refer by extension to objects other than those
directed denoted (Section 1.14) and with words that have double mean-
ings (Section 1.15). These techniques of glossing may be supplemented by
the use of separate definitions (Section 1.16). Further devices are used in
glossing special items such as proper names, the names of species, parti-
cles, indeclinables, and pronouns (Section 1.17).
When the word being explained happens to be a complex formation
arrived at through derivation, the formation may be analyzed using the for-
mulas available for each of the types of derived stems (Sections 1..\8 ff).
Words may also be explained through the use of the technical elements of
PiiI)inian grammar.
19
1.9 Simple synonyms
The words repeated from the text being commented on are most commonly
glossed by the insertion of a synonym after each. Simple glosses of this sort
may be provided for virtually every part of speech:
Ragh.M. 1.24 tasii,!" pitaTas tujanma-hetavo
kevalam utpiidaka evdbhuvan. But their were causes of birth
creaters merely of birth alone simply producers. [here eva glosses
kevalam]
21
1.11. Glossing by equivalent forms
Chapter 3. Giving the Meaning of Words 20
Ragh.M. 1.1 vande'bhiviidaye I honor I salute 1.10.2 Vigrahas
Ragh.M. 1.2 matilJ prajflii mind intellect
Ragh.M. 1.4 me mama my of mine
When an analysis of a grammatical complex (such as a nominal compound)
is given, the original word from the text is generally repeated after the
analysis, rather than before it (see Section 1.21).
Ragh.M. 1.29 nun"'!' dhruvam surely certainly
1.11 Glossing by equivalent forms
These inserted glosses are placed in grammatical apposition with the
word being glossed. Thus a gloss of a verb will agree with it in person and
number; a gloss of a substantive will agree with it in case and number, and
a gloss of an adjective will agree with it in case, number, and gender:
A simple gloss need not be an exact morphological copy of the original;
often a different construction with equivalent meaning is used instead. Two
types of glossing by equivalent forms are especially common:
Ragh.M. 1.27 vande'bhiviidaye I honor I salute
Ragh.M. 1.2 mohiid ajfiiiniit out of folly out of ignorance
Ragh.M. 1.36 snigdo agreeable (masc.)-sweet
1.11.1 Periphrasis
A periphrastic construction is used to bring out the precise meaning of a
morphological feature in the original: '
Ragh.M. 1.20 anumeyii anumiitu'!' inferable able to be
inferred
1.10 Location of glosses
Ragh.M. 1.26 yajfiiiya yajfia,!, kartum for sacrifice in order to
perform sacrifice
Usually the gloss follows immediately after the word quoted from the text
being commented on. There are two principal exceptions:
Ragh.M. 1.72 mucye mukto bhaviimi I am freed I become freed
1.10.1 Enclitics
Enclitics and other words closely dependent upon a preceding word must
usually be separated from the words they are glossing.
1.11.2 Participle
A past active participle may be used to gloss a finite verb:,
Ragh.M. 1.17 na ryanyur they did not trangress
they did oot go beyond
Ragh.M. 1.28 yathii iva as if medicine like
medicine [here iva glosses yatha]
Ragh.M. 1.21 jugopa ralqitaviin he guarded he protected
22
Chapter 3. Giving the Meaning of Words
1.13. More removed paraphrases 23
1.12 Expansions of glosses
A simple gloss is often expanded beyond the mere giving of a synonym by
including illustrative or explanatory material.
1.12.1 -iidi
Inserted examples are generally offered in the forms ending in -adi ("such
as," Section 2.44):
Ragh.M. 1.14 siiribhi/;J kavibhir viilmfky-iidibhih by wise men by
poets such as Viiltnlki
Ragh.M. 1.16 npa-gu'.'ai riija-gunais tejah-pratiiptJdibhih
kula-Sfla-diilqinyadibhis ca with the qualities of a king with royal
qualities such as valor and prowess, and such as good family, good
conduct, and politeness.
1.12.2 Implications
Commentators very frequently fill out a gloss by explicitly stating what is
merely implied in the original wording:
Ragh.M. 1.24 bhaya-hetubhyas triiniit because of
protecting because of protection from sources of danger
Ragh.M. 1.24 bhara1}tJd anna-piinddibhih posaniit because of
supporting because of nourishing with food, drink, etc.
Ragh.M. 1.25 pari'.'el1Jr diiriin parigrhnatah of the marrier of the
wedder of a wife
1.12.3 Implied contrasts
Similarly, commentators often spell out the implications of an expression
by making implied contrasts explicit:
Ragh.M. 1.7 yaJase kfrtaye .. , vijiiMu!,ii,!, vijetum icchanam, na tv
artha-sa1J1grahliya. prajayai sa'fltanaya grha-met!!'inii'!'
diira-parigrah{uJiim, na tu kiim8pabhogiiya. Of those who were
desirous of victory wishing to conqner for the sake of fame of glory, but
not for the sake of amassing wealth. Of those who were householders
married men for the sake of offspring of progeny, bot not for the sake of
the enjoyment of pleasure.
1.13 More removed paraphrases
When a commentator feels that glossing a word with a synonym or equiv-
alent form is not sufficient, he may provide a freer paraphrase. Such para-
phrases are usually followed by one of three phrases: (Section 1.13.1) ity
arthah, (Section 1.13.2) iti bhiivaiz, or (Section 1.13.3) iti yavat. For the
first two of these phrases other expressions are sometimes substituted (Sec-
tion 1.13.4), and additional formulas are used to mark comments serv-
ing to bring out the implications of the words being commented on (Sec-
tion 1.13.5) and the purposes of the author's statements (Section 1.13.6).
1.13.1 ity arlhah
The phrase ity arthaiz (lit., "such is the meaning," "the meaning is that ... ")
is used to mark a paraphrase that simply restates the literal meaning of the
text in another way. Such paraphrases are often added after a simple gloss:
Kum.M. 1.46 lalo nirbhedam. tat-sadrsam ity artha!r.
Not distinguished from it-not different from it. In other words, like it.
Kum.M. 1.43 low capa/ii, paribhramana-sT1ety arthah Unsteady
restless, Le., tending to roam about
Just as frequently they are used instead of a simple gloss:
Kum.M. 1.11 mandiim ... gati,!, na bhindanti na tyajanty ity arthah
their slow pace they do not break-in other words, they do not abandon it
24
Chapter 3. Giving the Meaning ofWords
1.13. More removed paraphrases 25
Kom.M. 1.18 niiga-vadhUpabhogyalfl ity
artha!,. who was to he enjoyed hy a Naga wife-i.e., who was to
marry a Naga bride.
1.13.2 iti bhiivaJ.!
The phrase iti bhiivaJ:z (iiI. "such is the idea," "the idea is that ... "! is used
to mark comments that go a step further by stating the underlymg tdea
intended by the author of the text being commented on.
Kum.M.1.15 anvi!j{a-mrgair margita-mrgaiJ;. sriintair iti for
deer who had been hunting deer. The idea is that they were tired.
KomM.3.67 /dmcit ... parilupta-dhairyaJ:z. na tu priikrta-janavad
atyanta-lupta-dhalrya iti bhiiva!,. With his calmness slightly ruflIed..
But not with his calmness exceedingly ruffled, as in common people-this
is the idea.
Because expressions that are difficult enough to require a free paraphrase
are also likely to require further explanation, it is quite common to find a
remark ending with ity arthaJ:z being followed immediately by a comment
ending with iti bhiivaJ:z:
Kom.M. 3.68 siicr-k[lii . .. lasthau. hriyii mukha1[l siicr-krtya sthitlty
arthah. no kevalll1Jl harasyaiva devyii apy udito rati-bhdva iii
She stood turned aside, i.e., she stood with her face averted out of
embarrassment. The idea is that the emotion of love had arisen not only in
Siva, but in the Goddess as well.
Kom.M.5.86 klesah phalena phala-siddhyii punar-navaliilf' vidhatte.
iirvavad apiidayatfty arthafl. sa-phalaJ; kleso na klesa iti
p.. .
bhiivah. Toil, through its fruit-through the accomplishment of Its
freshness. In other words, it makes one unburdened by
toil, just as one had been before. The idea is that fruitful toil is not toil.
Kom.M. 4.10 tava padavllfl margam aham pratipatsye. tviim
ato me nasti viciira iti Your path way
I shall resort to-i.e., I shall follow you. Hence you need not worry
about me-this is the idea.
1.13.3 iti yavat
The phrase iti yiivat (lit., 'Just this much [is meantY') is used after a para-
phrase that expresses the meaning of the original text more precisely, either
by substituting a more specific term for a wider one, or by reducing a pic-
turesque expression to plain language. iii yiivat may thus be translated
"specifically," "to be precise," "to put it plainly," etc., depending on the
contexl.
Ragh.M. 1.25 manffi!,o yiivat. Of a wise man
of a learned man. Specifically, of one who can recognize faults.
Kum.M. 3.5 punar.bhavaJ.t punarutpatti!1.. sa1J1sdra iti yiivat.
Reoccurring arising again. To be precise, transmigration.
Kum.M. 1.39 vedi-viiagna-madhyii vedivat krsa-madhyii.
tanu-madhylti yiivat. Altarground-narrow-waisted thin-waisted like
an aItarground. To put it simply, slender-waisted.
1.13.4 Other phrases
Several other phrases are frequently used to mark the end of a free para-
phrase. The most common of these are ity abhipriiyaJ:z ("such is the im-
port"), iti tiitparyaJ:z ("such is the overall sense"), and ity iisayaJ:z ("such is
the intention [of the author]"). These phrases are used in exactly the same
way as ity arthaJ:z (Section 1.13.1) and iti bhiivaJ:z (Section 1.13.2), with
similar meanings.
Kum.M. 5.73 yatOO na kriyate tatoo tvarn api
tasmai na gha.tasa iti tatparyiirthatJ.. The overall sense is that just as the
consecration of the sacred post used in Vedic ritual is not performed for
the executioner's stake in a cemetary, so also you are suited for him.
1.13.5 anena, etena
In another group offormulas used to mark comments on the author's inten-
tions, the remark of the commentator begins with the word anena or elena
("by this [expression or statement]"). The formulas anena ...
and anena ... iti vivakryate (both meaning "by this he means to say ... ")
are used with essentially the same force as the formula ity artha/.! (Sec-'
tion 1.13.1). The other formulas beginning with anena generally involve
an attempt to convey the subtler implications or suggestions of the author's
statement; e.g., anena . .. sueita/.! and anena . .. iti sucyate (both mean-
ing "by this .. , is hinted at"), anena . .. iti gamyate ("by this ... is made
known"), and anena ... iti dhvanyate ("by this ... is suggested"). All such
formulas should be self-explanatory once their general pattern has been
understood.
1.14 Meanings by extension
Kum.M. 1.15 kramtid faitya-saurabhya.miindyiiny
uktiini. The three modifying (phrases) express:coolness, fragrance, and
gentleness, in that order. ';.
27
1.14. Meanings by extension
The verbs and are used by commentators to refer to the
power of words to include in their meanings objects connected with the
object directly denoted by them.
Chapter 3. Giving the Meaning of Words 26
Ragh.M. 1.8 abhyasta-vidyiiniim. etena brahmacaryasramo
"Of those who studied the sacred lore"-by this he means the stage of
sacred studentship.
Ragh.M. 1.59 atharva-nidher ity anena purohita-krtyabhijnatviit
tat-karma-nirvahakatvarrr muner astfti sficyate. "Of him who was the
repository of the Atharva Veda"-by this it is hinted that the sage was
expert in the rites of the domestic priest and was therefore capable of
carrying out those rituals.
Kum. v: 1.37 athavii nzvi-sabdena nfvyil yad grathyate tad adho-viiso
lakryate. Or, by the word "waistknot" there is included by extension the
lower garment that is bound by that waistknot.
Kum. v: 3.10 madhus caitras tena ca vasanto lakfyate.
Madhu---Caitra, and this includes by extension the springtime.
Knm. V. 1.27 dipa-Sabdenatrajviilii-varti-taila_bhiijanatmaka/.!
samudiiya Here the word "lamp" implies by extension the '
aggregate composed of flame, wick, oil, and vessel.
1.13.6 References to speech
Similar observations are frequently made by any of the many ways of re-
ferring to the act of speaking on the part of the author of the text being
commented on.
Kum. V. 1.15 bhiigfrathf-nirjhara-fikarii!'iiIJI
gariga-prapiita-jala-ka!,iiniif!l saitya-pu!,yatvoktiJ:!. The
carrier the tosser of the sprays of the cataracts of the Bhagi"rathi of the
drops of water of the waterfalls of the Ganges-this is an expression (ukti)
of coolness and holiness.
1.15 Glossing of puns
When words of the text being commented on refer simultaneously to two
objects through double meanings, the commentator may organize his ref-
erences to these two sets of meanings either by using expressions meaning
"in (reference to) one ... in (reference to) the other place" or by referring
to the two sides with the word
Kum.M.2.29 sahasra-grahaIJam iisthlitisayartham. The mention of
the thousand (eyes of Indra simultaneously turned towards his guru) is
aimed at the abundance of his regard (for the guru).
Kum.M. 2.27 labdha.prati!!hii/.! labdha-sthitayaJ:!. labdhiivakiiSii ity
anyatra. Having obtained a foundation [in reference to the gods],
having obtained their positions. In the other [alternative, i.e., in
to rules of grammar], having obtained the opportunity for operation.
Ragh.M. 1.44 yupo niima sarrzskrtafl pasu-bandhiiya yupii
eva cihnani (That which is called) ayupa
("sacrificial post") is a special piece of wood consecrated for the tying up
of the sacrificial victim. Of which the marks are yupas, in those
(yiipa-marked ones).
Kum.M. 4.9 upaciira-pada'!l parasya rafijaniirrharrz yad
asatya-bhia1}Qrrz sa upaciiras tasya padarrz sthii1.!#m. kaitava-sthiinam iti
yiivat. An occasion of upaciira-an upaciira i;'an untrue statement
made for the purpose of pleasing someone else. An occasion-an
instance-of it. To put it bluntly, an instance of deceit.
Alternatively the commentator may simply insert the definition before or
after the portion of the running gloss in which the word in question is pre-
sented. In such instances he will often employ the particle nama ("by
name," "that which is called ... ") immediately after the word being de-
fined, in order to make it clear that he is pausing to give a definition:
29
1.16. Definitions
Ragh.M. 2.73 adarianena praviisa-nimittena ... candra-palqe,
adarianarp By nonappearance caused by his
being away from home.... In the case of the moon, the nonappearance is
caused by its waning.
Kum.M. 1.24 labdha udayo yaya sa labdhOdayii. utpannety
abhyuditety anyatra. By whom arising has been attained. she is
labdhOdaya, i.e., born [in reference to ParvatI]. In the other [i.e., in
reference to the moon], having risen.
Ragh.M. 4.33 phala,!, labham. prasava,!, ca. ... utkhiita/,l
sva-padlic cyavitai!z. anyatra, utpli,titai/.t. The fruit the profit [in
reference to the kings]. And in the case of the trees, their produce....
Dislodged unseated from their positions. In the other [Le., in reference to
the trees], uprooted.
28
1.16 Definitions
Ragh.M. 1.4 valflie kule. santiino varrziaJ:z. In
the vfl1!lsa in the family line. A vaTflsa is a series preserving continuity by
birth.
When a word in the text being commented on is used in a technical or
uncommon sense, the commentator may find it necessary to interrupt his
running gloss in order to give a formal definition of the word. In such
instances he may either offer his own definition (Section 1.16.1), or quote
an authoritative dictionary (Section 1.16.2).
1.16.1 Definitions otTered by the commentator
1.16.2 Definitions quoted from dictionaries
Most of the Sanskrit dictionaries quoted by the commentators follow one
of two standard formats. Some, such as the dictionary of Amarasimha
simply string synonyms together after the fashion of a thesaurus, so ali
the words appear in the nominative case:
Since a definition involves an independent sentence stating the meaning
of a word, it cannot reproduce the case and number of the word as used
in the original text. To solve this problem a commentator will sometimes
add a pronoun after the definition in order to "pick up" the defined word
and restore it to its original case, etc., so as to match the original text and
continue the running gloss:
Ragh.M. 1.7 tyagiiya-sat-piitre viniyogas tyiigab-tasmai for
donation---donation is presentation to a worthy recipient-for it
Ragh.M. 1.7 "tyiigo vihiipitarrz diinam" ity amarafl. "Donation, grant,
gift"-Amara [cited to define tyaga].
Others, such as the dictionary of Kesavasvamin, display the word to be
defined in the nominative case and its meanings in the case:
Ragh.M. 1.4 "vajrarrz tv a-St11 kulisa-sastrayofl, mQl}i-vedhe
ratna-bhede" iti kesavab "vajra, non-feminine (i.e., found both as
masc. and as neut.), in (the meanings of) thunderbolt and weapon, drill for
boring gems, type of jewel"-Kesava.
1.17 Treatment of special items
Kum.M.2.64 rati1J smara-priya. "ratitz smara-priya" ity omara!}.
Rati the beloved of Kama. "Rati, the beloved of Kama"-Amara.
Frequently the quotations from dictionaties add little to what the commen-
tator has already said on his own, and appear to be made simply for the
weight of the lexicographer's authority or to reinforce the memory of stu-
dents who were supposed to learn the dictionary by heart:
31
1.17. Treatment ofspecial items
Kum.M. 3.26 aSoko aSoka a species of tree
Kum.M. 2.59 ayaskiintena by lodestone by a
particular kind of mineral
1.17.2 Names of species
When words referring to species or particular typeS' of things are glossed.
the term is placed in compound after a word referring to a wider
class of things to make it clear that the word being glossed does not apply
to all members of that wider class.
Kum.M.2.21 piisa!J rajjur lariat lasso, a particular
type of weapou
Quotations from dictionaties interrupt the runrting gloss when introduced
near the place in the commentary at which the word to be defined is pre-
sented. Some commentators therefore save such quotations and lump them
together at the end of their comments on a verse or other portion of the text.
30
Commentators use special devices to mark proper names (Section 1.17.1)
and names of species (Section 1.17.2), and to explain the meanings of par-
ticles (Section 1.17.3), other indeclinables (Section 1.17.4), and pronouns
(Section 1.17.5).
1,17.1 Proper names
Proper names are marked as such, either by being followed by nama (used
as a separate, indeclinable word) or by being displayed as the first mem-
ber of a bahuvrfhi compound ending in a word meaning "name" (nama,
sa'l"jfili, akhya, etc.).
1.17.3 Meanings of particles
Many indeclinables in Sanskrit are function words whose mearting is pri-
marily grammatical. For these items it may be difficult to provide a syn-
onym or definition, and to explain their meaning a commentator must cite
the word and make a direct statement of the force with which it is u$ed.
For examples of how such statements are made see Section 1.16. Note par-
ticularly the device of expressing the function of a particle through a word
presented in the locative case (e.g., apii,l samuccaye, "api in the sense of
conjunction"; ali maryiidiiyam, "a in the sense of an exclusive limit").
Kum.M. 2.47 uccai?Jsravo niima by name
1.17.4 Other indeclinables
Kum. v: 2.47 The most excellent
horse named UccaiJ.1Sravas.
For other indeclinables, such as temporal adverbs, a commentator may sim-
ply indicate the precise sigrtificance of the word in the present context:
Kum.M. 2.1 tiirakelJa tiiraka-namna vajrat;akha-putre!Ia kenacid
asure!la. By Taraka by a certain demon named Taraka, the son of
VajraI)akba.
Ragh.M. 1.59 atOO prasnantaram next after the question
Kum.M. 3.59 katharrtcid atiyamena somehow with great efforl'
32 Chapter 3. Giving the Meaning afWards
1.17.5 Pronouns
Similar information is inserted to explain the significance of pronouns.
When the pronoun's antecedent is not in the text, the commentator may
supply it:
Ragh.M. 1.18 sa riijii he the king
Kum.M. 1.2 ya,!, himiilaya whom Himalaya
And for other pronouns the special force with which they are used may be
indicated:
'ddha k. -. m that famous head of hair
Kum.M. 1.48 ta,!, pras. '!' e a-pasa
Kum. V. 1.56 kendpi kiimena kaydpy ajiiiitecchayii with a certain
desire with some unknown wish
Megh.M. 1 kaicid a certain-unspecified by
name-yakfa.
Chapter 4
Analysis of Grammatical Complexes
1.18 Introduction
The opportunity to break down grammatical complexes into their con-
stituent parts arises constantly in commentaries on Sanskrit texts, since
most words in the language are made up of readily discemible elements.
Any inflected word, for example, contains both an inflectional stem and an
inflectional affix, and such words can be analyzed by identifying the stem
and naming the affix. The resulting analysis, however, will be a purely
technical one, since the affixes can be named only through the use of the
artificial terminology of Sanskrit grammar. When speaking of San.krit
commentaries, what is more commonly meant by the term vigraha is an
analysis in ordinary language, which is produced by using several separate
words to restate the meaning of a single complex form-Le., by "expand-
ing" the complex form into a phrase with the same meaning.
Because ordinary language provides no entire words that can serve as
direct substitutes for morphemes such as the inflectional endings, tltis sec-
ond type of analysis is possible only for words more complex than inflected
primary words, and it is therefore reserved for formations involving nomi-
nal or verbal stems that are themselves derived from simpler elements. In
analyzing these complex formations, it is usually possible either to sep-
arate the constituent elements and place them in a phrase in which their
relationship is explicitly stated (as in the analysis of many kinds of nomi-
nal compounds), or to produce a phrase that gives the same meaning as the
complex form by using ordinary words to express the force of the gram-
33
matical elements involved in the complex formation (as in the analysis of
forms derived thtough suffixation).
For these complex formations there are thus two kinds of analysis avail-
able: analysis in the technical language of grammar (sastrfya-vigraha,
"technical analysis;' or alaukika-vigraha, "extraordinary analysis"), and
analysis in ordinary, non-technical language (laukika-vigraha).
The siistrfya-vigraha proceeds by naming the sequence of stems and
suffixes involved, together with any inserted elements required by the se-
quence. For example, a genitive compound such as
("king's man") will be explained in a technical analysis as being made up
of the stem rajan with the termination of the genitive singular, -as (which
is dropped in the final form), followed by the stem with the ter-
mination of the nominative singular, -so The actual analysis will take the
form rajan + Nas + + sUo The extra letters added in the names for
the suffixes serve various purposes and disappear in the final form; here
the N in Nas indicates that certain things are inserted before this suffix
when it comes after a feminine stem, and the U in sU is simply to make it
possible to pronounce the name of the suffix. Similarly, the taddhita forma-
tion steyam ("thievery," from stena, "thief'), may be analyzed in technical
terms as stena +yaT + am. Here the T in the technical name of the taddhita
suffix yaT indicates which sort of pitch accent the suffix carries.
For these same formations the laukika-vigraha proceeds by expanding
each formation into a phrase having the same meaning:
("a man of the king's") and stenasya karma ("the activity of a thief').
This sort of analysis in ordinary language occurs constantly in Sanskrit
commentaries, and is obviously more readily accessible to students than
analysis in technical language, since it requires no knowledge of the artifi-
cialterminology of PilJ!.inian grammar. What it does require is familiarity
with the modes of complex formation susceptible to such analysis (Sec-
tion 1.19), the general principles of analysis involved (Section 1.20), the
manner in which the formulas of analysis are inserted in a running com-
mentary (Section 1.21), and the common formulas used in analyzing each
type of complex formation (Sections 1.27-1.75).
Full coverage of the terms required in technical grammatical analysis
and the procedures by which they are used by the commeotators is beyond
the scope of this handbook, but a brief overview of how such analyses are
usually worded by commentators is given below in Section 1.20.
1.20 General principles of analysis
For each type and subtype of complex formation susceptible to analysis in
ordinary language, there exists a basic formula of analysis that, while it
may be understood without a knowledge of the details of PiiI,tinian gram-
mar, is based directly on the treatment of these formations by Piinini and
his successors. For some formations, notably those derived by the
of taddhita suffixes (Section 1.35), the formula follows the exact wording
of the rules of PiiI,lini. For others, the formulas are desigued to express the
meanings assigned to the formation by PiiI,1ini.
35
1.19. Modes of complex formation
1.19 Modes of complex formation
5. or expressing the sense of several words related by
meaning or syntactical concord thtough the retention of only bne
of those words, e.g., "dogs" for "dogs and bitches" (Sections 1.73-
1.75).
4. samiisa-vrtti, or the formation of new nominal stems thtough the
compounding of nouns (Sections 1.43-1.72).
I. san-ady-anta-dhiitu-vrtti, or the formation of new verbal stems (end-
ing in the suffix san, etc.), i.e., of denominative verbs by adding
suffixes to nouns, and of stems of secondary conjugation by adding
suffixes to verbal roots (Sections 1.22-1.26).
2. krd-vrtti, or the formation of new nominal stems by primary deriva-
tion, i.e., by adding krt suffixes to verbal roots (Sections 1.27-1.34).
3. taddhita-vrtti, or the formation of new nominal stems by secondary
derivation, i.e., by adding taddhita suffixes to nominal stems (Sec-
tions 1.35-1.42).
An analysis in ordinary language is a restatemeni>in the "phrase mode"
(vtikya-vrtti) of the meaning expressed in a more compact mode by the
formation being analyzed. Sanskrit grammar recognizes five modes (vrt
ti
)
of the formation of derived words for which a vigraha in ordinary language
may be possible:
Chapter 4. Analysis of Grammatical Complexes 34
36
Chapter 4. Analysis of Grammatical Complexes
1.20. General principles ofanalysis 37
Some words are analyzed not in terms of formulas used for types or
subtypes of complex formations but in terms of the specification by the
grammarians of a meaning for that particular word. Such irregular words
are said to be nipiitita, "set down," as opposed to words that are
labdha, i.e., whose formation is arrived at by more general rules (see, for
example, Section 1.34).
To some extent the formulas actually used by many commentators are
drawn from a practical tradition of commentarial techrtiques rather than
directly from a close adherence to the detailed explanations of the gram-
marians. Thus for some formations (e.g., compounds, Sec-
tion 1.58) the formula usually encountered in the commentaries is more
straightforward and less strict than some grammarians would like, and for
others (e.g., samlihiira-dvandva compounds, Section 1.63) the commen-
tators often do not bother to include in their formula of analysis the full
details of the special elements of meaning involved in the formation. And
for some formations the full standard formula coexists with a simplified
formula used when the complex formation forms part of a larger complex
(as in the analysis of kannadhiiraya compounds serving as members of
larger compounds, Section 1.71).
When some of the less common formations are to be analyzed, the
degree to which the formulas used in a given commentary coincide with
the formulas offered by the granunarians may also be influenced by the
type of reader for whom the commentary is intended. Rudimentary lists
of simple formulas of analysis, collected in short handbooks such as the
Samiisacakra, join lists of verbal and nominal paradigms as the first items
memorized by students in the traditional system of Sanskrit-medium ed-
ucation, and commentaries intended for students in their early years of
study will naturally attempt to conform to the formulas presented in these
elementary handbooks.
Careful study of the basic formulas of analysis available for each type
of formation is valuable in several ways. First, a knowlege of the formulas
is useful in itself as an aid to a clearer understanding of the different types
of complex formations used in Sanskrit. Second, farrtiliarity with the nature
of the more important formulas is helpful in recogrtizing an analysis for
what it is, should it show up unexpectedly in an otherwise discursive work-;
of commentary. Finally, the ability to recogrtize the basic formulas easily
is necessary if one is to understand the sentences in running glosses into
which they are inserted, because the formula itself may often be obscured
in ways described in the following section.
For all these reasons the treatment of each type of complex forma-
tion in this handbook includes a statement of the basic formula of analysis
appropriate to that formation, drawn from the works of the grammarians,
alongside any examples taken from commentaries.
Of the five types of complex formation listed above, the first three
(derivative verbs, nouns of primary derivation, and nouns of secondary
derivation) are also susceptible to analysis in terms of the technical ele-
ments of Pfu:J.inian grammar in addition to analysis in ordinary language.
While both approaches depend upon the details of Pal)ini's treatment of
the language, the technical analysis goes beyond the ordinary analysis in
employing the specially coded terrrtinology used by Pfu:J.ini in referring to
grammatical elements. In the analysis of derived forms the most important
of these elements are the suffixes, and the Pfu:J.inian treatment of suffixes
involves three special problems.
First, Pfu:J.ini recognizes derivation by suffixation only, so that many
phenomena that we might not automically associate with suffixes are de-
scribed by the Pfu:J.inians as being due to the addition of some suffix. For ex-
ample, the form -libha that occurs at the end of the upapada compound an-
alyzed by Mallinatha and Hemadri in the examples used in Section 1.21.2
is a derivative of the verbal root bhii, and the vowel of the root appears to
have been shortened. For reasons of economy the way that Pal)ini expllnns
such a form is to say that it is the result of the addition of a short -a suffix,
before which the long -li of the root has dropped. Similarly, formations that
appear to consist of the bare root with no alteration (e.g., the final member
of veda-vid, "Veda-knowing," which is simply the root vid, "to know") are
explained by Pfu,tirti as the result of the addition of a "null suffix," i.e., a
suffix that completely disappears.
Second, the technical names of the suffixes generally contain extra let-
ters beyond the actual form of the suffix itself, the purpose of which is
purely indicatory. For example, the short -a suffix that forms the compo-
nent -libha is technically called Ka, and in this name the K is merely a
"dummy letter" serving to convey grammatical information abont the re-
sulting form. Such an indicatory phoneme (displayed here in uppercase)
is called an anubandha or it, and it does not appear in the final form. The
general purpose of K in this role is to indicate that when the suffix marked
38
Chapter 4. Analysis of Grammatical Complexes
1.21. Insertion of analyses in a commentary
39
with it is added to a root, the gUl)a-strengthening usually caused by the ad-
dition of a suffix doe not occur (e.g., when the suffix Ka is added to the
root budh, the resulting form is budha rather than bodha), and there is also
a special rule to the effect that the final vowel of roots in long -a is dropped
before suffixes with indicatory K (thus the form -abha rather than -abha).
Third, in the technical names of some suffixes even the actual form
of the suffix does not appear, being replaced by some substitute specified
by Pal)ini. For example, yu in the technical name of suffixes stands for the
longer form ana, and to this yu will be added indicatory letters covenng the
grammatical phenomena associated with the suffix in question. Thus LyuT
is the name of the krt suffix -ana that produces nominal stems in which
the acute accent on the syllable before the suffix (which phenomenon
is indicated by the L in the technical name) and for which the feminine
stems end in -I rather than the usual -a (which is indicated by the Tin the
technical name).
All three of these problems arise in the case of a suffix such as KviP,
the technical name for one of the null suffixes mentioned under the heading
of the first problem noted above. The actual suffix is vi, which is said to
always disappear, so that it is in effect replaced by zero. The K, once
again, indicates lack of strengthening, and the P indicates that roots ending
in short vowels will add the augment -t (e.g., loka-krt, "world-making,"
from the root kr).
The of these indicatory letrers are fully explained by Otto
von Biihtlingk in his Partini's Grammatik and (much more accessibly for
readers of English) by G. V. Devasthali in his Anubandhas of Partini. Both
books also give lists of the suffixes, both with and without the indicatory
letters, and some similar information is provided by M. R. Kale in A Higher
Sanskrit Grammar. A more thorough introduction to Pal)ini's methods is
available in the works by George Cardona listed in the Appendix. We
cannot reproduce this information here, but we can point out the two most
common ways in which these technical names of suffixes are used.
First, the actual technical analysis of a complex formation consists of
the technical name of the suffix, preceded by the ablative of the item to
which the suffix is added. Thus the word saisavam ("infancy"), formed
from siSu ("infant") by the addition of the short -a suffix technically called
aN (the J:I indicates v,ddhi-strengthening of the first syllable of the forma-
is analyzed sisor aJ:l. Further technical devices may used in identify-
ing the base to which the suffix is added; for example, one way of referring
to verbal roots is by adding the suffix -ti, so that the word manam, formed
from the root ma ("to measure"), may be analyzedby saying mater LyuT
("LyuT [i.e., the taddhita suffix -ana] after mati [i.e., the verbal root ma]").
Second, the meaning of these suffixes is indicated by preceding the
technical name of the suffix with its meaning in the locative case (cf. Chap-
ter 6). Thus for saisavam a commentator might say bhiive aN ("aN in
the sense of a state or condition"; see Section 1.28), and for manam, 'if it
is interpreted as having instrumental force (miyate aneneti mfinam, "one
measures with it, thus a measurer:' i.e., an instrument of measuring) the
comment might be kararte LyuT ("LyuT in the sense of an instrument"; see
Section 1.29.3).
This same technique is used in dealing with other grammatical ele-
ments as well. Verbal roots, for example, as they appear in the tradi-
tionallist (the Dhiitupa,tha, an appendix to Pal)ini's grammar), are provided
with indicatory letters of their own-e.g., kr, "to do," is technically called
PUkrtil, with the PU indicating that the root may take a certain suffix, and
the til indicating that the root takes atmanepada endings with the benefit of
the action accrues to the agent. The meaning of the root is indicated by a
word the locative case placed after the technical name of the root-e.g.,
PUkrN kararte, "kr in the sense of doing."
Commentators frequently use such technical remarks as supplements
to analyses in ordinary language or in place of them. In the quotations
from commentators offered in the handbook we will often include such
remarks as samples of this approach, but a full understanding of their use
requires a dedicated study of PilI)inian grammar.
1.21 Insertion of analyses in a commentary
The formulas of analysis described below in Sections 1.22-1.75 are in-
serted at the proper places in a running commentary in much the same way
as the simple glosses described in Section 1.9 (for detailed examples of
the arrangement of running commentaries and the insertion of explanatory
material, see Sections 1.76--1.80). In their placement within a commen-
tary the formulas of analysis differ from simple glosses only in that, as
the examples given in the following sections indicate, an analysis is usu-
40 Chapter 4. Analysis of Grammatical Complexes
1.21. Insertion ofanalyses in a commentary
41
ally placed before the complex being analyzed, while synonyms and other
simple glosses usually follow the words they explain (Section 1.10).
In practice commentators often do not bother to repeat the full original
form of the complex being analyzed (especially when it is a nominal com-
pound); in such instances the analysis itself replaces the repetition of the
original formation, or the analysis may conclude with a simple reference to
the original form rather than with its actual restatement (as in the phrases
tat tatha, tat tath8ktam, tat tatha-bhlitam, etc., used at the end of analyses
of bahuvrfhi compounds; Section 1.70).
The basic formula of analysis may often be obscured by the addition
of other material when the commentator finds it necessary either to gloss
the words used in the analysis (Section 1.21.1), or to indicate the inflec-
tion of the complex being analyzed as it appears in the original text (Sec-
tion 1.21.2).
1.21.1 Glossing the words used in the analysis
Within the basic framework of a standard formula of analysis, commenta-
tors may insert glosses of the words making up that analysis at any point,
using the same techniques that are employed in glossing separate words
in the original text. For example, the word an epithet of
fire, is a straightforward genitive compound, meaning "friend of
the wind," and its analysis, using the standard formula (Section 1.53.1) is
therefore simply sakhii "friend of the wind; wind-
friend." In employing this standard formula, however, the commentator
Mallinatha inserts synonyms after the first member in the analysis and af-
ter the repetition of the entire compound;
Ragh.M. 2.10 marutab viiyob sakhii marut-sakho 'gnib. Friend of the
wind, of air; wind-friend, fire.
This technique is taken a step further in the practice of using syn-
onyms in place of the parts of the original complex when analyzing the
complex. For example, the analysis given by Mallinatha of the compound
("impelled by the wind") in the same verse is not maruta
but vayuna for the original members of the
compound, displayed in the analysis with the same inflection and position
that the original members would have in the standard formula of analysis.
1.21.2 Indicating the inflection of the original
The formulas for the analysis of derivative and compound nouns as offered
by the grammarians generally present the noun that is being analyzed in
the nominative case. Of course such nouns may appear in any case in a
text being commented on, and in order to make the transition from the
standard formula to the inflection found in the original text a commentator
may follow either of two procedures in analyzing the noun.
One procedure is to modify the basic formula by inflecting the syntac-
tically predominant part of the analysis to match the inflection of the noun
being commented on. For example, where the text has raja-putrat, "from
the king's son," the commentator may give the analysis putrii/,
based on the nominative-case formula but reflecting the
ablative case of the word being explained. Such an analysis is called a
"analysis (using) the case-ending of the fin-
ished (form)."
The other procedure is to use the standard formula without modifica-
tion, and then to add a pronoun to pick up the formation and convert it
to the required inflection. Using this method the analysis of raja-putrat
would be rajna/:! putras tasmii/, "the king's son, from him." Such an anal-
ysis is called a prathamdnta-vigraha, "analysis ending in the nominative
case-ending."
Where a pronoun already OCCurs at the end of a standard formula, as for
example in the analysis of bahuvrfhi compounds, the same two options are
still available. Here the pronoun already present at the end of the standard
formula is the syntactically predominant part of the analysis, as it refers to
the same thing that the formation as a whole refers to. Thus the analysis
of the compound dfrgha-kan:zat, "from Long-ear," may be either dfrghau
kan:zau yasya tasmilt (tath8ktat), "whose ears are long, from him (who
is so-called)," or, with the addition of a further pronoun to pick up the
analysis, dfrghau kan:zau yasya sa (tathOktas) tasmat, "whose ears are long,
he (IS so-called), from him."
. Just as parts of a formula of analysis may be separaied by interven-
mg glosses (Section 1.21.1), this "pick-up" pronoun added at the end of a
nominative-case analysis may be separated from the body of the analysis
by any amount of intervening explanatory material. For example, com-
pound marut-sakhah used as an example in Section 1.21.1 is actually the
42 Chapter 4. Analysis of Grammatical Complexes
first member of a longer compound, marut-sakhdbham ("shining like fire,"
masculine accusative singular), which Mallinatha interprets as an upapada
compound (Sections 1.32, 1.59). He therefore uses a pronoun to pick up
marut-sakhab for the rest of the analysis of the longer compound, and then
he must add another pronoun at the end to place the whole compound in the
accusative case. But before adding this last pronoun he pauses to explain
which krt suffix is used to form the last part of the compound:
Ragh.M. 2.10 marula sakhii marut-sakho 'gnifl. sa iWlbhiitfti
marut-sakhiibhafl. "iitos copasarge" iti ka-pratyayafl. tam. Friend of
the wind, of the air: wind-friend, fire. (He) shines as he does, thus
wind-friend-shiner. The suffix Ka (i.e., -a, without strengthening of the
root to which it is added) by (Pa. 3.1.136) "And (there is the suffix) ka
after (a root ending in) long -ii, when a verbal prefix is used." Him.
Note that insertions such as this do not alter the basic formula of analy-
sis: they simply interrupt it. The same problem arises when a commentator
offers more than one explanation of a part of the analysis, for he will carry
on with the basic structure of the formula no matter how many alternatives
he inserts in a part of the formula. For example, when the commentator
Hemadri analyzed this same compound marut-sakhdbham, he not only in-
serted a quotation from a dictionary, but also could not decide whethet
marut-sakhab means Agni or Indra, and whether the longer compound is
an upapada compound, as Mallinatha thought, or a madhyamapadalopa-
bahuvrfhi compound (Section 1.66.2):
Ragh.H.2.10 maruto vayofl sakhii iti marut-sakho 'gnis tasydbhewibhii
yasya. indrasyeti va, "pulomtirir marutsakhii" iti dhana1'J1jayafl. tadvad
abhiitfti vii. tam. Friend of the wind, of the air, thus fire.
Whose appearance is like the appearance of his [i.e., of Agni's]. Or [like
the appearance] of Indra's: "Pulomari, Marutsakha [among the epithets of
Indra]," says [the lexicographer] DhanaI)ljaya. Or, (he) shines as he does,
thus [wind-friend-shiner]. Him.
In reading the examples quoted from commentaries in the sections that
follow, take care to separate the features of expression required in order to
match the inflection of the word being explained from those features proper
to the formula of analysis being exemplified.
Chapter 5
Derivative Verbs
1.22 IntrOduction
Of the five modes of complex formation recognized in Pfu,tinian grammar,
four result m denved nominal stems. The fifth involves the addition of
suffixes to verbal roots and to nominal stems to create formations that are
treated as derived verbal roots. The first of these "root-fOrming suffixes"
taken up by is the desiderative suffix -sa, called san by (the fi-
nal -n serves only to indicate that the resulting formation will have the anute
pitch accent on its first syllable). The formations so derived are therefore
called san-ady-anta-dhiltu ("roots ending in san, etc.," Pa. 3.1.5-3.1:32;
corresponding to "derivative or secondary conjugation," Whitney 996 ff.).
Some of the suffixes are added to items already technically consid-
ered roots :-vithout producing any change in meaning that requires anal-
YStS m ordmary language. This is true, for example, of verbs that use
the causative suffix without clear causative meaning, including the items
VIewed by Pfu)ini as roots of the cur-adi or tenth-class and by Western
grammanans as denominatives with causative rather than denominative
accent (corayati, cintayati, etc., Pa. 3.1.25; Whitney 607, 775, 1041 b,
1056), and it is also true of the roots that use the desiderative suffix without
desiderative meaning (Pa. 3.1.6-3.1.7; Whitney 1040)-jugupsate,
tltlk:;ate, clkltsatl, mfma'!'sate, bfbhatsate, dfdamsate, famsate. But these
suffixes also produce four large groups of to analy-
SIS m ordmary language: desiderative verbs (Section 1.23), intensive' verbs
(SectIon 1.24), causative verbs (Section 1.25), and denominative verbs
43
44
(Section 1.26).
Chapter 5. Derivative Verbs
1.24 Intensives
1.24. Intensives
45
1.23 Desideratives
These are the sann-anta-dhtitus, "roots ending in (the suffix) san," Pa. 3.1.7;
"desiderative verbs," Whitney 1026-1040. A straightforward desiderative
formation is analyzed in ordinary language by a formula using the verb
(icchati) after the infinitive of the root in question:
Klisikli 3.1.7 kartum icchati He wishes to do.
The same principle applies to the analysis of desiderative formations in
which the addition of a further suffix has produced a participle, an adjective
in -u, or a noun in -ii; the same suffix will be added to the root in the
analysis:
Kum.M.3.14 fpsitam iiptum desired sought to be obtained
Ragh.M. 1.2 taritum icchuJ;. desirous of crossing seeking to
cross
Ragh.M. 1. 7 vijetum icchiinarn of those desirous of
conquering of those seeking to conquer
Kum.M. 2. 7 icchaya through the desire of
creating through the wish to create
Certain idiomatic uses of the desiderative require special analysis:
Kasikii 3.1. 7 sanke kulam {pipaWfati kulam] I fear the bank
is going to fall-the bank is likely to fall
Klisikli 3.1.7 [Janke iva] iva I fear the dog is
going to die-the dog is likely to die
These are the yan-anta-dhatus, "roots ending in [the'suffix] yan," Pa. 3.1.22-
3.1.24; "intensive or frequentative verbs," Whitney 1000-1025. In the ba-
sic formulas for the analysis of intensives and frequentatives the finite sim-
ple verb is used with the addition of punaf:z punah ("again and again") for
the frequentative meaning or bhrsam ("intensely") for the intensive mean-
ing:
Klisikli 3.1.22 pacati piipacyate he cooks repeatedly
Kasikii 3.1.22 bhrsa1!l jvalatijajva/yate it blazes intensely
As in the analysis of desideratives (Section 1.23), the same basic for-
mula applies even when further suffixation occurs:
Kum.M. 1.23 jarigamyante bhrsaJ?'l gacchantftijarigama
They move all around-they go
intensely-thus moving creatures-gods, animals, humans, and so forth.
Here a krt suffix used in the sense of the agent (Section 1.29) is added to
the intensive formation in order to form a nominal stem.
Again as in the analysis of desideratives, certain idiomatic uses of the
intensive require special analysis: '
Kiisikii 3.1.23 ku.tila1!l krtimati cankramyate he moves crookedly (i.e.,
he goes in zig-zag fashion).
KiiSikii 3.1.24 garhita,!, lumpati lolupyate he breaks it in a
reprehensible way (i.e., he busts it all to pieces).
1.25 Causatives
These are the r;ij-anta-dhatus, "roots ending in [the suffix] nic," Pa. 3.1.26;
"causative verbs," Whitney 1041-1052. Straightforward forma-
tions may theoretically be analyzed by using the present participle the
root in question, followed by some verb of instigation; thus gamayati, "he
46 Chapter 5. Derivative Verbs
1.26. Denominatives
47
causes him to go," may be analyzed gacchantal!' prerayati, "he sets him
going." In practice, however, such formations are only very rarely ana-
lyzed in ordinary language; instead a commentator will offer the causative
form of a synonymous verb as a gloss or provide an analysis in the techni-
cal terminology of grammar.
As in the analysis of ihe derivative verbs described in the previous two
sections, certain idiomatic uses of the causative require special analysis.
An example:
Kasikii 3.1.26 ka,!,sa-vadham ka'!'sa'!l ghiitayati He tells the
story of the slaying of Karrsa-he has KaJ!lS3 slain.
1.26 Denominatives
These are ihe nama-dhatus, "roots [formed from] nouns," Pa. 3.1.8 ff.; "de-
nominative verbs," Whitney 1053-1068. Denominative verbs are formed
by ihe addition of suffixes to substantives and adjectives. They fall into sev-
eral groups according to the precise meaning imparted by the suffix. For
each group ihe standard formula of analysis in ordinary language is based
directly upon the wording of ihe portions of the rules of describing
ihose meanings; ihe commentators follow ihese basic formulas of analysis
quite closely, but sometimes omit the elements referring to finer details of
meaning. In the following examples, the basic formulas are taken from ihe
Kasiki/, and ihe elements sometimes omitted by commentators have been
placed in parentheses.
1.26.1 Desire to have
(Pa. 3.1.8-3.1.9):
Kasikli. 3.1.8 (iitmanaM putram icchati pumyati he desires a son (for
himself)
Ragh.M. 1.35 putrakamyaya putrecchaya through desIre
for a son through the wish for a son for themselves
In ihe last example the denominative stem has been subjected to further
suffixation in order to produce a krt formation (cf. Section 1.23).
1.26.2 Behavior toward
(pa. 3.1.10):
Kiisikii 3.1.10 putram ivacarati putriyati he treats (him) like a son
1.26.3 Behavior like
(Pa.3.1.11):
KaSlk53.1.11 iyena ivficarati syenayati it acts like a falcon
Ragh.M. 2.61 amrtam iWicaratfty amrtiiyamiinami it acts like the
nectar of immortality, thus acting like the nectar of immortality
1.26.4 Becoming something
(Pa. 3.1.12-3.1.13):
Kasika 3.1.12 (abhrso) bhrso bhavati bhrsayate (not having been
intense) it becomes intense-it itensifies
Ragh.M. 2.17 fyiimiini bhavantfti syiimiiyamiiniini they become dark,
thus darkening
For ihe basic formula and the optional omission of its first part, cf. the
analySIS of CVt formatIOns (Section 1.42.4).
1.26.5 Making something so
(Pa.3.1.21):
Kiisikii 3.1.21 karoti mUTp!ayati he makes it shaven-he
shaves it
Ragh.M. 3.35 iaghu,!, by the' one being about
to lighten it-by the one being about to make it light
Ragh.M. 5.73 malinayanti malinani kurvanti they dirty
make them dirty
48 Chapter 5. Derivative Verbs
1.26.6 Feeling something
(Pa.3.1.18):
Kasikfi 3.1.18 sukha'!l vedayati sukhayate he feels happiness
1.26.7 Doing or making something
(Pa. 3.1.17, etc.):
Kiisikfi 3.1.17 sabdaf!l karoti sabdiiyate he makes sound-he sounds
Kiiiikfi 3.1.19 namaf:l karoti namasyati he does honor-he honors
Often a particular noun will call for a more specific verb:
Kiisikii 3.1.25 satyam satyiipayati he tells the truth
Kiisikii 3.1.15 tapas carati tapasyati he performs austerities
Such meanings shade off into the next category.
1.26.8 Making use of something
(Pa. 3.1.15, etc.):
In this category belong many specific analyses applying to particular words.
A few examples follow.
Kiisikii 3.1.15 romanthaTJl vartayati romanthiiyate it turns over the
cud-it chews the cud
KiiSikii 3.1.20 [puccha,!, paryasyati] paripucchayate it tosses its tail
around-it wags its tail
KiiSikii 3.1.25 slokair upastauty upailokayati he praises with
eulogies-he eulogizes
Kiisikii 3.1.25 lomany anulomayati he strokes the hairs in
the direction in which they lie-he curries
Klisikii 3.1.25 varma1}Q sQ'!'lnahyati saJ!lvarmayati he girds with
armor-be armors

Chapter 6
Primary Nominal Derivation
1.27 Introduction
The formations called krt or krd-anta ("ending in krt suffixes;' Pa. 3.1.91-
3.4.117; "primary derivatives," Whitney 1143-1201) are the nominal stems
(including participles and verbals) derived through the addition of suffixes
to verbal roots. In general terms the category of krt suffixes includes all
suffixes added directly to verbal roots with the exception of the san-adi
suffixes ("root-fonning suffixes," Sections 1.22-1.26) and the tin suffuo<es
(the inflectional endings of finite verbs).
Akin to the krt suffixes are the suffixes called ul}-adi ("beginning
Ul}"), which serve the same function of deriving nominal stems from verbal
roots, but often do so through forced and improbable etymologies. They
are preserved in collections outside the grammar of Pat)ini, and although
Pi1t)ini makes reference to this group of suffixes (Pa. 3.3.1-2, 3.4.75), his
general policy is to take the nouns derived by ul}-adi suffixes as ready-made
forms, so that he does not concern himself with their derivation. The com-
mentators mention them rather infrequently in comparison with the suffixes
taught by Pi1t)ini, and when they do refer to them their techniques of anal-
ysis are the same as those used in analyzing krt formations. For example,
the first of the ul}-adi suffixes, Ul} (the purpose of the final -'.I is to indicate
that the vowel of the root to which it is added is subject to vrddhi strength-
ening), is said to form the word kiiru from the root kr, and this formation is
analyzed karotfti Silpf, "he makes, thus a maker, an artisan," exactly
as in the analysis of kartari krt formations (Section 1.29.1).
49
50
Chapter 6. Primary Nominal Derivation 1.28. Basic bhave meanings
51
As in the analysis of the other types of complex formations, the for- 1-
mulas of analysis in ordinary language used by the commentators to ex- }
plain krt formations follow very closely the treatment of these formations
in Pa'!ini's grammar. Thus the formulas of analysis may be divided by
their being applicable to either of two major categories of meaning: that of
the verbal action itself (Section 1.28), and that of the various participants
in an action (Section 1.29). Beyond these basic varieties of meaning, the
precise meaning of a formation may be narrowed by restrictions of relative
time (Section 1.30) or by any of a number of other restrictions of meaning
(Section 1.31). Furthermore, many k,t formations are possible only when
compounded with an accompanying word (Section 1.32). The various ver-
bal participles form special subsets of k,t formations, and some of these
require special treatment (Section 1.33). Finally, the explanation of cer-
tain formations depends not upon the basic formulas of analysis but upon
the specification by the grammarians of particular meanings for particular
words (Section 1.34). "
1.28 Basic bhiive meanings
A great many k,t suffixes are used bhiive, i.e., in the sense of the verbal ac-
tion itself. Most of the resulting formations are nouns of action correspond-
ing in meaning to the English gerund in '-ing'. Since the only straightfor-
ward method of duplicating the meaning of such formations in ordinary
language in Sanskrit is through the use of another verbal noun formed with
a different suffix, the explanation of these items in commentaries is usually
achieved either by giving an analysis in technical terminology or by simply
offering a synonym:
Ragh.M.5.45 bhange bhafijane in breaking-in (the action of)
breaking
Ragh.M.5.61 iigamena iigamanena by tbe arrival-by the arriving
Kum. V. 1.59 sa11'lmiirjana11'l saTflmiirgaJ,l (the action of)
sweeping-sweeping
The offering of a synonym in the form of another verbal noun using a
different suffix is also employed when the word being glossed happens to
be a participle with simple bhiive meaning:
Ragh.M. 2.54 prahrta,!, tu prahtiras tu but the stroke-but the
striking.
When, however, the formation being explained is an indeclinable ver-
bal form, more complicated formulas are required.
Continuatives are generally analyzed by using a verbal noun derived
from the same root, followed by a continuative form of the root kr:
Ragh.M. 3.51 prahasya prahiisa11'l k,tvii laughing performing the
action of laughing.
When this technique is used, it is often the continuative made from
kr, rather than the verbal noun, that indicates the presence of secondary
meanings such as that of the causative:
Ragh.M.5.62 praveSya prave.sa'!l ktirayitvti showing in causing him
to perform the action of entering
The rarer l}amul form of continuative is usually glossed simply by of-
fering the commoner form of the continuative as a synonym:
Kum.M. 4.26 stanau sambiidhya striking the
breasts-striking the breasts
Infinitives may often be explained through the use of a verbal nouh in
the dative case, but even where this is possible other more roundabout ex-
pressions are also allowed, which generally use verbal formations referring
to the agent rather than to the verbal act itself:
KiiSikii 3.3.11 {paktuTfl vrajati] piikiiya vrajati [he goes in order to
cook] he goes for cooking
Kiisikii 3.3.11 bhoktuTfl vrajati bhojako vrajati he goes in order to
eat-he goes as an eater
Kasikii 3.3.12 [asvan diitu'!' vrajati] Qsva-diiyo vrajatt [he goes in
order to give horses] he goes as one intending to give horses
KiiSikii 3.3.13 {kartuTfl vrajati] vrajati [he goes in order
to do] he goes (with the thought), "I shall do"
52 Chapter 6. Primary Nominal Derivation
1.29. Basie karaka meanings
53
1.29 Basic kiiraka meanings
The krt suffixes that do not refer to the verbal action itself refer to a par-
ticipant in that action. In the PaJ)inian system there are six categories of
such participation, called ktirakas. With few exceptions (Sections 1.30.1,
1.31, 1.32.4, 1.32.5) formations made with such suffixes are analyzed us-
ing formulas containing finite forms of the verb in question. When the
suffix is used in the sense of the agent (Section 1.29.1) or the object (Sec-
tion 1.29.2) of the action, an active or passive finite verb is sufficient to
indicate the thing to which the formation refers, since by PaJ)ini's rules the
verbal termination of an active verb is itself used kartari (in the sense of
the agent), and that of a passive verb is used karma,:,i (in the sense of the
object). The thing teferred to by the formation will then be the grammati-
cal subject of the finite verb. But where the suffix is prescribed by PilJ;1ini
in the sense of some other ktiraka (Sections 1.29.3-1.29.6), the category
of participation through which the thing referred to is connected with the
verbal action must be indicated by using some form of the pronoun idam,
exhibited in the grammatical case appropriate to the ktiraka in question.
None of these formulas, unlike those used in the analysis of deriva-
tive verbs and of nominal compounds, can be substituted directly for the
word that it analyzes, because in each of these formulas the syntactically
predominate element is a verb, while the word being analyzed is a noun.
Instead the formula in itself merely offers a clear explanation of the manner
in which the thing referred to particpates in the verbal action. Accordingly,
each of these formulas must be concluded with the particle iti followed by
the word being explained. For example, in order to explain that the word
ptieaka refers to the agent in the action of cooking, the basic formula is
simply paean', "he cooks"; but the full analysis is paeatiti "he
cooks, thus [he is] a cooker." Much less frequently a commentator will
avoid the need for the particle iti by using a relative pronoun: paeati sa
ptieakah, "he who cooks is a cooker." The same technique is occasionally
used in analyzing formations involving any of the ktirakas; compare the
following analyses:
Kum.C. 7.53 apidhfyate 'nenety apidhiina'!' kapii!am (something) is
covered by this, thus a covering, a leaf of a door
Kuru. V 7.53 apidhfyate sthagyate yena tad apidhiinam that
by which (something) is covered--eoncealed-is a covering, a leaf of a
door
Notice also that, as in the analysis of other types of complex forma-
tions, the commentator may insert glosses at any point in the analysis.
1.29.1 kartari
Formations made with krt suffixes used in the sense of the agent of an
action are analyzed with finite active forms of the verb in question. Where
the word being analyzed is plural, a plural form of the finite verb will be
used. The same basic formula applies whether the word being analyzed is
asubstantive or an adjective.
Ragh.M.4.12 tapatftitapanaJ;t it heats, thus heater, the sun
Ragh.M. 4.42 yudhyanta iti yodhii!} they fight, thus fighters
Ragh.M. 2.72 sukhayatiti it pleases, thus pleasant
When such formations contain a verbal prefix, two methods of analysis
are available. The prefix may be expanded, so to speak, into a separate
adverbial form:
Ragh.M. 12.68 safran iti he especially
frightens his enemies, thus terrifying
Kum. v: 1.15 niyata1J1jharati sravatfti nirjharalJ it
fiows-streams---continuously, thus a cascade
Or the prefix may simply be retained with the finite verb used in the
analysis:
Ragh.M. 1.38 utkiranti utkirii!} they scatter
upwards-toss upwards, thus scattering up
Kum. v: 5.44 upayacchate svfkarotfty upayantii pati he
takes-accepts as his own, thus taker, bridegroom
1.30. Temporal restrictions
55
1.29.2 karmafli
Fonnations made with krt suffixes used in the sense of the object of an
action are analyzed in much the same way as kartari fonnations, except
that the finite verb used in the analysis will be passive rather that active.
Ragh.M. 2.65 ktimyata iti kamo varafz it is desired, thus a desire a
choice
Ragh.M. 1.20 priirabhyanta itf priirambhiil} they are undertaken, thus
undertakings
Ragh.M. 2.4 prasuyata iti prasutil} santatiJ:t it is procreated, thus
procreation offspring
1.29.3 karafle
Fonnations made with krt suffixes used in the sense of an instrument in-
volved in an action are analyzed with a finite verb, usually passive, and a
fonn of the pronoun idam exhibited in the instrumental case.
Ragh.M.5.57 sa7[lmohyate 'neneri sa,!,mohanam (one) is infatuated
by means of this, thus infatuating
Kum.M.2.4 vibhajyate 'neneti vibhiigal} (something) is divided by
this, thus a division
Kum.M. 3.40 vihanyanta ebhir iti vighniil} pratyiihah (people) are
obstructed by these, thus obstructions obstacles
1.29.4 sampradiine
Fonnations made with krt suffixes used in the sense of the one for whose
sake an action are analyzed with a finite verb and a fonn of the pronoun
idam exhibited in the dative case.
Kum. V. 6.87 [6.88] arpyate 'sma ity (something) is offered!
to this, thus recipient i
1.29.5 apiidiine
Formations made with krt suffixes used in the a point of departure
involved in an action are analyzed with a finite verb and a fonn of the
pronoun idam exhibited in the ablative case.
Ragh.M. 1.2 prabhavaty asmiid iti prabhaval} kara1)am (something)
arises from this, thus a source a cause
1.29.6 adhikarw}e
Formations made with krt suffixes used in the sense of a location involved
in an action are analyzed with a finite verb and a form of the pronoun idam
exhibited in the locative case.
Kum. V 6.43 sa'!lcaranty asminn iti sa'!'caro margafi. (people) travel
on this, thus a road a street
Ragh.M. 5.40 dhfyate 'syam iti dhiinf (something) is put in it, thus a
receptacle
Kum.M. 1.7 likhyate yesu te lekhiil} patrikiih those on which writil1g
is done are writing sheets leaves
Notice that in the last example the commentator has used the alteruative
approach described in the introduction to Section 1.29.
1.30 Temporal restrictions
While most k.rt suffixes are used with present or temporally indefinite force,
a great many suffixes are assigned specific temporal meaning, and this
specification will be reflected in the analysis of the resuliing fonnations.
Most common is the specification of past time (Section 1.30.1), but one
also often finds present time assigned to participles that usually refer to
past time (Section 1.30.2), and some suffixes have future meaning (Sec-
tion 1.30.3).
56
Chapter 6. Primary Nominal Derivation
1.31. Other restrictions ofmeaning 57
1.30.1 bhiite
In addition to the various past participles, many other words are formed
with krt suffixes expressly assigned past meaning by PaI;lini; the specifi-
cation' bhute ("in the sense of what is past") applies throughout a large
segment ofPaI;lini's rules for krt suffixes (3.2.84-3.2.122) and appears spo-
radicallv elsewhere as well. Formations to which this specification applies
are glossed through the use of past participles (either active
or passive) rather being analyzed through the use of of finite verbs. In such
glosses the particle iti is not necessary, since the participle, by virtue of its .l
own nominal character, can be substituted directly for the noun bemg ex-
plained. For the same reason it is possible to indicate the case of the word
being glossed by the case of the participle used, so that the formatIOn need
not be exhibited in the nominative case, as is necessary with formulas of
analysis using finite verbs and iti.
Ragh.M. 1.44 yajvaniilfl Of those who have
sacrificed of those who have offered sacrifice in accordance with the rules
Kum.M.2.46 yajvahhi/] vidhivail By those who have
sacrificed by those who have offered sacrifice in accordance with the rules
Here the past active participle is used. For further examples and for
examples of past passive participles in the explanation of krt formations,
see Section 1.32.4.
1.30.2 vartamiine
Specification of present meaning is necessary only for formations that by ;
more general rules would otherwise have past or future meaning. By far the ,I
most common instance is the assignment of present meaning to the "past ,
passive participle." When this occurs the commentator wil1 either use the :
usual formula for the analysis of a krt suffix used karmalJi (Section 1.29.2). 'f
or will gloss the item with a .
Ragh.M. 5.2 srilyata iti sruta,!, vedasiistram it is heard, thus irota,
the Vedic scripture
Kum.M. 1.58 arcitam ... pujyamanam worshiped ... being
worshiped
1.30.3
Formations made with krt suffixes used with future sense, like those to
which present sense is expressly assigned, may either be analyzed with a
finite future verb and iti, or glossed with a future participle.
Kum V. 1.49 [1.50] bhavisyatfti bhavitrf she will be, thus future
Ragh.M. 5.32 samprasthital} san going forth being
about to go forth
1.31 Other restrictions of meaning
In addition to the temporal specifications just described, there are other
conditions of meaning (upadhi) that are applied to various krt suffixes by
paI;lini. Often the condition amounts to a restriction on which suffixes may
be used with a particular root in particular meanings. For example, the
suffixes mentioned in Pi!. 3.1.145-147 carry the restriction iilpini, "when
an artisan is being referred to," and as a result words such as nartaka,
"dancer," and gathaka or gayana, "singer," should be used only when
speaking of professional dancers and singers, and not when speaking of
just anyone who happens to dance or sing. Such restrictions of meaning
are not usually reflected in the analyses offered by commentators; although
they may often refer to the rule that lays down such a restriction, they will
generally be content with the usual formula of analysis-gayatfti
etc.
Some more general conditions of meaning, however, are regularly in-
corporated in the fonnulas of analysis used by commentators. By far the
most frequently encountered of these is the specification tac-chf/a (i.e., tat-
if/a). "having that (action) as one's nature or habit, characterized by that
(actiVity)," a condition that applies throughout a lengthy sequence of rules
in PaI;lini (3.2.134-3.2.179) and in some other rules as well. For example,
58
1.32. Accompanying words (upapada) 59
the word satya-vildi, "truth-speaking" (an upapada compound; see Sec-
tion 1.32), may be used accurately in reference to a man who habitually Or
characteristically speaks the truth, even if he does not happen to be speak-
ing at the moment, but it may not be used accurately in referring to an
habitual liar, even if he happens to be speaking the truth at the moment.
In explaining formations involving this condition of meaning, some
commentators simply supply Par,tini's specification tac-chfla after a finite
verb giving the general meaning of the fcOllation:
Ragh.H. 2.3 dayiilui} dayate tac-chila!) compassionate-he has
compassion, one who is characterized by that
Much more frequently, commentators follow the model of Par,tini's
specification by glossing the formation with a word in which -Sfla is placed
in compound with a verbal noun denoting the action in question.
Ragh.M 4.85 victorious characterized by victory
Ragh.M. 2. 72 ... sahana-sflaf:l forbearing ...
characterized by forbearance
Kum.M.l.23 sthiti-if[aJ}sthiivarii!J characterized by
standing still-stationary things, mountains, trees, etc.
Slightly different formulas may be used when the krt formation being
analyzed occurs at the end of a compound (Section 1.32.5).
1.32 Accompanying words (npapada)
Many lqt formations occur only at the end of nominal compounds when
expressing a given sense, and thus cannot be used in that sense without
a subordinate word (upapada) accompanying them as the first member of
the compound. To take as an example the formation that has proVided"
the name for the entire class of primary compounds, the word krt itself,
formed from the root kr, "to do, to make," occurs at the end of compounds .
in the sense of "who has done or made (something)"--e.g., mantra-krt,
"one who has made mantras." The word krt is not available in this mean-
ing outside such compounds, although it does appear as an uncompounded
word in another meaning (viz., as the technical term for suffixes of pri-
mary derivation). It is therefore not possible to ana}yze the compound as
a straightforward compound using the same words that make up
the compound, since one cannot say *mantrlil}iiJTl krt, as one could say, for
example, mantrilnil,!, in analyzing the genitive compound
mantra-sraHil. Instead the analysis of an upapada compound is usually
achieved by analyzing the krt formation at the end of the compound with
the standard formulas used for uncompounded krt formations and exhibit-
ing the rrpapada separately in the case that indicates its relationship to the
action expressed by the verbal form in that formula, rather than its rela-
tionship to a noun that expresses the sense of the final member of the com-
pound:
Ragb.M. 1.61 mantriin krtaviin mantra-krt One who has made
mantras-a mantra-maker
When, however, a compound ends with a lqt formation for which .the
rules of Par,tini provide a derivation as an independent word and do not
restrict it to use in compound with an upapada, we have a regular
compound, and a commentator may thus analyze the krt formation in the
usual way outside the compound, and then analyze the compound itself as
a straightforward
Ragh.M. 2.3 dharantfti dhara!). ... payasa,!, dhara!) paya-dhariii}
stanaf! they bear (something), thus bearers. ... bearers of
milk-milk-bearers, breasts
Occasionally it may not be clear which interpretation of a given com-
pound is more appropriate:
Kum.M. 1.20 jiiniititijiiaf!.... vedaniiniilJIjiiaJ.l [vedanii-jfia?z). athavii
vedaniif!l janiitfti vedanii-jiiaJ.l. He knows, thus a knower. . ,. a knower
of pain [3 pain-knower], Or: he knows pain, thus a
For upapada compounds in which the krt suffix is used bMve, i.e., in
the sense of the verbal action, the only way of analyzing the compound
in non-technical language is to substitute a freestanding synonym for the
1.32. Accompanying words (upapada)
61
krt formation and proceed with the standard analysis for a com- j
pound (usually a genitive with the first member displayed in
the objective genitive case in the analysis), just as the substitution of a
synonym is the usual way of explaining uncompounded bhiive krt forma-
tions (Section 1.28). For example, Pii 3.1.108 prescribes the formation ?
of the noun hatyti, "the action of killing," from the verbal root han, pro-
vided that the form is accompanied in compound by an upapada. The
compound brahma-hatyti means "the killing of a brahmin," and is analyzed'
as a straightforward genitive compound-brahmallo hananam-
with the substitution ofthe verbal noun hananam for hatyti. Because hatya
cannot occur without an upapada, the analysis *brahmal}o hatyii is impos-':;
sible.!
Most upapada compounds, however, do not involve bhiive krt forma-
tions as their final member, and are therefore analyzed with formulas us-
ing finite verbs. The most common situation is a kartari krt formation as ,j
the final member (i.e., a formation referring to the agent, Section 1.29.1),
with an upapada appearing in the accusative case in the analysis (Sec-
tion 1.32.1), but the upapada may also appear in some other case (Sec-
tion 1.32.2), and many other compounds have a krt formation in the sense
of some other karaka as the final member (Section 1.32.3). Krt forma-
tions that appear at the end of upapada compounds, like those functioning
as freestanding words, may involve temporal restrictions (Section 1.32.4)
or other conditions of meaning (Section 1.32.5), which will sometimes re-
quire formulas of analysis that differ slightly from those used in analyzing'
uncompounded krt formations.
1.32.1 kartari krt with accusative
Most upapada compounds have as their final member a krt formation re-
ferring to the agent of the action expressed by the verbal root to which the
krt suffix is added, and as their first member a word referring to the direct
object of that action. These compounds are analyzed by displaying the first
member in the accusative case, and using the standard kartari krt formula
(Section 1.29.1) in analyzing the final member.
Ragh.M. 2.71 hutam a.1niitfti hutfSo 'gniJ:l it eats the oblation, thus the
oblationeater, fire
Ragh.M. 1.85 mahfrJ1. kJiyate iti mahf-ksit. tasya mahi.ksitah He
earth, thus earth-ruler. Of hiJIl, of the
Kum. v: 5.22 mahlintam arhatfti mahfrha bahudhana it
deserves a great (price)-a high (price), thus costly, worth much money
As in the analysis of uncompounded forms, relative pronouns may be
used to avoid the use of iti. Compare the following two analyses:
Kum.M. 2.51 sena'!1 nayariti senti-nfS camii-pati(l ... ta'!l senii-nyam.
He leads an army. thus anny-Ieader, general ... [we wish to create] him,
[we wish to create] an army-leader.
Kum. V 2.51 sena'!! nayari yas ta'!! senii-nyam who leads an army,
him-an army leader
1.32.2 kartari krt with non-accusative
When the first member of the upapada compound does not refer to the
direct object of the verbal action involved in the second member, it will
appear in the analysis in some Case other than the accusative.
Kum.M. 1.50 kamena svecchaya caratiti kiima-caral} he roams at'
will-at his own desire-thus roaming-at-will
Kum.M. 3.37 palikaj jayata iti parika-jam it is born from the mud,
thus mud-born
RaghM. 4.72 guhasu serata iti guhii-sayiil} they lie in caves, thus
cave-lying
This class of compounds includes many in which the first member re-
tains its case ending within the compound, a phenomenon that does not
affect the basic formula of analysis:
Ragh.M. 5.72 stambe ramanta iti stambe-ramii/;l they
delight in a clump of grass, thus stambe-ramas, elephants
Kum.M. 1.10 vane carantiti vane-carii/;l they roam in the
forest, thus forest-roamers, mountain tribesmen
62
1.32. Accompanying words (upapada) 63
1.32.3 krt in the sense of another kiiraka
When the krt formation appearing as the final member of an upapada com-
pound is p;ssive, the standard formula for the analysis of a karma'}i k,t for- 1
mation (Section 1.29.2) is used, and the upapada isexhlblled III the case ,
needed to express its connection with the verbal aClIon. In the followlllg
example the upapada refers to the instrument, and thus appears III the Ill-
strumental case.
Kum. V 1.10 tailena puryanta iti taila-piiral). they are filled with oil,
thus oil-filled
The krt formation may have the sense of any of the other karakas as
well, and'for each the standard formula of analysis described for uncom-
pounded k,t formations (Section 1.29) will be used, with the upapadaap- .
pearin" in whatever case is appropriate to connect it wllh the verbal aclIon.
In the Ofallowing example, the krt suffix forming the final member of the
compound is used kara'}e, in the sense of the instrument (Section 1.29.3),
and the upapada refers to the object of the aclIon. Slllce a passIve verb IS
used in the analysis, the upapada shows up in the nominative case.
Ragh.M.3.10 pumiin suyate 'nenitipuTflsavanam a male is produced
by means of this, thus a maleproducing (rite)
1.32.4 Temporal restrictions
As in uncompounded krt formations, restrictions of the time referred to
may apply, the most common instance being restriclIonto past lIme
tion 1.30.1). These are ordinarily glossed wllh a past acave partIcIple, WIth
which the particle iti is not strictly necessary, and may be either included
or omitted.
Ragh.M.3.62 vrtra1Jl hataviin iti vrtro-hii he has slain Vftra, thus
V
Kum. V 6.3 smaraYJ1sasitaviin smara-iisanal;t who has chastised
Smara-the Smara-chastiser
Kum. v: 1.19 chinnavlin it; he has clipped the
wings, thus wing-clipper
Ragh.M. 1.23 ptiram anta1J1 limit-seers
who have seen the limit, the extremity
The past passive participle may appear instead, when the verb is pas-
sive or has a "past passive participle" with active meaning. As in other
upapada compounds, in the analysis of compounds involving intransitive
krt formations the upapada may appear in some case other than the ac-
cusative.
Ragh.M. 5.69 sarasi jiitaf;. sarasi-jalJ kama/af:!. born in the
lotus
Ragh.M.1.31 magadha-va'!lsejatti magadha-va'!lia-jii born in the
Magadha
Ragh.M. 1.85 jiitii,!, produced by
bathing in the sacred ford-sacred-ford-bathing-produced
1.32.5 Other conditions of meaning
As with uncompounded k,t formations, the most common condition, of
meaning applied to k,t formations in upapada compounds is that of tiicchllya,
"having that as one's characteristic or habitual action" (Section 1.31). H'!w-
ever, if the compounds ending in -fla that are used to gloss uncompounded
krt formations are to used in glossing upapada compounds then one must
include the upapada as the first member of the compound gloss, or, if the
upapada has been exhibited separately in order to make a remark, one must
either include the upapada in the compound gloss or use tat- as a substitute
for it there:
Ragh.M. 1.3 kavi-yaialJ-prarthf. kavrniiJ!l yasa/:t kiivyaninnti!lena
jiita'J'l tat-prtirthanti-sflaf;.. Seeking the fame of poets: the fame of
poets, produced by the making of poetry-characterized by the seeking of
it.
More commonly, such upapada compounds are analyzed by using the
infinitive of the verb in question, with the upapada as a separate object in
the accusative, followed by iflam asya, "whose nature it is to ... ," and iti.
i.34. irregular forms (nipatana)
65
Kasikfi 3.2.83 darsanfyam iitmiinaY!l manyate ... da,.sanfya-miinf he
considers himself good-looking ... (thus,) considering himself
good-looking
Ragh.M. 1.43 sva-niJ:t.svasam anukartLIf!l srlam asyeti
whose nature it is to imitate their own breath,
thus imitating their own breath
Ragh.M. 1.49 samidhas ca kustif!ls co phallini cfhartuJ!1 sf/am iti
sarnit-kusa-phalfhartiJ:t., toif! samit-kusa-phaliharai/;r whose nature it is
to bring fuel, kusa grass, and fruits-bringers of fuel. kusa grass, and
fruits; by them, by bringers of fuel, kusa grass, and fruit
Other conditions of meaning are analyzed with formulas dependent
upon the specification of meaning provided in the rules of PaI)ini. For
example, the suffix -in that is prescribed with the condition tticchllye for
compounds such as svani!rsvtisanuktirin in the example given above (by
pa. 3.2.78) is prescribed in the sense of the agent with the condition kartary
upamtine, "when (the upapada) is (another) agent serving as a standard of
comparison" for some other compounds (by Pa. 3.2.79). Such compounds
will be glossed by exhibiting the upapada in the nominative case, followed
by iva and the standard formula for the analysis of a kartari krt formation
(Section 1.29.1):
Ragh.M 4.4 dvirada iva . .. gacchatfti dvirada-giimi he walks ... like
an elephant, thus walking like an elephant
By a subsequent rule (Pa. 3.2.83), the same suffix can be added to the
root man, "to consider," in the sense of an agent when accompanied by an
upapada, with the condition of meaning titma-mtine (considering oneself
to be so). Accordingly, the formula of analysis for such compounds adds
the object titmtinam:
.,
Most participles are simply glossed with participles of the same form
from a root of meaning.
Ragh.M. 12.74 ia1!Jsatr bruvQ':lQ indicating telling
Ragh.M. 11.10 uhyamiina iva vahanena priipyamiilJa iva as if being
carried along as if being transported by a vehicle
The same procedure is followed in general for a past passive participle,
but other synonyms of different form are chosen if the participle is used
with active meaning, with present meaning, or in the sense of the verbal
action itself (bhtive).
Ragh.M. 3.5 iidrta iidrtaviin with respect having felt respect
Ragh.M. 12.75 Iajjitii lajjtivatf embarrassed with embarrassment
Kum.M. 4.17 upagiifjhiini iilinganiini embraces embracings
Gerundives are analyzed with several formulas designed to bring out
the speCIfic meaning of the participle in the context, most often by using
other gerundives together with infinitive forms, but sometimes with Com-
pounds beginning with verbal nouns.
Ragh.M. 1. 71 asahyii sotlhum asakya unbearahle unable to be bobe
Ragh.M. 1.86 ytiyayitu,!, yogya'!'ytijyam worthy of sponsoring a
sacrifice-fit for sacrificing
Ragh.M. 4.6 stutyam stotrfirham praiseworthy deserving praise
Ragh.M. 1.3 upahasyatiim being laughable
being an object of laughter
1.33 Verbal participles
:11
The category of krt formations includes verbal participles, which are treated :I
in several ways by commentators.
1.34 Irregular forms (nipiitana)
Formations that diverge from the general rules of the class to which they
belong, either lD thelf meaning or in their form, are often singled out by
the grammanans for an explicit individual specification of the formation
66
as irregular (nipiitana). Such words are said to be nipiitita, "set down."
Commentators frequently mention the rules specifying these forms, and
explain them either with simple synonyms or with a formula of analysis
based on the meaning of the form rather than on the general formula for the
class to which the form belongs. The following are examples of irregular
gerundives:
Kum. v: 3.18 bhavya eva bhavitaiva. ... bhavya-sabdaJ:! kanari
nipatitaJ:!. is surely to be will certainly be. '" The word bhavya is set
down as an irregular form (by Pa9-ini 3.4.68) in the sense of the agent
(rather than in the sense of the object, as one would expect).
Ragh.M. 11.8 bhinatti kii/am iti "bhidy6ddyau nade"
3.1.115) iti kyab-antau niptititau. It bursts its bank, thus a rushing
river. By 3.].115 (the words bhidya and uddya) are set down as
irregular gerundive forms (with active meaning, rather than passive
meaning as one would expect).
Kum.M. 3.57 a-garhyam. iti nipata(z.
uncensurable not deserving of censure. By Pa:r:tini 3.1.101 (the word
avadya) is set down as an irregular form (in the meaning "censurable,"
although the regular form in other senses of the word is anudya).
Chapter 7
Secondary Nominal Derivation
1.35 Introduction
The formations called taddhita or taddhitanta ("ending in taddhita suf-
fixes," Pa. 4.1.76-5.4.160; "secondary derivatives," Whitney 1202-1245)
are the nominal stems derived through the addition of suffixes to other nom-
inal stems. In general terms the category of taddhita suffixes includes all
suffixes added directly to nominal stems with the exception of the suffixes
used to form denominative verbs (Section 1.26), the strf suffixes (which
convert nominal stems to the feminine gender) and the sup suffixes (the
inflectional endings of nouns).
Most of the formulas used to analyze taddhita formations are taken
directly from specifications of the meanings for each suffix. In
rules the nominal stem to which a suffix may be added is generally
represented by the pronoun tat or an equivalent, and the analysis of any
particular formation simply substitutes the noun that serves as the base
in that formation for the form of tat prescribed by PiiI)ini. To take as an
example the me"lning that has given the taddhita class of suffixes its name,
Pa. 5.1.5 states "tasmai hitam," "good for that, beneficial to that," and any
formation derived by the addition of a suffix added to a base to produce
a word with this meaning can be analyzed by exhibiting the base in the
dative case, followed by the word "hita":
KiiSika 5.1.5 vatsebhyo hito (go-dhuk) beneficial to
calves-(a dairyman) good to calves
67
68 Chapter 7. Secondary Nominal Derivation
Kasik. 5.1.10 sarvasmai hilam siirvam beneficial to all-good for all
1.36. Dependentformulas ofanalysis
(instrumental)
69
Formulas of this type may be called dependent formulas of analysis
(Section 1.36). Other formulas that follow the specifications of Plil)ini in
precisely the same way are those in which the base to which the suffix is
added is exhibited in the genitive case (Section 1.37), in the nominative
case (Section 1.38), or as the first member of a nominal compound (Sec-
tion 1.39). Important subdivisions of taddhita suffixes are the possessive
suffixes (matvarthfya, Section lAO), the suffixes that have the force of par-
ticular case endings (vibhaktisaf!ljfiaka, Section 1.41), and the suffixes that
produce formations referring to the same object as the base to which the
suffix is added (sviirthika, Section 1.42).
As in other categories of grammatical complexes (cf. Sections 1.20,
1.34), some particular taddhita formations are individually specified by
Plil)ini as exceptional forms (nipiita), rather than being derived through
more general rules. Such specification is called nipiitana by the com-
mentators, and the words so specified are said to be nipiitita. Typically
they involve either unusual forms or, more often, very specific meanings,
that could not be known from the general rules for the suffixes with which
they are formed. For example, the word va0in, although made by adding
the possessive suffix -in (called in! by Plil)ini) to the noun va0a, has the
very specific meaning of a member of the first va0a or social class, i.e., a
brahmaciirin or student. Ihis meaning is specified in Pli. 5.2.134, "After
va0a in the sense of a brahmaciirin," in the section of his rules dealing
with this suffix.
Ragh.M. 5.19 van}f brahmaca.rf. "van:zino ity amaraJ:z.
"varniid brahmactiriIJi" itfnipratyayaJ:z. A van.rin a brahmaciin'n.
Amara says "van:zins are brahmaca.rins." The suffix inl by (Pa. 5.2.134)
"After van:za in the sense of a brahmacarin."
1.36 Dependent formulas of analysis
In these the suffix is treated as though it were the second member of a
dependent compound. Thus in the analysis the word to which
the suffix is added is placed in whatever case is required by its relation to
the word expressing the meaning of the suffix:
kulii/ena krtam-kauliilakam "made by a pattir"
dadnti sQ,!/skrram--dadhikam "prepared with yogurt"
nilvii ttiryam-niivyam "crossable by boat"
(dative)
iitmane hita!z-iitmanfna!;t "good for oneself'
gave hitam-gavyam "good for a cow"
(ablative)
nyiiyiid anapetam-nyiiyyam "not deviating from the law"
manor "born from Manu, human"
"slightly less than learned, Le.,
fairly learned" (cf. Section 1.42.3), Section 2.19.3
(genitive)
fifor "the state of a child, childhood"
P!thor bhiival:z-piirthavam "the state of being wide, wideness"
modification of wood, wooden
gaviif!l collection of cows, herd of cattle"
miitul:z bhriitii-miitural} "mother's brother, materal uncle"
riijfiiif!l yogyaf!l-riijanyakam "appropriate for kings"
(note the use of ayam in the following genitive examples:)
devasyayam-daiva!J "this is a god's, belonging to a god"
janasyayam-janakfyal} "this is the people's, belonging to the people"
70
Chapter 7. Secondary Nominal Derivation 1.36. Dependent formulas of analysis 71
(locative)
sadhu(i.-iititheyal} "good toward guests, hospitable"
sabhtiyiifJl sadhuJ:i.-sabhyaIJ "good in the court, polite"
"relating to a camel"
"produced in summer"
jiital:t-kauravakal) "born among the Kurus"
1.36.1 Accusative
When the word to which the suffix is added is displayed in the accusative
case in the analysis, the formula used is like that employed in analyzing
upapada compounds (Sections 1.32, 1.59):
vyiikarartam adhfta iti vaiyiikara1JalJ "he studies grammar, thus
grammarian"
dharmalfl caratiti dharmikal} "he practises righteousness, thus
righteous"
And this formula may be used with the first element in other cases as well:
urjupena taratity "he crosses by boat, thus boatman"
Examples from the commentaries:
Ragh.M. 1.25 arhantfti yat" iti
ya-pratyayaJ:i. .... tlin da'.'r!yiin They deserve punishment, thus
punishable. The suffix -ya by 5.1.36) "(The suffix) yaT after
da!/-rja, etc." .... Those criminals.
1.36.2 Instrumental
Where the other oblique cases are involved, the formulas are akin to those
used in analyzing case-tatpuruJas.
Kum.M. 3.64 pataftgena tulya1'(l patangavat salahhavat . .. "tena
tulyarJl kriya ced vatiJ:l " iti vati-pratyayaJ:i.. equivalent to a moth
moth-like like a moth ... The suffix vatl by 5.1.115) "There is
(the suffix) vatl in the meaning of 'like this,' if an action is referred to."
Ragh.M.4.31 navya naubhis tarya navigable erossable by
boats rivers
1.36.3 Dative
Kir.M. 13.69 atmane hiram iitmaninam.
"atma-visva-jana-bhogottarapadiit khaf:i.." beneficial to oneself good
for oneself. By (Pa. 5.1.9) "(The suffix) kha (i.e., -rna) after atman, viiva,
jana, and a word ending in bhoga."
Sisu.M.1.41 vi.vasmaijaniiya hitarrz visvajanfnam.
"litma-visva-jana-bhogottarapadiit khal.z." beneficial for all people
good for all people. By (Pa. 5.1.9) "(The suffix) kha (i.e., -rna) after
litman, visva, jana, and a word ending in bhoga."
Sisu.M. 14.4 he slina sarvahitatvat slirvaf:i., tatsambodhane.
0 (you who are) good for all
good-far-all because of being beneficial for all; (the vocative) in
addressing that. (The suffix is by Pa. 5.1.10) "After sanJa and puru:5a there
are (the suffixes) f:/a and ifhaN (i.e., -a and -eya, respectively)."
1.36.4 Ablative
Kum.M. 1.26 abhijanlid iigatena iibhijanena come from a good
family of good family
Kum. V. 5.29 1M30J iigatam brahmamayam, tena
[brahmamayenaJ vedotthena tejasii coming from brahma brahmanic,
with that [brahmaniel-arising from the Vedas-splendor
72
Chapter 7. Secondary Nominal Derivation 1.36. Dependentformulas of analysis
73
Ragh.M. 4.4 pitur ligatQ/?t pitryam. "pitur yat" iti yat-pratyayaJ:z..
si,!"hiisanam coming from the father paternal-the suffix yaT (i.e.,
-ya) by (Pa. 4.3.79) "yaT after pitr (in the sense of coming from
than-throne.
Ragh.M.5.9 janapadebhya agatairjiinapadailJ. "tata ligata" ity
ka{ialigarfyai!J with country from the country-the suffix al! by
(Pa. 4.3.74) "come from that"-straw
Ragh.M. 1.60 daivfnii,!, devebhya agattintiY!l
agatiinti'!l caura-bhayiidfniim. ubhayatrapi
"tata iigataJ:t" ity Q?!. "!i-hQ1}-an- "ity-adinii nfp. tipadiim of
disasters (both) divine coming from the gods, such as famine, and
human coming from humans, such as the danger of theft. In both there is
the suffix at:lby (Pa. 4.3.74) "come from that." There is (the feminine
suffix) NIP (i.e., -0 by (Pa. 4.1.15).
Kum. V. 2.3 arthad anapetli acalitli arthyii anadhyaropitfinhaly,.
"dharma-pathy-artha-nyayad anapete." not deviating-not
moving-from the facts factual without hyperbole. (The suffix -ya is
added by Pa. 4.4.92) "After dharma, pathin, artha, and nyiiya in the sense
of not deviating."
Kum.M.2.2 arthad artha-yuktiibhir ity
arthaly,. "dharma-pathy-anha-nyayad anapete" iti yat-pratyayaly,. not
deviating from meaning meaningful provided with meaning. (The suffix
-ya is added by Pa. 4.4.92) "After dharma, pathin, artha, and nyiiya in the
sense of not deviating."
Ragh.M. 2.55 nyiiyyii nyiiyii anapetii. yuktety
dharma-pathy-anhfidina yat-pratyayaly,. lawful not deviating from the
law; i.e., proper. The suffix yaT by (Pa. 4.4.92.) "After dharma, pathin,
artha, ... ."
1.36.5 Locative
Perhaps the most common locative fonnula of analysis is tatra bhavam,
"occurring in that, pertaining to that."
Ragh.M. 1.20 "prak" ity avyayena purva-janmocyate. tatra bhavtif;,
priiktaniiJ:z. "sayaf!1-ciraTJ1- "ity-iidina .tyul-pratyayaly,. By the
indeclinable priik ("before") a fonner lifetime is to. Pertaining to
that former. The suffix .tyuL (i.e., -ana, before which -t- is inserted) by
(Pa. 4.3.23) "after siiyam, ciram, ...."
Ragh.M. 1.36 priivT:ji bhavaly, pravT:jel}yaly,. "pravT:ja el}yall" ity
erya-pratyayaJ;,. tam Occurring in the
monsoon-monsoonal. The suffix -el}ya by (Pa. 4.3.17) "el}ya after
priivT:ja (in the sense of OCCurring in that)." That monsoonaL
Ragh.M. 1.88 vane bhavaf!1 vanya1!l kanda-mulildikam occurring in
the forest forest (food), such as bulbs and roots.
Kum.M. 3.23 haimavatam himavati bhavam. Himalayan occurring
in the Himalaya.
When the relevant rule of Pal)ini specifies any special condition of mean-
ing (called an upiidhi), the commentator will include that specification in
analyzing the fonnation. Here, for example, are two refinements on the
tatra bhavam fonnula;
Ragh.M. 2.70 priisthiinikalJl prasthiina-ka/e bhavam. tat-kii16citam ity
arthaJ;,. "kiila! !han" iti !haii-pratyayaJ;,. pertaining to starting a
journey occurring on the occasion of starting a journey; i.e., appropriate
for that occasion. The suffix !haN (i.e., -ika, by Pa. 4.3.11) "(The suffix)
.thaN after (a word referring to) an occasion."
Ragh.M.5.15 iira1}yakii aral}ye bhavii manu:jyii muni-pramukhiiJ;,.
"aral}yan iti vuii-pratyayaJ;,. wilderness peoplehumans
occurring in the wilderness, such as sages. The suffix vuN (i.e., -aka) by
(Pa. 4.2.129) "After ararya when referring to humans."
Another common locative fonnula is tatra siidhub, "good in that, good at
that." ,
Ragh.M. 1.55 sabhiiyii,!, "sabhiiyii iti
ya-pratyayah. Good at court courteous. The suffix -ya by (Pa. 4.4.105)
"(The suffix) -ya after sabhii."
74 eiulpter 7. Secondary Nominal Derivation 1.37. Genitive formulas of analysis 75
Kum.M.2.57 sa1J!yuge sadhuYJ! sli'!'yugfnam. "pratijaniidibhyaft khan"
iti khan-pratyayah. good at battle good at battle. The suffix khaN (i.e.,
-fna) by (Pa. 4.4.99) "After pratijana, etc., (in the meaning 'good at that.'''
Kum. V 2.57 sii'!lyugfnam ralJa-pa1:uj.itam. siidhv-arthe
khan." good at battle expert in battle. (The suffix
-fna) in the sense of "good (at that)" by (Pi!. 4.4.99) "After pratijana, etc.,
(the suffix) khaN (in the meaning 'good at tha!')."
Ragh.M. 2.30 sarafle slidhuJ:t sara1)yal}. "latra sadhuJ:z" ifi
Good at refuge providing good refuge. The suffix yaT
(i.e., -ya) by (Pa. 4.4.98) "Good at that."
Another locative fonnula is based on the meaning of enrollment or enlist-
ment:
Ragh.M.4.45 senayii'!' samavetiift sainyii!1. ... "seniiyii vli" fty
Enrolled in the army soldiers. The suffix t:'ya (i.e., -ya)
by Pa. 4.4.45 seniiya va, "After sena there is optionally (the suffix Nya
[i.e., -ya] rather than !haK [i.e., -ika] in the meaning of enrollment)."
Ragh.M. 4.74 sainikdl) sentiyiilJ1 samavettitJ
soldiers enrolled in the army. The suffix!haK (i.e., -ika)
from the section extending (from pa. 4.4.1) up to "vahati" (in pa. 4.4.76).
1.37 Genitive formulas of analysis
The single largest topic in the section on taddhita suffixes in P3.Jfini is that
of patronymics and related fonnations referring to descent. These are an-
alyzed with a fonnula placing the base in the genitive case: tasyapatyam,
"offspring of his" (Section 1.37.1). Genitive fonnulas are also used in an-
alyzing derivatives with abstract meaning (Section 1.37.2), and fonnations
with many other meanings as well (Section 1.37.3), as well as fonnations
expressing a general connection (Section 1.37.4).
1.37.1 tasyapatyam
Ragh.M. 1.1 parvatasyapatyaf!l piirvatf. "tasyapatyam" ity
ity-tidina rilp. offspring of the mountain (parvata)
Piirvati. (The suffix) at:' (i.e., -a by (Pa. 4.1.92) "Offspring of that"; (the
feminine suffix) NfP (i.e., -;) by (pa. 4.1.15).
Ragh.M. 3.5 magadhasya rajiio 'patyam str! miigadhf.
"dvyafi-magadha-kalinga-suramaslid a!l" ity Female
offspring of the king Magadha Miigadhi. The suffix at:' (i.e., -a) by
(Pa. 4.1.170) "(The suffix) at:' after names of two syllables and after
Magadha, Kalinga, and Suramasa."
Ragh.M.4.87 riijfio 'patyiini Iqatriyiis tlin riijiinyiin
"rlijaivaiuriid yat" ity aparyarthe Offspring of a
king-royal men, k!atriyas-them royal men. The suffix yaT (i.e., -ya) in
the meaning of offspring by (pa. 4.1.137) "(The suffix) yaT after rajan
and ivasura."
1.37.2 tasya bhiivalf
Ragh.M. 1.8 iitor bhlivaJ:l iaiSava,!, blilyam.
"prli1}abhrjjiiti-vayovacan6dgiitradibhyo 'fi" ity afi-pratyayaJ:l. The
state of an infant infancy. The suffix aN (i.e., -a) by (Pa. 5.1.129) "(The
suffix) aN after names of animal species, words for stages of life, and
udgiit[, etc."
Ragh.M. 1.8 yuno bhiivo yauvana'!l tliru1}yam. yuvdditviid
a[l-pratyaya[z. tasmin vayasi [Yauvane]. The state of a young
man-youth, adolescence. The suffix aJY (i.e., -a) because of belonging to
the yuvadi group (mentioned in Pa. 5.1.130). In that stage of life in youth.
Kum. V 1.4 dhiitavo 'tra sinduradayo raktiiJ:l, te vidyante yasya sa
dhiitumiin; tad-bhiivo dhiitumattii, tlim [dhiitumattiim). Minerals here
are red ones such as red lead. That which has them is mineral-rich. The
state of being that is mineral-richness. That mineral-richness.
Kum.M. 2.11 prakamasya priikiimyalfl yathii-kamatvam. The
condition of being willful willfulness, doing as one wishes.
1.38. Nominative formulas ofanalysis
77
1.37.3 Others
Similar genitive fonnulas are used with many other specific meanings. For
example:
activity (tasya karma):
Kum.M. 2.35 stenasya bhiivatz karma vii steya'!Z cauryam. "stenad yan
nalopas ca iti na-lopas ca." The condition or activity of
a thief theft, stealing. By (Pa. 5.1.125) "After stena there is (the suffix)
yaT (i.e., -ya) and the dropping of na."
Kum. \I. 4.9 kitavasya kamza kaitavam The activity of a deceiver
deception.
Ragh.M. 4.1 riijya1fl riijfiatz kanna prajti-paripiilanatrnakam.
purohitiiditviid yak. rule the activity of a ruler, consisting of protecting
his subjectsrule the activity of a ruler, consisting of protecting his
subjects. (The suffix) yaK (i.e., -ya) because of belonging to the
purohitadi group (mentioned in Pa. 5.1.128).
modification, product (tasya vikiirai1):
Ragh.M. 2.29 dhiitor gairikasya vikiiro dhatumayf tasyiif'!'l
[dhiitumayyiim] adityakiiyiim urdhva-bhumau. a modification of a
mineral, of red chalk-made of mineral, on it on the highland the upland
made of mineral.
Ragh.M. 5.2 hira1Jyasya vikaral.z hira1)mayam. da1}dinayana- "ityadinii t.
A modification of gold made of gold. An exceptional form
(see Sections 1.20, 1.35), by (Pa. 6.4.174) ... ."
ruler (tasya riija, etc.):
Ragh.M. 4.40 kalingalJ kaliTigana'!l raja. '::'
"dvyan-magadha-kaliliga- "ity-iidinii The KiiliIiga king :i
of the Kaliilgas. The suffix a{V (i.e., -a) by (Pa. 4.1.170) "(The suffix) a{V
after names of two syllables and after Magadha, Kalinga, and saramasa."
Ragh.M. 2.20 piirthivena prthivyii fsvare1}a. "tasyeivaraf:z" ity
By the ruler by the lord of the earth. The suffix a{V by
(Pa. 5.1.42) "Lord of that."
collection (tasya samuhal.z):
Kum. V 7.57 malanam samaha ma/yam. "pasadibhyo yal.z" ity anena
A collection of garlands garland wreath. The suffix -ya by
(Pa. 4.2.49) "ya after pasa, etc. (in the sense of a collection of that)."
1.37.4 General connection
In analyzing fonnations expressing general connection, the base is dis-
played in the genitive case and a fonn of the pronoun idam is added:
rasyedam, "this belongs to it, this is connected with it."
Ragh.M. 5.75 gangaya idam giingam this is of the Ganges Gangetic
Kum.M. 1.41 tasyii imau tadfyau biihii These two are hers her two
arms
Kum.M. 2.1 svayam-bhuvo brahma1}a ida'!l sviiyambhuvam This
belongs to the Self-bom-Brahma of the Self-born
Kum. V 2.1 svayambhar brahmti tasyedar:n sViiyambhuvam The
Self-born is Brahma; this is his of the Self-born
Occasionally the pronoun idam is omitted and a fonn of the word sam-
bandhin, "connected to," is used instead. .
Ragh.M. 4.35 vaitasf,!, sambandhinfT[l v[ttim. pra1Jatim fty
arthaft The behavior associated with a reed of a reed; i.e., bending.
Ragh.M. 1.13 lqatrasambandhf warriorly connected with a
warrior
1.38 Nominative formulas of analysis
In these the base to which the suffix is added appears in the nominative
case together with a further specification of meaning drawn from Panini's
rules, and the object referred to by the reSUlting fonnation is indicated by a
pronoun in the genitive case. The genitive pronoun may be either a form of
idam or a relative pronoun picked up by some fonn of tat. These
are thus akin to those used in analyzing bahuvrfhi compounds.
78 Chapter 7. Secondary Nominal Derivation 1.40. Possessive suffixes (matvarthrya) 79
Kum.M. 4.18 rtur asya prapta/:l artava,!, vtisantam. "rtor Q!l" ity
Q'.1-pratyaya/:l Its season has arrived seasonal vernal. The suffix af! by
(Pa. 5.1.105) "(The suffix) aN after r'u (in the sense of having arrived for
this)."
Kum.M. 4.27 duf:lkham asyil/:l saf!ljijtaJ!l dulJkhitii. sQ1[lHita-du/:lkhery
arthaJ:z. tdrakiiditvild itac. grief has occurred for her aggrieved
griefstricken. (The suffix) itaC (in the sense of having occurred for this)
because of belonging to the tdrakadi group (mentioned in Pa. 5.2.36).
Kum. V 2.33 [yavan.miitre!la] yar parimill}am asya yilvan. tata/.t svdrthe
miitriic. [By just which much.] It has which as its
much. After that there is (the suffix) miltraC in the sense of the base itself.
Kum. V 6.92 viviihaIJ prayojanaf!l yasyiif:l sii vaiviihikf That which
has a wedding ceremony as its occasion is nuptial
Ragh.H. 5.58 [M5.57] giindharva,!, gandharvo devata yasya tat That
which has a Gandharva as its deity is Gandharva.
Kum. V 2.17 catvilro 'vayava yasyiil:t sii "sankhyaya avayave
tayap." That which has four parts is four-part. (By pa. 5.2.42) "(The
suffix) tayaP after a numeral, in the sense of (having that many) parts."
1.39 Compound formulas of analysis
In addition to the use of compounds ending in -sambandhin as a substitute
for a genitive pronouns (Section 1.37), there are some meanings which are
expressed through compounds to begin with in Piil)ini's rules, and com-
mentators preserved those formulas in analyzing formations. The most
common of these are tad-anham ("for the purpose of that") and tad-arham
("appropriate to that," "deserving that").
Ragh.M. 1.44 piljd-vidhih, tad-artha,!, dravyam arghyam.
"piidarghabhyiiTJ'l ca" iti yat-pratyayaf:l. Honoring is the rite of
hospitality, an offering for that purpose an offering of honor. The suffix
yaT by (Pa. 5.4.25) "and after pada and argha (there is yaT in the sense of
being for that purpose)."
Ragh.M. 5.2 arghlirtham idam arghyam. "plidarghilbhyiif!l ca" iti
yat-pratyaya(t. pujiirthaT[l dravyam This is for purpose of honoring
an offering of honor. The suffix yaT by (Pa. 5.4.25) "and after pada and
argha (there is yaT in the sense of being for that purpose)." An offering
for the purpose of hospitality.
Ragh.M. 5.3 vidhivad vidhy-arham. yatha-sastram ity
"tadarham" iti properly appropriately to the rules; i.e.,
in accordance with the prescriptions. The suffix vatl by (Pa. 5.1.117)
"appropriate to that."
Kum.M. 3.76 anukampitum arhiim anukampyiim. "r-halor iti
Deserving to be pitied pitiable. The suffix JYyaT by
(Pa. 3.1.124) "(The suffix) NyaT to (words eoding in) -r or a consonant."
1.40 Possessive suffixes (matvarthfya)
Formations made with possessive suffixes are glossed with asya vidyate or
asyasti ("of him there is, he has"), placed after the word to which the suffix
is added, and followed by iti:
'syastfti dal}tJikaJ:z "he has a staff, thus staffbearer"
dhanam asyastfti dhani "he has wealth, thus wealthy"
giivo 'sya santfN gomiin "he has cows, thus cow-having"
The last word could also be analyzed giivo 'smin santfli gomiin, "there are
cows in it, thus cow-having"
Similar formulas are used in conjunction with the additon of a word in
the analysis to bring out the full meaning of certain other suffixes:
ghrta,!! pracuram asminn iti ghrtamayaJ:z "ghee is abundant in it, thus
made (mostly) of ghee"
jiinu pramii.rtam asyetijiinudagdham "the knee is its extent, thus
knee-deep"
1.41. VibhaktisaI]1jfiaka suffixes
81
vlJ;lii-vadana1J1 silpam asyhi vail}ikal} "his art is viI).a-playing, thus
vil]a-player"
Less commonly. any of these possessive formulas of analysis can be re-
placed with formulas using relative pronouns, as in the analysis of bahu-
vr,hi compounds (Sections 1.64-1.68); e.g., ghrta,!, pracura,!, yasmin sa
"that is ghrtamaya in which ghee is abundant."
1.40.1 Straightforward possession
Kum. V 3.69 bala,!, vidyate yalra tad balavat. vasa 'syastfti van
That in which strength is found is strong. He who has control is
self-controlled peaceful.
Kum. V 5.1 0 prftir vidyate yasyilsQu pral}ayf That which
has love-affection-is a lover
Ragh.M. 1.12 suddhir asyiistiti suddhimiin. tasmin suddhimati.
tad-anvaye tasya manor anvaye. It has purity, thus pure. In that in the
pure lineage of his in the lineage-the dynasty, of him--{)f Manu.
Ragh.M. 1.53 iihuti-gandho yesiim astfly iihuti-gandhlnalJ of which
there is the aroma of offerings, thus having the aroma of offerings
1.41
Kum.M. 1.55 sparSavan7} sukha-sparsiif:z.. mrdvfr ity arthal:z..
prasa1J1siiytirrz matup. having (a good) feel a pleasant feel; i.e.,
soft. (The possessive suffix) matUP expressing pr;ise.
Kum.M. 1.4 dhiitavaf:z. sindura-gairikiidaya 'sya santW dhatumtin.
nitya-yoge matup. tasya bhal'o dhiitumattii. tam [dhiitumattiim). It has
minerals-red lead, red chalk, etc.-thus mineral-rich. (The possessive
suffix) matUP expressing constant connection. The condition of being
that, mineral-richness. It mineral-richness.
Kum.M. 1.27 putriis ca duhitaras ca putriif:z.. "bhriltr-putrau
svasr-duhitrbhytim" ity te 'sya santfti putraviin. bhumarthe
matup. tasya putravato 'pi bahv-apatyasyapfty arthal:z.. Sons and
daughters-<:hildren. An ekaiesa formation (see Chapter 9) by
(Pa. 1.2.68) "(When) bhriitr and putra (come together) with svas, and
duhitr (respectively, the first alone is retained)." He has them, thus having
children. (The possessive suffix) matUP expressing the sense of
abundance. of him although having children, i.e., although having many
offspring.
Vibhaktisa,!,jiiaka suffixes
Kum.M. 3.65 tapa 'syastf'ti tapasvf. "as-miiyti-medhti-srajo vinifz" iti
vini-pratyayaf:z.. tasmai tapasvine. He has asceticism, thus an ascetic.
The suffix vin1 by (Pii. 5.2.121) "(The suffix) vin1 after words ending in as
and after mtiyti. medhti, and sra)."
Ragh.M. 5.52 vtico 'sya santfti viigmf vaktii. "vaco gminif:z" iti
gmini-pratyayaf:z.. He has speeches, thus eloquent a speaker. The suffix
gminl by (Pii. 5.2.124) "(The suffix) gminl after vac."
1.40.2 Additions of meaning
Ragh.M. 1.91 putri,!ii,!, sat-putravattim. prasaf!lsiiytim ini-pratyayaf:z..
having sons having good sons. The suffix in! expressing praise.
These are indeclinable suffixes having the meaning of a particular case end-
ing. Most common are -tas, with ablative meaning, and -tra, with locative
meaning. They are glossed by replacing them with the appropriate case
ending.
Ragh.M.1.66 mattaJ,t para,!, mad anantaram. "paficamytis tasil."
After me subsequent to me. (By Pa. 5.3.7) "(The suffix) taslL in place of
the fifth c"e ending."
Kum.M. 1.46 tataJ,t tasyiif:z piirvatytif:z. "paficamyiis tasil." From her
from that one-from Plirvatf. (By Pa. 5.3.7) "(The suffix) tas1L in place of
the fifth case ending."
82
Chapter 7. Secondary Nominal Derivation 1.42. Sviirthika suffixes 83
1.42 Sviirthika suffixes
.,
These suffixes are added to a base "in the sense of its own meaning," i.e.,
to produce a fonnation meaning the same thing as the base to which the
suffix is added. Often these suffixes result in no change in meaning at all
(Section 1.42.1), but more broadly the category also includes those suffixes
5-'
that impart a some connotation, resulting in diminutives (Section 1.42.2)
or related formations (Section 1.42.3) such as pejoratives and honorifics.
Also included here are the suffixes allowing nouns to be used in conjunc-
tion with verbs in the senses of becoming that or making something that
(Section 1.42.4), the samasanta suffixes (Section 1.42.5), which are added
to end of nominal compounds without changing the meaning of those com-
pounds.
1,42.1 With no addition of meaning
In glossing these fonns a commentator will simply use the base word fol-
lowed by eva to emphasize that the word fanned with the suffix means just
the same thing as the base to which the suffix was added.
Kum. V 6.43 tama tamisram darkness darkness
Kum. V 2.44 sampraty eva siimpratam. prajiidditviid Now now.
(The suffix at:' because of belonging to the prajfiadi group (mentioned in
pa.5.4.38).
Kum. V 2.33 {Yiivan-miitreIJa] yar parimalJam asya yavan. svarthe
miitriic. [By just which much.] It has which as its extent-which
much. After that there is (the suffix) miitraC in the sense of the base itself.
1.42.2 Diminutives
Kum. V 5.14 alpa. hrasvii vii vr/qa meager or short trees
saplings
Ragh.M. 1.51 ... "hrasvarthe
saplings short trees.... The suffix -ka in a diminutive sense.
1.42.3 Other additions of meaning
Kum. V 3.52 atisayena bahu exceedingly much mostly.
Ragh.M. 1.12 atisayena suddhiman suddhimattaral).
"dvivaeana-vibhajyopa- "ity-iidinii exceedingly pure
purer. The suffix taraP by (pa. 5.3.57) "(The suffixes taraP and fyasUN
in the sense of exceeding) when speaking of two things," etc.
Kum. V 4.35 soma iva "akhadibhyo yat." Like the moon
gentle. (By Pa. 5.3.103) "yaT after sakhti, etc. (in the sense of 'like')."
Ragh.M. 5.36 kumiira'!' kUmiirakalpam
skanda-sadrsam. ity-adinii kalpap-pratyayaf:i.. A
slightly incomplete Kumara virtually a Kumara. The snffix kalpaP by
(pa. 5.3.67) "In the sense of slighty incomplete ...."
1.42.4 cvi, etc
Cvi is the Pal)inian name for a particular suffix that converts nouns into
items functioning as verbal prefixes before the verbs bha and kr in the
meanings of "becoming that, not having been it before" (abhata-tadbhiiye,
pa. 5.4.50) and "making something that." Several other taddhira suffixes
are used in much the same way.
Kum. V 4.13 akathii kathii krta'!' kathfkrtam. cvif:i.. "asya evau" itftvam.
Not (having been) a story, make (into) a story reduced to story. (The
suffix) evif:i.. There is f- (in place of the final -a before this Cvi, which is
then dropped) by (pa. 7.4.32) "(There is i) in place of -a or -ti when Cvi
follows."
Kum.M.4.13 kathf-krtam akathii katha sampadyamanaTJl krta'!'
reduced to story made to become a story, not
having been a story-remaining in report alone.
Ragh.M. 1.30 parikhf-/q1ii1}
. .. abhuta-tadbhave evif:i. made to become moats, not having been
moats made moats. .. (The suffix) Cvi in the sense of becoming
something, not having been it before.
Kum.M.3.68 asiici siid sampadyamiina tiryakkrfii
becoming oblique, not having been oblique made oblique cast sidelong
1.42.5 Samiisiinta suffixes
These suffixes are added at the end of a compound, without affecting the
meaning of the compound. The most commonly mentioned is -ka at the
end of a bahuvrfhi (see Section 2.15), but there are others, including -a at
the end of certain in which the final member is a stem ending in
a consonant.
Ragh.M. 1.13 vyurjhaJ!! vipulam uro yasya sa vyutJhoraskalJ
kap." He who has a broad-wide-shoulder is
broad-shouldered. (By Pi;' 5.4.151) "(The suffix) ka? after (a bahuvrfhi)
ending in uras, etc."
Ragh.M. 1.81 saha patnyii vartata iti sa-patnikalJ san. "nady rtas ca"
iti kap-pratyayal} He is together with his wife, thus being with his wife.
The suffix ka? by (Pi[. 5.4.153) "And (there is ka? after a bahuvrfhi) that
has the feminine ending -f or -u or that ends in ('
Ragh.M. 1.63 ity
"safayur vai iti srutei}. "acatura- "ity-iidi-sutrel}lic-pratyayanto ,;;
nipataJ:z. ... brahma1}o varco brahma-varcasam. "brahma-hastibhyiif!l
varcasaJ:z" ity ac-pratyayaJ:z. The life span of a man a man's life span;
Le., a hundred years, because the Veda says, "A man has a lifetime of a
hundred years." An exceptional fonn [see Sections 1.20, 1.35] ending in
the suffix aC, by (Pi[. 5.4.77) "acatura-," etc. The splendor of brahma lbe
brahma-splendor. The suffix aC by (Pi[. 5.4.78) "(The suffix aC) after
varcas, when it follows brahma or hastin."
Chapter 8
Nominal Compounds
A. General Principles
1.43 Introduction
Most Sanskrit compounds, no matter how many words they are composed
of, can be analyzed in the first instance as compounds containing only two
members (cf. Section 2.13). In the longer compounds, of course, either or
both of these members may itself be a compound of any number of words;
it is then analyzed in precisely the same way, beginning by breaking it
down into two members. The only compounds that must be analyzed from
the beginning as containing more than two members are certain dvandva
and bahuvrfhi compounds (Sections 1.62, 1.65.5), but for these the basic
techniques of analysis are exactly the same as those used in analyzing two-
member compounds of these types. It is therefore convenient to begin by
describing the standard formulas used to analyze two-member compounds
in each important variety of Sanskrit compound (Sections 1.44--1.68), be-
fore explaining how these same formulas are applied to the more compli-
cated business of analyzing long compounds (Sections 1.69-1.72).
Which formula is used in analyzing a given compound depends upon
several factors-which of the four basic categories the compound belongs
to (Section 1.44), whether the compound form conveys additional elements
of meaning beyond those included in the general rules for each category
(Section 1.45), and whether the compound is an optional one or an oblig-
85
86 Chapter 8. Nominal Compounds 1.44. Basic categories of compounds
87
atory one-a distinction detennining whether the compound can be ana-
lyzed using the same words that make up the compound (Section 1.46).
On the basis of these distinctions the grammarians have evolved a se-
ries of standard fonnulas to be used in analyzing each particular type of
compound; we describe the fonnulas for the most important types in Sec-
tions 1.48-1.68.
1.44 Basic categories of compounds
In what follows we assume that you are familiar with the basic differences
separating the four major categories of Sanskrit compounds:
1. avyayfbhava compounds (Section 1.44.1; treated in detail in Sec-
tions 1.48-1.51),
2. tatpu"'ea compounds (1.44.2; details in Sections 1.52-1.60), includ-
ing the important subclass of karmadharaya compounds (1.44.3;
Sections 1.54-1.55),
3. dvandva compounds (1.44.4; Sections 1.61-1.63), and
4. bahuvrfhi compounds (1.44.5; Sections 1.64-1.68).
Only occasionally is a compound analyzed as falling outside these groups
(Section 1.44.6). The English names given to these categories in Whit-
ney's grammar are noted at the beginning of the sections listed after each
category.
For each of these basic categories there is a standard fonnula of analy-
sis that can be used to produce a vigraha for most of the regular compounds
within that category. The fonnulas differ from each other to reflect the dif-
,i
ferences in the location of the syntactical predominance (pradhiinya) in
compounds of different categories.
1.44.1 Avyayfbhiiva
The most straightforward fonnulas are those that apply to compounds in
which one member adds some specification to the other simply by stand-
ing in some oblique (Le., other than nominative) case relationship to it; in
analyzing such compounds all that is necessary is to separate the two words
and give them their proper case-endings.
In avyayfbhiiva compounds it is the first member that is syntactically
predominant and the second member that simply adds some specification
to the notion expressed by the first member; in the analysis it is therefore
the second member that is exhibited in some oblique case (namely, the case
governed by the first member): e.g., bahir-gramam, "outside the village,"
is analyzed bahir gramtit.
1.44.2
In simple vibhakti-tatpurusa compounds ("case-tatpu"'ea," "dependent de-
tenninative compound") it is the second member that adds the specifica-
tion; in the analysis the first member is therefore placed in some oblique
case: e.g., "fallen from heaven," is analyzed svargat
1.44.3 Karmadhiiraya
For karmadhiiraya compounds, the variety of tatpuruea in which the first
member adds specification to the second member by standing in apposi-
tion to it ("descriptive compounds"), it is often possible to proceed just as
straightforwardly, by simply exhibiting the first member in the nominative
case; e.g., rakta-Iata, "red vine," may by analyzed rakta lata. '
But for some karmadhiiraya compounds-particularly those composed
of two substantives, such as rajarei, "king sage"-such an analysis may
leave the connection between the two members vague. To make it clear
that both members refer to one and the same thing, such a compound is
analyzed raja ca sa reiS ca, "he is both a king and a sage," and for the
sake of consistency this fuller fonnula is usually extended to apply to all
karmadhiirayas of simple apposition. Thus even rakta-Iata will generally
be analyzed rakta ca sa lata ca, "it is both red and a vine."
1.44.4 Dvandva
For the other two categories of compounds-dvandvas and bahuvrfhis-
the situation is more complicated than the furnishing of some
to the predominant member by the other member: in both categories some
88 Chapter 8. Nominal Compounds 1.45. Additional elements of meaning
89
additional element of meaning is involved in the compound and must be
reflected in the analysis by the addition of an extra word.
In dvandva compounds this additional element of meaning is the no-
tion of conjunction. Syntactical predominance belongs to both members
equally, and both are therefore displayed in the nominative case, but to ex-,
press the idea of conjunction the word ca must be added in the analysis:
e.g., "Rama and is analyzed ramas ca ca,
"(both) Rama and Note carefully the difference between this for-
mula and the fuller form of the analysis sometimes used for karmadharaya .;

compounds: the pronoun that always appears in the formula for karma-
dhiirayas in order to make it clear that both members refer to the same'
thing is missing in the formula for dvandvas, in which the two members
refer to two separate objects.
1.44.5 Bahuvnni
In bahuvrfhi compounds the additional element of meaning is the notion of
some relationship-usually that of possession-through which the com-
pound refers not to the object expressed by either member, but to some
other object outside the compound. Thus neither member has syntactical
predominance; the predominance belongs to the outside object, and that
object must be reflected in the analysis of the compound by the addition
of a relative pronoun, the case of which will convey the relationship tho-
rugh which the compound refers to the outside object: e.g.,
"Long-ear," is analyzed dfrghau kaT7Jau yasya "of whom there are
long ears, he," "he who has long ears."
1.44.6 Kevala-samiisa
A few compounds are not included by the grammarians under any of these
four categories. Occasionally an author will produce an ad hoc compound
that, while its meaning is perfectly clear, cannot conveniently be explained
by any of PaI!ini's particular rules for the formation of compounds in each
of the four categories (e.g., "pure since birth," Riigh. 1.5),
and occasionally words that could as well have been considered sepa-
rate are viewed as a compound for technical reasons of grammar (e.g.,
vagarthav-iva, "like sound and sense," Ragh. 1.1). Such compounds are
called kevala-samiisas ("simply compounds") or sup-sup-samasas ("noun-
noun-compounds," in reference to PaI)ini's general rule for compounding,
2.1.4: "[sup] saha supii," "[An inflected noun] (rna/be compounded) with
an inflected noun"). Commentators often content themselves with merely
noting the fact of such a compound:
Kum.M. 4.21 vina-k{1ii. viyojitii satfty sup-superi samiisaf:i..
bereft i.e., being deprived of. (This is) a compound, by the sup-supii
(rule).
Fortunately, when such compounds are analyzed, their analysis is usu-
ally trivial, since it consists simply of separating the two members of the
compound-thus is analyzed iijanma and vag-
arthav-iva is analyzed vagarthav iva-so that they need not detain us fur-
ther.
Most compounds, however, fall within the major groups, and more of-
ten than not the basic formulas of analysis are adequate. Yet for Some types
of compounds further formulas are necessary, either because the compound
involves the addition of a further element of meaning (Section 1.45) or be-
cause one of the members of the compound is a word that cannot be used
outside a compound (Section 1.46).
1.45 Additional elements of meaning
The conveying of special elements of meaning through composition, with
the resulting necessity of adding words in the analysis to convey that mean-
ing when the compound is dissolved, operates not only throughout the
general categories of dvandva and bahuvrfhi compounds (as explained in
the previous section), but also in particular types of compounds within the
general categories. Each of these special types of compound has its own
standard formula of analysis; the details are given in the lists of formulas
below.
1.45.1 Comparison
For example, certain kinds of karmadharaya compounds involve the notion
of comparison, so that the word iva ("like") must be inserted in the
even though it does not appear in the compound (Section 1.55).
90
Chapter 8. Nominal Compounds
1.46. Optional and obligatory compounds 91
In particular instances it is necessary to insert a word offering clarification
of the precise connection between two members of a compound, without
abandoning the basic formula of analysis (e.g., Sections 1.53.2, 1.54.4).
1.45.2 Aggregation
Similarly, certain types of dvandva and compounds involve the
notion of aggregation, so that the samiihtira ("aggregate") is often em-
ployed in the analysis (Sections 1.56, 1.63).
1.45.3 Clarification of connection
1.46 Optional and obligatory compounds
A further source of multiplication of formulas for the analysis of com-
pounds is the fact that not all compounds can be analyzed using the same
words that make up the compound. Traditionally, this problem is ap-
proached through the distinction between optional and obligatory com-
pounds. In the examples of a simple avyayfbhtiva and a simple
given in Section 1.44, the words that serve as members of the compound
are capable of being used as separate words in a phrase to express the
same meaning that they express when in compound. In order to express
that meaning the compounding of the words is therefore optional. In many
compounds, however, the members of the compound cannot be used to ex-
press the same meaning in a phrase; such compounds are therefore called
obligatory (nitya-samiisa, "a1ways a compound," in the sense that the words
must always be in compound if they are to express that particular mean-
ing), and are described as a-sva-pada-vigraha ("having no analysis in their
own words"). There are several general reasons why a compound might be
obligatory:
1.46.1 Asvapadavigraha componnds
In practice the designation of a-sva-pada-vigraha is applied only to com-
pounds of the types mentioned in the following subsections. Strictly speak-
ing, however, the compounds that involve additional elements of meaning
-/"
and thus require that other words be added to the members of the com-
pound when an analysis is given (Section 1.45) are 8bligatory, in the sense
that the members of the compounds are not capable in themselves of ex-
pressing the full meaning of the compound through their use in a phrase.
By this interpretation, all dvandvas, bahuvrfhis, karmadhtirayas of com-
parison, and other such compounds are obligatory compounds. There is
evidence that this was the view of Pal)ini himself, and that the later tradi-
tion misunderstood his treatment of optionality (Paul As
a Variationist [co-published by the Poona University Press and the M.LT.
Press, 1979], p. 43). In any case, the tradition does not regard such com-
pounds as incapable of analysis in their own words: the standard formula
of analysis for each of these types of compounds does use the words that
make up the compound, even if it has to use something else in addition.
When a compound is considered a nitya-samiisa any rewording of it
is technically not a trne vigraha analysis but simply a description of the
meaning of the compound. After offering such rewording a commentator
may then point out that the compound is considered nitya:
Ragh.M. 1.18 bhatya arthaya bhiily.artha,!, vrddhy-artham eva.
"arthena saha nitya-samiisab sarvaliflgatii ca vaktavyii." only for
prosperity's sake for the sake of prosperity, for the sake of income. (By
the Vmika rule) "It should be stated that (in compound) with arrha there
is a nitya-samiisa (and it may be used to modify words) in every gender."
1.46.2 Conventional meanings
Some compounds have acquired by convention a meaning that is more spe-
cific than that expressed in a standard analysis. Just as the English com-
pound "bluebird" refers not to any blue bird buy only to certain species, so
the Sanskrit compound refers not to any black snake, but only
to a particular kind of cobra. Thus an analysis such as ca sa sarpas
ca, "it is a snake and it is black," while it may indicate the etymological
formation of the compound, does not accurately restate the compound's
full meaning. Commentators frequently point out that such compounds
are nitya-samiisas; strictly speaking their full meaning can never be repro-
duced in the form of a standard analysis, but can only be conveyed through
a synonym or definition of the compound as a unit.
92
1.49. Optional avyaylbhiivas 93
1.46.3 Bound forms
In practical terms the most important reason for the obligatory status of
a compound is that one of its members is an element that simply cannot I
stand as a separate word within a phrase, or at least cannot do so With the "i
same meaning that it has within the compound. In such instances some
substitute must be found for this element in the phrasal analysis, and these
compounds can therefore not be analyzed using their own words. Sev-
eral important groups of Sanskrit compounds are obligatory compounds in
this sense: avyayfbhiivas in which the first member is incapable of con-
veying the meaning in question as a separate preposition (Section 1.50),
in which the first member is a prefix or preposition used ad-
jectivally (Section 1.57), in which the final member is a verbal
noun that can exist only at the end of a compound (Section 1.59), and a few
others. For each of these groups the formulas of analysis that are required
differ markedly from the standard formulas for the broad category within
which the group falls; specific details are given under each section.
1.47 Irregularities
Irregularities in compounds usually involve oddities in the form of the com-
pound rather than in the formulas required in analyzing them. They will be t
touched on in separate sections under each type of compound (e.g., Sec- i.
tions 1.51, 1.60, 1.68).
B. Avyayfbhiiva Compounds
1.48 Introduction
Avyayfbhiiva compounds (Pi;' 2.1.5-2.1.20; "secondary adjective
used adverbially," Whitney 1313) are indeclinable adverbial compounds ".j
with a preposition or other indeclinable as first member,i
For avyayfbhiiva compounds the distinction between optional (Section 1.49) ;l
and obligatory compounds (Section 1.50) is particularly noticeable: the .j
..
two are analyzed in quite different ways.
1.49 Optional avyayfbhiivas
For these compounds the analysis simply separates"the two members and
places the second member in the case that is governed by the first: bahir-
gramam, "outside the village," for example, is analyzed bahir gramiit,
"outside of the village." But things are usually not this simple, because
in order for the compound to be optional-and thus capable of being an-
alyzed in its own words-the preposition or indeclinable that constitutes
the first member of the compound must be able to stand as a separate word
governing another word in a particular case-words that can do so are tech-
nically known as karmapravacanfya-and furthermore it must be capable
of doing so with the same meaning that it has in the compound. Some
prepositions that occur as the first member of avyayfbhiivas are karma-
pravacanfya in some meanings but not in others; e.g., anu-vanam ("near
the forest") and anu-galigam ("along the Ganges") are optional compounds
and may be analyzed anu vanam and anu galigam, but anu-ratham ("be-
hind the chariots") and ("in order of age") are obligatory
compounds that require special treatment. Similarly, for certain meanings
there are some prepositions that are karmapravacanfya and others that are
not: e.g., anu-vanam ("near the forest") is an optional compound, but upa-
kumbham ("near the pot") is obligatory.
When the preposition beginning the compound is available as a sepa-
rate preposition governing a particular case, the analysis is straightforward:
Ragh.M.l.5 [a-janma] ajanmanah. janmiirabhyety arthah. "ali
maryadlibhividhyoh" ity avyayfbhavah. [From birth] from birth, i.e.,
beginning with birth. This is an avyayfbhava by (Pli. 2.1.13) "a in the
sense of limit or inclusion (may be compounded with a following noun)."
1.50 Obligatory avyayfbhiivas
For each of these compounds in which the first member cannot stand out-
side the compound while expressing the sarne meaning, some other way
to convey the meaning of that member must be found. The forms of anal-
ysis used for such compounds differ greatly depending upon the specific
94
1.51. 1rregular avyayThhavas 95
fonn and meaning of the first member. To take the obligatory avyayf-
bhtivas mentioned in the preceding subsection: anu-ratham ("behind the
chariots") is analyzed rathiinii7[! pasciit ("at the rear of the chariots"), anu-
("in order of age") is ("according to
the sequence from the oldest"), and upa-kumbham ("near the pot") will be
analyzed kumbhasya samfpe ("in the vicinity of the pot"). Obviously no
general rule for the treatment of these can be given; usually the precise .,
wording used in analyzing these fonns is drawn directly from specifica-
tions found in individual rules in Pal)ini's grammar. Yet the more commOn
of these obligatory avyayfbhiivas occur rather frequently, so that the for-
mulas used to analyze them show up often in commentaries; it is therefore
useful to be familiar with a few of them (of course the second member will
vary, but whenever the first member is used with the same meaning as in
the compounds mentioned here, the basic fonnula of analysis will be the
same):
kumbhasya samfpe: upa-kumbham, "near the pot"
iitmanyadhik[1Ya: adhy-iitmam, "concerning the self"
rathiiniilp pasciit: anu-ratham, "behind the chariots"
saktim anatikramya: yathti-sakti, "according to one's ability"
"in order of age"
ekakiile: sa-cakram, "simultaneously with the wheel"
jalasya madhye: antar-jalam, "within the water"
dine dine: prati-dinam, "on each day"
Some examples from commentaries:
Ragh.M. 1.6 vidhim anatikramyayathii-vidhi. "yathtisi1driye" ity
. ... yathii-kiimam anatikramya . ..
;::
yathQpariidham apariidham anatikramya ... yathii-kiila'!l kalam 4
anatikramya not transgressing the rule in accordance with the law; an

avyayfbhiiva by (Pa. 2.1.7) "yathli (may be compounded with a following
A
noun) in a meaning other than similarity." ... in accordance with their
wishes not transgressing their desires . .. in accordance with the crime
not going beyond the crime . .. in accordance wi;th the time not going
beyond the (proper) time. >
Ragh.M. 1.44 padasya pasctid paiciid-aJ1he 'vyayfbhiivah.
after the foot following upon. An avyayfbhiiva in the meaning of .
"behind" [by PaI)ini 2.1.6].
Ragh.M. 1.91 vighnasyabhiivo 'stu. "avyayaf!l vibhakti-"
(Pli. 2.1.6) ;ty-lidindbhiivdrthe May there be no
obstacle an absence of an obstacle. An avyayfbhava in the meaning of
"absence" (by 2.1.6).
Kum.M. 1.48 asaf!lsayam. sandeho nastfty
An absence of doubt without doubt no doubt; Le., no doubt exists.
Kum.M. 1.60 ahany ahani praty-aham. on each day daily
Kum.M. 5.1 samfpe samalqa,!, purataJ:z. "avyayarrz
vibhakti-samrddhi-" ity-iidinavyayfbhiivaJ:t. in the vicinity of her eyes
before her eyes, in tront of her. An avyayfbhiiva by (Pa. 2.1.6) "An
indeclinable (is compounded with a following noun) in the senses of. '"
1.51 Irregular avyayfbhiivas
Most of the common irregularities are in the construction of the compound
itself rather than in the fonnula by which it is analyzed. There are, for
example, avyayfbhlivas in which the order of the members is reversed,
so that they end up with a governed prior member and thus appear to be
these are nevertheless analyzed like regular avyayfbhiivas: e.g.,
gmigii-pliram, "across the Ganges," is analyzed gaJigliyiiJ.! pliram. Con-
versely, it should be noted that there are that have governed final
members and thus appear to be avyayfbhiivas, either because the order of
the members is irregularly reversed-e.g., "having occurred
previously," analyzed parva7[! because it is logically con-
venient to interpret them as having reversed members (Section 1.53.3),
96
1.53. Dependent 97
or because they really do have governed final members but are neverthe-
less traditionally regarded as because they are used adjectivally
rather than adverbially (Section 1.57.3).
C. Tatpurusa Compounds
1.52 Introduction
The category of tatpurusa compounds (pal).ini 2.1.22-2.2.22; "determi-
native compounds," Whitney 1262-1291) has several important subtypes.
The varieties of tatpurusa distinctive enough to require separate formulas
of analysis will be treated under the following headings:
1.53 dependent compounds
descriptive compounds (karmadhiiraya):
1.54 simple karmadhiirayas
1.55 comparative karmadharayas
1.56 dvigu compounds
1.57 prepositional and related compounds
etc.)
1.58 negative compounds
1.59 upapada compounds
1.60 irregular
1.53 Dependent
Dependent often called case-tatpurusas
PilJ)ini 2.1.22-2.1.48; "dependent compound," Whitney 1264-1278), are
those in which the first member of the compound stands in some oblique
case relationship to the second.
In these compounds the first member simply adds some specification
to the second member by standing in some case-relationship to it. Accord-
ingly the analysis is usually quite straightforward (Section 1.53), but there
are slight complications when additional clarification is necessary (Sec-
tion 1.53.2) and when the second member is most conveniently interpreted
as a partitive genitive (Section 1.53.3).
1.53,1 Straightforward analysis
The two members are simply separated, with the second member appear-
ing in the nominative case and the first member in any oblique case. The
names by which commentators refer to these compounds indicate the case
in which the first member appears in the analysis:
(accusative
grama,!, gataIJ-grama-gataIJ ("gone to the village")
muharta'!' sukham-muharta-sukham ("pleasure for a momen!")
(instrumental
asina kalahah-asi-kalahaIJ ("a fight with a sword")
parasuna chinnaIJ-parasu-cchinnaIJ ("cut with a hatchet")
(dative
kUlp;lalaya hirUl.'yam-kunala-hirUl.'yam ("gold for an earring")
kuberaya baliIJ-kubera-baliIJ ("an offering for Kubera")
(ablative tatpurusa):
svargat patitaIJ-svarga-patitaIJ ("fallen from heaven")
vrkebhyo bhayam-vrka-bhayam ("fear of wolves")
(genitive
candanasya gandhaIJ-candana-gandhaIJ ("fragrance of san-
dalwook")
idhmasya pravrascanah-idhma-pravascanaIJ ("cutter o(fuel,"
i.e., a hatchet)
accusative (dvitfya):
saptamf-tatpurusa (locative
("skilled in dice")
mase deyam-masa-deyam ("payable in a month")
Note that the first member need not appear in the singular in the analysis,
but may have the grammatical number required.
In these formulas, the choice of cases is of course governed by the
Sanskrit idiom rather than by what one might expect in or any other
language: thus "wolf-fear," "fear of wolves,' 15. an ablatlve
not genitive, because in Sanskrit the ablative case IS used with
words expressing the source of fear.
Ragh.M. 2.10 iiciiriirthair liijair iiciira-liijai/.l with puffed rice for the
sake of custom with custom-puffed-rice
99
samidbhir edhai1:z. samiddharrt dfptam
kindled-lighted-with firewood-with fuel
1.53. Dependent
Kum. V. 1.56 [M57}
[samit-samiddham]
firewood-kindled
Kum.M. 1.51 mantrai1:z. putafJ'l salJ'lskrtam [mantra-pittam]
purified-refined-with mantras mantra-purified
dative (caturthf): Dative compounds are relatively infrequent
because of the limited scope of the dative case in Sanskrit. When
they do occur commentators tend to explain them not by displaying
the first component in the dative case, but by using the equivalent
(and perhaps less ambiguous) compound ending in -artha (see the
example on p. 91 and cf. Section 2.19.2):
98
Kum.M. 1.5 adhaJ:r-sanuni megha-ma1JrJalad adhas taflini gaUl,!,
praptam [adhal}-siinu-gatiimj. "dvitfya
sritatfta-patita-gauityasta-priiptapannaib" fti samiisafl. gone
to---reaching-the below-slopes-slopes below the circle of clouds
below-slope-gone. The compound is by (Pi.. 2.1.24) "A word in the
accusative case is compounded with srita, atfta, patita, gata," etc.
Kum. V 1.60 avacitiiny uccitiini baly-arthaf[l
pujiirthaf!l yayii sa [collected-offering-flowered] she by whom
flowers for the sake of offering-for the sake of worship--had been
collected-accumulated
ablative (paiicamf):
Kum.M.2.23 astaf!l niisam ita!; priiptaJ:r [astam-itaJ:l]. astam iti
makiirantam avyayam tasya "dvitfyii
iritiitfta-patita-gatatyasta-priiptiipannai/:r" iti samasaJ:r. gone
to-reached to-setting-destruction [setting-gone]. astam is an
indeclinable ending in m. Its compound is by (Pa. 2.1.24) "A word in the
accusative case is compounded with . ..."
Kum. v: 1.37 [M38] sitii itaraJ:r sitetara iti "paiicamf" iti yoga-vibhiig'tif
other than white white-other (i.e., black). The compound is
by separating "a word in the ablative case (is compounded)" (from the rest
of the sutta in Pa.2.1.37).
Kum.M. 2.30 sahasran nayanebhyo 'dhika,!, sahasra-nayaniidhikam.
more than thousands of eyes eye-thousand-more
Kum.M.3.41 latiigrhadvara'!' gone to
the vine-hut-door [vine-hut-door-gone]
instrumental (trfya):
Kum.V. 1.27 [M 28} prabhaya mahatfprabhii-mahatfiti
t(tfya-samasa[l. great with splendor splendor-great; this is an
instrumental compound.
Kum.M. 2.45 tasya ttirakasyaptittit samtipatter bhaytit
[tad-iipiita-bhaYiit] from fear of his-Taraka's-attack-encounter
from his-attack-fear
genitive
Ragh.M. 1.9 te;itiTJl raghu1,lii'!l gU1,lais tad-gu!'ai!'. by the virtues of
theirs-of the Raghus by their-virtues
100
Chapter 8. Nominal Compounds 1. 53. Dependent tatpuru1as 101
Occasionally the connection between the first and second member is not ,(
adequately indicated merely by the case of the first member as it appears. i
in the analysis, so that an additional word of clarification must be sup-
plied. This occurs, for example, in a group of instrumental deal- T
ing with combinations of foods (Pfu,lini 2.1.34-35): thus ':
("yogurt-rice") is analyzed dadhn6pasikta ("rice sprinkled with
Ragh.M.1.12 tad-anvaye tasya manoranvaye kule. in
his-Manu's-lineage in the Manu-lineage
Ragh.M. 1.24 vinayasya adhcmiit [vinayadhiiniit}
from the imparting---doing-of education-instruction from
education-imparting
Kum.M. 1.58 adrrr:ziiT[l nathaJ:t adri-niithal} lord of mountains
mountain-lord
Kum.M.2.25 [ogha-sa,!,rodhal,r] oghasya sa'!'codhah
praviiha-pratibandhaJ:!. current-obstruction
obstruction-ftow-obstacle-of the current
locative (saptam!):
Kum.M. 1.44 pravale biila-pallava upahitaf!l nihitaJ'!l [praviilOpahitam}
laid-put down-on a bud-young leaf bud-laid
Kum.M.2.45 bhuvananam alokane prftih [bhuvaniilokana-prftihl
joy in the seeing of the worlds world-seeing-joy
Kum. V 3.8 piidiinatalJ piidayar [palita] iinatas cara!la-patitaJ:t
fallen-bowing down-at the feet foot-fallen
Kum.M. 4.10 tvad-adhfna'!' tvayy adhfnam dependent on you
yon-dependent
Ragh.M. 4.5 samrajye samrtijya-kannafli
suzerainty-consecrated
consecrated-anointed-in suzerainty-being suzerain, overlordship
'-f-
YOgurt"). and gua-dhiinah ("molasses-barley") is analyzed gLldena misra
dhiinah ("harley mixed with molasses"). Commen.tators call such com-
pounds madhyama-pada-Iopa (see Section 1.54.4). '
Kuru. V 3.46 sutram bead-filled string
bead-string
Kuru. v: 3.56 bimba-tulyo 'dharo bimbiidhara iti saka-ptirthivaditvtit
samasah. cherry-like lip cherry-lip: the compound is Uustified) by
belonging to the saka-ptirthiva group.
1.53.3 Partitive genitive as second member
In compounds with partitive genitive as second member (ekadeii-
pa. 2.2.1-2), in which the
first member of a compound is a word delimiting a part of a sin-
gle whole expressed hy the second member. a logical prohlem arises. The
object referred to hy a is supposed to be a variety of the thing ex-
pressed hy the second member, with the first member simply adding some
specification, but (to give an English example) a halfpenny, for instance, is
not really any sort of penny at all. Accordingly, the second memher in stich
compounds is construed as a partitive genitive and the compound is
ered to have its members reversed (because the syntactical predominance
ends up belonging to the first member). Thus purva-kaYah ("fore-body") is
analyzed purva,!, kayasya ("front of the body"), and arddha-pippalf ("half-
pepper") is analyzed arddha,!, ("half of a pepper").
This same procedure-making the first member predominant and ex-
hihiting the second member in the genitive case in the analysis-is also
used for several other rather obscure groups of tatpurusas (see Pfu,lini 2.2.3-
5), and for a few irregular compounds (Pat).ini 2.2.31) typified by the word
("king-tooth," i.e., front tooth), which might appear to be a
karmadhiiraya but instead is analyzed dantana'!' raja ("king of teeth").
Some commentators interpret the raja-dantadi category rather hroadly,
to include compounds with reversed memhers that other commentators in-
terpret as falling outside all categories (i.e., as sup-sup-samasas; see Sec-
tion 1.44.6):
1.54.1 Adjective plus substantive
1.54 Karmadhiirayas of simple apposition
The karmadharaya compound composed of an adjective followed by a sub-
stantive (visesana-[piirva-pada]-karmadharaya, "karmadharaya with an
103 1.54. Karmadhiirayas of simple apposition
adjective [as first member]," Pa\lini 2.1.57ff.; Whitney 1280) is the sim-
plest and most common form of karmadhtiraya, and)ts analysis is done in
the manner already explained in Section 1.44: while compounds such as
nf16tpalam ("blue waterlily") and rakta-Iata ("red vine") may be analyzed
simply by separating their members-nflam utpalam and rakta lata-they
are more commonly analyzed with the fuller formula that uses a pronoun
to make it clear that both members refer to the same thing: nflaT[l ca tad
utpala'!' ca ("it is both blue and a waterlily") and rakta ca sa lata ca ("it is
both red and a vine"). Again, note that the presence of this pronoun (usu-
ally some form of tad or adas) in the analysis serves to distinguish these
compounds from dvandvas.
In practice, commentators often use the regular formula but sometimes
simply separate the two words and place them in apposition:
Kum.M. 1.2 sarve ca te sailas ca sarva-iailiil}.
"purva-ktildika-sarva-jarat-purii1:za-nava-kevaIaJ; samiinddhikara1:zena"
iti samiisaf:l. they are both all and mountains all-mountains. The
compound is by (pa. 2.1.47) "When preceding, the words eka, sarva,
jarat, purii1}a, nava, and kevala (are compounded) with a word in
apposition."
Kum.M. 2.32 mahiin asuro mahdsural}.
pujyamiinaif:i." iti tatpunqal:t. a
demon great-demon. The is by (Pa. 2.1.61) 'The words san,
mahat, parama, uttama, and (are compounded) with things to be
praised."
1.54.2 Adjective plus adjective
Karmadharaya compounds composed of an adjective followed by an ad-
jective "karmadharaya with adjec-
tives as both members") are usually analyzed using the formula just de-
scribed. Thus the analysis regularly presents both members as adjectives
describing the same thing, even though in most of these compounds (e.g.,
those formed by Pa\lini 2.1.68) the first member actually functions as an ad-
verb modifying the second member. For example, ("equally
white") is analyzed tulyas casau svetas ca ("it is both equal and white"),
Ragh.M. 3.54 aniisviidita-piirvam purvam anosvaditam. sup superi
samasaJ:t. untasted-before previously untasted. A compound by (the
general rule) "noun with noun,"
Ragh.H. 3.55 [M 54] [an-iisviiditapiirvaT[lJ piirvam asviiditam
asviidita-purvam. tato nafi-samtisaJ:!, riijadantadiJ:!.
[not-tasted-before:] previously tasted-befare-tasted, then a negative
compound. (The compound tisva.dita-purva belongs to) the ra.ja-dantadi
group.
Karmadhtirayas (Pa\lini 2.1.49-2.1.72; "descriptive compounds;' Whitney
1279-1291) are compounds in which the two members appear in
grammatical apposition in the analysis. Although some important types
of karmadhtiraya involve the additional notion of comparison and thus re-
quire special formulas of analysis (Section 1.55), in most karmadhtirayas
the first member simply adds some specification to the thing expressed by
the second member.
For all karmadharaya compounds of simple apposition a single ba-
sic formula of analysis applies, but minor differences sometimes OCcur
depending upon whether the compound contains an adjective as the first
member only (Section 1.54.1), as both members (Section 1.54.2), or as nei-
ther member (Section 1.54.3). For the latter group, in which both members
of the compound are substantives, special treatment is sometimes required
when further specification of the connection between the members is re-
quired (Section 1.54.4). Note that karmadhiirayas in which both members
are substantives often involve the notion of comparison (Sections 1.55.2,
1.55.3), and karmadhtirayas composed of substantive plus adjective nearly
always involve comparison (Section 1.55.1).
102
104
1.54. Karmadharayas ofsimple apposition 105
and ("palatably warm") is analyzed bhojya,!, ca tad ca
("it is both palatable and warm").
As with other simple karmadhiirayas, commentators may use the reg-
ular formula or simply place the two words in apposition:
Ragh.M. 1.83 pallavavat snigdha casau pii!alii ca
[pallava-snigdha-pa!aliij it is both glossy like a sprout and pink'
[sprout-glossy-pink]
Ragh.M.l.16 bhfmaiS ca kantais calbhfma-kantailJj both awesome
and lovable awesome-lovable
When both members of the compound clearly do function as adjectives
describing the same object, they are often past participles indicating con-
secutive actions involving that object (pa1).ini 2.1.49); for these compounds
a special formula of analysis is used: for example, sniitanuliptalJ ("bathed
and oiled") is analyzed adau sniitalJ pascad anuliptalJ ("first bathed, after-
wards oiled"), and plt6dgfn.zam ("drunk and vomited") is analyzed tidau
pita,!, pasctid udgfn.zam ("first drunk, afterwards vomited").
Another common type of karmadhiiraya combines two color adjec-
tives.
Kum.M. 3.36 casau siiratz sabalas ca
vaT1Jena" iti samiisaJ:t. it is both black and speckled-spotted, a
black-spotted antelope. The compound is by (Pii. 2.1.69) "a color with a
color."
Kum.M. 4.14 harita7Jl ea aru!lar.n ca [haritilru!Jam]. "va1?lo vaf?lena"
iti both green and pink [green-pink]. The compound is by
(Pa. 2.1.69) "a color with a color."
1.54.3 Substantive plus substantive
Many such compounds involve the added notion of comparison (Sections
1.55.2, 1.55.3). When merely apposition is involved, the standard formula .\\
for a karmadharaya is used: e.g., ("king-sage") is analyzed raja 4
ca sa rsis ca ("he is both a king and a sage"). Such compounds are not vety;
in Sanskrit, except when a title is compounded with a proper name
(in which case the title usually comes last-e.g., Sankaracarya, Bhoja-
raja-even though it might logically be expected to come first), because
it may be unclear from the compound in what the first member ap-
plies to the thing expressed by the second.
In practice commentators often analyze noun-noun karmadhiirayas sim-
ply by putting the particle eva after the first member (cf. riipaka-samasas,
Section 1.55.3) rather than using the casau ... ca formula:
Kum. V 3.27 navQrrz ciltam eva biil}aJ:t saras tasmin [nava-cufa-biil}e]
the arrow-shaft-that is a fresh mango blossom, on it on the
fresh-mango-arrow
Kum.M. 3.27 navarrz cutarrz cuta-kusumarrz tad eva biil}as tasmin
nava-cuta-biiIJe. a fresh mango--mango blossom, the arrow that it is,
on it on the fresh-mango-arrow
Ragh.M.1.58 riijyam evasramas tatra munim[riijyasrama-munim]
the life-stage that is kingship, a sage in it [a kingship-lifestage-sage]
Ragh.M. 4.5 padmatapatrel}a padmam evatapatrarrz tena with a
lotus-parasol a parasol that was a lotus, with it
Kum.M. 3.36 kusumam evaikarrz siidhiiral}aT[l piitrarrz tasmin
[kusumaika-piitram]. a single---common---cup that was a flower, in it
in the flower-single-cup
Kum.M. 3.53 evabharal}am
an ornament that is springtime-flowers a
springtime-flower-ornament
This last problem is particularly acute in karmadhiirayas composed of two
substantives when the first member cannot be applied, as itstands, directly
to the thing denoted by the second member. In analyzing loose compounds
of this sort it is necessary to add some explanation of the precise con-
nection between the members. This is done by placing the explanatory
word in compound with the first member in the analysis: for
deva-briihma!,alJ Cdeity-bralunin") is analyzed deva-piijako brahma':falJ
107
1.55. Karmadharayas of comparison
Kum.M. 1.13 candra-marfcibhir iva gauraiJ.z subhraiJ.z
[candra-marfci-gauraiJ.tJ .... "upamiintini stimiinya-vacanaib iti
samiisaJ.z. pale-white-as moon-beams moon-bearn-pale. The
compound is by (Pa. 2.1.55) "Standards of comparison [are compounded]
with words expressing shared properties."
Ragh.M. 1.13 stilo iva prtitJ1sur unnataJ.z siila-prii'!lsu!.t.
tall-lofty-like a sal tree sal-tall
kumudam iva syenl-kumuda-syenl ("white like a waterlily-lily-
white")
analyzed by simply separating the members of the compound and inserting
between them the particle iva ("like"):
ghana iva ("dark as a cloud--eloud-dark")
("deity-worshipping brahmin"). Commentators frequently refer to cOrn- ',,)
pounds of this type as "siikapiirthivadi compounds," since they are grouped ";
together (by a viirtika on Pal)ini 2.1.60) in a list beginning with the cOrn_ }i
pound ("vegetable-king" or "era-king," analyzed siika-bhojl
["vegetable-eating king"] or siika-priyah ["a king fond :."
of vegetables," or, according to some, "a king beloved of an era"]),
Note that these compounds are similar to those such as
(Section 1.53.2): in both groups the connection between the two mem_
bers requires some explanation, and in both groups this explanation is sup-
plied by adding a word between the members deva-
that can be interpreted as standmg m compound With
the first member (gua-misrii deva-piijako briihmalJaM. Accord_
ingly, commentators often refer to such compounds as uttara-pada-lopa-
samiisas ("compounds in which the second word [of the first member]
has been dropped"), or, much more commonly, as madhyama-pada-lopa-
samiisas ("compounds in which the middle word has been dropped").
106
Megh.M. 1 chiiya-pradhiinas taravas chayii-tarava!t.
siikapiirthivaditviit shade-rich trees shade-trees. The
compound is because of being in the siika-piirthiviidi group.
Kum.M. 3.29 [biilindu-vakrii1}i] biilendur iva vakral}i
[young-moon-crooked] crooked as a young-moon
Acommentator may express the comparison through the use of the suffix
-vat rather than the particle iva:
1.55 Karmadhiirayas of comparison
These are of two basic types. In one type (Section 1.55.1) the first mem-
ber is a substantive representing the standard of comparison (the upamiina)
and the second member is a substantive representing the common property
(siimanyavacana). In the second type both members are substantives: the
first represents the object of comparison (upamita) and the second repre-
sents the standard of comparison (upamiina). This second type can be ana-
lyzed in two ways, depending upon whether the underlying comparison is
viewed as a simile (Section 1.55.2) or as a metaphor (Section 1.55.3).
1.55.1 Upamiinapiirvapada karmadhiiraya
These formations (upamiinapiirvapada-karmadhiiraya, "karmadhiiraya in
which the first member is a standard of comparison," Pal)ini 2.1.55), are
Ragh.M. 1.83 pallavavat snigdha casau piilala ca
[pallava-snigdha-pii!alii] it is both glossy like a sprout and pink
[sprout-glossy-pink]
1.55.2 Upamita-samiisa
In analyzing these (upamita-samiisa, upamiinottarapada-karmadhiiraya,
"karmadhiiraya in which the second member is a standard of comparison,"
Pa. 2.1.56), the particle iva is inserted after the second member. Because
these compounds are made up of two substantives, a pronoun may option-
ally be used in the analysis to make it clear that both nouns refer to the
same Object.
'ya,!, vyiighra ("[this is] a man like a
tiger-man-tiger")
108
1.56. Dvigu compounds 109
mukha,!, candra iva-mukha-candral1 ('"a face like the moon-moon_
face")
Ragh.M. 1.12 raja indur iva riijendu upamita,!,
vyaghrddinii samasaJ:z. a king like the moon moon-king best of kings.
The compound is by (pa. 2.1.56) "The object of a simile [compounded
with] a word such as 'tiger.'''
Kum,M. 2.47 hayo ratnam iva haya-ratnam a horse
like ajewel a horse-jewel best of horses
Ragh.M.4.61 mukhiini padmtinfva mukha-padmiini.
faces like lotuses face-lotuses. A simile compound.
Ragh.M. 5.73 dantiil1 kosii iva danta-kosiil} teeth like buds
teeth-buds
1.55.3 Riipaka-samiisa
In karmadhiirayas in which one substantive is equated with another, either
metaphorically or literally (riipaka-samasa, "compound of metaphorical
identification"; avadhiirar1.apiirvapada-karmadhiiraya, "karmadhiiraya in
which the first member is a limitation"), what is involved here is simply a
different interpretation of the compounds in the preceding group, by view-
ing the comparison involved as a metaphorical equation (riipaka) rather
than as a simile (upama). The formula of analysis inserts eva ("actually,"
"nothing but ... ") after the first member:
eva vyaghral1-purusa-vyaghral1 ('"a tiger that is actually a
man-man-tiger")
mukham eva candral1-mukha-candral1 ("a moon that is nothing but
a face-face-moon")
Kum.M. 3.30 cuta-praviila a lip that
is actually a mango-shoot a mangoshoot-lip
Kum.M. 3.39 lalii eva vadhvas [lalii-vadhiibhyaJ.!] wives that
are actually vines, from them from vine-wives
1.56 Dvigu compounds
The category of dvigu compounds (Pa. 2.1.52; Whitney 1312) includes
several types of compounds in which the first member is a numeral. There
are special restrictions on the use of numerals and words signifying di-
rections as the first member of a karmadhiiraya. These items may form
the first member of simple two-member descriptive compounds only when
the compound constitutes a proper name (Pal)ini 2.1.50). Such compounds
may be broken down using the standard formula for the analysis of karma-
dhiirayas, although the resulting analysis will simply indicate the gram-
matical formation of the compound rather than restating its full meaning,
since the analytical phrase is not in itself a proper name:
sapta ca Ie, ("they are seven and are
tbe Seven Sages")
piirvii casau ("it is both eastern
and
Example from a commentary:
Kum.M. 1.16 sapta ca te ca "dik-sailkhye
safljfliiyiim" itf samiisaJ;. they are seven al)d are sages seven-sages. The
compound is by (pa. 2.1.50) "Direction and number words (are
compounded with a noun in apposition) in the sense of a proper name:'
These items may also be used as the first member of a karmadhiiraya com-
pound under three more complex conditions (PliJ.1ini 2.1.51); when the
compound involves an additional element of meaning equivalent to one
of the senses of the suffixes used to form words by secondary derivation
(Section 1.56.1), when the whole compound is itself the first member of
a longer compound (Section 1.56.2), and (c) when the compound involves
the added notion of aggregation by functioning as a collective noun (Sec-
tion 1.56.3). Compounds beginning with numerals under any ofthese.three
conditions are called dvigu, and special methods of analysis are required
for each of the three conditions:
110
Chapter 8. Nominal Compounds 1.56. Dvigu compounds
111
1.56.1 Taddhitartha-dvigu
1.56.3 Samahara-dvigu
Ragh.M. 3.45 trayli/yii'!l!okiiniiJTl niithas
"taddhitdrtha-" ityiidinottara-pada-samtisafz. lena [tri-loka-nathena].
lord of the three worlds-three-world-Iord. (This is) a compound with
another word following in compound, by [Pii. 2.1.51] "To form a taddhita
derivative," etc. By him, by the three-world-lord.
Kum.M.2.17 [catur-mukha-sammtii} caturbhir samfrita satf.
"taddhitartha "ity-ddinottara-pada-samiisah. samtihiire catunnukhfti
syiit. four-mouth-uttered being uttered mouths. (This is) a
compound with another word following in compound, by (Pii. 2.1.51)
"(Number and direction words are compounded with a noun in apposition)
to form a taddhita derivative, or when another word follows in
composition, or when an aggregate is denoted." If an aggregate were
denoted [the form] would be caturmukhf."
paficasu samskrta/:l-pafica-kapiila/:l ("prepared in five pots"
-five-pot")
1
Here the added word sarrzskrta/:l, together with the locative case of the "
other words in dependence on it, serves to indicate the additional element }
of meaning. The word "dvigu" ("two-cow") that is used as the name of"
this whole class of compounds is in itself an example of this particular 1
subclass and is analyzed dvabhyarrz gobhyarrz krftam ("bought with twoE
cows"). Note that the sarne sort of analysis is used when the compound
begins with a word signifying a direction:
In the "dvigu involving the sense of a taddhita suffix" the two members of
the dvigu compound are placed in apposition, and something is added to "
indicate the additional element of meaning, using the techniques of anal- "
ysis appplicable to words derived by the use of taddhita suffixes (Sec-
tions 1.35 ff.):
purvasyarrz siilayarrz bhava/:l-paurva-sala/:l ("produced in the east-
ern hall, pertaining to the eastern hall--east-haller")
1.56.2 Uttarapada-dvigu
The "dvigu with a second member" is analyzed like any other longer com-
pound (Section 1.69 ff.); the two members of the dvigu itself are therefore
placed in apposition, using the shorter formula for karmadhiirayas com-
monly employed in the analysis of longer compounds (Sections 1.54.1,,;;
1.69): l'
J
pafica gavo dhanarrz yasya: pafica-gava-dhana/:l ("whose wealth is!
five cows: five-caw-wealth") ..
'"
The sarne principles apply when the first element is a word signifying a
direction:
purva salii priya yasya-purva-salii-priya/:l ("to whom the eastemi
hall is dear--east-hall-Iover") ;
In the "dvigu of an aggregate," or collective dvigu, once again the two
members of the dvigu appear in apposition in the analysis. The added
notion of aggregation can be indicated in either of two ways:
(I) pafica gavah samahrtti/:l-pafica-gavam ("five cows taken together-
five-cowl -aggregate]")
trl'.'i bhuvanani samahrttini: tribhuvanam ("the three worlds taken
together: three-world[-aggregate]")
or (2) paficanarrz gavarrz samahiira/:l-pafica-gavam ("an aggregate of five
cows-five-cow")
traya'.'arrz bhuvananarrz ("the aggregate of
the three wOrlds-three-world")
Compare with this the analysis of collective dvandva compounds (Sec-
tion 1.63).
In this particular subclass of compounds, words signifying directions
cannot be used as the first member.
112
Chapter 8. Nominal Compounds 1.57. with inseparable first member
113
Kum. V 2.17 catu0iif!! mukhiiniil!! samiihiiraJ:t catur-mukham.
plitradil:z.. an aggregate of fouf mouths four-mouth. (But for a different
interpretation see Mallinatha's comment on this word, quoted under
Section 1.60 a.)
Kum. v: 2.30 sahasraf!l nayaniini samahrriini sahasra-nayanam.
"taddhitarth6ttarapada-samiihiire ca." "dvigur ekavacanam." A
thousand eyes collectively thousand-eye. (Pa.2.1.51:) "To form a taddhita
derivative, or when another word follows in compound, or in the sense of
an aggregate," (Pa. 2.4.1:) "A dvigu is singular in grammatical number."
Kum.M. 6.93 trayiilJiim ahniif!l samahiiras try-akab
"taddhitarth6ttarapada-samiihiire ca" iti sarniisal:z.
tac" iti tac-pratyayaJ:t. dvigutviid ekavacanam. "riitrahnahiib pUf!lsu" iti
tasmat try-ahiid urdhvam an aggregate of three
days-triduum. The compound is by (Pa. 2.1.51) "to form a taddhita
derivative, or when another word follows in compound, or in the sense of
an aggregate." There is the suffix -a (in place of the final -an of ahan) by
(Pa. 5.4.91) "After ratri, ahan and sakhi there is TaC (at the end of a
compound)." There is singular number because it is a dvigu. There is
masculine gender because of (Pa. 2.4.29) "ratri, ahna, and ahaJ:t (at the
end of a are) in the masculine." After it after a triduum.
1.57 with inseparable first member

Many compounds-most of them karmadhiirayas--cannot
be analyzed in their own words because the first member of the compound
is a prefix or other item that cannot stand independently as a separate word
(Whitney 1281 ff.). These inseparable words fall into three broad groups:
individual prefixes with prescribed meanings (Section 1.57.1), words in-
cluded in the list of verbal prefixes (priidi) (Section 1.57.2), and other items
placed in compound with a following verbal form (Section 1.57.3).
1.57.1 ku, su, etc.
Most of the inseparable prefixes that are not included in the list of verbal
prefixes occur only in certain specific meanings, so that in the analysis of
compounds in which such prefixes are used a standard synonym may be
used in place of the prefix. For example, the prefixes kii- and kad- are
replaced by kutsita ("censured") when they mean "bad," and by f,<ad when
they mean "slightly, partly" (Pa\lini 6.3.101 ff.):
("bad person")
fsat ("semi-human")
kutsito ("bad horse")
("lukewarm")
Similarly, the standard synonym kutsita is also used for the prefixes ku-
(Pa\lini 2.2.18) and kim- (pilJ.1ini 2.1.64) when they mean "bad," and the
standard synonym is also used for the prefix kava- (Pa\lini 6.3.107)
when it means "slightly", e.g., in analyzing the compounds
("bad person"), ki,!,-sakhii ("bad friend"), and ("lukewarm").
Ragh.M. 1.2 dustara'!' taritum asakyam. "ity-adina
khal-pratyayah hard-to-cross unable to be crossed. The suffix KHat
(i.e., -a), by (Pa. 3.3.126) "In connection with f,ad, dUh, and su ...."
Ragh.M. 1.67 kavopJam yathii tatM ... iti
ku-sabdasya kav/ldesal.z. luke-warm so as to be slightly warm.... In
there is the substitution of kava for the word ku.
Ragh.M. 1.84 ki'!lcid "kava'!l iti ca-kiirtit
ktidesa!:t. luke-wann somewhat warm. There is the substitution of ka
[for kava] because of the word "also" in (Pa. 6.3.107) "And kava also,
when follows."
1.57.2 with fixed meaning
The traditional list of verbal prefixes (priidi, Pa\lini 1.4.58) includes a few
items that are rarely prefixed to finite verbs; these are treated like the pre-
114
1.57. with inseparable first member
115
fixes in Section 1.57.1. The prefix dur- ("bad"), for example, is often re-
placed by the standard synonym dus!a in the analysis of a simple descrip-
tive karmadhi'iraya: e.g., janar-dur-janar ("a depraved person-
bad-person"). However, because these prefixes frequently have a rather
vague meaning, the word that replaces them in the analysis may be chosen
from a large list of synonyms: thus su-, when it means "good" (i.e., when "
it is used as an adjective modifying a following noun), may be replaced by :::
sadhu, samfcfna, sobhana, subha, piijita, etc., and when it means "well" Ii
(i.e., when it is used as an adverb modigying a following adjective) it may:!
be replaced by samyak, atfva, etc. ,I
Often more than one explanation is possible in analyzing these com-
pounds. In the first example below, the commentator first interprets the
compound as an upapada then decides it is a pradi ,
Kum. V. 1.1 adhirdjata ity adhirdjah prabhuh pacddy-ac. adhir "
aisvarye. [nagfidhiriija iti] [yadi va] adhikas casau raja

[cal fae." He rules over, thus
overlord-master. (The krt suffix) aC [i.e., -a, which is added after the
roots] pac, etc. (in the sense of an agent, hy Pi;:' 3.1.134). (The prefix) adhi 1
(is used) in the sense of lordship. mountain-overlord is a genitive !
compound. [Or rather] he is both superior [and] a ;
king-over-king. (The suffix) TaC (i.e., -a, in place of the final -an of,l
-rdjan, by Pli. 5.4.91) "TaC after rdjan, ahan, and sakhi (at the end of a l
compound)."
Kum.M. 1.1 adhiko raja adhiriijalJ. superior king over-king
Megh.M. 4 iobhanam tigata1J1 svtigatam. good coming welcome
Kum.M.4.45 sU!j.thu caritiirthani [su-caritilrthiini] ... well
expressed well-expressed
1.57.3 Dhiitu/opa
A similarly straightforward analysis is sometimes used for compounds be,;
ginning with a prefix that is commonly used with finite verbs (e.g., adhi.,
rajar, "over-king," may be analyzed adhiko rajii, "a higher king"), but,.
more usually such compounds are analyzed by replacing the prefiX wltb.$'
the past passive participle of some verb, joined together with the prefix
in questlOn. The particular verb chosen may diffef.even in two analyses
of the same compound; e.g., pra,vatah ("strong wind") may be analyzed
prakrHo vatar or prabhiito vatah. Such compounds are called dhatu-Iopa
("involvmg the loss of a verbal root"--cf. Section 1.54.4).
. When the first member of a prepositional compound is expanded by
usmg the past passive participle of a verb in this way, the second member
of the compound may appear in any of the first five cases in the analysis
(viirtikas on PiiI)ini 2.2.18):
(nom.) pragatar pitamahar-prapitamahah ("previous
grandfather-great-grandfather")
(ace.) atikranto malam-ati-malar ("gone beyond the necklace
[i.e., surpassing it in beauty]-beyond-necklace")
(instr.) kokilaya-ava.kokilar ("called down to by the
koil-down-koil")
(dat.) parimlano 'dhyayanaya-pary-adhyayanar ("[too] exhausted
for study-beyond-slUdy")
(ab!.) ("gone out from
Kausiimbi--{)utside-Kausiimbi")
Note that in the analysis of such compounds ("prepositional compounds
With governed final member," Whitney 1310) the first member is syntacti-
cally predominant, despite the general rule that in tatpurusas the syntacti-
cal predominance belongs to the second member. In co'nstruction these
c.ompounds are thus akin to avyayfbhavas, but they are used as adjec-
tives rather than as mdeclinable adverbs. Whitney and some other Western
scholars therefore group them together with bahuvrihis under the heading
of secondary adjectIVe compounds (Whitney 1292 ff.), but because they do
not fulfil the basic requirement that a bahuvrfhi must refer to something
other than what is referred to by either of the members of the compound
(somethmg IS ali-mala, for example, is ati in a particular way), the
Sanskrit tradition regards them as that form exceptions to the
general rule by which it is the first member that adds some specification to
the second.
116 Chapter 8. Nominal Compounds 1.57. Tatpurusas with inseparable first member
117
Ragh.M. 2.9 unmadiiniim utka.ta-madtiniim high-impassioned
highly impassioned
Note also that, unlike other with an inseparable first mem_
ber, with a governed final member cannot be regarded as
karmadhi'irayas; once a substitute is provided for the inseparable members,
they are analyzed exactly like ordinary dependent compounds (vibhakti_
Section 1.53.1), except that the order of their members is re-
t
versed.
Ragh.M. 12.73 atiastra/:l sastrlil}y atikriintiil}
beyond-weapon
gone beyond weapons

derivatives of the unit composed of gati plus verb, rather than as nominal
compounds-e.g., "adorned," is viewed as the past participle
of the verb alam-kr rather than as a compound of aiam and krtah-so that
commentators regularly explain such items by giving a synonym or a tech-
nical analysis rather than an analysis in ordinary language. When it is
necessary to bring out the meaning of the prefix, the formation may be an-
alyzed by using some substitute (usually an adverb or a word in an oblique
case functioning as an adverb) for the prefix; this technique is used to ex-
plain the meaning of the prefix whether the prefix occurs in conjunction
with a finite verb, a participle, or a noun derived from a verbal root:
prakarse!,a jvalati-pra-jvalati ("it blazes strongly")
Ragh.M. 3.41
kriinttidyarthe dvitfyayti" iti super-sensory gone beyond the
senses. The compound is by (the Vartika rule) "ati, etc. (are compounded)
with a word in the accusative, when the sense is 'advanced,' etc."
Different commentators may choose different ways of expanding the pre-
fix:
Kum. V 1.45 [M 46] praviite ville. in a strong-wind in a
protracted wind
Kum.M.1.46 praviite prabhuta-vtita-sthale in a strong-wind in a
powerful-wind-place
Kum.C. 1.46 pragato VQto yatra tatpraviitam. where there is an
advanced wind, that is strong-wind(ed)
1.57.4
These formations (PaI)ini 2.2.18, Whitney 1090ff.) involve prefixes com-
pounded with verbal forms. When the prefixes pra-, etc. (Section 1.57.2)
re used in conjunction with a verbal form they are called gati (PaJ.lini
1.4.60), and this same designation also applies to other items used in a
similar way (PaJ.lini 1.4.61-79); e.g., iirl-kr ("to accept"), asta,!,-gam ("to
set, to go down"), ala,!,-kr ("to adorn"). For practical purposes nominal
forms derived from verbs with a gati as the first member are regarded as
madhye hitam-antar-hitam ("placed in the middle-interposed")
jfiiinam-vi-jfiiinam ("knowledge in terms of particulars-
particular knowledge")
Agerund may also be used for this purpose:
Kum.M. 2.28 sambhuyagatliJ;. samiigatiil} coming having gathered
come together
In dealing with derivative nouns, however, the prefix may also be repl"l::ed
with an adjective rather than an adverb:
jfiiinam-vi-jfiiinam ("special knowledge").
One gati-formation that requires special analysis and that appears quite
frequently is the compounding of a word ending in -I (technically said to
be formed with the suffix called cvi, PaJ.lini 5.4.50) and some form of the
verb kr or bhii. The analysis of such formations brings out the full meaning
of a thing's becoming or being made something that it had not been before:
("[ha'Cing earlier been]
non-black, it becomes black")
("having become black")
sampadyate ta,!, ("he makes it black")
118
sampadyamana!, ("made black")
1.59. Upapada compounds
aswid anyatz-an-aSval:z ("other than a horse-non-horse")
119
Similar fonnulas of analysis are used for other fonnations involving the
same notion; e.g., mithya-karoti ("he falsifies") may be analyzed mithya
na bhatva mithya sampadyate tat karoti ("he makes it become false, not
having been false [before]").
Analyses of cvi- fonnations have already been described under the
heading of taddhita suffixes; see Section 1.42.4.
1.58 Negative
Negative PaI)ini 6.3.73-77; Whitney 1288.1),
sometimes called "alpha-privative" compounds, begin with a- or (before
vowels) an-. They constitute a subclass of karmadhtirayas with inseparable
prefixes as the first member, and they require special fonnulas of analysis.
Commentators frequently content themselves with a direct statement of the
compound's meaning by using the sentence "[it is] not ... ," with or without
the verb "to be," followed by iti:
na brahmalJa ity a-brtihma1JaJ:z ("not a brahmin, thus non-brahmin")
nasva ity an-asval:z ("not a horse, thus non-horse")
brahma1}o na bhavatfty a-briihmalJafl ("[he] is not a brahmin, thus
non-brahmin")
asvo na bhavatfty an-asval:z ("[it] is not a horse, thus non-horse")
Actually, however, such compounds have a more precise meaning than is
indicated in this fonnula of analysis: a non-brahmin, for example, is not
anything that is not a brahmin, but rather someone who is like a brahmin
(by being a member of the same more general category-Le., human) and
yet different from a brahmin, and similarly a non-horse is an animal that
is other than a horse. To bring out this meaning commentators just as
frequently use a stricter fonnula in analyzing these compounds:
briihmQ1:uld anyai:z-a-brahma1}ai} ("other than a
brahmin-non-brahmin")
Note that, as this fonnula indicates, negative tatpurusas are anomalous
in that the syntactical predominance belongs to the first member rather
than to the second (cf. with governed final members, Sec-
tion 1.57.2).
Commentators often simply use the negative particle na in analyz-
ing such compounds, especially when it is a verbal activity that is being
negated:
Kum. V 1.10 na raila-para akrtrimatvlit not oil-filled
non-oil-filled nonartificiaI
Kum.M.3.50 na Iqaratfty alqaram aviniisinam it does not perish,
thus non-perishing neverending
1.59 Upapada compounds
These compounds PaI)ini 2.2.19, etc.; Whitney 1286-
1287) have as their last member a bound fonn, i.e., a word that occurs only
at the end of a compound. In compounds that cannot be analyzed
in their own words because the final member cannot stand as a
word with the same meaning outside the compound, the usual situation is
that the second member is a verbal derivative (i.e., a noun fonned from a
verbal root by primary derivation-see the sections on krt suffixes Sec-
. '
tions 1.27-1.34) that can appear with the meaning in question only when
placed in compound with a preceding subordinate word (upapada). For ex-
ample, "pot-maker," is an obligatory upapada-compound,
because there is no independent word with the general meaning
of "maker," whereas "water-bearer (i.e., cloud)" or "milk-
bearer (i.e., breast)," is an ordinary (and optional) because
can stand as a separate word meaning "bearer."
Note that a compound such as satya-vadin, "truth-speaker," is also
an upapada compound even though there is an independent word vadin,
because vadin as a separate word does not have the general of
"speaker," but rather more specific meanings, such as "debater." Thus a
120
Chapter 8. Nominal Compounds 1.60. 1rregular 121
compound like payo-charah can be analyzed using the standard fonnula
for ordinary (payasii,!, charaJ:!, "bearer of water [or of milk)"), "

while compounds like kumbha-kiiraJ:! and satya-viidin require more round-
e
about modes of analysis. ,
The usual fonnula involves replacing the second member with a finite ;"
verb, and displaying the first member separately in the required case. Be- "

cause this produces a sentence rather than a phrase in apposition with the i
compound being analyzed, the analysis is followed by iti, just as in the
analysis of negative tatpurusas (Section 1.58):
kumbha,!, karatfti kumbha-kiiraJ:! ("he makes a pot, thus pot-maker")
satyam vadatfti satya-viidf ("he speaks the truth, thus truth-speaker")
nagara,!, gacchatfti nagara-giimf ("he goes to the city, thus city-
goer")
As the examples show, the final member in an upapada-compound usually
has active meaning, and the first member usually represents the direct ob-
ject of that action. Where the second member has passive meaning, the
first member may appear in an oblique case in the analysis, and the second
member may be replaced either by a finite passive verb or by the equivalent
participle (in which case no iti is needed):
asvena krfyate stty asva-krftf ("she is bought with a horse, thus
horse-bought")
manoJ:! jiiyata iti manu-jaJ:! ("[he] is born from Manu, thus Manu-
born")
asvena krftii-asva-krftf ("bought with a horse-horse-bought")
manoJ:! jiitaJ:!-manu-jaJ:! ("born from Manu-Manu-born")
An alternative and simpler way of analyzing upapada compounds is
to replace the second member with a synonym that can stand as a sepa-
rate word; the compound can then be analyzed using the standard fonnula
for ordinary dependent For example, mantra-krt (mantra,!,
karotfti) can be analyzed mantrli'.lGrrz "creator of mantras."
Note that the basic fonnulas most commonly used in analyzing up-
apada-compounds are equivalent to the fonnulas employed in analyzing
krt-suffix fonnations in general (Section 1.27 ff.). One deficiency in these
formulas is that upapada-compounds often involv<;'an additional element
of meaning, in that the action referred to in the second member must be a
characteristic or habitual one-a nagara-giimin, for example, is not some-
one who just happens to be going to the city at this moment (as the analysis
nagara,!, gacchatfti might imply), but rather someone who habitually or
regularly goes to the city, whether he is going there now or not. To meet
this problem, commentators sometimes prefer to analyze these compounds
by using the word mam ("[characteristic or habitual] conduct, practice"),
which is either employed in conjunction with an infinitive in a formula
akin to that used in analyzing bahuvrfhi compounds (Section 1.64 ff.), or is
simply placed in a bahuvrfhi compound with a verbal noun:
nagararrz gantu'J'l Sflam asyiti ("his practice is to go to the
city, thus city-goer")
nitya'?l snatu'!l sflam asyeri nitya-snayr ("his practice is to continually
bathe, thus continual-bather")
nagara-giimf; nagara-gamana-fflaf:i. ("city-goer: having the practice of
going to the city")
("sweet-speaking: havina the
practice of speaking sweetly")
For examples of commentarial analyses of upapada compounds see the
treatment of this topic under the heading of krt suffixes in Section 1.31.
1.60 Irregular
The common irregularities affecting the analysis of fall into two
general groups.
1.60.1 compounds
One group contains the compounds that are analyzed as being composed
of words that appear in the compound in an inexplicably mutilated form.
Kum.M. 1.4 viirf1}ii'!l viihakii baIahakiil}.
carriers of water clouds. (The replacement of vitri by bala and of viihaka
by iihaka) is all right because (the compound belongs to) the P!sOdari1di
group.
Ragh.M. 3.57 datMfriiyiitJ1 te sarpiif:l.
siidhuf:l. Those who have poison in the
fang-toeth-are fang-poisons serpents. (The replacement of iiSi/:z by iis!)
is all right because (the compound) belongs to the group.
Ragh.M. 3.19 dyaur aka te divQukaso
stidhuft. [divilukasam] those whose dwelling is heaven are
heaven-dwellers-gods. (The replacement of dyauh by diva) is all right
because (the compound) belongs to the group. Of them of the
gods.
123
1.60. Irregular
Kum.M. 1.28 manasa sasydstfti . .. sakandhv-iiditviit
siidhul;. The desire of the mind is intelligence; he who has it is
intelligent ... (The compound is) all right because of its heing (in the
group) sakandhv-tidi.
The Pa\linian system of grammar speaks of other types of irregularities
as well. For example, a Vartika on Pa. 6.1.94, whiplI deals with situations
in which the second of two vowels is the single substitute for both, estab-
!ishes a group of irregular compounds such as sakandhu (supposedly from
saka and andhu) in which this replacement occurs unexpectedly. Commen-
tators may sometimes disagree even with themselves on how to explain ir-
regular sandhi combinations in compounds; so Mallinatha, who explained
the word manfsin as falling within the group in the last exam-
ple, places it (probably more exactly) within this sakandhv-adi group in a
commentary on another poem:
Ragh.M. 1.35 dam-panjiiya-patf. jiiyii-sabdasya dal1} iti
nipiitaniit siidhuf!. married couple husband-and-wife. (The compound)
is all right (even though one would expect the formpati-jiiye) because of
the exceptional use of dam as a substitute for the wordjiiyii within the
riija-dantddi group.
Phonetic alterations within a compound can occur in many categories
of compounds, and commentators make mention of other lists of irregu-
larities as well, including the riija-dantadi group mentioned above in Sec-
tion 1.53, which is used in the following example to justify a dvandva with
supposedly reversed members and an irregular replacement:
Once the supposedly original form of the mutilated member of the com-
pound has been restored, the analysis is straightforward along the lines
of the usual formulas. For example: balahakaf:! ("cloud") is said to be
..
a compound of vari (which has collapsed into ba) and vahaka (with its,
initial v changed into I), and is analyzed varilJa,!, vahakaf:! ("bearer of
waters"); smasanaf:! ("cremation grounds") is said to be a compound of
sava (which is replaced by sma) and sayana (changed into sana), and ._
is analyzed savana'!' sayanaf:! ("resting-place of corpses"). Commenta-
tors call such compounds because they are grouped together
(Pa. 6.3.109) under the example a bahuvrfhi compound said to
be made up of and udaram.
122
Ragh.M. 3.59 vyiipnotfti vi vyiipakam ojo yasya sa tasya
bi4aujasalJ,. indrasya. siidhul; it is active all over-it
pervades-thus ubiquitous, pervasive; of him whose strength is
(pervasive)--<Jf Bi<:iaujas ofIndra. (The replacement of vie' by bie'a) is all
right because (the compound) belongs to the P!sOdari1di group.
Ragh.M. 1.11 manasa l:<;ilJo dhfriil;, vidvii'!1sa iti yiivat.
siidhul;. Having the wish of the mind intelligent-wise.
To be precise, learned. (The compound is) all right because of its being (in
the group) P!sodarOdi.
1.60.2 Mayiiravya1flsakiidi compounds
The second group involves the compounds that are analyzed as having the
syntactical predominance in the first member rather than the second, so that
the order of the two members seems to have been reversed. We have al-
ready mentioned a few of these (see Section 1.51 and the sections referred
to there, and Section 1.53.3). In the same category belong compounds
beginning with prapta or apanna--e.g., prapta-yauvana ("[a girl] who
has reached adolescence")-which might logically appear to be bahuvrfhis
Ragh.M. 1.49 vanantarad anyasmiid vaniit from other-forest from
another forest
Among the other oddities included in the mayiiravya1'[lsakadi category
are compounds containing indeclinables of various kinds:
Kum. v: 1.56 [M57] anyii milrtir miirty-antaram another form
other-form
125
1.61. Introduction
Ragh.M. 1.50 bhtiga eva bhaga-dheyo "riipa-narna-bhogebhyo
dheya-pratyayo vaktavyaJ;" iti vaktavya-siitriit sviibhidheye
A portion (bhagadheya) is simply a portion
a part. The suffix dheya (is added) in the sense of the base itself (i.e., :
without changing the meaning of the word to which it is added), by the
Vmika rule "It should be stated that the suffix -dheya (may be added)
after riipa, ntima, and bhoga."
D. Dvandva Compounds
Ragh.M. 1.17 rekhti pramtiflam asyeti rekhiimiitra'!'. apfty
"pramii1;e dvayasac- "ity-tidina mtitrac-pratyayaJ;. it has the size of a
line, thus a mere line. The meaning is "even slight amount." The suffix
rnatraC (is added by Pa. 5.2.37) "-dvyasaC, daghnaC, and rnatraC in the
sense of size.
1.60.3 Suffix-formations resembling compounds
Finally, it should be noted that, in some formationS that look like com-
pounds, what appears to be the second member is treated by the gram-
matical tradition as a suffix rather than as a noun in compound. This is
trUe, for example, of -riipya in words such as samariipya ("formerly in the
possession of an honest man") and devadattariipya ("formerly belonging
to Devadatta") (PaI)ini 4.3.81), and of -maya (technically called mayar,
piil,lini 4.3.82, 4.3.143 ff., 4.3.148 ff.) in words such as asmamaya ("made
of stone") and vrfhimaya ("filled with rice"). In analyzing such formations,
commentators use the formulas appropriate for taddhita-formations (Sec-
tion 1.42 ff.) rather than the formulas applicable to nominal compounds
(cf. Section 2.19).
1.61 Introduction
Dvandva compounds (PiiI)ini 2.2.29, etc.; "copulative compounds," Whit-
ney 1252-1258) are divided into two major sub-categories of: simple cop-
ulative compounds (Section 1.62) and copulative compounds that involve
the added notion of aggregation (Section 1.63).
Chapter 8. Nominal Compounds 124
Kum. V. 1.14 yad-rcchO.-sabda icchfJrtho [The
compound] yad-rccha ("what(ever)-wish," meaning a whim, belongs to
the mayura-vyaf!lSakiidi group.
(analyzed, for example, prapta1'[l yauvanat!, yaya sa, "she by whom ado-
lescence has been reached") but which are traditionally (following PaI)ini
2.2.4) analyzed as accusative in which the order of the mem-
bers has been reversed and in which the second member is the direct object
of the first: thus prapta-yauvana is analyzed prapta yauvanam, "(being)
arrived at adolescence."
Compounds with reversed members that do not fall under specific rules
(as the last example does) are called mayiiravya1'[lsakadi by the commenta-
tors, because they are grouped together (PiiI)ini 2.1.72) under the example
of the irregular karmadhtiraya compound ("peacock-
cunning"), which is supposed to mean "cunning peacock." This group
includes many other irregularities as well.
Of the types of compounds in this group involving in which
the first member is syntactically predominant, two are especially notewor-
thy because of their great frequency: compounds ending in -an/aram (cf.
Section 2.19.1)--e.g., anyo raja-rajilntaram ("another king")-and com-
pounds ending in -matram (cf. Section 2.19.9--e.g., cid eva-cin-matram
(')ust consciousness, pure consciousness"). For these two types of com-
pounds the syntactical predominance of the first member is indicated by
the standard formulas of analysis using anya and eva.
Ragh.M.5.15 na vidyate kiTJlcana yasyety a-ki,!,canalJ.
mayiira-vyalJlsakdditvttf of whom there is not anything
anything-less. (This is) a tatpuru,a (which is all right) because of
belonging to the mayiira-vya'!2sakiidi group.
126
i.64. introduction 127
1.62 ltaretara dvandvas
!taretara dvandva compounds (dvandva of "one and the other"; Whit-
ney 1253.1) are cumulative compounds in which each member contributes
equally to the total grammatical number (dual or plural) of the compound.
For these compounds all that is necessary to accomplish an analysis is to
separate the members and add ca after each to express the notion of con-
junction; the standard formula is that already mentioned in Section 1.44:
ramas ca krsnas ca-rama-krsnau ("Rama and KrsI).a"). The same formula
applies no how many' the compound has: devtiS ca gand-
hantlii ca coraglis ca
("gods and fairies and humans and serpents and ogres").
Ragh.M. 1.1 ptirvatf ca paramesvaras ca piirvati-paramesvarau.
both PiirvatI and the Lord Parvati-and-the-Lord
Kum.M. 2.41 chedtif ca ptitjjs ca cheda-piitiis [cheda-piitiiniim]
both cuttings and failings, of them of cnllings-and-fallings
Kum. V 2.7 str[ ca pumtirrzs ca strf-pu'!'sau. acaturtiditviid ac.
Woman and man woman-and-man. (There is the suffix) aC (i.e., -a,
which replaces the final vowel and consonant of pumiin) because of
belonging to the list beginning with a-catura (in Pa. 5.4.77).
1.63 Samiihiira dvandvas
In this type of compounds ("dvandva of a combination," Pa.2.4.2; Whit-
ney 1253.2), the two members are thought of as combining to operate as a
unit, so the compound is always neuter singular (whereas the itaretara-
dvandva takes as its grammatical number the combined numbers of its
members). More often than not commentators analyze these compounds
using the simple formula applied to itaretara-dvandvas: e.g., hasta-padam
may be analyzed simply hastas ca padas ca ("hand and foot"). When
is felt desirable to indicate in the analysis the added notion of aggregation,
the techniques used are similar to those that are applied in the analysis of
collective dvigu compounds (Section 1.56.3), so that two options are avail-
able:
hastas ca pMas ca tayoJ.z samahtiraJ.z ("hand and foot, the combina-
tion [or aggregate] of the two")
hastas ca padas ca samahrtau ("hand and foot taken together")
Kum.M. 1.11 SrOf}ayas ca payodhartis ca irol}ipayodharam both
hips and breasts hips-and-breasts
Kum. Y." 2.5 cara'!l cacara'!! ca caracaram both moving and
unmoving moving-and-unmoving
E. Bahuvnni Compounds
1.64 Introduction
Compounds of the Bahuvrfhi category (PiiI)ini 2.2.23-2.2.28; "possessive
compounds," Whitney 1293-1308), are exocentric compounds referring
to something outside the compound itself that is described by the com-
pound. The distinctive feature of the formulas used to analyze
compounds is the relative pronoun that serves to express the relationship
through which the compound refers to that outside thing denoted by nei-
ther of the compound's members individually. The other relationship that
must be indicated in the analysis is that between the members of the com-
pound, but in this the compound is nearly always closely analogous to one
of the types of To take the example used in Section 1.44--
dfrghau kan:zau yasya sa dfrgha-kan:zaJ.z ("he who has long ears is Lang-
ear")-the relative pronoun yasya indicates that the thing referred ta by
the whole compound is masculine in gender and singular in number, and is
connected with long ears through a relationship expressed by the genitive
case, while the relationship between dfrgha and kan:za, as shown by the
placing of the two words in apposition in the analysis, is exactly parallel
to the relationship between the two members of a simple
dhtiraya (Section 1.54.1). In terms of the ways in which they are analyzed,
1.65 Appositional bahuvnnis
129
1.65. Appositional bahuvrIhis
(abl.) uddhrta adana yasyiilz sii uddhrtdudanii [sthalf]
("that (dish) from which the rice has been removed is removed-
riced")
genitive pronoun:
(dac) upahrtalz pasur yasmai sa upahrta-pasulz [rudralz]
("He [Rudra] for whom a beast has been sacrificed is sacrificed-
beasted")
(gen.) pftam ambara,!, yasya sa pftdmbarah [hari!l]
("He [Hari] of whom there is a yellow garment is yellow-garmented")
(loc.) vfriilz yasmin sa [griimalz]
("that [village] in which there are heroic men is hero-manned")
Note that except when the pronoun is in the genitive or locative the first
member of the compound is usually a past passive participle. The genitive
is most common, followed by the instrumental and locative; in bahuvrfhis
of most types other than this simple appositional type only the genitive
occurs.
t
"/
bahuvrfhi compounds can be divided into three basic groups: those that
,1
correspond in structure to kannadhiiraya compounds (Section 1.65), those'
that correspond in structure to dependent (Section 1.66), and
those special bahuvrfhis that cannot be directly compared in structure with
any type of (Section 1.67).
128
In these compounds "appositional pos-
sessive compounds"), the relationship between the members is parallel
to that between the members of a karmadhiiraya, so that in the analy-
sis the two members will usually be displayed in simple apposition (Sec- "II
tion 1.65.1), with modifications for bahuvrfhis that correspond to special
varieties of kannadhiiraya: negative compounds (Section 1.65.2), preposi-
tional compounds (Section 1.65.3), and a special form of the latter, com-
pounds with saha or sa as the first member (Section 1.65.4). The basic
principle of apposition applies also to bahuvrfhis analyzable as having
more than two basic members (Section 1.65.5).
1.65.1 Simple appositional bahuvrfhis
In the analysis of these compounds the two members are simply placed
in apposition, as in the analysis of kannadhiirayas of the adjective-plus-
substantive type (Section 1.54.1), or, more rarely, of the simple substantive-
plus-substantive type (Section 1.54.3). differences in the analysis arise not
in the treatment of the two members themselves, but in the case of
relative pronoun used to indicate the relationship thorugh which the entire
compound refers to some outside object; depending upon the precise nature
of this relationship the pronoun can appear in any of the six oblique cases
(in order to make the meaning of the compounds clearer in the following
examples, a word denoting the outside object to which the entire compound
refers is added in square brackets):
(ace.) priiptii atithayo ya,!, sa priiptdtithilz [griimalz]
("that [village] which guests have reached is reached-guested")
(instr.) Udho ratho yena sa iit.fha-rathalz [anadviin]
("that [ox] by which a cart has been pulled is pulled-carted")
Ragh.M. 1.15 sadrsa iirambha{! karma yasya sa tath6ktah
he who has commensurate undertaking-activity-is
so-called [commensurate-activitied]
Ragh.M. 1.55 siistram eva tattvavedakaf!l pramii1}QJ!l
yasya tasmai to him for whose eye-authority for
determining the truth-is policy-theory to the policy-eyed one
Kum.M. 1.53 sobhanii danta yasyiif.z sii su-dati she who has beautiful
teeth is beautiful-toothed
Kum.M. 2.20 ku!Z.thitii asrayo yasya tat kU{I!hitiiSri kU!Z!hita-ko!i. that
which has blunted COrners is blunt-cornered
instrumental pronoun:
Kum.M. 1.6 hato' dvipii gajii yais tesiim hata-dvipilnilm kesariniim
. . .. ..
si'J'lhiiniim of those by whom elephants-pachyderms-have been
killed of the killed-elephant ones
130
1.65. Appositional bahuvn1J.is 131
Kum.M. 1.1 [kampita-devadiiruM kampitii devadiiravo yena sa
tatht5kta(t. shaken-pine: he by whom the pines are shaken is so-called
Kum.M. 1.25 lahdha udayo yayii sa labdhodaya. utpanniity arthah
she by whom birth is obtained is obtained-birthed, Le., born
Kum.M. 2.22 apaviddha tyakta gada yena sail apaviddha-gadaJ.!. he
by whom the mace waS discarded-abandonect-is abandoned-maced
locative pronoun:
Ragh.M. 1.18 sahasra1J1 yasmin karmalJi tad yathii tathii
sahasra-gul}a'r' sahasradhti. in which action there are a thousand
repetitions, that; in such a way, a thousand-fold a thousand times over
1.65.2 Negative bahuvnnis
Negative bahuvrfhis (naii-bahuvrfhi compounds) are related to the nega-
tive (Section 1.58). The standard fonnula of analysis uses the
present participle avidyamiina:
avidyamiinii bhiiryii yasya so 'bhiiryaJ.! ("he is a-bhiirya [wifeless]
of whom a wife does not exist," "he is wifeless who does not have a
wife")
avidyamiinaJ.! putro yasya so 'putraJ.! ("he is a-putra [sonless] of
whom a son does not exist;' "he is sonless who has no son")
Other fonnulas of equivalent meaning are also used; thus a-putraJ.! ("son-
less") may also be analyzed putro nasti yasya or putro na bhavati yasya
("of whom there is no son").
Ragh.M. 1.30 Qvidyamiinam anyasya rtijiia!t siisanarrz yasyiis ttim
an-anya-sasaniim urvlm her for whom the rule of another king does not
exist the un-other-ruled earth
1.65.3 Priidi-bahuvnni compounds
Prepositional bahuvrfhis correspond to (Section 1.57.2),
and in their analysis the prefix that constitutes the first member of the com-
pound is replaced by a past passive participle with the prefix in question,
just as in the analysis of prepositional tatpurusas:
prapatital[l pa,,:,al[l yasya sa pra-pa,,:,aJ.! ("that is away-leaved whose
leaf has fallen away")
udgatii kandharii yasya sa ut-kandharaJ.! ("that is up-necked whose
neck is lifted up")
Ragh.M. 2.22 puraJ;t-phaliini purogatiini pratyasanniini phalii'!l
tiini hi. ahead-fruited those whose fruits are located ahead-in the
near future
Kum. V 4.1 vigato dhavaJ:z patir yasyiiJ:z sii vi-dhavii she whose
husband-spouse-is gone away-perished-is husbandless
Kum.M. 3.75 ucchirasatz unnata-siraso mahataJ:z high-headed
uplifted-headed-great
Note in the last example that the commentator simply inserted the expan-
sion into the compound without bothering to offer an analysis.
1.65.4 Saha-bahuvrihi compounds
Bahuvrfhis beginning with saha ("with") or its substitute sa- might be ex-
pected to be analyzed along the lines of with inseparable pre-
fixes as the first member (thus sa-putraJ.!, "with his son," might be ana-
lyzed sahitaJ.! putro yasya, "whose son is together [with him]"), and some
commentators use such a formula. But many such compounds are tech-
nically limited to instances in which both the thing referred to by the the
entire compound and the thing denoted by the second member are equally
involved in some action--e.g., sa-putro gataJ.!, "He went with his son"-
so that some commentators prefer the stricter and simpler formula saha
132
Chapter 8. Nominal Compounds 1.66. Non-appositional bahuvnllis 133
"together with his son." Although this analysis makes the com- ,1
pound look like a dependent such compounds are still classed .,
as bahuvrfhis.
In practice commentators using the saha formula usually employ it in
connection with a form of the verb vartate:
Ragh.M. 1.81 saha patnyii vartala iti sa-patnikaJ:l san. "nady-rtas Cil"
iti he is with his wife, thus being with-wifed. The suffix i
kaP (i.e., -ka, is added by pa. 5.4.153) "and (after a bahuvrfhi) ending in J
the feminine endings I or or in r"
Ragh.M. 2.23 sa-diirasya diirair arundhatyii saha vartamtinasya
with-wifed being with his wife ArundhatI
These compounds must be distinguished from those bahuvnllis in which .
sa- stands not for saha but for samana ("same," "shared") or sadria ("sim-
ilar"). Examples are "having the same social class" (samano
va00 yasya sa sa-va0aM and "having the same teacher" (lit.,
"having the same font").
1.65,5 Anekapada-bahuvrfhi compounds
Aneka-pada or "multiword" bahuvrfhis must be analyzed as consisting of
more than two members. Ordinarily compounds containing many words-
with the exception of certain dvandvas (see end of Section I .62)-can be
analyzed in the first instance as compounds of only two members, by com-
bining words in such a way that either or both of the two basic members
is itself a compound. But because of the numerous rules governing the
retention of feminine endings and the addition of suffixes in bahuvrfhi"
compounds, it is sometimes necessary to analyze an appositional bahu- .
vrfhi compound as being composed of three or more basic members. In .
the compound ("having a spotted old cow"), for example,
we cannot combine the first two words ("having a spotted-and-old cow")
because the form of the compound would then have to be
and we cannot combine the last two words ("having a spotted old-cow")
because the form would have to be citra-jarad- We are there-.
fore stuck with three members, but the resulting analysis is fortunately 7;
straightforward-the three are simply put in apposition: citra jarat! gaur ,"'
yasya, "of whom there is a spotted old cow." Similarly dfrghii tanvfjQlighii
yasya sa "he is Long-thin-shapk who has a long thin
shank."
1.66 Non-appositional bahuvnnis
Afew types of bahuvrfhi compounds are analyzed as or non-
appositional. For these few bahuvrfhis the precise way in which the first
member of the compound adds some specification to the second member
is most conveniently demonstrated by displaying one of the two members
in an oblique case in the analysis. This sort of analysis is used for two
particular types of compound:
1.66,1 Locative as second member
iaSf iekhare yasya sa ("he is moon-crested who has a
moon on his crest")
cakra,!, panau yasya sa ("he is discus-handed who has
a discus in his hand")
Example:
Kum.M.3.10 piniikaf! yasya sa . ... tasya
[piniika-piilJe1}j. in whose hand is (the bow) Pinaka, he is
Pinaka-handed.... Of him [of the Pinaka-handed one].
1.66.2 Genitive as first member
unrasyeva mukha1'[l yasya sa ("he is camel-faced who
has a face like a camel's")
candrasyeva kantir yasya sa candra-kiintiJ:z ("he is moon-splendored
who has splendor like the moon's")
In analyzing this last type of compound, many commentators prefer a fuller
fonnula:
134
1.67. Special bahuvnnis 135
mukham iva mukha1J'l yasya sa u:;.tra-mukhal} ("he is
camel-faced who has a face like a camel's face")
candrasya kiintir iva ktintir yasya sa ("he is
moon-splendored who has splendor like the moon's splendor")
Such compounds are considered another form of uttara-pada-Iopa-samiisa
(see end of Section 1.54.4), since the compound can be interpreted as
standing for candrakiinti-kiintiJ:!, etc.
Ragh.M. 1.8 muniniif[l vrttir iva vrttir ye:;ii'!l te:;ii'!l [muni-vrttfniim]
of those who have a life like the life of sage of the sage-lived ones
Ragh.M. 1.13 vf:;Qsya skandha iva skandho yasya sa tatka
... uttara-pada-lopf bahu-vrfhilJ he whose shoulder
is like the shoulder of a bull is so [bull-shouldered.... A bahuvrfhi with
loss of the latter member (of the first word).
Kum.M. 1.11 asviiniirrt mukhiinfva mukhiini yiisiirr: ttil}
samiisaJ:t. those who have faces like
the faces of horses are horse-faced Kinnara women. The compound is
like "camel-faced."
Kum.M.3.54 taru'.liirkasya raga iva raga yasya tat [tarulJiirka-riiga'!'l
biiliirkiiru'.lam ity arthaJ:t. upamiinapurvapado bahuvrfhir
uttarapadalopas ca. that which has redness like the redness of the
young sun is young-sun-rednessed, i.e., as pink as the young sun. A
bahuvrfhi with a standard of comparison as the first member, and with loss
of the latter member (of the first word).
Kum.M. 4.3 ivakrtir yasya tat pUTUfakrti that which
has a shape like the shape of a man is man-shaped
Ragh.M. 3.55 kumiirasya skandasya vikrama iva vikramo yasya sa
tathoktaJ:t [kumiira-vikramalJ]. He whose power is like the power of
Kumara-of Skandha-is so-called [Kumara-powered].
1.67 Special bahuvnnis
The standard technique of using a relative pronoun inan oblique case while
analyzing bahuvrfhis does not apply in three peculiar types of bahuvrfhis,
involving numerals (Section 1.67.1), intermediate points of the compass
(Section 1.67.2), and indeclinable descriptions of reciprocal combat (Sec-
tion 1.67.3). The other peculiarities observed in the analysis of bahuvrfhis
correspond for the most part to the irregularilies found in com-
pounds (Section 1.68).
1.67.1 Saf!lkhyii-bahuvrfhi compounds
Sarnkhyii-bahuvrfhis ("bahuvrfhi of numerals, Pal)ini 2.2.25) fall into three
main groups, for each of which the peculiar form of the standard analysis
is self-explanatory;
(I) dasiiniim iisanniiJ:!-iisanna-dasiiJ:! ("around ten")
dasiiniim aduriiJ:!-adura-dasiiJ:! ("not far from ten")
dasiiniim adhikiiJ:!-adhika-dasii!, ("more than ten")
dasiinii
rn
samipe ye vartante te upa-dasiiJ:! ("those which are in the
vicinity of ten are 'around ten''')
The first member of this sort of compound must be one of the four listed
here. Note that these are akin to genitive with reversed members
(with the last, "upa- ... " being like an avyayfbhiiva, and glossed in the
same way as upa-kumbha, Section 1.50).
(2) dvau vii trayo vii--<ivi-triih ("two or three")
trayo vii catviiro vii-tri-caturaJ:! ("three or four")
Note that these are akin to dvandva compounds. Commenlators sometimes
leave out one of the vii words;
Ragh.M.5.25 dvi-trii1,li dve [vii] trflJi vii. [either] two or
three.
(3) dviJ:! (iivrttii/:l! dasa--<ivi-dasiiJ:! ("ten [repeated]lwice," i.e., twenty)
These are like our compounds "two score," etc.
be the first member is a past passive participle; those compounds of this
type in which the reversal is optional are said by the commentators to be-
long to the ahitdgny-adi group (see the last example>below). Mutilation of
members (e.g., inprsodaradi compounds-see Section 1.58) also does not
affect the basic fonnula used in analyzing the compound, and the same is
trUe of the numerous suffixes added at the end of bahuvrfhi compounds.
136
Kum.M. 3.1 trir iivrtta dasa parimii1}am iti tridaSiin devan.
"sarikhyayavyayasanntldurildhika-sarikhyiiJ:l sankhyeye" iti bahuvrfhi.
their extent is ten repeated thrice, thus thirty gods. A bahuvrfhi by
(Pa. 2.2.25) "An indeclinable, iisanna, adara, adhika, and a numeral [is
compounded] with a numeral, in the meaning of a number."
;.69. Introduction 137
1.67.2 Dig.bahuvnlli compounds
Dig-bahuvrfhis ("bahuvrfhis of direction," PaI)ini 2.2.26) fonn names of
the intennediate points of the compass:
piirvasyas ca disor (yad) antaralam (sa)
(dik) ("[that direction is] southeastern [which is] the interval between
the southern and eastern directions")
purvasyiii ca disor antariilam-uttara-purva. ("the inter-
val between the North and the East-Northeast")
Ragh.M. 1.23 dharme ratir yasya tasya [dharma-rateM of him
whose love was for the law of the law-love one
Kum. V 2.18 yuga-prarnli[lli blihavo tebhyo yuga-biihubhya
to those whose arms are pole-length to the
pole-armed ones to those whose arms reach to their knees.
Kum.M. 3.72 bhavasya netrrlj janma yasya sa bhava-netra-janmii.
"avarjyo bahuvrfhir vyadhikara[lo janmady-uttarapadaJ:z" iti vamana!}.
He whose birth was from the eye of Bhava is Bhava-eye-birthed.
Vamana says that a nonappositional bahuvrfhi need not be avoided if the
last member is a word such as "birth."
1.67.3 Reciprocal combat
Bahuvrfhis describing reciprocal combat (PaI)ini 2.2.27) are indeclinable
compounds; in their analysis the members are displayed in the locative
when the word grhftva ("grasping") is added, and in the instrumental when
the word prahrtya ("striking") is added:
ca grhftveda,!, yuddham pravrttam-keidkeii ("this
battle proceeded grasping each other by the hair-hair-to-hair")
dar,zair dar,zais ca prah[1yeda,!, yuddha,!, vrttam-aar,zddar,zi ("this
battle occurred striking each other with staves-staff-against-staff") .
:J
Kum.M. 2.47 deha-baddham baddha-deham [i.e., baddha deho yena}.
rniirtimad ity arthaIJ iihitagny-iiditviin para-nipiitaJ:z.
body-assumed assumed-bodied [i.e., by whom a body has been assumed}.
The meaning is embodied. Irregular reversal of members because of
belonging to the group iihitfigny-iidi (iihitfigni, "established-fired," by
whom the ritual fire is maintained, is a compound that also appears in the
reversed form agny-iihita).
F. Longer Compounds
1.69 Introduction
Compounds containing more than two words can usually be viewed as
consisting of two members, one or both of which happens to be a com-
pound in itself. The principle exceptions are dvandva compounds, where
Irregular bahuvrfhis 1.68

These correspond in general to the types of irregularities found in the fonn,
;{
of on which the bahuvrfhi may be said to be based. Reversal i
of members is most common in bahuvrfhis in which what would nonnally!
138 Chapter 8. Nominal Compounds 1.70. General principles
139
the number of members in the compound does not fundamentally alter the
method of analysis (Section 1.62), and certain rare bahuvrfhi compounds
(Section 1.65.5). Longer compounds are analyzed by breaking them down
into their parts, and analyzing those parts as well, using essentially the
same techniques as in the analysis of simple compounds containing only
two words. The differences from the treatment of simpler compounds con-
sist primarily of techniques used to avoid the tedious repetition of parts of
the overall compound. The general principle is one of omitting as much as
possible of this repetition with the help of relative pronouns (Section 1.70),
and this principle is supplemented where useful by the use of simpler for-
mulas of analysis for some types of compounds or by other special tech-
niques (Section 1.71). Commentators often show individual preferences in
their choice of the available options (Section 1.72).
1.70 General principles

'0j
In analyzing a longer compound the standard formulas of analysis, if used $
in their full form for each of the compounds forming part of the overall
compound, would generate a tedious amount of repetition, and commenta-
tors often manage to omit much of this repetition, using relative pronouns
where necessary to preserve an indication of how the parts fit together. We
may take as a simple example a compound in Kumarasaf!lbhava 1.60, tac-
(i.e., meaning "the beams
of the moon on his (Siva's) head." While an English translation of longer
compounds such as this usually starts from the end of the compound, the
analysis in Sanskrit usually starts from the beginning; thus tat is the first
member of the genitive compound which in turn is li
the first member of the locative compound tacchiras-candrafi., !
which in turn is the first member of the entire genitive com- '?
pound A full analysis of the compound, using the .!
standard formulas for these types of compounds, would be as follows:
(1) rasya sirah: rac-chirah,
tacchirasi candrab: tacchirai-candra/:L,
tacchirafcandrasya tacchiraScandra-pildiiJ:t
his head: his-head,
the moon on his-head: his-head-moon,
the beams of the his-head-moon: his-head-moon-beams
Of course such an analysis is unnecessarily cumbersome, and scarcely
any commentator would be so tedious (but see the first example in Sec-
tion 1.72). The repetition of the compounds arrived at in each step of the
analysis can be avoided by using a pronoun to replace each compound
when marking its case as the first member of the next larger compound:
(2) tasya sirah: rac-chirah,
ta5min candraJ:t: tacchirai-candra/:l,
tasya piidii/:l: tacchiraicandra-piidiib
his head: his-head,
the moon on it: his-head-moon,
its beams: his-head-rnoon-beams
But even this single mention of the form of each intermediate compound
is more than most commentators find necessary. It can be omitted by rely-
ing on these pronouns to indicate the connection between the parts of the
entire compound. When this is done the only further necessity is to make
it clear what each such pronoun refers to, and this may be done by supply-
ing relative pronouns. If the part being referred to by the pronoun is itself
one of those types of compounds analyzed by a formula ending with a pro-
noun whose antecedent is clear--e.g., a bahuvrfhi ending with "... yasya
problem is automatically taken care of. In our example, each
of the constituent compounds is a in which the last member is
syntactically predominant; for these it is therefore possible simply to insert
a relative pronoun in apposition with the last member of each successive
part:
(3) tasya yac chiras tasmin yas candras tasya piidiif:i.
the beams of that moon which is on that head which is his
This process of simplification is carried to its extreme in the analysis of the
compound offered by the commentator Mallinatha:
(4) tasya girisasya sirasi candrasya piidiif:i.
140 Chapter 8. Nominal Compounds 1.71. Special techniques 141
the beams of the moon on his-Siva's-head
(For the inserted gloss girisasya see Section 1.21.) The only problem
with this simplest form of analysis is that, while it indicates the syntac-
tical connections between the individual words in the whole compound,
it leaves unclear the precise arrangement of these words into subordinate
compounds within the whole-in this instance, it may be unclear whether '
the final compound is composed of plus or of
plus piidiih. The analysis offered by the commentator
Caritravardhana avoids this problem by using a single relative pronoun:
(5) tasya sambhoJ:t iirasi ya candras tasya piMa/}
the beams of that moon which is on his-Siva's-head
This makes it clear that he divides the compound into
plus
Relative pronouns of this sort are frequently used to avoid what would
otherwise result in uncertainty or misapprehension:
Kum.M.1.16 [agra-saro-ruhii'}i] ogre upari yilni sartlf!"lsi ruha!}i
padmiini [highest-pond-growing:] growing in those ponds that are
highest-on top---lotuses. [Here the pronouns are necessary to prevent
the reader from taking saroruha as a compound.]
Ragh.M. 1.58 [iitithya-kriyii-siinta-ratha-Iqobha-parnramam]
iitithyasya [yo} kriyii raya siinto yo?! parisramal:z. sa yasya
sa ta'!' tathoktam [hospitality-rite-pacified-chariot-jolt-fatigue:] he is
so-called who has that fatigue, due to chariot-jolts, which is pacified by
that [which is] the rite of hospitality. [Hete the tayii after kriyii is to
prevent the reader from taking iitithyasya with anything other than kriyii.]
Ragh.M. 1.51 [iilaviililmbu-piiyiniim] iilaviil9u yad
ambu tat-piiyiniim [trough-water-drinkers:] drinkers of that water
which is in troughs-in watering places.
Ragh.M. 1.4 /q'ta-viig-dvare krta'!1 riimiiya'}iidiprabandha-rUpii yii viik
siiiva dviira'!1 pravdo yasya tasmin. in the one having a
made-speech-door: in that which has a door--entrance-made that is
nothing other than that which is speech, in the fonn of the Riimiiyafla, etc.
Kum.M. 1.47 [sva-ciipa-saundarya-madam) sva-ciipa-saundaryefla yo
tam [own-how-heauty-pride:] that prid9'which is due to his
own-bow-beauty
1.71 Special techniques
The sort of straightforward analysis just described works well for com-
pounds made up of a string of dependent but it can be inter
rupted when an element within the larger compound is one that requires a
more complicated formula. Commentators do what they can to streamline
the process of analysis, usually by employing a simpler formula of analysis
than that applied to compounds in isolation (Section 1.71.1), and if this is
not possible they will analyze a constituent compound separately from the
larger compound (Section 1.71.2).
1.71.1 Simpler formulas of analysis
By far the most common instance of simplified formulas is in the analysis
of karmadhiirayas forming part of a larger compound; in this situation they
are often analyzed simply by placing the members of the karmadhiiraya in
apposition. In Kumiirasa,!,bhava 4.29, for example, the bow of is
described as "having tender blossoms as its arrows,"
and each of the major commentators on the poem gives as the analysis of
this pelaviini puspiiny eva yasya tat, "that which has arrows that
are tender blossoms," even though in isolation the karmadhiiraya at the
beginning of the compound would always be analyzed pelaviini ca tiini
ca (Section 1.54.1), rather than pelaviini
1.71.2 Separate analysis
Often the formula required for the analysis of a constituent compound can-
not be fit smoothly in the analysis of the larger compound, and some part
of the larger compound must then be analyzed separately, before begin-
ning the continuous analysis of the whole. In Kumiirasa,!,bhava .1.11,
the kinnara women are described as "tor-
mented by their burdensome hips and breasts," where
142 Chapter 8. Nominal Compounds
1.72. Choice of options
143
is a sanu'ihara-dvandva compound (Section 1.63). The commentators first
analyze this part separately, and then treat it as a unit in the analysis of the
larger compound:
Kum.M. 1.11 sro1,1ayas ca payodharas ca srOf.li-payodharam; durvahe,a
durdhare1)Q pfiti1{l hips and breasts:
hips-and-breasts; tormented-oppressed-by
burdensome-heavy-hips-and-breasts
Notice that the first member of this whole compound, durvaha ("bur-
densome, difficult to carry") is itself a compound, and could be given an
analysis of its own--e.g., krcchrel}ohyata iti durvaham, "it is carried with
difficulty, thus difficult-to-carry"-which would then be converted to the
instrumental case for the analysis of the rest of the larger compound by
adding tena after its analysis. But commentators very frequently choose
not to bother with the analysis of items whose construction is obvious, and
will often simply provide a synonym instead, as Mallinatha has done here
(durdharena).
1.72 Choice of options
The analyses of any longer compound offered by different commentators
may differ, aside from actual differences of opinion on the grammatical
construction of the compound, depending on the frequency with which
each commentator chooses not to analyze constituent parts of the whole or
to insert explanatory glosses, and also depending on the extent to which
each commentator chooses to use the inserted relative pronouns described
in Section 1.70. A simple example is Kumiirasa'!'bhnva 1.10, where phos-
phorescent plants are described as
"having light clinging in the interior of cave-houses." An anony-
mous commentary (inserted in a conflated version of Vallabhadeva's com-
mentary, BORI Devanagatf MS No. 379 of 1887-91) analyzes this com-
pound in the most tedious way possible, with no inserted glosses or ex-
planatory material:
ki'!l-visi:'l!ii darf-grhotsanga-n4akta-bhiisal}. darya eva
grhaf!l darf-griulm. utsarigo darf-grh6tsarigaft.
darf-grh6tsange bhtiso yasti'!! tlib. Vlhat sort of plants? Having
cave-house-interior-clinging-light. Houses that are actually caves:
cave-houses. The interior of the cave-houses: Whose
light is clinging in the cave-house-interior, those.
The major commentators use much more streamlined methods of anal-
ysis, with a judicious number of synonyms thrown in:
Kum.C. 1.10 krdrfya
darya eva guhii eva grhiis utsarige bhiiso yastif!l
tiiJ:z. What sort of plants? Having
Those that had light clinging-penetrating-in the interior of those houses
that were actually caves-actually grottoes.
Kum.M. 1.10 daryaJ:z kandarii eva grhiis
sa'!'krtintii bhiiso yiisii'!' tiiJ:z. Those that had light
clinging-reflected-in the interiors-insides---of those houses that were
actually caves-grottoes.
Kum. V 1.10 daryaJ:z kandarti eva grhiis tad-utsarige lagnii
bhiiso yiisii'!' tiif:. Those that had light clinging-adhering-in the
interior of those houses that were actually caves-grottoes.
Here the three commentators differ somewhat in the choice of syn-
onyms they offer, but in the actual analysis of the compound the commen-
tator Caritravardhana differs from Mallinatha only in interpreting utsanga
as singular rather than plural. Vallabhadeva also prefers the singular here,
and deviates in a minor way from the other two commentators in using the
more compressed form tad-utsange in place of utsange.
The style of each commentator is often marked by consistent prefer-
ences in these matters. For example, a preference for singular or plural,
wherever there is a choice, characteristically separates Vallabhadeva from
Mallinatha throughout their commentaries: when they differ on this point,
it is consistently Vallabhadeva who gives the singular and' Mallinatha who
gives the plural.
To illustrate further these choices and to introduce several other as-
pects of the analysis of long compounds, here are three long compounds
with analyses offered by these two commentators. In the quotations from
144 Chapter 8. Nominal Compounds 1.72. Choice of options
145
the commentaries, all inserted explanatory material and glosses have been
deleted, so as not to distract attention from the analyses themselves. No-
tice that in the analysis of longer compounds such as these, Vallabhadeva
differs from Mallinatha in another way as well: Vallabhadeva consistently
uses the technique of inserting relative pronouns with following correla_
tives, while Mallinatha consistently avoids the use of pronouns wherever
they are not absolutely necessary.
First, a long compound (Sisupiilavadha 4.20): vilambi-gharqii_J
"the grace of a mighty elephant girded
by a hanging pair of bells." Vallabhadeva analyzes this as vilambi yad
tena pariviirito yo tasya lflii, "the grace of
that elephant-lord which is girded by that bell-pair which is hanging." The
analysis that Mallinatha's gives is vilambinii pariviiritasya
lflii, "the grace of an elephant-lord girded by a hanging bell-
pair." Note that neither commentator bothers to analyze or
viirm}endra.
Second, a long bahuvrfhi compound (Sisupiilavadha 3.80): uttiila-tiilf-
vana-sQrrzpravrtta-samfra-Sfmantita-ketakfkiib. "(regions) where the screw
pines are parted by the breeze moving forth from the tall grove(s) of palms."
Vallabhadeva's analysis is uttiila,!, yat tiilf-vana,!, tasmiit sa'!'pravrtto
samfras tena ketakyo yatra te "Those (regions) are
so-called in which (there are) screw pines parted by that breeze which is
moving forth from that palm-grove which is tall." Mallinatha's analysis
is sQfJ1/Jravrttena samfrera sfmantitii!z ketakyo te
"Those (regions) are so-called in which (there are) screw pines
parted by the breeze moving forth in tall palm-groves." Again, neither
commentator bothers to analyze tiilf-vana. Notice also that while Vallab-
hadeva interprets vana as ablative singular, Mallinatha presents it as loca-
tive plural-an instance of a characteristic difference already described.
The formula te ("they are so-called"), and the simpler te tatM
("they are thus," see the next example), are used very frequently at the end
of an analysis requiring a pronoun to restitute the whole compound after
the use of a relative pronoun (as in the analysis of bahuvrfhi compounds)
or iti (as in the analysis of some upapada compounds). These formulas
are designed to avoid the necessity of repeating the whole compound at
the end, and are used even with simple two-member compounds. Thus the
bahuvrfhi compound dfrgha-karna ("Long-ear") can be analyzed dfrghau
kan,zau yasya sa dfrgha-karnah, or dfrghau yasya sa tathOktah, or
simply dfrghau yasya sa tatM. .
Finally, an example of a long upapada compoundCSisupiilavadha 5.12):
vanyebha-mastaka-nikMta-nakhdgra-mukta-muktiiphala_prakara-bhafiji,
"(cave-houses) possessing [-bhafiji, neut. pI. of -bhaj] heaps of pearls re-
leased by claw-tips that had been dug into the heads of wild elephants."
Vallabha's analysis is vanyebha-mastake1u nikhatiini yiini naklulgriini lair
muktiini yiini muktiiphaliini te1ii,!, prakariis tiin bhajantiti tatM tiini, :'those
(cave-houses are) so, as they possess those heaps of those pearls which
have been released by those claw-tips dug into wild-elephant-heads." Malli-
natha's analysis is vanyebMnii,!, mastake1u nikMtair nakhdgrair muktiin
muktiiphala-prakariin bhajantiti tatMkliini, "those (cave-houses are) so-
called, as they possess pearl-heaps released by claw-tips dug into the heads
of wild-elephants." Here Vallabha does not analyze vanyebha-mastaka,
Mallinatha does not analyze muktiiphala-prakara, and neither analyzes
vanyebha or nakhdgra.
Chapter 9
Formations
1.73 Introduction
("remainder of one") formations (PiLI.2.64-1.2.73) are those in
which one noun is used in the dual or plural to stand for more than one
noun of identical form but different signification (e.g., "Smiths" for "[Bob]
Smith and [Tom] Smith"), one noun (the masculine form) is used to stand
for both male and female members of a pair (Section 1.74), or a neuter
adjective is used to modify both a neuter noun and a non-neuter noun (Sec-
tion 1.75). For every kind of formation the formula of analysis
is exactly parallel to that of a simple itaretara dvandva compound (Sfc-
tion 1.62).
1.74
Here the name of the male of a pair stands for both the male and the female
(e.g., "dogs" for "dogs and bitches").
Ragh.M. 1.1 miitii ca pita ca pitarau mother and father-parents.
Ragh.M. 1.40 mrgya ca mrgiis ca mrga!, does andhucks-deer
Kum.M. 1.27 putrtis ca duhitaras ca putriilJ sons and
daughters--<:hildren
147
148 Chapter 9. Formations
1.75
Here a neuter adjective applies simultaneously to a neuter noun and a mas-
culine or feminine noun.
Kum.M. 1.2 bhtisvanti ca [ratntini] bhtisvatya ca
bhiisvanti dyutimanti shining [jewels (neuter)] and shining [herbs
(feminine)]-shining (neuter), radiant.
Chapter 10
Indicating the Construction
A. Arranging the Words of the Text
1.76 Introduction
Sanskrit commentaries are built up on a framework consisting of a rear-
rangement of the words of the text being commented on. The original text
is called the mula ("root"). In arranging the words of the mula the com-
mentators use either of two basic approaches, representing radically dift;er-
ent ways of indicating the construction of the text: the katha,!,bhutinf ap-
proach (Section 1.77), and the anvayamukh! approach (Section 1. 78). The
explanatory material offered by the commentator is superimposed upon the
framework made up by the repeated words of the mula, using techniques
that are the same in both approaches (Section 1.79). Several variations on
the two approaches are possible (Section 1.80).
1.77 The kathaTflbhiitini approach
In this approach the commentator begins by giving the basiC "skeleton sen-
lence" of the original text, and then gradually fills in each of the remaining
parts of the mula by asking questions that bring out the ways in which
those parts fit into the construction of the original text. A commentary us-
ing this approach is called a kathaf!lbhutin, CJ,kti) because one of the most
149
150
Chapter 10. Indicating the Construction
1.79. Insertion of explanatory material 151
frequent questions used in this style is "katharrz-bhUta?" ("of what na-
ture?", "being how?", "what kind of ... ?"), or its equivalent
("distinguished by what?"). To illustrate this approach we may use as the
mula verse 2.1 of the Kumarasarrzbhava of Kalidasa:
tasmin viprakrtti kale tlirakel}Q dil'flukasab
turtistiha'!' pltrodhaya dhlima svayafJ1bhuvam
At that time, being oppressed by Tar-aka, the gods (lit.,
"heaven-dwellers"), placing (Indra) before them, went to the
abode of the Self-born (Brahmil).
An anonymous commentary on the poem (inserted in the conflated recen-
sion of Vallabhadeva's commentary in the BORl Devanagari MS 379 of
1887-91), using the katharrzbhutinfapproach, arranges these words as fol-
lows:
divaulwsaJ:! sviiya1Jlbhuvaf!l dhtima ... yayuIJ kif!l-visiHiiJ:z?
viprakrtiiJ:z kena? tiirake1}Q. kva? tasmin kale .... ki1'[l krrv
ii
?
purodhiiya kam? turiisiiham ....
The gods went to the abode of Svayambhii. How characterized?
Oppressed. By whom? By Taraka. At what? At that time. Doing what?
Placing before them. Whom?
Notice that the "skeleton sentence" need not be restricted to just the subject
and verb. The commentator will generally include at least the direct object,
and may also include items that are only loosely connected with the basic
sentence (e.g., interjections and vocatives), as well as any simple items that
can be dealt with easily and unobtrusively.
1.78 The anvayamukhl: approach
In this approach the commentator goes straight through all the words of tbe
mula in one long string, merely rearranging them so as to place them in the
most easily understandable prose order. In practice this usually results in
some version of a "standard prose word-order," the essentials of which are
recorded in a verse in the Samasacakra, an elementary handbook used at
the beginning stage in the traditional system of Sanskrit education:
puraskrtyQ tad-anantaram
kart!-kanna-kriya-yuktQ'!I ewd -
The word order is characterized by subject, object, and verb, placing
modifiers before what they modify.
In addition, vocatives usually appear first in such an arrangement, and rela-
tive clauses precede correlative clauses. A commentary using this approach
is called an anvayamukhr (tfka) because it arranges the words "by means of
the normal construction" (anvaya-mukhena). For example, the commenta-
tor Mallinatha uses this approach to arrange the words of the verse quoted
above as follows:
tasmin kale . .. tiirake[lQ ... ... divaukasal.z ... turtistiham ...
purodhiiya . .. sViiya'!1bhuvam... dhiima,.. yayui.z.
At that time, by Taraka oppressed, the gods, placing before, to
Svayarrbhl1's abode went.
1.79 Insertion of explanatory material
The basic arrangement of the words of the mula, arrived at by either of
the two approaches just described, serves as a framework upon which \:he
commentator superimposes a mass of explanatory material. The glosses,
analyses, and other comments are simply placed alongside the words of
the mula in their appropriate locations, using the techniques explained in
preceding sections. The resulting interspersions often obscure the basic
framework made up of the words repeated from the mula. To illustrate this
problem, here is the actual beginning of Mallinatha's commentary on the
verse we have been using as an example (the words repeated from the mula
are in bold type here):
tasminn iti. tasmin kale tiirake'.'Q tiiraka-niimnii
vajrar:tkha-putrerJa kenacid asureJ:lQ viprak[tii upaplutii divam okal.z
sthiinarrt te divaukaso devatz ...
"At that," (etc.)---At that time at the time of Parvatl's service by Taraka
by a certain demon named Taraka, the son of VajraI).akha oppressed
152
Chapter 10. Indicating the Construction
1.79. Insertion of explanatory material 153
distressed, of whom heaven is the abode-place-those heaven-dwellers
gods ....
In the Sanskrit the disjoining effect of all this inserted matter is lessened
somewhat by the connective powers of inflectional endings, and in giving
this example we have helped matters further by underlining the words of
the mtila. In many printed editions of Sanskrit commentaries, however, the
customary lack of typographical emphasis and detailed punctuation makes
the basic framework blend in with the mass of comments, so that the obsta-
cles to an awareness of the underlying framework are intensified. Despite
these difficulties, it is essential to remember that the words of the mala
at least readily discernible substitutes for them) have been arranged in a
particular way by the commentator. If the commentary is viewed merely
as a collection of unconnected clusters of explanation, the value of the un-
derlying framework as a guide to the construction of the mala will be lost.
As a practical device it is often helpful to seek out the words repeated from
the mula and underline them, if the edition being used does not already set
them apart in some way.
The explanatory portions of a commentary are inserted in precisely the
same way in a commentary using the kathaTf'lbhutini' approach. For ex-
ample, the anonymous commentator on KumiirasaTf'lbhava 2.1 cited above
fills out his basic framework as follows (note that in either approach the
commentator first identifies the passage about to be commented upon by
citing the prati'ka, the first word or two, of the passage in question):
tasrninn iti. divaukaso deviiJ:i. sviiya'!'bhuva'!' dhiima sthii.na'!l yayur
gatti/}. kifJ'l-visiHiiJ:i.? viprakrtii apakrttiJ;. kena? tiirakel}a. kva? tasmin
kale fsvara-tapo-vidhiina-kiile. kif[l krtvti? purodhiiya agre krtvti. kam?
turiisiiham indram.
"At that," (etc.j-The heaven-dwellers gods to Svaya'!'bhii's abode
place went had gone. How characterized? Oppressed mistreated. By
whom? By Taraka. At what? At that time at the time of the Lord's
performance of austerities. Doing what? Placing before putting in front.
Whom? Indra.
This is followed in the anonymous commentary by a mass of grammat-
ical argument, including the citation of rules of Pal)ini and the analysiS
of the complex formations sviiyaTf'lbhuvam, and turiisiiham.
While most commentators place analyses in ordinary language alongside
the formations being analyzed, comments using technical language and ar-
ouments on technical matters are treated like quotations from dictionaries
c
(see Section 1.13): some commentators insert them in the relevant place
in the running commentary, and others bunch them together at the end of
the commentary on the verse or section of the original text. Comments
on technical matters concerning the verse or section as a whole (e.g., the
explanation of the meter in which the verse is written or of a poetic figure
involving the whole verse) are nearly always placed at the end, unless they
are worked into the introduction to the comment on that verse
or section.
Bunching the technical comments together at the end of the commen-
tary on a passage has the effect of reducing the interruption of the flow
of the running commentary. Even when this is done, the indication of the
construction of the mula can still be considerably obscured by the insertion
of explanatory material, and this problem is more acute in an anvayamukhi'
commentary than in a kathaTf'lbhutinf commentary, because a commentary
using the kathaTf'lbhiitini' approach sets forth the construction of the original
text in a much more obvious way to begin with. While the anvayamukhf
approach thus produces a commentary that is more difficult to read until
one has had some experience with its characteristic procedures, it is impor-
tant to learn how to deal with its intricacies, because in practice most cdm-
mentators rely primarily on this style in preference to the kathaTf'lbhutini'
approach, which, in addition to being rather more tedious, came to be re-
garded as old-fashioned.
As a further example to help make clear the essential difference be-
tween the two approaches and the ways in which explanatory material may
be inserted in the basic framework, we may take as our mUla the opening
verse of the Sisupiilavadha of Magha:
sriya!} pati!} srfmati siisitum jagaj
jagan-niviiso vasudeva-sadmani
vasan dadarSiivatarantam ambariid
dhira1Jya-garbhdnga-bhuvaf!l munif!l hari!}
Hari, the husband of Sri, the abode of the world, dwelling in the glorious
house of Vasudeva in order to govern the world, saw descending from the
The earliest of the commentators on the great Sanskrit ki'ivyas whose works
have survived, Vallabhadeva, comments on this using the old katha,!,bhatinf
approach. His remarks begin as follows (the numbers in square brackets
indicate the sections of this handbook in which you will find explanations f
of some ofthe devices being used here):
The same verse is subjected to the anvayamukhf treatment by the more
recent commentator Mallinatha, who begins as follows:
juncture of the Kali and Dvapara ages. But the Lord is the ancient sage
who resides in the asrama at BadarI, so why would;he dwell there? He
tells us-in order to do what? In order to govern the world in order to
protect, to rescue, the universe. This (statement) refutes (the notion) that
the Lord is dependent upon actions. What sort of Hari? The abode of the
world the abode, receptacle, of the worlds. They abide in him, thus (he is)
an abode. What sort (of sage)? Descending from the sky flying down
from the heavens-this modifies the sage. The word anga-bhu (lit.,
"arising from the body") is metaphorically applied to a son as
well, just as the word sarasi-ja (lit., "born in water") is applied even to a
lotus growing on dry land....
155 1.79. Insertion of explanatory material
sriya iti- ... sriyo lalqmya patilJ. anena rukmil}f-rupayii sriyii sameta iti
siicitam [Section 1.13.5J, "riighavatve 'bhavat sftii rukmil:zf
iti jagatiim iidhiira-bhuta1:z
[Section 1.84.3]. iti yiivat [Section 1.I2.3J.
tathapijagallokaf[l iiisitu'!l niyantufJl
inmati lalqmfyukte vasudeva-sadmani vasudeva-riipi1}a1:z kasyapasya '
vesmani vasan hanr ambariid avatarantam.
indra-sandesa-kathanartham iti bhiiva1:z [Section 1.13.2]. hira1}yasya
garbho hira1}ya-garbho brahmal)a-prabhavatviit, tasyanga-bhuvaf[l
tana-jam. athavii tasyangad avayaviid utsangakhyiid bhavatfti
[Section 1.32.2, Section 1.59] ta,!,
[hiraf,lya-garbhiinga-bhuvaf!lJ munim. naradam ity artha1:z. "utsangan
niirado jajiie dalqo svayaf[lbhuva1:z" iti bhagaviit. dadarsa.
kadiicid iti [Section 1.83]. ...
"Of Sri," (etc.)-... The husband of Sri of Lak!mi. By this it is
suggested that he was with Sri in the fonn of RukmiI).l, because the
says, "She was Sita when he was Raghava, and RukmiI).I in
his birth as KJe9a." The abode of the world the receptade of the
universe. To be precise, the one in whose belly the entire universe is
situated. Nevertheless in order to govern the world to control the w.orld
by restraining the evil and favoring the good in the glorious endowed
with prosperity house of Vasudeva home of Kasyapa in the fonn of
Chapter 10. 1ndicating the Construction
sky the sage (Narada) who is the son of Hira9yagarbha.
sriya iti. harir muni,!, dadarsa itl sambandhafi. [Section 1.82.2]. nanu
[Section 2.49] harayo dasa sa,!,khyiitii
vartante-" VQsava- vtiyava/:l /
turaftga-si'!lha-sftii'!lsu-yamiis ca haraya data." leo 'SQU hanT ity
liha-sriyal} ity arthii/:t [Section 1.13.1]. munirrt
kam ity tiha-hiraJ)ya-garbMtiga-bhuva,!, brahma-tanayaf!l naradarn ity
arthaJ:l. yady api brahma-[tanaya-]tvencisau radii hari!Zti rna] jiiiiyi, yena
kenacic chabdena nirdeie kartavye muni-sabdaJ:! [Section 1.6.2]
prayuktal; kavinii jfiiita eva, tasya cditad VQcanam. kva
vasudeva-sadmani surasena-putra-grhe inmati sarva-samrddhi-yukte
vasan kali-dviipara-sandhiiv ity arthaJ:t. nanu bhagaviin
puriitanamunir badary-iisrama-niviisf kim-iti tatr6vasety iiha-kif!l
kartuf!ljagac chasitu'!' tri-bhuvanaf!l ral0ituf!l piilayitum. etena
karmiidhfnatvaf!l bhagavato nirastam. kfdrso hariJ:tjagan-niviisalJ
bhuvaniiniif!l nivasaJ:t adharaJ:t. nivasanty asminn iti nivasaJ:t
[Section 1.29.6]. kfdrsam ambarad akasad avatarantaf!l patantam, muner
[Section 1.82.3]. angabhasabda upacariin miinase 'pi putre
vartate. yatha sarasija-sabdaJ:t sthalakamale 'pi. ...
"Of Sn," (etc.)--The construction is "Hari saw the sage." But there are
ten Haris who are enumerated-'The sun, a monkey, a frog, Indra,
the wind, a horse, a lion, the moon, and Yama are the ten Haris." Which
one is this? He tells us-the husband of Sri the spouse of i.e.,
Vi!9U. What sage? He tells us-the son of Hiral}ya-garbha the offspring
of Brahma, i.e., Narada. Although at that time Hari did not recognize him
as N1:irada, since he had to be designated with some word or other the
word 'sage' is used, which the poet knew him to be, and this is his (i.e.,
the poet's) utterance. When did he see him? Where? Dwelling situated in
the house of Vasudeva in the home of the son of Silrasena, (which was)
glorious endowed with all prosperity. What is meant is (that it was) at the
154
156 Chapter 10. Indicating the Construction 1.80. Variations on the two approaches 157
Vasudeva dwelling residing in the form of Hari descending
from the sky-the idea is that (he was coming) for the purpose of
delivering a message from Indra. The fetus (garbha) of the golden one
Brahma, because he was produced from the
golden egg. His son, offspring [the son of Hira'.'yagarbha]. Or-arising
from an ariga-a part of the body, namely the lap-of him, born from a
part of the body of HiraJ).yagarbha. That sage, i.e., Narada. Because the
Bhagavatapurfu:Ia says, "Narada was born from the lap of the Seltborn,
and from the thumb," He saw. Supply "at a certain time." ...
Here the words of the mula, shown in bold type in the text and translation,
are presented in the nonnal prose word order of Sanskrit, although the con-
tinuity of this word order is impossible to convey in the English translation.
Notice that Mallinatha first analyzes hirOl}ya-garbhiiliga-bhuvam as a gen-
itive tatpuruea, and then decides that it is an upapada compound. When
a commentator gives more than one possibility in such instances, the last
alternative mentioned is ordinarily the one that he prefers.
1.80 Variations on the two approaches
Commentators frequently use one of three variations on the basic ways of
arranging the words of the mula.
The first variation involves a combination of the two basic approaches.
Quite often a commentator will use something akin to the kathalJ'lbhutinf
approach to break the mula into two or more large parts, but will use the
anvayamukhf method in dealing with each of these parts.
In the second variation, which occurs most often in conjunction with
the anvayamukhf approach, the commentator arranges the whole commen-
tary in standard fashion, but omits the repetition of the words in the mula.
When this is done the glosses of the omitted words appear where they
would be if the words of the original text were repeated in the commentary-
to put it another way, the gloss often takes the place of the repeated word
from the mula in giving the basic construction of the original text.
In the third variation, the commentator begins by giving the bare an-
vaya of the original-i.e., the arrangement of the words of the mula in the
order that fonns the framework of an anvayamukhf commentary-without
inserting any explanatory material. Once this has been done, the commen-
tator begins again and gives a full commentary using any of the approaches
or variations mentioned above.
As examples of the approaches and variations just described, here are
seven actual commentaries on a single text. The mula is verse 4.3 of the
Klllniirasar[lbhava:
ayi jfvita-niitha jfvasfry abhidhay6tthitaya taya puraJ:1
dadrse kfitau hara-kopiinala-bhasma kevalam
[When the love-god Kama was burned to death by the fire from the third
eye of Siva, Kama's wife RatI fainted. Now she has regained
consciousness, and-]
"Ah, lord of my life, are you alive?" Speaking thus, by her, rising up, was
seen before her on the ground only the ashes from the fire of Siva's anger,
in the form of a man.
(I) Example of a kathalJ'lbhutinf commentary (once again, the anony-
mous commentary inserted in the conflated recension of Vallabhadeva's
commentary):
ayfti. tayii ratyti kama hara-kopanala-bhasma dadre drf!am.
PUTufdkiiram. kasytirn? prthivyiim.
katham? puro 'gratal:z. ki'!l-viiif!ayii ratyii? utthitayii. ki'J'l krtvii? ity
abhidhaya uktvti. iti ifi kim? he]Tvita-natha tva'!lJ7vasi ki'J'l na? ayi iti
komalamantrafJe.
"ayi," (etc.)-By her by RatI, Kama, the ash of the fire of Siva's anger,
was seen was seen. Characterized by what? Having the form of a man
having the shape of a man. On what? On the ground on the earth. How?
Before in front. By Rati, being characterized by what? Rising up. Doing
what? Thus speaking uttering. What is "thus"? Oh lord of my life you
are living, aren't you? "Ah" (is used) in tender address.
(2) Example of an anvayamukhfcommentary (again, Mallinatha's SalJ'l-
jfvini):
ayfti. ayi iti "ayi iti
jfvasi kaccid ity abhidhiiyotthitayii tayii ratyti puro 'gre k#tau
158
Chapter 10. 1ndicating the Construction
1.80. Variations on the two approaches 159
iviikrtir yasya tat kevalam ekaf(l
dadrse na tu iti
"ayi," (etc.)-"Ah" (is used) in the sense of a question: "ayi in (the senses
of) interrogation and conciliation"-Visva. Ah lord of my life, you are
alive? You are living, I hope? Speaking thus, rising up, by her by RatI
before in front on the ground that whose form is like the form of man in
the form of a man, only alone the ash of the fire of Siva'8 anger was
seen was seen. But not a man-this is the idea.
(3) Another example of an anvayamukhf commentary, this time with
a bit less explanatory material added (the Dhfraraiijika of Rama Govinda
SiddhfultavagIsa):
ayijrviteti. ayi iti sambodhane. hejlvita-niitha tva'!ljivasity abhidhiiya
uktvti utthitayii tayii puro 'gre sQ'!'sthtinaTJ1
yasya tiidrsafJl harawkopanala-bhasma kevalaf!l dad[se ity
"ayijfvata-," (etc.)-ayi (is used) in addtess. Oh lord of my life you are
alive? Thus speaking uttering rising up, by her before in front in the
form of a man of which the form, the configuration, is like a man's, of
that sort only the ash from the fire of Siva's anger was seen-this is the
construction.
Notice that this commentator ignores the word in the mula.
(4) Example of a commentary using a mixture of the katha,!,bhutinf
approach and the anvayamukhfapproach (Caritravardhana's
ayfti. tayii ratyii puro 'gre akliro yasya tat.
hara-kopanala-bhasma harasya kopiigner bhasma kevala,!, dad[se
na tu kiinta iti kifJlbhataya taya? ayi iti komaliimantra':le
bho jfvita-niitha priiIJa-niithajfvasfty abhidhaya ukIVii utthitayii.
"ayi," (etc.)-By her by Rat! before in front in the form of a man that
whose form, shape, is like a man's. The ash of the fire of Siva's anger
the ash of the anger-fire of Siva only was seen was seen. Supply "but not
her beloved." By her, being of what sort? Ah used in tender address, Oh
lord of my life lord of my life are you alive? Thus speaking uttering
rising up.
(5) Example of an anvayamukhf commentary with omission of many
of the words of the mula (Vallabhadeva's Paiijika, with the omitted words
supplied in square brackets): >
ayiti. [ayi] kaccit [jlvita-niitha] pranesvarajfvasfty [abhidhiiya] ukIVa
utthitayii tayii kevala,!,
{hara-koptinala-bhasma] dadre na tu
priyafz. matsyasyeva dagdhasya tasyiikaro na
"ayi," (etc.)-[Ah] I hope [lord of my life] lord of my life yon are living?
Thus [speaking] uttering rising up, by her only [in the form of a man]
in the shape of a man [the ash from the fire of Siva's anger] the ash from
the fire of Siva's anger was seen was seen. But not her beloved. When he
was burned, his shape, like that of a fish, was not destroyed.
(6) Example of a sophisticated commentary that offers explanations of
only a few of the words in the mula (the PrakiiSika of
Note that reads [feminine, and thus a genitive
rather than [neuter, and thus a bahuvrfhi modifying
-bhasma]:
avadhtinasya kiiryam aha-ayiti. "ayi prasne sanunaya" iti halayudhafz.
iti pa.thafl. hara-kopilnala-bhasmeti.
hara-kopdnale bhasmfbhatti. kevalam iti, na kaScid apy abhasmf-bhutas
tad-avayavas tat-paricchado vety arthaJ:t. etad vibhav6dbodha-para'!l
vlikyam.
He tells us the effect of her attentiveness-"ayi," (etc.). "ayi in
conciliatory inquiry"-Hahiyudha. The (correct) reading is "the form of
a man." ash of the fire of Siva's anger" modifies this.
"Entirely"-the idea is that not a single limb of his, and not a single
garment, had not been reduced to ashes. This is a statement dealing with
the awakening of the factors (connected with the mood of compassion).
(7) Example of a commentary giving the bare anvaya without comment
before proceeding to a full commentary (the VivaraIJa of NaraYaJ)apaJ)dita,
who closely follows AruJ.lagirinatha both in his readings and in his com-
ments):
160 Chapter 10. Indicating the Construction
1.81. Introduction
161
vilocanayor phalam aha-ayitf.
jrvita-natha tva'!! jfvasy ayity abhidhaya, utthitayti taya puraJ:z.
dadrse.
jivita-niitha priil}esvara tvaJ!l jrvasi. ayi iti
avyayam. "ayi siinunaye praine" iti halayudhaJ;.. jfvasi kaccid fty arthaJ:t.
ity abhidhiiya uktvii utthitayii krtotthiinayii. puruplkrti-mtltra-darsanalJ'l
hetuJ:z.. ata eva sa-tvaram utthiinam. tayii ratyii [pura1}l
agre bhamau akiiro dadrse
ayam abhipriiyaJ:z.-hara-kopiinala-ni.Jpatananantaram eva moham
upagatii ratir priyaarfrasQJ!lsthiina'!l
purvavad evavalokya satvaram utthilya tat-samfpam upagamya purastiid
eva dadarstti.
kfdrSf punaJ:z sti ity tiha-kevalalfl hara.kopiinala-bhasma
iti. kevalafJ1 vti na kaseid afJ1sas
tatrdbhasmastid-bhuto 'bhud iti bhtivaJ:z. eva keeit
parhanti. tan-mate hara-kopdnala-bhasma dadrsa ity
anena slokena karuJ:lasya vibhtivodbodho
He tells us the result of her making her eyes attentive-"ayi," (etc.).
Lord of my life, you are living, aren't you? Thus speaking, by her, rising
up, before her on the ground a man's form was seen.
Lord of my life lord of my life, you are living? ayi is an indeclinable
used in the context of conciliatory inquiry. "ayi in conciliatory
inquiry"-HaHiyudha. The meaning is "You are alive, I hope?" Thus
speaking uttering rising up performing the action of rising up. That she
sees only the form of a man is the cause for her doubt concerning whether
he is alive or not. Hence her rising up quickly. By her by Rat! [before] in
front on the ground on the earth a man's form the form, shape, of a man
was seen was seen.
What is intended is this-immediately after the fire of Siva's anger had
gone forth, Rati fainted. After regaining consciousness somewhat, she
the shape of the body of her beloved, just as before. and rising quickly she
went near him and saw up close the fonn of a man.
Of what further nature was that fonn of a man? He tells us---entirely
ash from the fire of Siva's anger. Entirely-the idea is that not a single
portion was seen, either in the limbs of his body or in their garments, that
had not been completely reduced to ashes. Some read In
their opinion the construction is "the ash from the fire of Siva's anger was
seen in the form of a man.
This verse displays the awakening of the factors connected with the
mood of compassion.
B. Other Techniques
1.81 Introduction
Although the methods of arranging the words of the text described in Sec-
tions 1.76-1.80 constitute the primary means of indicating the construction
of the text being commented on, commentators also make use of several
other techniques to indicate features of syntax and construction. Among
these are various ways of directly pointing out connections between words
(Section 1.82), fonnulas for supplying words needed to complete the mean-
ing of a sentence (Section 1.83). and devices for marking words belonging
to particular syntactical categories (Section 1.84). Much syntactical infor-
mation is also conveyed through the use of the technical terminology of
paI).inian grammar.
1.82 Direct statement of connections
1.82.1 anvaya
The word anvaya ("going together," construction) can refer not only to
the construction of an entire passage (as in the name of the anvayamukhf
style of commentary, Section 1.78), but also to the construction of any
part of the whole, and commentators frequently mark a specification of the
construction of the original text by adding the formula ity anvaya/:!, "thus
(is) is the construction." For an example, see the next-to-Iast sentence in
the commentary of quoted at the end of Section 1.80.
The portion of a passage most commonly singled out for this treatment
is the basic "skeleton sentence" of the original text, so that the fonnula
ity anvaya/:! is often used in connection with the katha'!!bhatinf style of
commentary (Section 1.77). For example, in explaining his statement tatra
dravyii':!i prthivy-ap-tejo-viiyv-iikiisa-kiila-dig-iitma-manii'!JSi navdiva
("Among these the substances are nine only---earth, water, fire, wind, space,
time, direction, soul. and mind," Tarkasangraha 3), Annambhatta makes
the following remark in his own commentary on the text, the Tarkadfpikii:
162
Chapter 10. Indicating the Construction 1.82. Direct statement of connections 163
tafra dravyadimadhye dravyiilJi navaiva ityanvayaJ:z.. kant ttini? ityata
iiha-prthivl ity-iidi. The basic construction is: "Among these among
the substances, etc. the substances are nine only." What are they? In
reply he says-----earth, etc.
1.82.2 sa'!tbandha
The tenn sambandha ("connection") and related fonns (salflbadhyate, sa'!'-
bandhanfva, etc.) are used in expressly stating the connection of one word
with another word or words. These terms appear in commentaries in either
of two situations.
One situation is that in which the commentator must explain in his re-
marks on one portion of a work that a word is to be construed with a word
or phrase occurring in an earlier or later portion of the work (cf. the for-
mula iti sesah, Section 1.83.1). This often happens when several individual
verses of poem are syntactically connected (Section 1.82.4):
Ragh.M. 1.5 so 'ham. "raghal}iiYfl anvayaf!! ity uttare1}a
sambandhab . I, being so. The connection (of this) is with what follows
(four verses later in the poem), "shall proclaim the dynasty of the Raghus."
Ragh.fo.,1. 1.36 listhitDv tirat)hau. "jagmatuJ:t" iti pilrve1}Q sambandhal}.
Having seated themselyes having taken their seats (in the vehicle). The
connection is with what came before (in the preceding verse), "they went."
The other situation is that in which a word in the portion of the
being commented on is to be construed with more than one word in that
portion:
Kum.M.2.10 tatka iitmiinam iitmanii eva-iitmany every atnipi
saf!lbadhyate-svasminn eva srjasi. Likewise your self by the self
alone-"in the self alone" (occurring in a later clause in the same verse) is
to be construed here as well-in yourself alone you create.
Ragh.M. 11.11 sievire iti sarvatra sambadhyate. (The verb) "served"
is to be construed (with the subject in each clause) throughout (this verse).
Kum. v: 2.9 apir virodhe sarvatra (The particle) api
in the sense of contradiction is to be construed with (the first word of each
clause) throughout (this verse).
The tenn yojya ("to be connected") is used in much the same way.
1.82.3
The word ("modifier") is used in explaining which word an ad-
jective or adverb is to be construed with:
Ragh.M. 1.83 sandhyiiyiim apy etad yojyam. This modifier
(viz., "glossy and pink as a fresh shoot") is to be construed with the dawn
as well.
Ragh.M. 1.14 meriiv api tulyani. The modifiers apply
equally to Meru as well.
Ragh.M. 1.18 bhuty-artha'!' vrddhy-artham eva . ...
caitat. Only for the sake of prosperity for
the sake of well-being.... And this is a modifier of the action of taking.
As the first two examples indicate, this tenn is used in reference to ad-
jectives chiefly when one adjective is to be construed with more than one
noun.
1.82.4 Connections between verses
When a work composed in verse contains a sequence of verses that are
syntactically connected (either by grammatical opposition or by the use of
relative clauses), a commentator may call attention to such a sequence in
two ways. In the introduction to his comments on the first verse in the
sequence the commentator will usually specify the number of verses so
connected through such statements as sloka-dvayena .. _ aha ("through the
use of a pair of verses [the author] says ... "), sloka-trayel}a ... aha, etc.
The commentator may also make use of the technical tenns for con-
nected sequences of various lengths. These tenns are used most frequently
at the end of the comments on the final verse of the sequence, in order
to mark the close of the group of connected verses, but they may also be
llsed in the introduction to the first verse in the sequence as well. The most
frequent of these technical tenns are yugmaka or yugma (a pair of syn-
tactically connected verses), sa,!,danitaka (a group of three such verses),
kalapaka (four verses), and kulaka (five or more verses)_
164 Chapter 10. 1ndicating the Construction 1.83. Supplying words needed to complete the sense
165
Kum.M. 4.42-4.43 stipiivadhir api tenaiv6kta fty iiha
sloka-dvayena-... yugmakam. [In the introduction to verse 42:] That
he himself also stated a limit to the curse (the poet) tells us in a pair of
verses-... [At the end of the comments on verse 43:J (this has been a)
yugmaka.
Ragh.M. 1.5-1.9 eVGlJ1 raghu-vQ'!1se tad-van:zanaf!l
"so 'ham" slokaifz kulakenaha-...
kulakam. [In the introduction to verse 5:] Having thus been granted
entry into the dynasty of the Raghus, (the poet) tells us that he proposes to
describe it. with a kulaka in five verses beginning "I, being 50"-... [At
the end of the comments on verse 9:] (this has been a) kulaka.
1.82.5 Connections between statements
In addition to specifying the syntactical construction of a passage, com-
mentators often tty to help the reader understand the connections between
the parts of the passage by making small asides designed to indicate the
logical chain of thought involved. In the following example the commen-
tator neglects to repeat many of the wotds being commented on, being
content to let the synonyms he supplies for them take their place in reveal-
ing the structure of the original sentence [see Section 1.80, example (5)];
here the original words are inserted in square brackets.
Kum. V. 3.63 krta-praniimii ca [sii] asau [bhavena] harena
an.anya-bhiijam a-siidharana'!' [patim] bhartiira'!' [iipnuhi]labhasva ity
[abhihi!ii] agadyata. etac ca [tathyam] satyam eva. devena pascad
udvahitatviit. na caitad ascaryam. [hi] yasmiid lSvarii!'iim lujhyaniilJl
vyiih[laya piike phala-kale viparitam anyam eva arrha,!, na
na phalanti. devas cesvaraJ:t.
And [she] that woman, having bowed to him, [was addressed] was
spoken to [by Bhava] by Siva, "[Obtain] acquire [a husband] a spouse
enjoyed by no other not shared." Aod this quite [accurately] truly.
Because afterwards Siva married her. Nor was this surprising. [For]
because the utterances the statements of lords of rich persons do not
foster do not result, at ripeness at the time of fruit,in an opposite an
entirely different sense. And Siva is a lord.
For further examples see the passage from Sisu. V. /.1 quoted in Section 1.79.
This technique is essentially another way of achieving the goal of the ques-
tions asked in the katha,!,bhatinl style of commentary (Section 1.77).
1.83 Supplying words needed to complete the sense
When a wotd must be supplied in the original text, two situations are possi-
ble: if the wotd is simply to be understood in the passage without actually
occurring elsewhete in the text, a commentator will use the formula iti
when the supplying the word (Section 1.83.1), but if the word is to
be carried over from a preceding part of the text, formulas involving several
technical terms are used (Section 1.83.2). See also Section 1.82.2.
1.83.1 iti
The formula iti ("thus the remainder," "this temains [to be sup-
plied]") is placed aftet words understood in the original text and supplied
by the commentary. It may be translated "supply ... ":
Ragh.M. 12.71 tena pathii uttfrya sagaram iti sefaJ:t.
By that path by the route of the bridge having crossed over. Supply "the
sea."
Kum.M. 2.56 tat-tapa!, vare!'a vara-danena samitalJl hi. mayeri
For by a boon by the granting of a boon bis ascetic power was quelled.
Supply "by me."
Ragh.M. 1.35 vas4!hasyasrama,!, jagmatui}.
They both went to the hermitage
of Supply "because of their need for a means of obtaining a
son."
1.83.2 Words carried over
Wotks of the sutra type achieve much of their conciseness by assunting
that a word used in any sutra will be supplied in each succeeding sulra in
which it is needed. This "carrying over" of a word is called anuv[lti, and
a word so carried ovet is said to be anuvrtta; finite vetbal forms are used
as well: anuvartate ("it remains over"), or simply varrate ("it is present").
The cessation of this process of carrying over, which occurs at the point at
which the word is no longer needed to complete the meaning of a sutra,
is called nivrtti, and a word that has ceased being carried over is said to
be nivrtta. The authors of commentaries on works in which anuvrtti and
nivrtti operate frequently use these technical terms, most often at the very
beginning of their commentary on the sutra in question:
adjectives is the subject and which is the predicate. Commentators some-
times meet this difficulty by marking the subject with a relative pronoun
and the predicate with the correlative pronoun. For example, the sentence
lokdntara-sukhaf!l pu"yam (Ragh. 1.69) could theoretically mean either
"pleasure in the next world is merit" or "merit is pleasant in the next world
(i.e., leads to happiness there)." Mallinatha's commentary marks pu"yam
as the subject:
166 Chapter 10. Indicating the Construction 1.84. Devicesfor marking features ofsyntax
167
over (from a previous rule). "ktva" is no longer carried over (from this
point on).
Kiisikii 1.2.26 veti vartate, se iti ca. "vii"" is present (in this sutra by
anuvrtti), and also "se('
Kasikii 8.1.54 pilrvaf(l sarvam anuvartate gaty-artha-lofarrz varjayitva.
Everything previous (Le., everything in the previous sutra, whether
actually read there or present by anuvrtti) is carried over, except
"gary-artha-Io!."
Other terms are occasionally used to point out similar phenomena even
in works other than those of the surra type:
Kum.M. 2.11 tvarn ity "You" is carried over (from the
previous verse in the poem).
1.84 Devices for marking features of syntax
Commentators use special devices to mark the subject and predicate (Sec-
tion 1.84.1), adjectives that are used attributively rather than as predicates
(Section 1.84.2), substantives used attributively (Section 1.84.3), locative
absolute constructions (Section 1.84.4), words used adverbially (Section
1.84.5), and vocatives (Section 1.84.6).
1.84.1 yat ... tat marking subject and predicate
The relatively free word order of Sanskrit, together with the frequent omis-
sion of linking verbs, often makes it difficult to tell which of two nouns or
sukha-karam. That which is merit is pleasant pleasure-producing in
the other world in the next world.
1.84.2 sat marking attributive adjectives
Another way of distinguishing attributive adjectives from predicate adjec-
tives is through the use of the present participle sat (masc. san, fern. sari),
which is placed after adjectives that are used attributively. For example, the
sentence prfta kamadughti. hi sa (Ragh. 1.32) could theoretically be inter-
preted in several ways; Mallinatha's commentary makes it clear that prfta
goes with the subject sa:
Ragh.M. 1.32 hi yasmiit kiira1}iit sa pritii satf ... kama-dughii
bhavati. For from which as cause she, being pleased gratified, ... is 'a
granter of wishes.
Similarly, the words ... (Kum. 1.5) might
be taken to mean "the siddhas are tormented by rains," until this device
makes it clear that the adjective is not a predicate but serves as a
modification of the subject:
Kum.M. 1.5 v!1{ibhir udvejitiil} samal; . .. srngiil)Y iisrayante.
Being tormented troubled by rains, they resorted to its peaks.
1.84.3 -bhuta marking attributive substantives
The word bhuta may be placed in compound after a substantive to make it
clear that it is being placed in apposition with some other noun.
168 Chapter 10. Indicating the Construction 1.84. Devices for marking features of syntax
169
This device is often used avoid any difficulties the reader may have in
linking two nouns of different gender, as the addition of -bhuta converts
an attributive substantive into an adjective that may be inflected in any
gender; Bhagavadgliii 7.5, for example, has prakrti,!, viddhi me pariim
jrva-bhutiim, "know my other nature, which is the sou!." For more on this,
see Section 2.17.
In commentaries this device is particularly useful when a pronoun of
the same gender as the attributive noun might seem to refer to it, but in fact
refers to some other object that is being equated with the thing denoted by
that substantive:
Kum.M. 3.74 vighnam antariiya-bhiJ.taJ?1 tatp kiimam. Him Kama the
obstacle being an impediment.
Kum,M. 3.19 surii1)iif!Z vijayabhyupiiye jayasyoptiya-bhilte tasmin hare.
Against him against Siva the means of victory being the means of
winning for the gods.
Without this device the phrases commented on in these examples might be
taken to mean "that obstacle" and "against that means of victory."
1.84.4 safi marking the locative absolnte
The present participle sat, when used in the locative case, serves to mark
locative absolute constructions. For example, mahaty avarodhe rnahaty api
(Ragh. 1.32) could mean either "even in his large harem" or "even though
his harem was large"; Mallinatha's commentary makes it clear that the
latter is the intended meaning:
Ragh.M. 1.32 avarodhe antafzpura-varge mahati sary api even on his
harem-on the group of women in his seraglio-- being large
The participle may be in any gender:
Ragh.M. 1.89 asyiim ... pi/iimbhasi satyiim apaI,J pibeI,J. you must
drink water. .. on her . . , being one by whom water is drunk.
For beginning students this device is especially useful in marking items
that they might otherwise interpret as finite verbs in the third personal sin-
gular present indicative:
Ragh.M. 1.78 sTotasi pravahe nadati sari the current-the flow of
water-being roaring
1.84.5 yathii tatha marking adverbs
The formula yatka (bhavati) tatka is used to mark words used adverbially:
KumM. 1.32 catasro 'srayo yasya ac catur-asram anyuniinatirikta1J1
yatha tatha sobhata iti caturasra-sobhi. That which has four comers is
four-cornered (Le., syrnrnetrical)--having neither too little nor too much
(in any part)-it is beautiful in that way, thus symmetrically beautiful.
Ragh.M. 1.18 sahasra'Tl gW}a yasmin kannalJi tad yatha tatha
sahasradha. In which there are a thousand repetitions,
that, in such a manner: in a thousand ways a thousandfold.
Ragh.M. 1.40 adiira,!, samfpaTJ'l yatha bhavati tatha. not far
in a way that is so.
1.84.6 he marking vocatives
Commentators regularly use the particle he before words in the
case to mark them as such, even if the word is used with no accompanying
inteIjection in the original text:
Kum.M. 2.51 he vwho svamin 0 lord master
Kum.M. 5.36 hepiirvati 0 Piirvafi
Chapter 11
The Fifth Service
2.1 Introduction
Those portions of a Sanskrit commentary that are not specifically devoted
to glossing the words of the text are usually concerned with discussing the
contents and implications of the text. Often these portions constitute the
major part of a commentary on a philosophical or scientific text, but only a
small part of a commentary on a literary text. Traditionally, discussion of
meaning takes the form of a dialogne in which the commentator allows op-
ponents to raise various objections against the text or against his interpreta-
tion of it and then proceeds to refute these objections. This format is based
both historically and stylistically on oral debate (see Section 2.48), and its
traditional importance may be seen in the term chosen for the fifth service
that a full commentary is expected to provide, i.e., "the
answering of objections." A distinctive style of Sanskrit prose evolved for
such discussion, a style that became the standard both for commentaries
and for independent works of philosophy and science.
This bharya style presents many features of syntax and tenninology
that may be unfamiliar to beginning students. The present chapter deals
with those features that seem to pose the most problems for students, i.e.,
those that occur fairly often and have been found by experience to be the
most puzzling. Some general observations on the discussion of meaning
are given in Sections 2.2 and 2.3; the rest of the chapter has been arranged
by bringing together (as far as possible) those items that are similar in form
or function. Particular words or constructions may be sought in the index.
173
175
2.3. Style
ter 14) and of causal sentences (i.e., sentences that express a reason or
cause; Chapter 15), and peculiarities in the use ofi[i (Chapter 16) and in
the use of various particles and pronouns (Chapter 20).
Chapter 11. The Fifth Service 174
Note: although Part Three deals specifically with commentaries, all
but Section 2.39 is applicable also to independent works composed in the
bharya style.
2.2 Content
Discussion of the meaning of the text is less lintited in content than the
analysis of the words themselves. Ordinarily discussion proceeds by a se-
ries of objections and refutations that lead to confirmation of the commen-
tator's view. In this discussion the commentator may exantine the implica-
tions of the text, defend the views represented in the text against those of'
rival schools, defend his interpretation of the text against other interpreta-
tions, and introduce his own insights (though the last are rarely claimed
the commentator as his own, and are usually argued to be implicit in the
text itself). He may support his arguments by citing examples and com-
parisons (Chapter 18) or by appealing to the authority of other texts, other
teachers, and common experience (Chapter 17). In learning to read such
discussion it is essential to learn the conventional terms used to mark dif-
ferent sections of the commentaty. For example, certain terms are used to
raise an objection, to reply to an objection, or to introduce an alternative Co:
interpretation (Chapter 19). Familiarity with these terms often perntits af'
reader to follow the flow of an argument even when the details of the argn-':'
ment are not yet understood. It is also important to remember that glosses;:
of the text may be interspersed among long sections of discussion. These I
glosses may be more easily recognized as such if the words of the text are
kept in mind as the commentary is read.
2.3 Style
The prose of commentaries is characterized by nontinal constructions em- i
ploying abstract nouns (Chapter 12), long compounds (Chapter 13), and ',:.
relatively few finite verbs. These constructions permit the concise ex-':
pression of complex ideas and so are well suited to the writing of com- f
mentaries. They also create the greatest single difficulty for the beginning
student who is not yet accustomed to this style. Other characteristic fea- ,.
tures that the student must learn include conventions of word order (Chap- .
Chapter 12
Abstract Nouns
2.4 Introduction
Nominal abstract nouns (i.e., abstract nouns derived from other nouns) are
widely used in commentarial prose. This section discusses the fonnation
and sigrtificance of such nouns, the common constructions that employ
them, the uses of the different cases in these constructions, and two id-
iomatic constructions that involve the instrumental and the accusative. (See
also: Coulson 140-42, 154-58; Speijer 235-39; Whitney 1237, 1239.) ,
2.5 Formation
2.5.1 fii and tva
The most common nominal abstract nouns are those fonned with the suf-
fixes til (feminine) and tva (neuter). These may be added to any adjective
or substantive to fonn abstract nouns meaning "being such, the quality of
being so and so" (Pi\. 5.1.119).
suklatii, sukLatvam being white, whiteness
aivata, asvatvam being a horse, horseness
Note that til and tva are identical in meaning.
177
178 Chapter 12. Abstract Nouns
2.7. Common constructions
179
2.5.2 -bhiivaJ.! and vrddhi derivatives
Nominal abstract nouns also include compounds formed with
-bhiivaJ.!, "state, being," and neuter vrddhi derivatives. For example. from
guru, "heavy," are formed guru-bhiivah and gauravam, both meaning "be- ,
ing heavy. heaviness," and both synonymous with gurutvam and gurutii.
Abstracts in to. or tva are more common in commentarial prose, and so
they are the subject of the following paragraphs. However, in
-bhiivaJ.! and vrddhi derivatives are occasionally used in the same constrnc-
tions described for abstracts in tii and tva.
indrasya liighaviit::;: indrasya laghutvtit because Indra is swift (causal
ablative)
maitryelJa gacchati ::;: mitrataya gacchati he goes as a friend
(predicative instrumental)
tasya pufrabhiive 'pi::;: tasya putratve 'pi although he is a son (locative
absolute)
prapadyate = devatiirrz prapadyate he becomes a god
(accusative with verb of motion)
2.6 Significance
2.6.1 Used by necessity
In some cases the abstract notion conveyed by the abstract noun is essential
to the meaning of the sentence.
SBh 2.2.17 pa!asya tu suklatvddayah pa!a-desa
abhyupagamyante na tantu-deSllh but the qualities of the cloth,
whiteness and so on, are admitted to reside in the cloth, not in the threads :'"
Here the is suklatvam and the sentence requires an abstract noun
that means "being sukla." Note that abstract nouns formed by adding ta,
tva, or -bhiivaJ.! to a dvandva compound may denote relationship, e.g., pitr-
putratvam, "being father and son," i.e., "the relation of father and son."
Here too the abstract notion is essential to the meaning of the sentence.
SBh 2.1.4 na ca prakrti-vikara-bhavo and the
relation of material cause and effect is not seen (in two things) when they
are different from one another ..
SBh 1.1.1 feJa-feJitve 'dhikrtadhikiire va pramaJ;labhaviid
because there is no proof that the inquiries
into dharma and into brahman are related as subordinate and principal or
as qualified and qualification
2.6.2 Used by choice
But in most cases the abstract notion conveyed by the abstract noun is not
essential to the meaning of the sentence. Instead abstract nouns are used
to form nominal constrnctions when the same meaning could be expressed
by a different constrnction that does not require the abstract noun. It is
this syntactical use of the abstract noun that is taken up in the remaining
paragraphs of this section.
2.7.1 With subjective genitive
Abstract nouns in tii or tva are usually constrned with a separate subject;ive
genitive. This constrnction may provide a complete statement, where the
abstract noun expresses the predicate.
indrasya rajatii of Indra there is kingness: Indra is king
pa.tasya nflatvam of the cloth there is blueness: the cloth is blue
A Sanskrit sentence of equivalent meaning can be obtained by dropping
the abstract suffix and making the subjective genitive the new subject.
indrasya rajata = Iudra is king
pa.tasya nflatvam = nflaJ:! pa.taJ:! the cloth is blue
It is important to realize that these constrnctions are simply different ways
of expressing the same idea. The abstract suffix does not imply an
lion of thought lacking in the alternative constrnction.
Abstract nouns in tii or tva may be used as a member of a compound.
ramasya raja-putrata visrutd Rama's king-san-ness is well-known: it is
well-known that Rama is the king's son
180
2.7.2 In compound
Chapter 12. Abstract Nouns
2.7. Common constructions
181
tad-riijatii-hetub his-kingness-reason: the reason for his being king
iitma-nityatvam self-eternalness: the etemality of the self
But more often tii and tva are added to whole compounds in the sense of
"being that which the compound denotes." The compound may be a tat-
bahuvrfhi, or dvandva.
riija-putratti king-son-ness: being the son of a king
mahii-balatvam great-strength-ness: being possessed of great strength
(bahuvrfhi)
moha-sokatvam delusion-grief-ness: being delusion and grief (dvandva)
In analyzing such constructions it is essential to determine the type of the
original compound (i.e., bahuvrfhi, or dvandva), and it is helpful
to keep in mind the alternative constructions without the abstract suffix.
riimasya riija-putratii ::: riija-putro rlima Rama is the kiug's son
indrasya mahii-halatvam = maM-bala indrafz Indra is possessed of
great strength
hetor moha-sokatvam ::: moha-sokau hetub the cause is delusion and
grief
2.7.3 Causal phrase within a sentence
The nominative of an abstract noun in tii or tva may be used (as above)
with a subjective genitive to fonn a complete statement. More often such
a phrase is employed as part of a longer sentence where the case ending
of the abstract noun indicates the relationship between the phrase and the
other words of the sentence.
deva-patir indro mahii-balatvti{ Indra is lord of the gods because of
[his] great-strength-ness: Indra is lord of the gods because he possesses
great strength
avidyiiyii darsayati he shows ignorance's
suffering-cause-ness: he shows that ignorance is the cause of suffering
durkhasyavidya-hetutvaf!l darsayati he shows suffering's
ignorance-cause-ness: he shows that suffering has ignorance as its cause
The nominal construction consisting of abstract noun and subjective geni-
tive is usually translated into English by a clause. The subjective genitive
is often omitted in the Sanskrit when it is easily understood from the con-
text; in the second example above we must supply tasya or indrasya with
maha-ba1atviit. Note the importance of correctly analyzing the original
compound. In the third example above the compound is a in the
fourth it is a bahuvrfhi. Sometimes an author will distinguish a bahuvrfhi
compound by adding the optional suffix ka (e.g., in the fourth example !'ne
could say avidyii-hetukatvam; see Section 2.15), but usually the compound
must be analyzed on the basis of context.
2.7.4 iii clause without abstract suffix
Because the word iti may be used to quote a statement or cite a reason
(see Chapter 16), the above constructions can be expressed equiValently by
means of an iti clause without the abstract suffix.
ramasya riija-putratii viSruta =riija-putro rlima iti visrutam
deva-patir indro mahii-balatvat = sa mahii-bala iti
avidyiiyii du!}kha-hetutvaf!l darsayati = avidyeti darsayati
du!}khasyavidyii-hetutvaf!l dariayati = avidyii-hetuf!l dul:tkham iti
darsayati
182 Chapter 12. Abstract Nouns 2.9. Other cases
183
An understanding of these equivalent constructions can be very helpful in
trying to understand long and complex sentences using abstracts in ta or
tva.i
2.7.S Two genitives
As was mentioned above, abstract nouns in ta or tva are usually construed
with a separate subjective genitive that may have to be supplied from the
context. If another genitive is present, it is usually placed in compound
with the abstract noun.
riimasya riija-putratii ; putraTJl Raffia is the kiog's son
Here the subjective genitive is ramasya, and rajfiaiz (raja- in compound)
is a possessive genitive. But occasionally the other genitive appears sepa-
rately, particularly if emphasis is placed on its case and number, since these
are obscured in compounds.
SBh 1.1.1 sarvasyatmatva.c ca brahmastitva-prasiddhiJ.1. and the
existence of brahman is well known because it is the self of everyone
Here sarvasya is a partitive genitive. and the subjective genitive (which
must be supplied) is brahmar;aiz.
2.8 Causal ablative and instrumental
Sanskrit pennits the use of both the ablative and the instrumental cases to
express a reason or cause, and abstract nouns in twit, tvena, and rayii are
very common in this sense. The ablative of ta is rare, probably because it
is identical in form with the genitive (its use is prohibited by Pa. 2.3.25).
SBh 1.1.2 tatrtiivt1f!l sati brahma-jfiiinam api vastutantram eva
this being the case, knowledge of brahman also
depends only on the thing itself, since it has for its object an existing thing
SBh 1.1.2 na, vediinta-vtikya-kusuma-grathanarthatviit siitriil:ulm no,
since the siltras have the purpose of stringing together the flowers that are
the Vedanta passages
SDS p. 3 na ctisya {sukhasya] duJ:lkha-sambhinnatayii
eva ntistfti mantavyam nor should it be thought that it [pleasure] is not
the aim of man because it is mixed with pain
SBb 1.1.4 mantriilJiiJ'!l tvetyiidfniiJ'!l kriyii-tat-siidhantibhidhiiyitvena
karma-samaviiyitvam uktam and it has been said that the mantras,
tvii and so on, are based on action because they declare action and the
means to action
Several such abstract nouns may be combined in one sentence to express
coordinate or serial causes; see Section 2.26 for a discussion of such con-
structions. Note also the idiomatic use of the instrumental discussed in
Section 2.10.
2.9 Other cases
The following are fairly typical examples of abstract nouns in ta or tva used
in cases other than the ablative or instrumental.
2.9.1 Nominative as subject of sentence
SDS p. 2 atha kathaJ'!l paramevarasya ni!tsreyasa-pradatvam
abhidhfyate but how can it be said that the Supreme Lord bestows
niJ:lSreyasa?
2.9.2 Accusative as object of the sentence
SBh 1.1.1 yasmiid veda evtigni-hotr-iidfniirrz sreyaft-siidhaniiniim
anitya-phalatiirrz dariayati because the Veda itself shows that the means
to success, the agnihotra and so on, have transient fruits
2.9.3 Dative of purpose
GBh 2.20 hantir atra viparilJiimilnhe 'punar-uktatiiyai the
verb hanti should be regarded here as signifying "transformation," to
avoid redundancy
2.9.4 Genitive to modify another noun
In the last example above note the addition of the suffix ttl to the demon-
strative pronoun idam to fonn idanttl "this-ness, bei?g this."
184 Chapter 12. Abstract Nouns
SBh 1.1.7 tan na [iitma-sabdasya] anekarthatvasyiinyiiyyatviit that is "
wrong, because of the impropriety of [the word arman's] having more ,"""
than one meaning
2.11. Accusative with verb ofmotion
2.11 Accusative with verb of motion
185
2.9.5 Locative in a locative absolute construction
(See Section 2.29.)
SBh 1.1.2 nanu
eva but if brahman is an existing thing, then it must be the object of
another means of knowledge
2.10 Predicative instrumental
The instrumental of an abstract noun in ttl or tva is often used idiomatically
to complete the predicate of a sentence. The instrumental expresses "in
what quality or capacity" and can usually be translated by "as."
riijatvena vrta Indea was chosen with kingness: Indra was
chosen as king
mitratvena riima'!l Janamt I know Rama with friendness: I know Rama
as a friend (i.e. to be a friend)
Usually the context will distinguish this construction from the use of the
instrumental abstract to state a reason (see Section 2.8). Further examples:
SBh 1.1.1 yadi tarhi lake brahmatmatvena prasiddham ast; if in
that case brahman were well known in the world as the self
SBh 1.1.2 srutyiiiva ca sahtlyatvena tarkasytibhyupetatvlit and
because STUff itself accepts reasoning as an aid
SBh 1.1.4 na hi ftistram idantayii brahma
for the siistra does not wish to teach brahman as this,
as an object
The accusative of an abstract noun in tii or tvam may be used idiomatically
with a verb of motion to mean "to become."
riijatiiTJ1 gacchati he goes to kingness: he becomes king
nflatvaTJ1 yati it goes to blueness: it becomes blue
In the same way the accusative may be used with the causative of a verb
of motion or with a verb signifying Uto lead," "to bring," etc., to mean "to
cause to become, to make into."
rajatiim indra'!1 gamayati he makes Indra go to kingness: he makes
Indra a king
nflatva'!l paraTJ1 nayati he brings the cloth to blueness: he makes the
cloth blue
Chapter 13
Compounds
2.12 Introduction
Long and complicated compounds are common in most Sanskrit commen-
taries. This section deals with how to analyze long compounds, how a com-
pound may be related to the other words of a sentence, three devices used
to clarify ambiguous compounds, the use of tat, yat, and kim in compound,
and several common terms used idiomatically at the end of a compound.
(See also: Coulson 107-11; Speijer 193.3, 197,213 b, 228-31; Whitney
1222 h.) .
2.13 Analyzing long compounds
(See also Sections 1.69-1.72.)
2.13.1 Identification
First one must detennine the type of the original compound and its re-
lation to the other words of the sentence. Long and difficult compounds
are usually t t u r u ~ s or bahuvrfhis (avyayfbhiivas are rare and dvandvas
are usually straightforward). Sometimes a compound can be identified on
grammatical evidence alone: The gender and number of a t t u r u ~ are
those of its final member. A bahuvrfhi is an adjective and its gender and
number are those of the word (either expressed or understood) that it mod-
ifies. bahuvrfhis are sometimes distinguished by the suffix ka (see Sec-
187
2.13.2 Dissection
tion 2.15). When such evidence is inconclusive, the compound must be
identified On the basis of context.
natural unit. Such units should be looked for when the simple procedure
described above does not yield a satisfactory sense. A longer example
isjanmiJdi-sarva-bhiiva-vikiira-varjitam, "devoid ofall transformations of
state, such as birth and so on":
189
2.14. Relation to words outside the compound Chapter 13. Compounds 188
Long compounds can often be analyzed as compounds of two elements,
where the second element is the last member of the compound and the
first element is everything that precedes the last member. By a series of
such dissections, a long compound can often be reduced to several shorter
compounds. For example, dhanniJdhanna-jiiiina-phaltlbhyudayalz, "suc-
cess that is the fruit of the knowledge of dhanna and adhanna," can be
analyzed as follows:
dharmiJdhannajiiiinaphala-abhyudaya kannadhiiraya
dhanniJdhannajiiiina-phala genitive
dharmiJdhanna-jiiiina genitive
dharma-adhanna dvandva
The shorter compounds from which the original long compound is con-
structed are usually dvandvas or rarely are they bahuvrfhis
(except for a few very common types of bahuvrfhis, e.g., those ending
in -iidi). Thus adjoining words within a loog compound are usually to be
construed as coordinate (dvandva: e.g., dhanniJdharma) or as appositional
(kannadhiiraya: e.g., phaliJbhyudaya) or as connected by a case relation-
ship e.g., jiiiina-phala).
2.13.3 Grouping
janmadisarvabhiivavikiira-varjita instrumental
janmiJdi-sarvabhiivavikiira kannadhiiraya
sarva-bhiivavikiira karmadhiiraya
bhiiva-vikiira genitive tatpurusa
janma-iidi bahuvrfhi
2.14 Relation to words outside the compound
In theory, a word standing outside of a compound may be construed with
the compound as a whole and not with some prior portion of it. A word
that is to be construed with a prior portion must be incorporated into the
compound.
jagaj-janmiidi-kaTal}QfJ1 brahmadhigamyate brahman is known to be
the cause of the origin and so on of the world
Here brahman is construed with the entire compound and jagat is con-
strued withjanmiJdi within the compound. But in actual practice this rule
is often broken, and a separate word or phrase (especially one in the geni-
tive case preceding the compound) is construed with a prior portion of the
compound.
The analysis outlined above must be modified for some compounds. Some-
times two or more members form a natural group and must be so treated
when the compound is resolved into its two basic elements.
indraputrapautrii/:z. = indra-putrapautra./:z. the sons and grandsons of
Io<ira
riimasitasarli/:z. = rama-sitasarii/:z. the sharp arrows of Rama
The original compound can still be resolved into two elements; however,
the second element here is not a single word but a compound that forms a
SBh 1.1.3 jagaro janmiidi-kiiral}Q,!, brahmMhigamyate
Such constructions are a mixture of the sentence mode of expression (viikya-
vrlti) and the compound mode of expression (samiisa-vrtti). For example,
"knowledge of the eternality of the self" might be expressed:
iitma-nityatva-vidyii compound mode
iitmano nityatvasya vidyii sentence mode
iitma-nityatvasya vidyii mixed mode: theoretically acceptable
iitmano nityatva-vidyii mixed mode: theoretically unacceptable
190 Chapter 13. Compounds
2.16. tat to mark a dvandva
191
The last construction occurs less often than the preceding three, but it oc-
curs often enough to warrant consideration when the usual interpretations
of a passage do not yield a satisfactory sense.
2.15 ka to mark a bahuvnni
With a few minor exceptions the suffix ka may be added optionally to any
bahuvrfhi compound (Pa. 5.4.154). ka does not alter the meaning of the
compound and is often added to obtain a stem more easily inflected, e.g.,
maha-yasaska from mahii-yaSas. ka is also used to mark a compound as a
bahuvrfhi, especially where, because the gender and number of the com-
pound are inconclusive and the context might admit alternative interpreta-
tions, the bahuvrfhi might be mistaken for a
bhakti-nimittarrz jiiiinam knowledge is the cause of devotion
or, knowledge has devotion as its cause (bahuvrfhi)
bhakti-nimittakaJrl jiilinam knowledge has devotion as its cause
(bahuvnhi)
In the first example above only the context can determine whether the com-
pound is a or a bahuvrfhi; in the second example the compound
must be a bahuvrfhi. ka may be added to a bahuvrfhi before the abstract
suffixes ta and tva; in this case ka is particularly helpful because the gender
and number of the original compound are no longer evident.
jiilinasya bhakti-nimittatvarrz prasiddham it is well-known that
knowledge is the cause of devotion or, it is well-known that
knowledge has devotion as its cause (bahuvrfhi)
jiilinasya bhakti-nimittakatvaTJl prasiddham it is well-known that
knowledge has devotion as its cause (bahuvrfhi)
A further example:
GBh 1.1
dharme when dharma is being overcome by adharma of which the
cause is a lessening of discriminatory knowledge
2.16 tat to mark a dvandva
The pronoun tat may be used in a compound to a dvandva (either of
the entire compound or more often of part of the compound) where other-
wise a would ordinarily be understood. In this case tat repeats a
prior member of the compound. For example, karma-phala-tyagaf.! would
ordmanly be understood as "renunciation of the fruits of actions," and
karma-phala taken as a genitive But karma-tat-phala-tyagah
means "renunciation of actions and their fruits." Here karma-tat-phala is
dvandva compound of karma and tat-phala, and tat refers to the preceding
word karma. Further examples:
SBh 1.1.1 adhyilso superimposition of the
object and its qualities
PPp. 33 for the sake of the
nature of superimposition and the possibility of it
2.17 -bhiita to mark a karmadhiiraya
The past participle -bhuta may be used within a compound to show that tWo
adjoining members are to be construed as a karmadharaya (Le., as appo-
sltIonal) where otherwlse one might construe them as a vibhakti-tatpurusa
(i.e., as connected by a case relationship) or as a dvandva (Le., as coordi-
nate).
destruction of the cause of ignorance; or,
destruction of the cause that is ignorance
destruction of the cause .that is ignorance
In the first example only the context can determine whether avidya-karana
means "the cause of ignorance" (genitive tatpurusa) Or "the cause that is
ignorance" (karmadhiiraya); in the second example aVidya-bhuta-kttrana
must mean "the cause that is ignorance." Further examples: .
The following words are common in an idiomatic'sense at the end of a
compound. See Section 2.44 for bahuvrfhis in -adi.
192
Chapter 13. Compounds
GBh 2.10 prii!Ifnti'!' foka_mohiidi_sQ'!lsiira_bfja_bhuta_dosodbhava_
karaIJa-pradarsanarthatvena vyakhyeyo this section should be
interpreted as serving to show the cause of the origin of that defect that is
the seed of saf!lStira, Le., grief, delusion, and so on, for living beings
2.19. 1diomatic terms
2.19 Idiomatic terms
193
SBh 1.1.19 phalasya
because the fruit that is the obtaining of brahman-who is
the tail, the support-iS indicated (in the text)
2.18 Pronouns in compound
2.19.1 -antara
difference," often fonus neuter compounds
meamng 'a dIfferent, another."
rajiintaram a different king, another king
grhiintara!li different houses, other houses
-artha, m.. "purpose," often forms masculine compounds used
adverbially in the accusative singular to mean "for the sake of, in order to."
dhaniirtham for the sake of money
atmiiikatva-pradarsandrtham in order to show the oneness of the self
Note that the same meaning may be conveyed by the dative case of ;he
word or compound to which -arrha is added.
The pronouns tat, yat, and kim appear frequently in compound (yat and
kim as first member, tat as any member except last). In compound tat, yat,
and kim (the stem forms) may stand for any gender and number. Thus tad-
may be "his color," "her color," "its color:' "their color," "the color
of those two," etc. The compounds may be of various types; for example:
tad-rupa/:t its color or, having that as its color (bahuvrfhi)
yad-rupaJ:!. the color of which Of, having which as its color
(bahuvrfhi)
the color of what or, having what as its color
(bahuvrfhi)
Commentators usually gloss such compounds with the word anya;
rajiintaram would be glossed anya raja.
2.19.2 -artha
e.g.,
A special use of tat is mentioned in Section 2.16. Note that when yat is in
compound, its correlative pronoun need not be in compound.
GBh 2.16 buddhir na vyabhicarati tat sat
having-which-as-its-object a cognition is not in error, that is sat: that thing
is sat when a cognition having it as its object is not in error
bahuvrfhi compounds containing kim are often used to ask a question.
SBh 1.1.2 punas tad brahma bUI
having-what-as-its-definition is that brahman: but what is the definition of
that brahman?
tat-siidhaniirtham = tat-siidhaniiya in order to prove that
2.19.3 -kalpa
-kalpa, m., "manner, way," forms bahuvrfhi compounds with the meaning
"almost, nearly."
sa mrta-ka1pah he is nearly dead
raja-ka/po Devadatta is almost a king
Note: Pi\. 5.3.67 explains -kalpa as a taddhita suffix (kalpaP in
terminology) and commentators may identify it as such.. .
194
2.19.4 ogata
2.19.7 -jata
2.19. 1diomatic terms 195
-gata, the past participle of gam, may be used in a compound to
mean "being in or on," where no previous motion is implied.
lqetragatti the trees in the field
Such compounds may also mean "relating to, concerning,"
bhtiryii-gatii cintti worry for his wife
In both instances -gata conveys a meaning usually expressed in other con-
structions by the locative case.
2.19.5 .garbha
-garbha, m., "womb, foetus," forms bahuvrfhi compounds meaning "hav-
ing inside, containing, filled with."
ghatam a pot filled with milk
a sack containing jewels
2.19.6 -jana
-jana, m., "person, people," forms masculine compounds that
may denote either an indefinite individual or a collection.
pitrianalf a father; or, fathers
In longer compounds -jana often pluralizes a preceding member that might
otherwise be understood as singular.
mitranugrhitah helped by a friend; or, helped by friends
mitra-jananugrhital} helped by friends
-jiita, n., "collection," forms neuter compnunds denoting a col-
lection of things.
ktirya-jiitam the aggregate of effects
In longer compounds -jiita may pluralize a preceding member that might
otherwise be understood as singular.
deva-prasiida!} the grace of the god; or, the grace of the gods
deva-jiita-prasiidaJ:! the grace of the gods
2.19.8 -purva
-piirva, "previous," when modifying a past participle in a com-
pound may be placed after the participle.
sankalpita-purva = purva-sankalpita previously imagined, already.
imagined
2.19.9 -matra
-miitrii, "measure," forms bahuvrfhi compounds whose literal meaning is
"having ... as its measure, having the size of."
grha-miitro gajaT:t an elephant the size of a house
Often such compounds have the special meaning "measured by, being no
more than, mere, only." In this case the compound often appears as a neuter
singular substantive.
SBh 1.1.1 deha-matra,!, atma the self is just the
body qualified by consciousness
196
When -miitra has this meaning commentators usually gloss the compound
with the word eva; e.g., deha-miitram would be glossed deha eva. Occa-
sionally compounds in -matra have the opposite meaning: "the full mea-
sure of, the whole class of, any."
buddhily. an idea referring to any quality
Thus -miitra can generalize as well as particularize.
Note: Pli. 5.2.37 explains -miitra as a taddhita suffix (miitraC in Pfu)inian
terminology) and commentators may identify it as such. See Section 1.60.3.
2.19.10
-visesa, m., "distinction, difference, peculiarity," may form masculine tal'purus,a:'l
compounds meaning "a particular" or "a distinguished."
a particular mountain
distinguished beauty
in and in -bheda, m., "difference" may also mean "a
kind of, a variety of."
a kind of flower
malJi-visefaJ:z a variety of jewel
Chapter 14
Word Order
2.20 Introduction
There are relatively few constraints on the order of words in Sanskrit prose,
and some commentators of a more literary style vary their sentence struc-
ture often. But some observations about word order can be helpful, espe-
cially in reading long sentences. This section discusses the order of subject
and predicative nouns, the order of attributive and predicative nouns, the
scope of the word iti, and the underlying syntactical construction of long
sentences. See Section 2.26 for the conventions of word order in causal
sentences. (See also: Coulson 46-49; Speijer 16-17, 31.)
2.21 Subject and predicative nouns
2.21.1 Predicative nouns
Predicative adjectives agree with the subject in number, gender, and case,
while predicative substantives agree in case but not necessarily in gender
or number.
udiiro rtimaJ:! Rama is noble
mitra'!1 riimaJ:r. Rfuna is a friend
A predicative noun usually precedes the subject, especially in short
sentences or when the predicate is emphasized.
197
tan mitram he is a friend
The same order is usually followed in sentences with predicative ac-
cusatives.
2.21.2 Pronoun subjects
Pronoun subjects deserve special mention. The pronoun often agrees in
gender and number with a predicative substantive.
199
2.22. Attributive and predicative nouns
2.22 Attributive and predicative nouns
2.22.1 Word order with subjects
In these examples attributive nouns precede and predicative nouns fol-
low the respective subjects, asarfratvam and adhyiiso.
In general, attributive nouns express something already known or estab-
lished about the subject, while predicative nouns assert something new.
Thus the context is usually sufficient to distinguish the one from the other
when both occur in the same sentence. But sometimes word order is also
helpful, especially in long sentences. Some authors characteristically place
attributive nouns before and predicative nouns after the subject.
SBh 1.1.1 evam ayam antidir ananto naisargiko 'dhytiso
mithyti-pratyaya-rapatz kartrtva-bhoktrtva-pravartakatz
thus this beginningless, endless, natural
superimposition takes the form of false knowledge, causes [the self] to be
an agent and enjoyer, and is evident to everyone
SBh 1.1.4 ata
asarfratvarrz nityam iti siddham therefore it is proven that bodilessness,
which is different from the fruit of actions to be performed and is called
is eternal
SBh 1.1.1 asmat-pratyaya-gocare cid-iitmake ... adhytisatz
superimposition on the which is the object of the notion of I and
whose nature is intelligence . ..
2.22.2 Word order with other nouns
This same order, i.e., attributive qualifiers-substantive-predicative qual-
ifiers, may also be used for substantives that are not the grarnmmatical
subject of the sentence.
Chapter 14. Word Order
SBh 1.1.4 ata na brahma-vidya therefore
knowledge of brahman does not depend on the activity of man
SBh 1.1.4 nityas ca sarvair abhyupagarnyate
and is accepted as eternal by all those who uphold the doctrine of

198
SBh 1.1.4 evamiidyiib srutayo
eva/ma.,jfiiinasya phala,!, darsayanti scriptural passages such as these
show that the fruit of self-knowledge is simply the cessation of obstacles
to
But it is also quite common for the subject to precede a predicative
noun, particularly if the subject is emphasized or if it has just been men-
tioned in the previous sentence.
SBh 1.1.4 jiftinG'!' tu pramiil:lajanyam. pramii1}Q'll ca
But knowledge arises from apramiilJa.
And apramiil}Q has for its object a thing as it really is.
Thus the general rule that a predicative noun precedes the subject un-
less the subject is emphasized is often but not always followed. Subject
and predicative noun are best distinguished on the basis of context: usually
the subject has been mentioned or implied in previous sentences and the
predicative noun contains some new information about the subject.
The pronoun may either precede or follow the predicative noun.
Sometimes the pronoun is placed in the middle of a predicate of two or
more words.
PPp. 79 nanu mahad etad indra-jiilam but this is a great deception
Here asmat-pratyaya-gocare expresses a quality that was
mentioned previously, while cid-iitmake expresses a new quality. The dis-
tinction might be indicated in English as follows: "the which is the
object of the notion of I, has intelligence as its nature; superimposition on
it ..."
200 Chapter 14. Word Order
2.24. Analyzing long sentences
201
2.22.3 Markers of attributive qualifiers
Attributive qualifiers may be marked as such by the present participle sant
placed after the qualifier (compare the use of sant to gloss attibutive qUal-
ifiers; see Section 1.84.2), or by the past participle bhiita placed in com-
pound with the qualifier (Section 1.84.3. Usually sant and bhiita serve
to make the construction clearer and can be omitted without changing the
meaning of the sentence.
SBh 2.1.4 tathedam api jagad acetana'!l sukhawdufrkhawmohiinvitaf!J sad
acetanasyiiiva sukha-dufrkha-mohdtmakasya kiira!lasya kiirya'!l bhavitum
arhati in the same way this world also which is unconscious and
possessed of pleasure, pain, and delusion must be the effect of a cause that
is unconscious and whose nature is pleasure, pain, and delusion
SBh 1.1.5 tri-guJ}atvtit tu pradhiinasya
sattva'!' pradhtinavasthiiyiim api vidyate but since the pradha.na
consists of the three gU1}QS, sartva, which is the cause of all knowledge,
exists also in the pradhiina state
Here sad shows that acetana'!' sukha-dul.zkha-mohlinvitam qualifies ja-
gad and bhiita shows that sarva-jfi.iina-kiirara qualifies sattvam. In many
cases sant has the special meaning "although being."
SBh 1.1.5 pradhanasyacetanasyaiva sata/:! sarvajiiatvam upacaryate
in the Vedanta texts omniscience is figuratively
ascribed to the pradhiina, although it is really unconscious
Often this meaning is confirmed by adding the particle api to sant; see
Section 2.54.2. Occasionally bhiita marks the predicate of the sentence
instead of an attributive qualifier.
PPp. 183 avidyti ca sa1'J'lsiira-hetu-bhutii and ignorance is the cause of
sa1'J'lsiira
2.23 Scope of iti
The word iti is normally placed immediately after the word or words that
are quoted by it. Thus iti marks the end of a quotation, but in most cases the
beginning must be determined on the basis of context. Sometimes words
placed at the beginning of a sentence for the sake .of emphasis are to be
construed with words after the iti clause; in this ca;e care must be taken
not to include these beginning words in the iti clause. Negatives such as na
ca and nahi are especially common in this construction. In the following
examples quotation marks have been added to show the extent of each iti
clause.
SBh 1.1.1 ta1'J'l ke cid "anyatranya-dharmadhyiisa" iti vadanti some
say that it is the superimposition of the qualities of one thing on another
thing
PPp. 356 na ca "sa na pratfyata" iti yukta1'J'l vaktum nor is it
reasonable to say that it is not known
SBh 1.1.4 nahi" 'sarfrady-iitmlibhimiinino dU/:lkha-bhayadimattvaf(l
iti tasydiva veda-pramiirta-janita-brahmiitmlivagame
tad-abhimiina-nivrttau tad eva mithyli-jiiiina-nimitta1'J'l
du/:!kha-bhayiidirnattvaf(l bhavati" iti sakya1'J'l kalpayitum for it cannot
be imagined on the ground that someone who wrongly identifies his self
with his body, etc. is seen to possess pain, fear, etc., that when the same
person has knowledge of the self as brahman produced by the pramiilJO '
which is the Veda and when the mistaken identification has ceased, he
should have that same possession of pain, fear, etc. which was caused by
false knowledge
The last example shows that one iti clause may include another: The
first iti quotes the words sarfrddy-iitrndbhimiinino . .. and this clause
expresses a reason. The second iti quotes all the words between nahi and
sakya,!, kalpayilum and this clause expresses what it is that cannot be imag-
ined.
2.24 Analyzing long sentences
Unusually long and difficult sentences are best approached by first identi-
fying their underlying syntactical construction. Once the skeletal construc-
tion is understood, the remaining words of the sentence can be construed
202
Chapter 14. Word Order 2.24. Analyzing long sentences 203
correctly. Sometimes it is helpful (when possible) to underline the words
comprising the skeletal sentence. For example:
SBh 1.1.3 kim u vaktavyam "anekaitikhii-bheda-bhinnasya

rg-veditdy-tikhyasya sarva-jiiiinlikarasya aprayatnenaiva lfltl-nytiyena
yasman makato bhiitiid yone1;J sambhava!z 'asya
mahato bhatasya niJ:lsviisitam etad yad rgveda' tasya
mahata bhiitasya niratisaya,!, sarvajiiatva,!, sarva-aktimattva'!'ca" ill
The skeletal construction of this sentence (underlined above) is:
kim u vaktavyam ... iti how much more must it be said that ...
Thus the final iti quotes every word of the sentence after vaktavyam. Within
this long quotation, the following skeletal construction (in boldface above)
may be seen:
sarva-jiiiinakarasya ... yasmiin mahato bhiltiid yonel} sambhavaJ:t ...
tasya mahato bhiltasya niratisaya1J1 sarvajiiatvarrz sarva-saktimattva,!, ca
from which great being as the source there arose that mine of all
knowledge, that great being possesses supreme omniscience and
omnipotence
To this skeletal construction the remaining elements of the sentence
may be added. The first is a series of qualifiers of sarva-jiiiinfikarasya:
aneka-sakha-bheda-bhinnasya
deva-tiryan-manu.rya-van;iisramiidi-pravibhiiga-hetor
rg-vediidy-iikhyasya divided according to various branches, the cause of
the distinctions of gods, animals, men, class and stage of life, etc., and
known as the Rig Veda, etc.
The second is an adverbial phrase modifying
aprayatnenaiva lrlli-nyiiyena without any effort at
all, in the manner of sport, like a man's breathing out
The third is substantiation of the entire statement by a quotation from
scripture:
asya mahato bhutasya niJ:tsvasitam etad yad rgveda ityiidi-iruteJ:t
because of scriptural passages such as "from that great being this was
breathed forth which is the Rig Veda"
The whole sentence may now be translated:
How much more must it be said that that great being possesses supreme
omniscience and omnipotence, from which great being as the source there
arose-without any effort at all, in the manner of sport, like a man's
breathing out-that mine of all knowledge that is divided according to
various branches, that is the cause of the distinctions of gods, animals,
men, class and stage of life, etc., and that is known as the Rig Veda,
etc.-as we know from scriptural passages such as "from that great being
this was breathed forth which is the Rig Veda."
2.25 Introduction
Chapter 15
Causal Sentences
(ablative phrase)
(iti clause)
(relative pronoun)
(locative absolute)
(causal particle hi)
indrasya rajatvat
rajendra iti
yasmad
indre Tajni sati
raja
Sentences that express a reason, cause, or motive are essential for interpre-
tation and discussion and so are common in almost every commentary. The
constructions most often used to state a reason are ablative and instrumen-
tal phrases, clauses formed by iti, clauses introduced by a relative pronoun
or adverb, locative absolute constructions, and clauses introduced by the
causal particle hi. For example, each of the following may convey the idea
"because Indra is king": '
Note that commentators often vary these constructions in successive sen-
tences and occasionally combine two or more of them in the same sentence.
SBh 1.1.2 kartcivye hi nanubhaviipeksiistfti sruty-iidfniim eva
priimii.!lya1J1 syiit. ca For if the
object [of the knowledge of brahman] were something to be
accomplished, then there would be no dependence upon direct experience;
consequently the srufi and so on would be the only means of knowledge.
And because the origination of something to be accomplished depends
upon man.
205
(See also: Coulson 170-71; Speijer 365-68; Whitney 303.)
2.26.1 Ablative causal phrases
2.26 Ablative and instrumental phrases
207
2.26. Ablative and instrumental phrases
SBh 1.1.4 kriya.-samaviiyabhiiviic ditmana!:z kartrtvdnupapaneJ:t and
because the self cannot be an agent since there is no inherence of action
(in it)
instrumental cases alternate (the ablative being used for the nearest cause).
Examples of the former:
Chapter 15. Causal Sentences 206
Nouns in the ablative or (less often) the instrumental case may express
a reason or cause. Abstract nouns in fa or tvam are especially common
in this construction (see Section 2.8). The noun may be a single word
or a compound. The reason may be expressed by the noun alone or by
a nominal phrase of which the noun is a member. The noun or nominal
phrase may form a complete statement by itself, or it may stand at the
beginning, middle, or end of a longer sentence.
SBh 1.1.1 pasvadibhi1 and because there is no difference
from cows, etc.
SBh 1.1.4 ato vediintiiniim iinarthakyam akriyarthatviit therefore the
Vedanta texts are meaningless because they are not concerned with action
SBh 1.1.7 bhrtye tu svami-bhrtya-bhedasya pratyalqatviJd upapanno
gaufla but in the case of a servant the word litman may be
used in a secondary sense because the distinction of master and servant is
obvious
In the first example above the nominal phrase pasviidibhis is a
complete sentence; the conclusion, stated earlier in the passage, must be
supplied by the reader. In the second example the compound akriyiirthatvi1t
substantiates the assertion made in the prior portion of the sentence. In the
third example the nominal phrase svttmi-bhrtya-bhedasya pratyalqatvi1d,
inserted in the middle of the sentence, gives a reason for the assertion made
by the rest of the sentence.
2.26.2 Series of causes
In sentences citing a series of causes, the more removed cause is generally
placed before the nearer cause. Two conventions of case are common:
either the whole series of causes is put in ablative cases, or the ablative and
SBh 1.1.4 heyopiideya-sunya-brahmatmatavagamiid eva
because of the
accomplishment of the aim of man, which results from getting rid of all
afflictions, which in turn results only from the knowledge that brahman
who is devoid of anything to be avoided or sought after is the self
Examples of the latter:
SBh 1.1.2 na, sambandhagrahal}iit no, because
the connection is not grasped, since [brahman] is not an object of the
senses
SBh 1.1.4 sarva-gatatvena nityapta-svari1patviit sarvel}a brahmal}aJ:z
because brahman is by nature always present to everyone, since it is
omnipresent
,
Note the importance of reading the entire Sanskrit sentence before
ning to translate. Often the most natural English translation reverses the
order of the Sanskrit, beginning with the nearest cause and proceding to
the most removed cause.
2.26.3 ca in causal series
In sentences citing several causes it is essential to understand correctly the
conjunction ca. ca may connect the sentence to the preceding sentence, as
in the first example in Section 2.26.2. In this case ca is placed after the
first word of the sentence. ca may also connect causes within the sentence,
and here one must distinguish between coordinate causes (connected by
cal and serial causes (where one is the cause of the other). Sometimes
the meaning of ca cannot be determined by its position alone and must be
determined by context.
208
Chapter 15. Causal Sentences
2.27. iti clauses 209
2.27 iti clauses
Some of the most common of such verbal nouns are listed below, along
with the meanings that they often have in this construction.
Other tenns used in the same construction include smarar:za and sravaJa
(see Section 2.41) and prasQliga and apatti (see Section 2.50.3).
absence; because there is no
acceptance; because we accept that
fitness; because it is reasonable that
seeing; because we see that
reknown; because it is generally
known that
Ii tness; because it is reasonable that
possibility; because it is possible that
establishment; because it is established
that '
yoga; -yogat
sambhava; -sambhavilt
siddhi;
SBh 1.1.2 api evantar-bhtiva iti
janma-sthiti-niiftiniim iha grahal}flm the other transformations of state
also are included among just these three; therefore the text mentions birth,
subsistence, and destruction
abhava; -abhilvilt
abhyupagama; -abhyupagamat
upapatti;
darsana; -darsaniit
prasiddhi;
SBh 1.1.5 api ca pradhtinasyanekatmakasya pari?ltima-samblzaviit
ktiralJatvopapattir moreover the heterogeneous pradhtina can be the
cause because of the possibility of transformation (in it)
SBh 1.1.4 farfra-sambandhasyiisiddhatvtid dharnu3.dharmayor
because it is not establisheEtthat dharma and
adharma are performed by the self, since the connection of the body [with
the self] has not been established
One of the many uses of iti is to cite a reason or cause (see Chapter 16 for
other uses). In this case iti, whose basic meaning is "thus, in this way,"
comes to mean "thinking thus, on these grounds, for this reason." Usu-
ally the words quoted by iti precede the assertion or conclusion that they
substantiate.
SBh 1.1.23 tad ayukta,!, sabda-bhedtit ca
SGf!lsayanupapatteJ:t that is not right since a doubt is not possible
because of the difference of words and because of the topic
SBh 1.1.1 na ttivad ayam
ca pratyag-titma-prasiddhe/:t first of all, it [the self] is not
a non-object in an absolute sense, because it is the object of the notion of I
and because the inner self is well known since it is directly experienced
ktirya-dar.saniit because of seeing the effect: because the effect is seen,
because we see the effect
In some cases the same meaning may be expressed by replacing the verbal
nOun with an abstract noun formed by adding ta or tva to the past passive
participle.
In the following examples note the use of the words abhilva, upapatti, asid-
dhi, and sambhava.
SDS p. 3 dehiitirikta iltmani pramilflabhiiviit because there is no
of a self separate from the body
SBh 1.1.22 naisa doso bhatakiiiasyapi vilyv-ildi-krame1}a
fault does not apply because the elemental ether
also can be a cause through the series beginning with air
because of the seen-ness of the effect: because the
effect is seen, because we see the effect
2.26.4 Verbal nouns expressing cause
Verbal nouns formed by adding the suffixes a, ana, -ana or ti to the root
are often used as the final member of a compound that expresses a reason
or cause. Often the resulting nominal construction is most conveniently
translated into English by a clause.
In the first example above ca connects sabda-bhediit and prakara,!iic; both
are causes of In the second example, the particle
ca connects and pratyag-atma- .
here is the cause of
This construction may be made even plainer by the addition of a word such
as kiiral}a; e.g., iti tasmiit kiiral}iit "for that reason." iti krtvii, which often
means "thinking thus," may also mean "therefore, for this reason."
Sometimes iti is followed by a pronoun or adverb that serves to distinguish
this use of itl from other common uses. For example, itl tasmiit, itl lena, tty
"therefore" (lit. "from that" where "that" refers to the content of the
iti clause).
GBh 2.10 Iqatram karma
atyantarrz kriiram api sva-dhanna iti
krtvii nddhannliya the action of a warnor, characterized by battle and
by violence to his teachers, brothers, sons, etc., though ex.tremely cruel, is
his own dharma; therefore it does not lead to adharma
211
2.29. Locative absolute
SBh 1.1.4 napi satJ1skliryo yena vylipiiram nor is
something to be purified, by reason of which it would depend on '
action
SBh 1.1.6 tathli pradhlinam api mahad-lidy-liklirel}a niyamena
pravartate yasmlic cetanavad upacaryate in the same way the
pradhlina also acts according to rule in the form of the mahat and so on;
for this reason it is figuratively said to be intelligent
When the reason precedes the conclusion, the conclusion is usually intro-
duced by a correlative pronoun or adverb such as tasmiit, tena, tat,
tatah, or tiivatii. The pronoun or adverb functions as a conjunction and has
the meaning "therefore."
Occasionally the relative pronoun or adverb is placed at the end of the
clause expressing the reason and the correlative is omitted.
GBh 2.14 sukha-duJ:zkhe punar niyata-nlpe yato na vyabhicarataJ:z but
pleasure and pain are constant in nature because do not vary
PPp. 258 yasmlid brahmavliptiJ:z tena tat jfilinenaptum
since obtaining brahman is the aim of man, therefore it
[brahman] is what one most desires to obtain through knowledge
Chapter 15. Causal Sentences 210
SBh 1.1.4 tasmiin na slistra-pranulf.lakatva'!'l
brahmal}ab sambhavatfty araft sva-tantram eva brahma
siistra-pramtiJ:wkam therefore it is impossible that brahman should be
made known by the s(istra as the object of an injunction to know;
consequently brahman is made known by the stistra as an independent
entity only
2.28 Relative pronouns and adverbs
A reason or cause may be expressed in a clause introduced by a relative
pronoun or adverb such as yasmiit, yena, yat, or yiivatii. The relative
in this case functions as a conjunction and has the meaning "because."
Such a clause may form a complete sentence that substantiates a previous
assertion.
One way to understand this construction is to supply the correlative before
the conclusion: the first example becomes ... yasmiit tasmiic cetanavad
upacaryate "because ... , therefore it is figuratively said to be intelligent";
the second example becomes ... yena tena vyiipiiram "because
. .. , therefore it would depend on action." See Chapter 20 for further
discussion of pronouns and pronomial adverbs.
SBh 1.1.14 itasca prlicuryarthe maya.t. yasmlid linanda-hetutvaJ!l
brahmal}o vyapadisati srutir hyevanandayatfti. And for this reason
the suffix maya means abundance. Because the scriptural passage "for he
alone causes bliss" shows that brahman is the cause of bliss.
Or the clause may substantiate an assertion made earlier in the same sen-
tence.
2.29 Locative absolute
Commentators often use the locative absolute construction to express causal
and conditional clauses. Usually the locative absolute precedes the main
assertion of the sentence. The basic significance of this construction, i,e.,
"this being the case," must be variously interpreted according to the con-
text.
212 Chapter 15. Causal Sentences 2.30. Causal particle hi
213
Note: included here are constructions involving a single abstract noun
in the locative case. While these are not strictly locative absolutes, they
are used in the same way (such locatives are regarded as nimitta-saptami
"locative of motive" by the Indian grammarians, though they may be con-
verted to locative absolutes without changing the meaning by adding sati).
2.29.1 Expressing a cause
The locative absolute may express a reason or cause, in which case it means
"given the fact that, since."
SBh 1.1.1 aham-mamabhimana-rahitasya
pramiitrtvanupapattau because the
prarniilJas could not function, since one who does not identify "I" and
"mine" with the body, senses, etc. cannot be a knower
2.29.2 Expressing a condition
The locative absolute may express a condition, in which case it means "if,
when."
SBh 1.1.1 hi satsupriig apt dharmajijiiiisiiyii
urdhvQrrz ca sakyate brahmajijfiiisitUf!ljiiiitu'!l ca for when these [the
desire for etc.] exist, it is possible to desire to know and to know
brahman, both after and even before the desire to know dharma
The condition may be hypothetical, in which case an optative often appears
in the conclusion.
SBh 1.1.2 sati brahma1JQ idarrz brahmalJa
sambaddha1'[l ka.ryam itl grhyeta for if brahman were an object of the
senses then this [world] would be perceived as connected with brahman as
its effect
2.29.3 Concessive locative absolute
The addition of api gives the locative absolute the concessive meaning
"even though" (see Section 2.54.2).
SBh 1.1.4 na ca 'pi
and, even though it has the nature
of an accomplished thing, brahman is not an object of perception, etc.
2.29.4 Abbreviated locative absolute
The locative absolute often appears in abbreviated forms such as tatraiva'!'
sati, eva'!' sati, and tatM sati. Each of these means "this being the case,"
where "this" refers to some previously stated condition. Note that the loca-
tive absolute in general consists of a "subject" and a "predicate," i.e. words
that would become the subject and predicate if the locative absolute were
made into an independent sentence. For example, in rtime gate "Rama be-
ing gone" the subject is rtime and the predicate is gate (the independent
sentence being gata rtimal.z). In these abbreviated forms the predicate con-
sists of the participle sant and the adverbs evam or tatM, while the subject
is either omitted or represented by tatra (equivalent here to tasmin; see
Section 2.55).
SBh 1.1.4 eva'!l ca saty athdto brahma-jijfiaseti
prthak-sastrdrambha upapadyate and this being the case, the
undertaking of a separate sastra which has that for its object, with the
words athdto brahma-jijfiasa, is appropriate
SBh 1.1.5 tatM sati yan-nirnittam pradhanasya tad eva
sarvajfia'!l mukhya'!l brahma jagatal:z. if that were the case
then the very cause of the pradhiina's being a seer would be the
omniscient and primary brahman who is the cause of the world
2.30 Causal particle hi
Areason or cause may be introduced by the particle hi "since, for." Usually
hi is placed after the first word of the sentence or clause in which it appears.
214
Chapter 15. Causal Sentences
SBh 1.1.1 SQf1JO hi puro'vasthire adhyasyati
since everyone superimposes an object on another object that is in front of
him
SBh 1.1.1 na hfndriyany anupiidiiya
sambhavati for without the senses the functioning of perception and so
on is not possible
Chapter 16
Uses ofiti
2.31 Introduction
The particle Iti "thus, in this way" has many uses in Sanskrit commentaries.
These uses all derive from the basic function of iti; it marks the preceding
word or words or sentences as a quotation of some sort. "Quotation" is
used here in a broad sense; the iti clause need not consist of words actually
spoken or thought and may describe more generally a reason, purpose, sit-
uation, and so on. This section discusses typical constructions employing
iti and the use of iti in quotation, explanation, and comment, as well' as
constructions employing the infinitive (both with and without iti) and the
use of iti to mark the end of a passage. Quotation marks have been added in
the examples to show the extent of each iti clause. See Section 2.27 for the
use of iti to cite a reason or cause and Sections 2.44.3 and 2.45.3 for ityiidi,
itivat, and ityiidivat. The various uses of iti in glossing are discussed in
Part One. (See also: Coulson 203-D5, 230-35; Speijer 383-90,491-99.)
The words quoted by iti usually form a complete statement. If a single
word is quoted it normally appears in the nominative case.
"devadatta" iti tl1f!l vadanti they call him Devadatta
Generally the iti clause stands first in the sentence. In some cases words be-
fore the iti clause are construed with words following iti, and it is possible
215
216 Chapter 16. Uses of iti 2.33. Quotation 217
for one iti clause to include another; see Section 2.23 for such construc-
tions. The words immediately after iti normally indicate the relationship
of the iti clause to the rest of the sentence. However iti is also used alone,
without other explanatory words, to mark words spoken (= ity ukrva) or
thought (= iti marva) or more generally to give an explanation of some
sort.
"indra upagacchati" iti matvii paliiyate = "indra upagacchati" iti
paliiyate he flees, thinking "Indra is coming"
Occasionally demonstratives such as evam and etad are added redundantly
to iti.
SBh 1.1.12 "tQsmiit SQf(lsiiry evanandamaya iitmii" ity evarrt priipte
therefore the self consisting of bliss is just a transmigrating being-the
piirvapa4a being thus established
SBh 1.1.4 tasmiid "aham brahmasmi" ity etad-avasiinii eva sarve
sarviilJi celara,:,i pramii'pjni therefore all the injunctions and
all the other prarniilJas come to an end in the realization "I am brahman"
Here ity eva'!' prapte =iti prapte and ity etad-avasana =ity avasana.
2.33 Quotation
2.33.1 iti with source
iti may quote the actual words of a person or text, or the views held by a
particular person or school of thought. Usually iti is followed by a word
or phrase giving the source of the quotation. A verb mearting "to say," "to
think," etc. may be added but is not necessary.
SBh 1.1.2 "srotavyo mantavya" iti the scriptural passage "[the
self] is to be heard, is to be thought about"
SBh 1.1.5 nityanumeya" iti sankhya
manyante the Sankhyas believe that the connections of the pradhana
with the must necessarily be inferred
2.33,2 iti with ablative
Often the word giving the source of a quotation'ls put into the ablative
case. Here the ablative does not signify that the words are quoted from that
source; rather, it means "because ... says ...."
This construction is very Common in appeals to scriptural authority.
SBh 1.1.5 yat tujiiiina,!, manyase sa "sattvtit saiijayate
jiianam" iti smrteJ:z but what you think is knowledge is actually the
quality of sattva because smrti says that knowledge arises from sattva
2.33.3 iti with partial quotations
A commentator may wish to refer to a long passage without quoting the
passage in its entirety. In this case iti is often used to quote selected sen-
tences from the passage. These quoted sentences may be connected by
expressions such as the following;
... ity arabhyaha ... iti beginning thus ... the text goes on to say ...
tathii ... iti similarly the text says ...
ante ca ... iti and the text concludes ...
. .. ityiidi... ityantaTJl the beginning with. .. and'
ending with ...
An example:
SBh 1.1.2 ki,!, punas tad vediinta-viikya,!, yat sutre1}eha
"bhrgur vai varu1}ir varu!Ulf!l pitaram upasasara. adhfhi bhagavo
brahma" ity upakramyaha "yato va imani bhiltiini jayante. yena jiittlni
jfvanti. yat prayanty abhisa,!,visanti. tad vijijiiiisasva." tasya ea
"anandad dhy eva khalv imiini bhiltanijayante.
anandena jatani jfvanti. ananda1[l prayanty abhisa'?lvi{anti" iti.
But which Vedanta text does the siltra wish to indicate here? Beginning
with "Bhfgu Viiruni indeed approached his father Varuna and said ',Sir,
teach me brahman' ," the passage goes on to say "from which indeed these
beings are born; by which, when born, they live; into which, when dying,
2.34 Explanation
SBh 1.1.7 taSmac iJultma-sabda" iii niscfyate
therefore it is settled that the word atman here refers to a conscious entity
the past passive participle (neuter singular) are most often used in this con-
struction. Note that these two are identical in meajling; e.g., iti niscitam
=iti niscfyate. Occasionally one finds the synonymous construction iti
niScitafJ1 bhavati.
219 2.35. Comment
Chapter 16. Uses of iti
they enter---desire to know that." And its concluding sentence is "from
bliss only, of course, these beings are born; by bliss. when born, they live;
into bliss, when dying, they enter."
218
An iti clause is often used to explain a particular term or concept. Some-
times the iti clause anticipates what a person ntight think or say in the
situation being described.
SBh 1.1.1 "aham ida,!, mamedam" iti naisargiko 'ya'!1[oka-vyavahiira!}
there is this natural worldly practice expressed in the thought "this is I,
this is mine"
SBh 1.1.1 sarva hy titmastitva'!' pratyeti na "niiham asmi" iti for
everyone knows the existence of the self; no one thinks "I am not."
Occasionally the iti clause follows the tenn that it explains.
SBh 1.1.1 na diyam asti niyama!} "puro'vasthita eva
adhyasitavyam" iti nor is there an invariable rule that an
object can be superimposed only on another object that is in front of us
SBh 1.2.1 punar apy anyiini vakyany
sandihyame "kitJ1 paratJ1 brahma pratipa.dayanty aha svid arthtintaratJ1
ki,!, cid" iti but other passages, in which the characteristic marks
of brahman are not clearly set forth, give rise to doubt: do they refer
to the highest brahman or to something else?
Here the meanings of and sandihyante are specified by the subse-
quent iti clauses.
SBh 1.1.11 tasmat "sarvajiiaf[l brahma karalJ-af!l nacetana1'J1
pradhanam anyad va" iti siddham therefore it is established that the
omniscient brahman is the cause of the world, not the unconscious
pradhtina nor something else
SBh 1.1.14 yo hy anyan anandayati sa pracurtinanda iti prasiddha'!1
bhavati for it is well-known that one who causes others to be blissful
must himself abound in bliss
A gerundive, an adjective, or a substantive may also be used with iIi.
SBh 1.1.4 "yasya tu yathii-pilrvatJ1 saf!lsaritvaf!l nasav
avagata-brahmatma-bhava" ity anavadyam it is irrefutable that one
who is a transmigrating being as before has not comprehended that
brahman is his self
PPp. 163 "napi jiiiinantaram utpannam" iti nor does another
cognition arise-this is the difference
Note that some of the phrases commonly used in glossing, e.g., ity
(see Section 1.10) or ity (see Section 1.44), employ the same con-
struction.
2.35.2 Commonly used comments
2.35 Comment
2.35.1 Comment with iti
Some of the most common comments with iti are listed below. In some
cases more than one fonn is used; e.g., iti viruddham, iti virudhyate, and
iti are all used to mean "this is a contradiction."
An author may comment on a statement by adding iti and a word or phrase
that conveys the desired comment. The passive (third person singular) and
ity anavadyam this is irrefutable
ity abhyupagatam this we accept
Uniersilat Hamburg' Asien-Afrika-Institut
AbtellulIg fijr K"ltlH und GlIschichte
220
Chapter 16. Uses of iti 2.36. Constructions with the infinitive 221
ity abhyupetam
fty avasitam
ity upapannam
iti gamyate
itf
itf niscitam
iti prasiddham
iti btidhitam
iti bhaviryati
iti yuktam
iti viruddham
iti sambhiivyate
iti siddham
itf sugamam
itf
this we accept
this is settled
this is reasonable
this can be inferred
this is seen (in ordinary experience)
this is settled
this is generally known
this is ruled out
this might be
this is reasonable
this is a contradiction
this is possible
this is established
this is clear
this is clear
devadattaJ:! saknoti gha[Q7[l kanum it is possible for Devadatta to make
a pot (active)
devadattena sakyate gharaf!l kartum it is possible for a pot to be made
. by Devadatta (passive)
In the examples below the infinitive has a passive sense and may be ren-
dered in English by a passive infinitive (e.g., "to be made").
SBh 1.1.25 katha1J1punas "chando'bhidhiiniin na brahmabhihitam" itf
sakyate vaktum but how can it be said that brahman is not denoted
because the meter is denoted?
SBh 1.1.4 ata sa [titmii] na kena cit pratytikhytitum takyo
va netum therefore it [the self] cannot be denied by
anyone, not can it be made subordinate to an injunction
2.36 Constructions with the infinitive
The following constructions with the infinitive are used both to govern an
iti clause (compare comments with iti in Section 2.35) and by themselves.
2.36.1 Infinitive with sak and yuj
The infinitive is often used with the passive or the gerundive of sak "to be
able," or with the passive or the past passive particle of yuj "to fit, to be
suitable."
iti sakyate vaktum =- itf sakya'!1 vaktum this can be said, it is possible to
say this
itf yujyate vaktum == itf yuktaTJ'l vaktum this can reasonably be said, it is
reasonable to say this
Note that the infinitive in Sanskrit has only one form, and acquires an active
or pasive sense according to the mode of the accompanying verb.
SBh 1.1.1 ... cddhyiisal}
tad-dhanniil)ii'!l ca 'dhyiiso mithii" iti
bhavitufJ1 yuktam it is reasonable that superimposition of the object and
its qualities on the subject and conversely superimposition of the subject
and its qualities on the object are false
SBh 1.1.7 na tu siistram apramiil}af(l kalpayitufJ1 yuktam but
it is not reasonable to imagine that the faultless siistra is not a
Note that sakyam and yuktam sometimes appear as impersonal indeclin-
abIes in this construction; e.g., in the second example above sakyo agrees
with sa, but the sentence would also be correct if sakyam were substituted
for sakyo.
2.36.2 Infinitive with ark
The infinitive may be combined with the verb arh "to be obliged to" to give
the sense "ought, should." The infinitive of bhii is particlilarly common in
this construction; e.g., iti bhavitum arhati "this must be the case."
SBh 1.1.12 para evatmanandamayo bhavitum arhati the [sell]
consisting of bliss must be the highest self
The infinitive sometimes expresses aim or purpose.
2.36.3 Infinitive expressing purpose
Compare this construction with the common use of iti to mark the end of a
chapter or an entire work.
223 2.37. End ofa passage
iti srfmac-chtirfraka-mfmti'!tsti-bhtirYe srf-arikara-bhagavat-ptidakrtau
prathamiidhyiiyasya thus ends the first ptida of the
first adhytiya of the commentary on the Srfmac-chiirfraka-mfmii'!lsti
composed by Sn-sailkara-bhagavat-pada
SBh 1.1.4 niirhaty evaf!l bhavitu'!J karya-vidhi-prayuktasyaivQ
brahmaflQI; pratiptidyamiinatvtit it does not have to be so, since
brahman is taught only in so far as it is connected with injunctions to
action
Chapter 16. Uses of iti 222
riimaf{t agacchat he went to see Rama (i.e., in order to see
Rama)
This sense of purpose may be conveyed by an infinitive governing an iti
clause.
PPp. 30 "nyonya-dharmiif ca" iti Prthag dharma-graha1JlllJ'l
"dhanna-miitrasyiipi kasya cid adhyasa" iti darsayitum the separate
mention of the word dharma in the phrase anyonya-dharmiii ca serves to
show that there may also be superimposition of some quality only
2.37 End of a passage
iti sometimes marks the end of a section of text. This usage is particularly
common in passages expressing view (see Chapter 19).
SBh 1.1.1 tat punar brahrna prasiddham aprasiddhaf!l va sytit. yadi
prasiddhalJ'l na jijiiiisitavyam. athtiprasiddhalJ'l naiva fakya'!l jijiitisitum
iti. ucyate... But that brahman must be either well-known or not
well-known. If it is well-known then one need not desire to know it; if it is
not well-known then it is impossible to desire to know it. We reply ...
A much longer example is found in SBh 1.104, where a discus-
sion of many lines is concluded as follows:
tasmtit ftistra-pramtir:zakarrz
brahmtibhyupagantavyam iti. atrtibhidhfyate . . . Therefore brahman is
to be accepted as proven by the flistra only in so far as it is the object of
the injunction to know. Here we reply ...
Chapter 17
References
2.38 Introduction
The most common reference to another person or text is a direct quotation
followed by iti and the name of the person or text quoted (see Section 2.33).
This section discusses other terms and constructions that are often used in
making references. These will be taken up according to the nature of the
reference: a commentator may refer to the text or the author of the text he is
commenting on, to his own work, to various recognized authorities, ~ to
persons holding a particular view. The traditional terms mula and mulakiira
are used in this section to refer to the text being commented on ~ the
author of the text being commented on, respectively. See Section 2.57.2
for the use of etat and idam to refer to preceding and subsequent passages
ofa work.
2.39 Text and author of text
2.39.1 Verbs of taking
erbs of taking-(
Derivatives of the roots grah "to grasp," parigrah "to embrace," and
upiidii "to take" may indicate what is specifically mentioned or referred
to in the mula. For example: upiidiina "mention of or reference to in the
mula," grhfta "mentioned or referred to in the mula."
225
2.39.3 Author of text commented on
The malakara may be referred to directly by name, by an honorary title
such as bhagavan, or by a descriptive term such as satrakara "author of
the satras" or "author of the
2.39.2 prakrta, aprakrta
The past participle prakrta "being the subject of discussion" may refer to
what is being talked about in a passage of the mala. Conversely a-prakrta
points to what is not the subject of discussion.
iisarikate he raises a doubt
227 2.39. Text and author of text
2.39.4 Author as understood speaker
udiiharati he cites as an example
iirabhate he begins
upasafJ'lharati he summarizes
GBh 2.48 yadi karma-phala-prayuktena na kartavyarrz karma katharrz
tarhi kartavyam fty ucyate if action is not to be performed by a person
impelled by the fruit of action, then how is it to be performed? the
mulaktira explains (in the following verse)
GBh 2.23 kasmiid [atma] avikriya evety aha why is [the self]
changeless? the millakiira explains (in the following verse)
The malakara is often the unexpressed subject (aragent, in a passive con-
struction) of a verb in the third person singular. Commentators use such
verbs (especially in the avataral}a or brief introduction to a verse or other
portion of the mala) to explain the intentions of the malakara in a given
passage. For example, aha "the malakara says" often introduces a quote
from the mala.
PP p. 135 tat-siidhaniirtham iiha pratyagiitma-prasiddher iti in order
to substantiate that, the millakiira says "because the inner self is
well-known"
Note that ucyate is also used to introduce the commentator's own reply to
an objection (see Section 2.50.1). Some other common verbs used in this
construction are listed below (he = malakara).
Other phrases employing aha include ity asmikYaha "having raised this
doubt the miilakara says" (see Section 2.49.4), aha "the malakara
gives an example," and uttaram aha "the malaMra replies." Note that aha
sometimes means "an objector says" and serves to introduce an objection
(see Section 2.49.6). ucyate "is said by the malakara" may introduce a
quote from the mala, especially where this answers a question or objec-
tion.
SBh 1.2.1 sarvafJ1 khalv ida,!, brahmhi sva-sabdemiiva brahmopiittam
brahman is directly mentioned by its own name in the passage "of course
all this is brahman"
Chapter 17. References
SBh 1.1.26 itas caivam abhyupagantQvyam asti purvasmin vakye
prakrtarrz brahmeti and for this reason we must accept the fact that
brahman is the topic of the previous passage
SBh 1,1.12 kim ihanandamaya-sabdena param eva brahmocyate
yat-prakrtaT!J satya'!1 jniinam anantaf!l brahmeti does the word
iinandamaya here refer to the highest brahman that is the subject of the
passage "brahman is true, knowledge, endless"?
SBh 1.1.2 kimlalqa'}a",. punas tad brahmety ata iiha bhagaviin
siltraktiraJ:t but what is the definition of that brahman? the blessed
author of the siltras explains (in the following sutra)
SBh 1.1.20 tac CQ gha!ate and that makes
sense if the reference in the mala is to the Supreme Lord
SBh 1.1.22 tasmi'id iiktisa-sabdena bhutiiktifasya graha1}am therefore
by the word "ether" there is reference in the mula to the elemental ether
erbs of taking-)
226
2.39.5 Participles of verbs of saying
SBh 2.1.6 itf ca kevalasya tarkasya
vipralambhakatvQl'J'l and the siltrakiira will show the
fallaciousness of mere reasoning in the sutra
The future tense refers to a passage of the mula subsequent to the one
presently under consideration; e.g., vakerati "the mulakiira will state,"
"the mulakiira will show."
229 2.40. References to the commentary
SBh 1.1.3 kim-arthaf(l tarhfdaf(l sutraf(l yavata
evilivarrz-jiitfyakaf(! iiistram udliharatli ilistra-yonjtvarrz brahmaflo
darsitam then what is the purpose of this satra, in as much as by
adducing such a iiistra on the previous sutra it was shown that the slistra
is the source of brahman?
2.39.7 Auto-commentary
Here the present participles udiiharatii and pratijiinatii properly refer to
sutrakiire/'}a, the unexpressed agent of darsitam and sucitii, respectively.
Note that neither the Sanskrit nor the English translation requires mention
of the word sutrakiira.
PPp. 137 nanu brahma-vidyiim anartha-hetu-nibarh01}Trrz
sucitii objection: by asserting that
knowledge of brahman destroys the cause of evil, it is shown that
ignorance is the cause of evil
When an author writes a commentary on his own work (a so-called "auto-
commentary") he normally refers to the rna/aMra as if the mulakiira were
another person. Thus the third person is used (not the first person, as one
might expect), and the other terms and constructions described above' are
employed where appropriate.
variiayati he describes
darsayati he shows
prapaficQYQti he explains in detail
pariharati he refutes
Participles of verbs meaning "to say," "to show," etc. may refer to the
mulakiira; e.g., ukta may mean "stated by the mulakiira," may
mean "to be stated by the mulakiira," and may mean "intended
by the mulaMra" (lit., "what he means to say").
228
SBh 1.1.1 brahma ca asya yata iti
and the definition of brahman will be stated by the sutrakara in the satra
"from which there is the origin and so on of this (world)"
SBh 1.1.31 triwvidham iha brahmopasanaTJl vivaJqitam a three-fold
meditation on brahman is intended here (Le. in this passage) Note that
such participles may also refer to the commentator's own work (see
Section 2.40.2).
2.39.6 Instrumental of present participle
The instrumental of a present participle may refer to the mulakiira in pas-
sive constructions where the agent (the mulakiira) is not actually men-
tioned.
2.40 References to the commentary
2.40.1 First person plural
A commentator often refers to his own work with a verb in the first per-
son plural. The past and future tenses refer to preceding and subsequent
passages of the commentary.
SBh 1.1.4 nityam asarfratvam akarma-nimittatviid ity avociima we
said earlier that bodilessness is eternal because it does not have action as
its cause
SBh 1.1.1 yathii cilyam arthaf:t vediintiiniirrz tathii
asyiirrz siirfraka-mfmaf(lSiiyal'J'l and how this is the
meaning of all the Vedanta texts we shall show in this
230 Chapter 17. References 2.41. Recognized authorities
231
2.40.2 Participles
Participles of verbs meaning "to say," "to show," etc. may refer to the
commentator's own work. The past participle ukta is particularly common
in such references; e.g., ukta "previously stated," yathiJkta "as previously
stated," ukta-nyiiyena "in the manner previously stated."
ORb 2.21 nanCtkta evatmano 'vikriyatvQ'!l
objection: it was previously
said that the particUlar cause of the impossibility of all actions (in the self)
is the fact that the self does not change
SBh 1.1.5 brahma ca sarvajfiaf[l sarva-sakti
jagad-utpatti-sthiti-ntisa-kiira!lam iry uktam and it has been said that
brahman is the omniscient. omnipotent cause of the origin, subsistence,
and destruction of the world
Note that such participles may also referto the mulakiira (see Section 2.39.5).
2.40.3 Relative clause
A relative clause introduced by yat and concluded by iti may quote some
earlier statement of the commentary. This construction is especially com-
mon where an earlier point is raised again in the course of discussion. For
example, yat tuktam '" iti or simply yat tu . .. iti may mean "what was
said previously, i.e., ...."
SBh 1.1.4 yat tu hey6ptideya-rahitatviid upadesananhakyam iti
dosafi. but as for what was said before, i.e., that the teaching is
purposeless because it is free of anything to be avoided or sought
after-that fault does not apply.
When the correlative tat is used it refers to the entire statement quoted by
yat and iti.
SBh 1.1.5 yat tl1ktarrz jfitinena sarvajfiarrz pradhtina'!1
bhavifJatfti tan nopapadyate but what was said before, i.e., that the
pradhiina may be omniscient because knowledge is a quality of
sattva-that is unreasonable.
Note that the same construction (but not with ukta) may be used to quote
something other than the commentator's own wodc; e.g., a rival interpreta-
tion. .>
SBh 1.1.4 yad api ke cid
prayrtti-nivrUi vidhi-tae -vyatireke kevala vastuvtidr vedabhtigo
niistfti tan na and what some say, i.e., that no portion of the Veda
declares a mere thing separate from injunctions to act or to refrain from
acting or from something subordinate to such injunctions-that is wrong
2.40.4 Reference to earlier topic
An abbreviated reference to earlier commentary may be made by mention-
ing the topic of the earlier discussion.
vtiyunii vytikhytite this is explained by vtiyu: this is explained by what
we said concerning viiyu
2.40.5 agre
The adverb agre "in front of" means "later on, below" when used in r,fer-
ence to the commentator's own work.
evam agre thus it will be explained later on
2.41 Recognized authorities
Special terms may be used to refer to recognized authorities. The choice of
such terms varies according to the subject matter of the work, the particular
commentator, and the nature of the authority referred to. A few examples
are given below.
Derivatives of iimnii refer to what is taught in the sacred or traditional
texts; ity iimniiyate "this is taught in the sacred texts," iimniita "taught in
the sacred texts."
Derivatives of drs refer to what is seen in ordinary iti
drsyate = iti "this is seen in ordinary experience," iti drstam loke
232
"this is seen in the world," -darsaniit (final member of compound) "because
it is seen that... ,"
The word nyiiya refers to a popular maxim: iti nyiiyiit "because of this
popular maxim."
Derivatives of prasidh refer to what is generally known in the world:
iti prasldhyate = itl praslddham "this is generally known," ItI praslddha,!,
loke "this is generally known in the world," -praslddhe!, (final member of
compound) "because it is generally known that. ..."
Derivatives of smr refer to what is taught in smrti: itf smaryate "this
is taught in smrti," -smara,:,at (final member of compound) "because smrti
teaches that. .. ."
Derivatives of sru refer to what is taught in srutl: ItI sruyate "this
is taught in sruti," -srava,:,iit (final member of compound) "because sruti
teaches that. .. ."
2.42 Genitive in reference to particular views
Persons who hold a particular view may be referred to by a word in the
genitive case, where the genitive is construed with a word or phrase that
expresses the view in question. The reference may name the persons di-
rectly.
SBh 1.1.4 yatha prthivy-adi jagan-nityatva-viidfnii'!' yatha ca
slirikhyanti1J1 gW}liJ:z like the earth and so on for those who believe that
the world is eternal, and like the gu,:,as for the SiiI\khyas
Or a general reference may be made to anyone who holds such a view.
SBh 1.1.4 yasya tiltplidyo molqas tasya mlinasQrrz vlicika1J1 kiiyikaf[l va
kiiryam apelqate for whomever is something to be produced, for
him it depends on a mental, verbal, or bodily effect
Commentators normally gloss such a genitive with the word mate "in the
view." For example, in the examples above siirikhyiiniim would be glossed
siirikhyiinii,!, mate "in the view of the Sailkhyas," while yasya would be
glossed yasya mate "in the view of whomever."
Chapter 18
Examples and Comparisons
2.43 Introduction
Certain terms are commonly used to introduce examples and comparisons.
bahuvrfhis in iidi often give specific examples of a more general term.
Comparisons may be drawn by adverbs formed with the suffix vat or by
phrases containing iva or yathii. Such comparisons may give a corroborat-
ing example (e.g., "this situation is exemplified in this instance") or may
comprise the main statement (e.g., this situation is like that situation") .. See
Section 2.53.7 for the use of tathii hi and tad yathii. (See also: Coulson 49-
50,195,214-15; Speijer 229.1, 241, 430, 470 a; Whitney 1107.)
2.44 Bahuvnnis in lidi
2.44.1 Compounds in -iidi
iidi m. "beginning" is very common in bahuvrfhi compounds whose literal
meaning is "of which ... is the first, beginning with ... " Sometimes the
prior member of the compound is in fact the first element of a series and
the compound is best translated literally. In such cases abahuvrfhl in anta
m. "end" often expresses the final element of the series.
SBh 1.1.19 when the parts
beginning with the head and ending with the tail have been mentioned
233
234
Chapter 18. Examples and Comparisons 2.44. BahuvrThis in adi 235
More often the prior member of the compound is simply a characteristic
or prominent member of the group that is denoted by the compound as a
whole. Here a less literal translation of tidi is more appropriate.
indradayo devliJ;. the gods, IndIa and so on; IndIa and the other gods
akiisitdfni bhiUani the elements, ether and so on; the elements, such as
ether
Here bahuvrfhis in tidi give examples of the general terms they modify:
Indra is a particular god and ether is a particular element. Often the general
term is unexpressed and the bahuvrfhi is used as a substantive.
tigatii riimiidayaJ:z Rama and the others came
Here mentions a particular individual (Rama) and denotes (in
addition to Rama) the set of individuals associated with Rama. Note that
the bahuvrfhi is often singular in this construction, even though it refers to
more that one thing.
SBh 2.1.24 nahi viiyur akliso baliid dadhi-bhavam
lipadyate for neither air nor ether is compelled by heat and so on to turn
into curds
bahuvrfhis of the same meaning as those in tidi may be formed by tidika
(see Section 2.15), the related word tidya, and other words meaning "be-
ginning" such as prabh.rti.
indradayo devii/:z = indrddika deviif:z ;;;; inddidyii devaJ;. =
indraprabh[1ayo devil/;.
2.44.2 Series of compounds
Several bahuvrfhis in tidi may be used in a single statement in order to
make the sentence logically complete. Often it is not necessary to translate
each tidi by "etc." or "and so on," and a more natural translation can be
obtained by considering the purpose of the tidis.
SBh 2.1.1 sarvajfiaJ:z sarvesvaro jagatQ utpatti-kiiral}Q'!l
mrt-suvanylidaya iva gha.ta-rucakiidfniim the omniscient Lord of all is
the cause of the origin of the world, just as clay gold, for instance, (are
the causes of the origin) of a pot and a gold ornament
SBh 1.1.1 pasviidayaJy sabdfidibhiJy srotradfntirJ1 sambandhe sati
sabdadi-vijfiline pratikale jate tato nivartante when there is COntact of
sense objects such as sound with sense organs such as the ear and when an
unpleasant cognition of sound, etc. has arisen, cows and other animals
move away from that
In the first example above the tidis simply show that other related exam-
ples could be named if desired, and so the tidis may be rendered by "for
instance." In the second example each bahuvrfhi in tidi implies a more or
less well defined set of items (e.g., sro/radi "the ear and so on" refers to
the five sense organs), and so the tidi compounds may be rendered by the
general terms that they imply.
2.44.3 ityiidi
The compound ity-tidi (lit., "of which the first is thus") may be used, in
direct quotation to indicate that the quotation is one of several that might
be named. In this case ity-tidi has the sense "quotations such as. " ." Like
other bahuvrfhis in tidi, ity-tidi may modify another word in the sentence,
or the word that it modifies may be unexpressed, or ity-tidi may occur in
compound.
SBh 1.1.4 kim-arthtini tarhy titmli vli are srotavya ity-tidrni
vidhi-cchtiytini vacantini then what is the purpose of passages that
resemble injunctions, such as "the self indeed, my dear, is to be seen, is to
be heard"?
SBh 1.1.12 evam tiktisas tal-lingtid so it is to be
regarded in passages such as tikiisas tal-Ungtit
SBh 2.1.7 sarvaTJ'l taTJ'lparlidtidyo 'nyatratmana}y sarva'Jl
vedetytidi-srava1}lit because of sruti passages such as "everything has
deserted him who knows everything in something other than the self'
236
Chapter 18. Examples and Comparisons
2.45. Suffix vat 237
Sometimes evam is added redundantly to form expressions such as ity-
evam-lidi (= ity-lidi) or iti caivam-lidi (= ity-lidi cal.
2.45 Suffix vat
2.45.1 Adverbs in -vat
The suffix vat (neuter accusative singular of the possessive suffix vant) of-
ten forms adverbs meaning "like, as." Such adverbs may be made from any
noun, e.g., dehavat "like the body." The grammatical relationship between
the noun and the other words of the sentence is not indicated in the form of
the adverb and must be understood from the context. The correct use of vat
when it means "like, as" is limited by PaJ.lini 5.1.115-116 to constructions
that can be paraphrased in one of the following ways:
brlihmanavad vartate :;;;: brahmaneIJa tulyaf!l vartate he acts like a
(paraphrase by instrumental of noun with tulyam-similarity of
action must be denoted)
rnathurtivat srughne :;;;: mathurtiyam iva srughne the
wall in Srughna, like the one in Mathura (paraphrase by locative of noun
with iva)
devadattavad yajiiadatwsya gavall :;;;: devadattasyeva yajiiadattasya
Yajiiadatta's cows, like Devadatta's (paraphrase by genitive of noun with
iva)
However in actual practice adverbs in vat are often used incorrectly, espe-
cially in popular literature; e.g., mlitrvat para-diirlin paiyati = miitaram iva
para-dlirlin paiyati "he looks on another's wife as he looks on his mother"
(paraphrase by accusative of noun with iva).
2.45.2 With standard of comparison
Adverbs in vat often express a compatison where the noun to which vat is
added is the standard of compatison (upamlina).
SBh 1.1.4 sa ca dagdhendhanagnivQt svayam ev6pasamyate and that
extinguishes itself, like a fire whose fuel has been consumed
Here dagdhendhanagnivat = dagdhendhanagnir iva (note that vat is used
correctly here, since dagdhendhanagnivat svayam ev"pasamyate can be
paraphrased by dagdhendhanagninli tulya,!, svaya;" ev"pasamyate, and
the point of compatison is the act of extinguishing itself). Often the suffix
vat is added to bahuvrfhis in lidi to indicate that the noun mentioned in the
bahuvrfhi is one of several that might be named for the sake of compatison.
dehiidivat like the body and so on; as is the case, for example, with the
body
Sometimes the noun to which vat is added is construed with another word
of the sentence (see Section 2.14 for a parallel construction in which the
prior member of a compound is construed with a word outside the com-
pound).
SDS p. 2 tebhya eva
mada-saktivac caitanyam upajiiyate consciousness arises from just
those [four elements] which are transformed into the form of the body, just
as the power of intoxication (arises) from yeast and so on
2.45.3 itivat, ityiidivat
The adverbs itivat "as is the case in this quotation" and itylidivat "as is
the case in quotations such as this" are used when direct quotations are
presented for compatison.
SBh 1.1.4 napi viiyur VQVQ
priil}a viiVQ saTJ'lvarga itivat nor is [knowledge of the self as brahman]
based on connection with a particular action, as is the case in the passages
"wind indeed is the absorber," "breath indeed is the absorber"
SDS p. 6 mama sarfram iti vyavahiiro sira ityiidivad
the expression "my body" is figurative, as is the case in expressions such
as "Rahu's head"
238 Chapter 18. Examples and Comparisons
2.46 iva and yathii
Comparisons are often expressed by iva "like, as" or yatha "like, as." If the
standard of comparison (upamiina) is one word, iva is placed after it; if it
consists of several words, iva is often placed in the middle of them. yatka
usually precedes the standard of comparison and is particuarly common
when the standard of comparison is expressed by a verbal clause.
SBh 2.1.4 nanu dharma iva brahmal}y apy agamo bhavitum
arhati objection: the sacred texts must be independently authoritative
with regard to brahman also, just as (they are) with regard to dhanna
SBh 1.1.4 svatma-dhanna eva sa1'Jls tiro-bhuto kriyayatmani
SQf!lskriyamiine 'bhivyajyate yathildar.se
sQ'J'lskriyamtine bhiisvaratva'!1 dhannaJ.1 is a hidden quality of
the self that is manifested when the self is purified by action, just as the
quality of brightness (is manifested) when a mirror is purified by the act of
rubbing
Comparisons may be made more explicit by using the correlatives yatka ...
tathii (or evam or ittham) ... . If the subject under discussion is compared
to something else, then the former is normally treated as the subject of
comparison (upameya) and placed in the tatka clause.
SBh 1.1.4 yatha svargiuii-kiimasyllgnihotriidi-sadhana'!1 vidhryata evam
amrtatva-kiimasya brahma-jfiana'fl vidhryate just as means such as the
agnihotra are enjoined for one who desires heaven and so on, so
knowledge of brahman is enjoined for one who desires immortality
Sometimes this construction is abbreviated by replacing the tatka clause
with tadvat, lit., "like that," i.e., "so it is here (in the case under discus-
sion)."
SBh 1.1.1 yatha rajasau gacchatfty ukte sa-parivarasya rajno gamanam
ukta,!, bhavati tadvat just as the sentence "there goes the king" implies
the going of the king with his retinue, so it is here
Chapter 19
Terms of Discussion
2.47 Introduction
Discussion of the meaning of the text being commented on usually takes
the form of a dialogue between the commentator and one or mOre oppo-
nents who raise objections to his interpretation. The parts of the dialogue
are marked by certain conventional terms that may be roughly divided into
those used to present an objection and those used to reply to an objection.
Certain terms also mark an alternative interpretation in cases where .the
commentator is willing to admit another view.
2.48 Dialectic style
2.48.1 Oral debate
Analysis of the meaning and implications of the text being commented on
(as opposed to analysis of the words of the text, discussed in Part One)
usually appears as a dialogue whose form is based, both historically and
stylistically, on oral debate. The view of the commentator and his school
is known as the siddhiinta, lit., "the demonstrated conclusion." Oppos-
ing views are included under the term lit., "ihe prior view."
The piirvapakea may be the established doctrine of another school, or it
may be any question or doubt anticipated by the commentator. The dia-
logue that leads to proof of the siddhiinta may take several forms. Often
a problem is posed, a is put forth, and a debate develops be-
239
tween the parvapakJin (the upholder of the parvapakJa) and the siddhiintin
(the upholder of the siddhiinta, i.e., the comentator) that eventually leads
to confirmation of the siddhtinta. Sometimes the siddhiinta is stated first,
only to be attacked by the purvapakJin and defended in the ensuing de-
bate. In longer arguments the siddhtinta may be confronted with a series
of purvapak1as. In this case a commentator often permits one purvapak1in
to debate another (or several others) before the siddhtinta is finally estab-
lished. Such passages deserve careful reading since the reply to an ob-
jection may not represent the commentator's own view but may be the
reply of one purvapak1in to another. Note that when the text is a collec-
tion of sutras, some sutras may express a purvapakJa but lack any of the
terms described in Section 2.49 that ordinarily mark an objection. How-
ever purvapak1a sutras (both with and without the special terms that mark
an objection) are usually clearly identified as such by the commentators.
SBh 1.1.1 ca pratyag-iitmano
and you (Le., the siddhantin) say that the inner self which is free
of the notion of you is not an object
SBh 1.1.5 ida,!! tavad bhavan katharJ1 nitya-jniina-kriyatve
sarvajnatva-hanir iti first of all, you (Le., the are to be
asked this: why should [brahman's] always being engaged in cognitional
activity cause it to lose its omniscience?
241
2.48. Dialectic style
2.48.3 Direct address
The influence of oral debate on the dialectic style of Sanskrit commentaries
is seen most clearly when the siddhtintin and the piirvapak1in address one
another directly. Second person pronouns and second person verbs are not
uncommon in such passages.
Chapter 19. Terms of Discussion 240
2.48.2 and siddhiinta
The terms purvapakJa and siddhiinta are occasionally used directly.
SBh 1.1.4 the word tu (in the
sutra) serves to rule out the
When the piirvapak1a is the well known view of another school it may be
identified as such.
SBh 1.1.5 kii1:uldiis tv etebhya eva vakyebhya fsvara1J1 nimitta-kiira1)am
anumimate 'IJu'!lS ca samaviiyf-kiirw:zam the followers of Kal).ada on the
other hand infer from these same passages that the Lord is the efficient
cause and the atoms are the material cause
More often the piirvapak1a and siddhtinta are identified by the terms dis-
cussed in Sections 2.49 and 2.50.
The word pak1a "side, alternative" may refer to any particular view.
SBh 1.1.5 ttini [vediinta-viikyiinij 'pi yojayituT[l
sakyante those [Vedanta passages] can also be interpreted on the view
that the pradhiina is the cause
2.48.4 Example of dialectic style
The following passage is an example of the dialectic style and illustrates
some of the terms discussed in Sections 2.49 and 2.50.
PPpp. 100-101
mithyiiivdhankartrtvam ivopadhana-nimitto
lohitimli. Therefore the self's thinking "I am the agent," caused by tIle
coloring (of the self) by the mind, is false, just as the red color in a crystal
caused by proximity [to a redjapa flower] is false
katha'!1 spha!ike lohitimno mithyatvam. But how is the red color
in the crystal false?
ucyate yadi spha!ika-pratisphiilitii niiyana-rasmayo japii-kusumam
upasarpeyus tadii tad eva lohita'!1 na hi
pratyayo napi svasrayam
rupa-mlitra'!1 pratibimbita'!1 kva cid upalabdha-purvam. We
reply: if the ocular rays reached the japa flower after bein"g deflected by
the crystal, then they would cause us to grasp that red color that inheres in
the thing it qualifies (i.e., the flower). For visual cognition has never been
seen to consist of color alone. Nor has mere color ever been found be
reflected without reference to its substrate.
242
Chapter 19. Terms of Discussion 2.49. Objection 243
nanv abhijiitasyevQ padma-riigadi-ma1}er japii-kusumader api prabhii
vidyate raya vyiiptatviit spha.tiko 'pi lohita ivQvabhiisQte. Objection:
light exists in the japa flower and so on also, just as it does in noble gems
such as a ruby; the crystal also apears to be red because it is pervaded by
that (light).
tathiipi svayam alohito mithyaivQ lohita fty iipadyeta. Even so the fact
remains that what is not red in itself (Le., the crystal) falsely appears to be
red.
atha prabhiiiva lohitiivabho.sate na spha!ika ili. Then the light alone
appears red, not the crystal.
sauklyam api rarhi spha!ike prakiiseta. In that case whiteness also
would appear in the crystal.
atha prabhayapasiiritQrrr tad iti eet. Now if it is objected that it (i.e., the
whiteness) is driven out by the light.
sa tarhi katha'!l syiit. na ca rapi-dravya-sa1J1yogiid
viiyor api tathiitva-prasangi'it. na prabhii-nimitta'!llauhitya1J1
tatrotpannam uttara-kiilam api tathti-rapa-prasangiit. How in that case
could (the crystal) which would have no color be visible? Nor (is the
crystal visible) because of a connection with another substance possessing
color, since the same would apply to air also. Nor is redness produced in it
by the light, since that would involve (the crystal's) retaining the same
color (i.e., red) even after (thejapii flower is removed).
2.49 Objection
The following terms are commonly used to raise an objection. Normally
the objection is a but occasionally some of the same terms are
used for a counterobjection by the siddhiintin. iti often marks the end of
the objection (see Section 2.37).
2.49.1 nanu and atka
The most common term for introducing an objection is the particle nanu
"but, objection:." nanu stands at the beginning of the first sentence of the
objection, and is especially common in objections consisting of several
sentences.
PP p. 84 nanu cid-titmana{l kathafJl but how
can the subject whose nature is consciousness be an object?
PP p. 55 nanu na vytipakam svapna-sokadiiv
asambhavtit objection: this definition does not extend far enough since
it does not include dreaming, grief, and so on
The particle atha "now, then, but" introduces a new idea or a new topic
(see Section 2.53.1) and may introduce an objection. atha also stands at
the beginning of the sentence.
PP p. 171 atha ko 'ya,!, tarko niima but what is this so-called
reasoning?
PPp. 333 atha sabda-janyasyiliva jiiiinasyCibhyiiso vidhfyata iti then
repeated thinking of the cognition produced by sabda is enjoined
2.49.2 Interrogatives
Objections are often phrased as direct questions introduced by interroga-
tives such as katham, kva, or kim. Especially commOn is katha,!, punar
"but how?"
SBh 1.1.1 katha,!, pratyag-iitmany 'dhyiiso
but how can there be superimposition of the
object and its qualities on the inner self which is not an object?
kim combined with the instrumental has the idiomatic meaning "what is the
use of ... ?" This construction may be used to find fault with an opponent's
statement.
SBh 1, 1.4 yadi ki'!l tava tena sytid iti even if [an.
ordinary thing] is taught [in the Veda]-what is the use of this to you?
SBh 1.1.4 tasman na brahmaIJaf:z sastra-yonitvam iti prapta ucyate
therefore the siistra is not the source of brahman-the being
thus established, the sutrakiira replies (in the following sutra)
SBh 1.1.12 ki1'(l tiivat priipta1?1 brahma1}o 'rthantaram amukhya
sytit first, what is the purvapal0a? the [self]
consisting of bliss must be something other than brahman, not the primary
self
2.49.5 priipta and
The past participle prapta "obtained," in the sensE! of "obtained at first
view, concluded superficially," often marks a Thus prapta is
"obtained as the and iti prapte is "this being obtained as the

244 Chapter 19. Terms of Discussion
2.49.3 iti cet
An objection, particularly a short one, may be marked by the concluding
words iti eet, lit., "if thus," i.e., "if this objection is raised, if you say this."
Usually the objection concluded by iti cet is followed in the same sentence
by the response to the objection.
SBh 1.1.4 asarfratvam eva dharma-karyam itf cen na tasya
svabhtivikarvar if it is objected that bodilessness itself is produced by
dharma, then we reply no, since that (bodilessness) is natural
SBh 1.1.4 brahmalJaJ:t tistra-yonitvanupapattir iti cen
navidyii-kalpita-bheda-nivftti-paratviic chiistrasya if you say that if
brahman is not an object it will be impossible for the tistra to be its
source, then we reply no, because the itistra is concerned with discarding
the distinctions fabricated by avidyii
,
I
,j
.J
2.49. Objecnon 245
2.49.4 ity iisankya
Derivatives of the root iiSallk "to suspect (wrongly)" often mark an in-
terpretation that appears to be reasonable but is incorrect. For example:
ity iiSallkya "having raised (or anticipated) this doubt," iiSallkyate "is sus-
pected, is wrongly held to be true."
PPp. 248 tatra krama-niyamartho 'thasabda ity asankyOha
phala-jijfiiisya-bhedac ca the word atha in the passage signifies order
of succession-having raised this doubt, the author of the text replies
"because of the difference of fruits and objects of inquiry"
SBh 1.1.1 ata eva na brahmasabdasya jaty-adyarthantaram
iisankitavyam therefore one should not suspect that the word brahman
has another meaning such as class
SBh 1.1.3 janmadi kevalam anumanam upanyastam ity iisankyeta tam
asankii1J1 nivartayitum idarrz sutrafJ'l pravavrte sastra-yonitviid iti one
might suspect that the origin and so on [of the world] were put forth [in
the sutra] merely as an inference-in order to rule out this doubt, the
sutrakiira has undertaken this sutra "stistra-yonitviit"
The past participle "objected, put forth as an objection" may be
used in the same way.
SBh 2.1.6 tasmad evedaT[l brahma1)o jagad ca
na brahma-prakrtikam ity iilqipte pratividhatte therefore this world is
completely different from brahman and because it is different it cannot'
have brahman as its material cause-this objection being raised, the
sutrakiira replies (in the following sutra)
2.49.6 Verbs of speaking
The third person of verbs meaning "to speak" may introduce an objection,
e.g., aha "an objector says," apara aha "another says," atra ke cid
"on this point some say." But the same expressions may also introduce
another interpretation acceptable to the commentator (see Section 2.51.2),
and aha without an expressed subject may refer to the auth0f of the text be-
ing commented on (see Section 2.39.4). The meaning of these expressions
is usually clear from the context.
PP p. 199 aha mii bhaj jiiiina-dvayam an objector says: let there not
be two cognitions
2.49.7 Clauses begiuning with na ca '"
SBh 1.1.4 na kanna-brahma-vidyti-phalayor
here we reply no, because of the difference between the
fruit of the knowledge of karma and the fruit of the knowledge of brahman
SBh 1.1.28 tatra prasiddher prat;w iti prapta ucyate
pralJa-sabda1[l brahma vijneyam the word prana means air because it is
well-known in that (meaning)-the thus established, we
reply that the word prtif/a must be understood to denote brahman
247
2.50. Reply
SBh 1.1. 6 atrtiha yad uktaf!l nacetana'!l pradhiinalfl jagat-ktirattam
fk#t,-tva-sraval}iid iti tad anyathdpy upapadyate here an objector says:
what was said before, i.e., that the unconscious pradhana cannot be the
cause of the world because sruti teaches that [the cause] is a seer-that
can be explained in another way
Chapter 19. Tenns of Discussion 246
A brief objection may be expressed in an iti clause introduced by na and
followed by a gerundive, e.g., na ca '" iti mantavyam "nor should it be
thought that. " ," na ca ... iti vacyam "not should it be said that. .. ,"
or "na ca . .. iti codanfyam "nor should it be urged that ... " (see Sec-
tion 2.23). The reason for the prohibition usually follows.
SBh 1.1.19 tatra katha1Jl sva-pradhiinatvaf!l sakya1'[t
vijniitutrl prakrtatviid iti briimaf:z. how can it be known that brahman is
the chief thing in the passage? we reply: because it is the subject of
discussion
SDS p. 8 na ca vytipti-jiitina1J1 stimanya-gocaram iii mantavya1J1
vyaktyor aviniibhiivo.bhiiva-prasangiit nor should it be thought that
knowledge of concomitance has the generic properties as its object, since
that would involve the absence of concomitance of the two individuals
2.50 Reply
The following terms are commonly used to answer an objection. In most
cases it is the siddhantin who replies, but occasionally some of the same
terms are used when the pilrvapal0in responds to an objection.
2.50.1 Verbs of speaking
The most common term for introducing the siddhantin's reply to an ob-
jection is ucyate, lit., "it is said," Le., "this is said in answer, we reply."
Similar expressions include atrdbhidhfyate = atrocyate "here we reply, on
this point we reply" and "we reply." These expressions are particu-
larly common when the siddhanta is introduced after a long or
after a debate between two or more The reply itself may be
formulated in the commentator's own words or it may contain a quote from
the text being commented on. In the latter case it may be more appropriate
to render ucyate by "the author of the text replies" (see Section 2.39.4).
250.2 tan na and other denials
The reply to an objection may begin with an explicit denial of it, e.g. tanna
"not that," tad asat "that is not right," tad ayuktam "that is not reasonable"
nditad evam "this is not so." Or a prohibition may be placed within an
clause, e.g., tan rna bhild iti "lest that be the case," rna bhad Mankeri or
tan mtlankfti "lest this doubt arise." The pronoun tat in such expressions
may refer to a preceding yat (see Section 2.40.3) or may simply refer to ,the
objection in general.
SBh 1.1.31 nditad evam uptisti-traividhytit this is not so because of
the threefoldness of meditation
OBh 2.18 anyathti prthivy-adivad api nityatva1[l sytid iitmanas tan rnii
bhad iti nityasyaniifina ity iiha otherwise the self's eternality would be
like that of the earth and so on-lest that be the case, the author of the text
uses the words "eternal" and "unperishing"
Sometimes the opponent's view is rejected in a more dramatic fashion.
SDS p. 7 tad elan mano-rtijya-vijrmbhanam so this [i.e., the argument
you have presented] is just building castles in the air .
248 Chapter 19. Terms of Discussion
'i"""
..
I
2.50. Reply 249
2.50.3 prasatiga and iipatti
Derivatives of the roots prasanj "to become attached to" and apad "to faU
into misfortune" are often used to reject a proposition by pointing to its
undesirable consequences. For example: prasanga and apatti "undesir-
able implication," prasajyate and apadyate "is implied as an undesirable
consequence."
SBh 1.1.22 na diyam iiklisa-sabda ubhayoJ:t siidhiiral}aJ:t sakyo vijiiiitum
anekarthatva-prasangiit nor can the word tikliia here be understood to
have both (meanings) equally, since that would involve (a single word's)
having more than one meaning
SBh 1.1.4 yady iilma sviiiraya-kriyayii vikriyetiinityatvam iitmanaJ.t
prasajyeta if the self were altered by an action inhering in it then
impennanence of the self would result
SBh 1.1.1 tata jfllitam evty ajijfiiisyatvafJ1 punar iipannam then
[brahman] must be known and so [brahman's] not being an object of
inquiry would again result
When an assertion is made, the reason for rejecting its alternative is of-
ten expressed in a phrase containing prasanga or apatti and introduced by
anyathii "otherwise."
SBh 1.1.19 teniinandamayasya brahmatvam anyatha.
prakrtahiinaprakrta-prakriyii-prasarigiit therefore the [self] consisting
of bliss is brahman, since othenvise we would make the mistake of
abandoning the subject of the passage and accepting what is not the
subject of the passage
2.50.4
The word "fault, defect" refers to a flaw in an argument or doctrine.
The reply to an objection often begins na or or ndya,!,
"this fault (does) not (apply)," i.e., "the fault proposed by the objec-
tor does not hold true for our doctrine." Similarly na dusyati means "does
not harm our position."
SBh 1.1.12 yat ti'iktam annamayady-amukhyatma-praviiha-patitatviid
iinandamayasytlpy amukhyatvam iti ndsau iin9ndamayasya
sarvantaratviit but as for what was said before, Le., that the [self]
consisting of bliss also cannot be the primary [self] because it belongs to a
series of secondary selves beginning with the [self] consisting of
food-this fault does not apply, because the [self] consisting of bliss is the
innennost of all
SBh 1.1.4 ato 'vidyii-kalpita-sarrzsiiritva-nivartanena
nitya-muktdtma-svarupa-samarpal}tin na
therefore the fault of impennanence does not belong to because by
discarding the appearance, fabricated by avidyii, that (the self) is a
transmigrating being, (scripture) teaches that the nature of the self is
eternally free
2.50.5 syiit and bhavet
The optatives syat and bhavet "might be" may be used to state that the
objection might have been valid if the objector had not overlooked an iJ;n-
portant point, e.g., syad etat "this might be the case," syad etad evam "this
might be so." The reason why the objection is not valid then follows.
SBh 1.1.4 sytid etad evarrz yadi rajjusvampa-srava1}a iva
sarrtstiritva-bhrtintir brahmasvarilpa-srava1}a-mlitre1}a
nivarteta na tu nivartate this might be so, if the error of
transmigratoriness were dispelled by merely hearing about the true nature
of brahman, just as the error of the snake (is dispelled) when one hears
about the true nature of the rope; but it is not dispelled
SDS p. 7 syiid etat. syiid mano-ratho yady anumiinddeb priimii1}yarrt
na sytit. asti ca prtimti1}yam . This might be. This would be our wish if
inference and so on had no validity; but they do have validity.
Note that in some authors syat may refer forward and may mean "let us
suppose the following."
250
Chapter 19. Terms of Discussion
1
'_"";
.. "
I
2.51. Alternative interpretation 251
2.50.6 yady api and other concessives
If the respondent is willing to grant part or all of an objection, expressions
such as yady api . .. tathdpi ... "even if ... still ... ," kiimam "granted,"
satyam "true," baham "yes, surely," astu "let it be," and sarvathiipi "but
in any ca,e" may be employed.
SBh 1.1.4 atrapare yady api sastra-pramii1}aka,!,
brahma tathapi sa.stre1}Q brahma
samarpyate here others raise an objection: even if the ia.stra is the
means of knowing brahman, still brahman is taught by the siistra only as
an object of the injunction to know
PPp. 73 nanu na jfvo brahma1}o 'nya fty uktam. biit;iham ata evtirthiij
jfve brahma-svarnpa-praktis{icchadikavidyii ka/pyate. Objection: it has
been said that the jfva is not different from brahman. Yes-for this very
reason Qvidya is known by implication to be that which obscures the light
of the true nature of brahman in the jfva.
SBh 1.1.24 astu tarhy atrivrt-krta'!l prathamajam in that case let
the light (in the passage) he the first-born (light) not yet made tripartite
SBh 1.1.25 asmin brahmaivabhihitam iti na chando 'bhidhiinam.
sarvathapy asti purvasmin viikye prakrta'!l brahma. On this view
brahman only is denoted and so the meter is not denoted. But in any case
brahman is the subject of the previous passage:
2.50.7 tarhi
In discussing the implications of a statement or objection the particle tarhi
"in that case," i.e., "if we accept that to be true," is often used.
PP p. 128 anupakurvann api tarhi sahakiiry in that case the
auxilIary would be required even though it does not render assistance
SBh 1.1.19 annamayiidfniim api tarhi in that
case the [self] consisting of food and the other [selves] also would have to
be brahman
I
2.51 Alternative interpretation
2.51.1 Second interpretations
After giving one interpretation of the text being commented on or of some
point raised in discussion of the text, a commentator may introduce a sec-
ond interpretation with particles such as athava or yadvii, both of which
have the sense "or else, on the other hand." Frequently (but not always) the
second interpretation is favored by the commentator. In the first example
below a second interpretation of the sutra "siistra-yonitviit" is introduced;
in the second example a second interpretation is given of the word jyotis in
the sutra "jyotis carm;ujbhidhiiniit".
SBh 1.1.3 athavii yathoktam rg-vedadi fa-stra"!' kiiraf.laTJl
pramij-!lam asya brahma!,o yathiivat-svarapadhigame or else the siistra
consisting of the Rig Veda and so on as described above is the source, i.e.,
the cause, i.e., the authority for knowledge of the true nature of that
brahman
SBh 1.1.24 yadvii naya'!l evdnugriihake
tejasi vartate 'nyatrdpi prayoga-darsaniit or else the wordjyotis here
does not denote the light that aids the function of the eye, since we see that
it is used in other meanings also
2.51.2 Citing others
Expressions such as apara aha "another says" and atra ke cid iihuJ:! "on
this point some say" often introduce an objection (see Section 2.49.6). The
same expressions introduce an alternate interpretation (usually not the fa-
vored one) when the interpretation is allowed to stand unrefuted by the
commentator.
1
I
I
I
Chapter 20
Particles and Pronouns
2.52 Introduction
In Sanskrit prose, particles play an important role in connecting sentences
and in altering the meaning of a sentence. This section discusses some of
the particles used for these purposes, and also the use of pronomial adverbs
as pronouns, some peculiarities in the use of correlative pronouns and ad-
verbs, and uses of the four common demonstrative pronouns. (See also:
Coulson 61-{j3, 76,171-77,191-95,207-14; Speijer 261-90, 394-490.)
2.53 Particles connecting sentences
The following particles often connect a sentence (or clause) to what has
been said before, especially when they appear at the beginning of the sen-
tence (or clause). Frequently this connection is crucial to the meaning of
the passage. Note that particles such as kiT[l ca, lit., "and what?" are prop-
erly rhetorical questions but are best rendered in English by less literal
translations. See Section 2.30 for the use of the causal particle hi.
2.53.1 atka
atha "now, then, but" introduces a new topic or a new idea.
SBh 4.1.1 trfrye 'dhya.ye parapartisu vidyiisu siidhanasrayo vicarafJ
prayeru'ltyiigiit. atheha caturthe phaltisraya i i m i ~ y t i . In the third
adhyiiya discussion was centered primarily on the means (of knowledge)
253
Compare the common use of atha to begin a chapter or an entire work.
atOO dvitfyo 'dhyayal} now begins the second adhyaya Often the new idea
introduced by atha is an objection (see Section 2.49.1). In conditional
sentences atha may be used with yadi to propose an alternative, i.e., yadi
atha ... "if ... but if ...."
PPp. 257 bahavas ca sambandha-vise-!a!l. tatravasyam anyatarnal}
pratipattavyo 'nyathti . And there are many kinds
of relationship. Of these one must necessarily be accepted because
otherwise discourse would be impossible.
When a number of items have been mentioned in a previous sentence, tatra
may begin the next sentence with the meaning "ofi;hese things just men-
tioned." In this case tatra is the equivalent of and is regularly glossed
as madhye "among these."
254
Chapter 20. Particles and Pronouns
with regard to the higher and lower knowledges. Now in this fourth
(adhyaya there will be discussion) of the fruits (of knowledge).
1
I
2.53. Particles connecting sentences 255
PP p. 123 yadi na [artha-kriyam] kuryad na
paramfinha.vastv atha kuryiin na tarhi sthiiyf if it does not perform
[useful activity] then it is not a real thing since it fulfills the requirements
of a non-existent thing; but if it does perform (useful activity] then it is not
permanent
Occasionally atOO is used in place of yadi or cet to state a condition.
SBh 1.1.4 atha prakarQ/:uintara-bhaytin naitad abyupagamyate tathiipi
sva-viikya-gatopiisaniidi-kanna-paratvam if this is not accepted for fear
of introducing a new topic. then even so (it must be admitted that the
Vedanta texts) are concerned with actions such as meditation contained in
those very texts.
2.53.2 tafra
tatra often refers to a recently mentioned quotation or fact in the sense "in
that passage" or "in that case" or "with regard to that" and so on. When
tatra refers to a recently mentioned substantive it is the equivalent of the
locative of the pronoun tat (see Section 2.55).
SBh 1.1.1 tatratha-sabda anantaryarthal} parigrhyate the word atha
in the passage is used to signify immediate succession
SBh 1.1.23 vayu-vikiirasya pafica-vrtte!:z pra7}asyopadanaf!l yuktam.
tatra hi prasiddhatara!:z prii7}a-sabda!:z . It is reasonable that the
reference is to the five-fold prlirta that is a modification of air, since the
word pra7}a is better known in that sense.
PPp. 90 tatra na tavat prathama!:z kalpo pramii7}a-phale
visvasyanavabhasa-prasangiit of these, first of all, not the first
alternative, because if the fruit of pramti7}a were insentient then nothing at
all would appear
2.53.3 tavat
ravat may be used adverbially in the meaning "first, first of all." In this
sense it usually introduces the first of a series of points or alternatives and
may be combined with tatra (see the last example above). Generally tavat
is placed after the first word of the sentence.
SBh 1.1.22 kuta!:z sa'!1saya!:z. ubhayatra prayoga-darianiit. bhuta-vife-!e
tlivat suprasiddho loka-vedayor brahma7}Y api leva cit
prayujyamiino drfyate. Why is there a doubt? Because we see its use in
both meanings. First, the word aktisa is well known in both worldly
discourse and in the Veda in the sense of a particular element; but in some
cases we see that it is used to denote brahman also.
SBh 1.2.1 na tavad anlyastva'!1 jyiiyastvaf!l cobhayam ekasmin
samiisrayituf!l sakyaf!l virodhiit first of all, smallness and greatness
cannot both reside in one thing, because of the contradiction
2.53.4 afra and iha
atra and iha often refer to a recently mentioned quotation or fact in the
sense "in this passage" or "in this case" or "with regard to this" and so
on. If the reference is to a recently mentioned substantive, atra and iha are
equivalent to the locative of the pronoun idam (see Section 2.55).
Similarly ity arra may be "in this passage (quoted by iti)."
The expressions iMpi "here too, in this case also" and iha tu "but here, but
in this case" show that the situation just described is similar to or different
from the subject of discussion.
2.53.7 tathii hi and tad yathii
257
2.53. Particles connecting sentences
SBh 1.1.2 tasmiijjanmadi-sutralJ1 nanumlinopanylisartham. ki1J1 tarhi
vedlinta-vlikya-pradarianartham . Therefore the sutra beginning with
the words janmiidy does not serve to put forth an inference, but rather to
point out a Vedanta text.
SBh 1.1.4 na tu tathli brahma[lQ sambhavati
but brahman cannot thus be subordinate to injunctions to meditate
2.53.6 tu, etc.
ta, ki,!, tu, para,!, ta, punar, and the like share the meaning "but" and of-
ten introduce a slightly different idea or (more strongly) a modification or
correction of what has been said before. A more emphatic antithesis is
introduced by ki,!, tarhi "but rather, instead."
I
I
j
1
J
SBh 1.1.23 ata iha priif}a-fabdena katarasy6ptidtinaJ!l yuktam iti
bhavati samsaya/:I therefore there is a doubt: which of the two
(meanings) is signified by the word prii1}Q in this passage?
SBh 1.1.22 kif!! punar alra yukta,!, bhfitiikiisam iti but what is right in
this case? [the word akasa refers to) the elemental ether
SBh 1.1.4 yatha ca vliva gautamagnir viiVQ gautamdgnir ity
atra agni-buddhir manasi bhavati and just as in the
passage "man indeed, 0 Gautama, is a fire; woman indeed, 0 Gautama, is
a fire" there is a mental meditation upon man and woman as fire
256
PPp. 249 iha tu nitya-nirvrtta1J1 purzqa-vyapiirdnapelqa'!1 hrahma
jijfiiisyam but in this case the object of inquiry is brahman, eternally
accomplished and independent of man's action
Clarification or illustration of what has been said before is introduced by
tatM hi (lit., "for thus") and tad yatM (lit., "that is like this"). Both have
the sense "that is to say" or "for instance."
iha may also signify "in this world" (= iha lake).
2.53.5 ki,!, ca, etc.
kiTJ1 ca "moreover," api ca "moreover," anyac ca "and another thing," apara1'J1
ca "and a further thing," tatM ca "and similarly," and the like all serve to
add something to what has been said before.
PPp. 49 kirrr ea rajata-nlpe1Ja eet pari1Jata k4fram iva
dadhi-rape1Ja tadli 'pi moreover, if shell
were transfonned into silver as milk is transfonned into curds then it
would remain just so (i.e., as silver) even when the defect is removed
SBh 1.1.4 api ca brlihma1Jo na hantavya iti eaivam-lidyli nivrttir
upadiiyate moreover, abstinence from action is taught in passages such
as "a Brahmin is not to be killed"
SBh 1.1.1 prlik ca tathli-bhutiitma-vijfilinlit pravartarniinafJl ilistram
nativartate. tathli hi brlihmQ/}o yajetetylidfni
slistriiflY litmani van:uiirama-vayo ain'tya
pravartante. And the iiistra, operating before such knowledge of the
self has arisen, does not go beyond having as its object those affected by
avidyii. That is to say, passages of the ilistra such as "a Brahmin should
sacrifice" operate on the assumption that there is superimposition on the
self of particulars such as class, stage of life, age, state, and so on.
SBh 1.1.1 adhyiiso niimatasmif[ls tad-buddhir ity avoelima.< tad yatha
wiham eva sakalo veti
blihya-dhannlin iitmany adhyasyati . We have said that superimposition
is the notion of one thing in another thing. For instance, a man
superimposes external qualities on his self if he thinks "I am ill" or "I am
well" when his son, wife, and so on are ill or well.
2.53.8 Words for "therefore"
SDS p. 3 tac evatmii therefore the self is just
the body qualified by consciousness
SBh 1.1.1 tasmad eva pratyalqadfni pramiil}tini
siistrii1}i CQ therefore the pramiil}Qs such as perception and the iistras
have as their object those affected by avidyii
A conclusion may be introduced by a pronoun or adverb such as tasmiit,
tena, tat, tatal;!, atal;!, or ital;!. The pronoun or adverb functions as a con-
junction and has the meaning "because of the preceding, for this reason,
therefore."
259 2.54. Particles of emphasis
women This meaning is strengthened in the expression eva kevalam 'just
this and no more."
PP p. 107 atraha bhavatu na vastv-antaram. tad eva tad iti tu na
l0amyate. Here the objector says: we grant that [the image in a mirror]
is not a different thing [from the object reflected]; but we cannot accept
that it is the same thing (Le., that the two are identical).
SBh 1.1.19 tatha safi tad eva brahrnilnandamaya tirmiivayavftad eva ca
brahma puccha,!, this being the case, that very
brahman would be the self consisting of bliss, the whole, and that same
brahman would be the tail, the foundation, a part
SBh 1.1.1 brahma-codanii tu avabodhayaty eva kevalam but
the scriptural imperatives concerning brahman only instruct a man and no
more
When combined with the pronoun tat, eva means "the very one, the
same one."
I
j
1
I
.j
1
PPp. 140 ba(iham pratyag-d!Sy api sambhaven
nditilvatii tat-sambhavaJ:z sidhyate. lena nidarsanfya/:z safz. Granted that
avidya as defined above is possible even in the case of the inner self; still,
this much does not establish the actual existence of it (i.e., Qvidyii).
Therefore that is to be demonstrated.
258
The force of the pronoun or adverb may be strengthened by addition of the
particle eva (see Section 2.54.1); e.g., tasmiid eva or ata eva "for this very
reason." The reason itself may be expressed in a clause introduced by a
relative pronoun or adverb; see Section 2.28 for such constructions.
2.54.2 api
api has two common meanings when modifying the immediately preceding
word: "also, too" and "even, though." The intended meaning is usuatly
clear from the context.
2.54 Particles of emphasis
2.54.1 eva
eva lays stress on the immediately preceding word and may be rendered as
'just, only, merely." It conveys the sense 'just this-no more, no less."
SBh 1.1.2 na vastu-yathiitmya-Jiiiina,!, puru,a-buddhy-apekeam ki,!,
tarhi vastu-tantram eva tat. Knowledge of the true nature of a thing
does not depend upon the notions of man; instead, it depends only on the
thing itself.
SDS p. 3 anganady-tilinganiidi-janya'!l sukham eva the
aim of man is just the pleasure that arises from such things as embracing
SBh 1.1.5 nan-ktaTJ1 sarva-jfiiina-saktimattvena sarvajfiaTJ1
tad api nopapadyate . Objection: it was said that [the pradhiina] might
be omniscient because it possesses the capacity for all knowledge. That
also is not reasonable.
SBh 1.1.1 nanu 'pi brahmtJ!lo jijiidsa-karmatvaTJ1
na virudhyate objection: even if the genitive of general relation is used
in the text, brahman's being the object of the desire to kno:w is not
contradicted
With regard to the first meaning, note the use of api (either alone pr in
combinations such as api cal at the beginning of a sentence to connect the
sentence to what has been said before (see Section 2.53.5). With regard to
260 Chapter 20. Particles and Pronouns
1
!
2.56. Correlative pronouns and adverbs 261
the second meaning, note the use of api with the locative absolute to mean
"even though" (see Section 2.29.3) and the common expressions yady api
"even if" and evam api = tathapi "even so, still, nevertheless" (see Sec-
tion 2.50.6).
SDS p. 31 yad vedyate yena vedanena tat tato na bhidyate whatever
is made known by a cognition, that is not different f!?ID that (cognition)
PPp. 211 adhftavedasya yo 'rtho 'vagatas tatriiiva vipratipattayaJ:z
santi there are conflicting views with regard to that meaning learned by
one who has studied the Veda
2.55 Pronomial adverbs used as pronouns
Pronomial adverbs formed with the suffixes tas and tra (also the adverbs
kva and iha) may refer to persons and things as well as places and are
sometimes equivalent in both construction and meaning to the correspond-
ing pronouns of suitable gender and number in the ablative or locative case.
In the first example above yatra = yasmin and both yatra and yad corre-
spond to tat (= tayoM; in the second example = tasmiit and corre-
sponds to yena; in the third example tatra = tasmin and corresponds to yo

2.55.2 Adverbs as attributives
2.55.1 Adverbs corresponding to pronouns
The adverbs mentioned above may act as attributives of singular substan-
tives in the ablative or locative case.
The following adverbs may be used in place of the corresponding pro-
nouns:
2.56.1 Repeated correlative
2.56 Correlative pronouns and adverbs
The same relative pronoun or adverb may be repeated to give a general
sense, e.g., yo "whoever" and yatra yatra "wherever." In this case the
correlative pronoun or adverb is also repeated.
from that place
from some cause
in this pot
in this world
tato deiiit = tasmild deiilt
kUfas cit kiira1:uit = kasmiic cit kiirm:uit
atra gha!e = asmin gha!e
iha loke = asmin loke
I
= ablative of kim
= locative of kim
= ablative of yat
= locative of yat
= ablative of tat
= locative of tat
= ablative of idam
= locative of idam
= locative of idam
= locative of kim

kutra

yatra

latra

alTa
iha
kva
Note that the relative adverbs and yatra may be correlated with
the pronoun tat and the demonstrative adverbs and tatra may be cor-
related with the pronoun yat.
SBh 1.1.12 yatra yatra vibhuty-iidy-atisayaJ:t sa sa isvara ity uptisyatayii
codyate in whomever there is pre-eminence of so on, he is
enjoined as the one to be worshipped as the Lord
SBh 1.1.1 yarra yadadhyasas tad-vivekagraha-nibandhano bhramaJ:z
the superimposition of one thing on another is an error based on the failure
to grasp the difference of the two
SBh 1.1.24 tasmild yad yat kasya cid avabhilsaka,!, tat taj
jyotiJ:t-sabdenabhidhfyate therefore whatever illuminates something is
denoted by the wotd jyotis
262 Chapter 20. Particles and Pronouns 2.56. Correlative pronouns and adverbs 263
2.56.2 Multiple pairs
Sanskrit often employs more than one pair of correlatives in the same sen-
tence. The correlatives (or all but one) usually have an indefinite sense.
There is no requirement that the demonstratives appear in the same order
as the relatives, and either may occur in compound. Such sentences are
often difficult to translate into English since English rarely uses more than
one pair of correlatives in the same sentence. It may be helpful, especially
upon first reading, to represent each pair of correlatives symbolically by
letters such as "A," "B," and so on.
SBh 1.1.1 yatra yad-adhyiisa tat-krtena dO$e/}Q gUlJena vQIJu-miitre/}opi
sa na sambadhyate when there is superimposition of A on B, B is not
connected even in the slightest degree with any fault or virtue created by
A; or, when there is superimposition of one thing on another, the latter is
not connected even in the slightest degree with any fault or virtue created
by the former
5DS p. 25 yad yadii yat karoti tat rada latm samartham when A
causes B at time C, A is samartha of B at time C; or, whenever one thing
causes another, the fonner is samartha of the latter at that time
In the first example above the correlatives are yatra = sa (B) and yad =
tat (A); in the second example the correlatives are yad = tat (A), yadii =
tadii (C), yat = tatra (B).
2.56.3 yatha with tatha
A clause introduced by yathii may express the object of a verb meaning
"to say," "to show," and so on. The correlative tathii is normally used if
the yathii clause precedes the main verb. In this construction yathii means
literally "in what way" and can often be translated by "how."
PPp. 137 yathfikaiasyalqa-vyiipiiram antarapy aparolqatii tathii
darsayirYiima/:I we will show how the ether is directly known even
without the functioning of the eye
SBh 1.1.5 yatha tu tarkel}1jpi brahma1}Q eva kiira!Jatva'!l nirvOfjhu'!l
sakyate na pradhiinadfniirrr tatha na
J
j
I
1
j
t
I
J
I
asyeryevamiidina but the sutraktira will explain in siltras such as na
asya how by reasoning also the of brahman alone
can be established, and not that of the pradhiina and so on
Note that in both examples a sentence of equivalent meaning can be ob-
tained by omitting yatha and replacing tathii with iti.
2.56.4 yat meaning "that"
The pronoun yat is sometimes used as a conjunction to introduce a sub-
ordinate clause. In this case yat means "which is the fact that" or simply
"that." A correlative to yat is sometimes used, and iti is sometimes added
redundant!y to the end of the yat clause.
SBh 1.1.4 alankiiro hy ayam asmiika,!, yad brahmatmiivagatau satyti1'[l
sarva-kartavyatii-hiiniJ:! krta-krryata eeti for this is our glory, that when
there is knowledge of the self as brahman all duties cease and everything
is accomplished
PP p. 317 ato '[ankara eva sutrii1}iirrz yad anekarthata niima therefore
it is indeed an embellishment of sutras that they have more than one
meaning
In the first example the yat clause refers back to the correlative pronoun
ayam and is concluded by iti. In the second example the yat clause refers
to the preceding word alankara. Sentences of equivalent meaning can be
obtained by converting the yat clauses to iti clauses (i.e., by omitting the
yat in both examples and adding iti to the end of the second), and in fact
such subordinate clauses are far more often formed with iti than with yat
(see Chapter 16). Note that when such a clause is formed with iti it tends
to precede the word or words to which it refers, but when it is formed with
yat it tends to follow.
2.56.5 Repeated anya
The notion "the one ... the other ... " or "some ... others ... " may be
expressed by repetition of the pronoun anya.
264 Chapter 20. Particles and Pronouns 2.57. Demonstrative pronouns 265
SBh 1.1.23 yathii pituJ.1 piteti prayoge 'nyaJ:z pita 'nya!}
just as in the expression "the father of the father"
one father is indicated by the genitive case and the other is indicated by
the nominative case
Similar expressions include ke cit (or eke) '" apare (or anye) ... " some
'" others ... ."
2.57 Demonstrative pronouns
2.57.1 Scale of proximity
The uses of the pronouns idam, etat, adaJ:z, and tat are distinguished by the
Indian grammarians in the following kiirikii:
idamas tu samfpatara-varti caitado rilpam
adasas tu tad iti vijiinfyiit
One should know that idam refers to what is near, etat to what
is closer at hand, adaJ:z to what is remote, and tat to what
is not visible to the speaker.
The words idam, etat, and adaJ:z are normally deictic pronouns, i.e.,
they point out a person or thing and distinguish it from others of the same
class. Often they point to something in the speaker's presence. In this case
idam and etat designate something near the speaker and usually correspond
to the English demonstrative "this." Their areas of meaning tend to overlap,
though etat is more emphatic and is often translated by "here" (e.g.,
devadattaJ:z "here is Devadatta"). adaJ:z designates something remote from
the speaker and possibly not visible to him, and usually corresponds to
the English demonstrative "that." These three pronouns may also point to
something just mentioned or about to be mentioned; in particular etat often
points to what has just been said and idam to what is about to be said (see
Section 2.57.2 tat is properly an anaphoric pronoun, i.e., it refers back to
a preceding word in the context. It is therefore the usual correlative of
the relative pronoun yat when the relative clause precedes the main clause
(when the relative clause follows the main clause the correlative is often
idam or etat or is omitted altogether). When tat is used in apposition with
-!
;
.{
1
1
1
1
1
!
1
I
j
a substantive it is usually translated by the definite article "the" (e.g., sa
briihmm:wJ:z "the Brahmin"); when it appears alone tt is usually translated
by the appropriate personal pronoun (e.g., saJ:z "he" or "it"). In cases other
than the nominative and accusative, idam may be used anaphorically as a
synonym of tat (e.g., asya = tasya "his" or "its"). idam is enclitic when
used in this way.
2.57.2 Reference to earlier and later passages
In conversation and discussion etat usually refers back to what has already
been said while idam usually refers ahead to what is about to be said.
ORh 4.20 tadfyaf!l karmakannaiva sampadyata ity etam arthaf!l
darsayi,ryan aha being about to reveal this meaning, i.e., that the action
of such a man becomes non-action, the author of the text says (the
following verse)
GBh 8.19 saf!lsiire vairiigya-pradarianiirthaf!l cedam iiha and in
order to demonstrate dispassion toward saf!lsiira the author of the text says
this (i.e., the following verse)
In the first example etam artham refers back to the preceding iti clause;' in
the second example idam refers ahead to the verse of the text about tO,be
quoted.
2.57.3 tad etat and tad idarn
tat is sometimes used in apposition with etat or idam to make an emphatic
reference to something recently discussed; e.g., tad etat or tad idam "this
one here (etat or idam) that we have been talking about (tat)."
SBh 1.1.1 tam etam evam-IalqatJam adhyiisa1J1 patJ4itii avidyeti manyate
this superimposition as defined above scholars regard as (lvidya
SDS p. 6 tad etat sarva1J1 samagrahi all of this is summarized (in the
following verses)
GRh 1.1 tad idarrz gfta-sastra'!'
samastavediirtha-sara-sangraha-bhiltaf!l durvijfieyartham this
266 Chapter 20. Particles and Pronouns
gfta-sQstra which is a summary of the essence of the meaning of the entire
Veda and whose meaning is hard to grasp
1
!
,
When tat is used in the predicate and the subject is etat or idam, the sense
is "this one (etat or idam) is the one that we have been talking about (tat)"
or "this is the same one."
so 'yaf!l briihmal}a(t this is the Brahmin; or, this is the same Brahmin
Note that the pronoun subject (ayam) may be placed in the middle of the
predicate (sa see Section 2.21.2.
I
j
j
l
I
1
I
I
I
1
I
j
j
I
I
I
I
I
Appendix
Useful Reference Works
The following reference works may be useful in reading Sanskrit commen-
taries and other works composed in the style.
1. Abhyankar, Kashinath Vasudev and J. M. Shukla. A Dictionary of
Sanskrit Grammar. 2nd rev. ed. Gaekwad's Oriental Series 134.
Baroda: Oriental Institute, 1977.
Dictionary of Sanskrit grammatical terms, including both technical
terms of P3J).inian grammar and more general grammatical terms.
2. Apte, Vaman Shivaram. The Student's Guide to Sanskrit Composi-
tion. 24th ed. Pune: Anath Vidyarthi Griha Prakashan, 1960.
Helpful observations on syntax, the uses of particles, and many spe-
cific points of grammar and composition.
3. Biihtlingk, Otto. Piil}ini's Grammatik. 2nd ed. 2 vols. 1887. Kyoto:
Rinsen Book Company, 1977.
Translation ofPiiI).ini's grammar; the second volume includes several
indices that are very useful for looking up technical terms found in
commentaries.
4. Cardona, George. Piil}ini: A Survey of Research. New Edition.
Delhi: MotHal Banarsidass, 1998.
A critical survey of research on Pii,!ini and Pii,!inian grammar, which
provides not only a detailed bibliography but a reasoned treatment of
work to date on every relevant topic.
267
5. Cardona, George. Piil}ini: His Work and Its Traditions. Volume I:
Background and Introduction. Second revised and enlarged edition.
Delhi: Motilal Banarsidass, 1997.
A very thorough and careful presentation of PilJ)ini's approach and
techniques, providing a rich and invaluable description of the entire
system.
6. Coulson, Michael. Sanskrit: An Introduction to the Classical Lan-
guage. Oxford: Teach Yourself Books-Hodder and Stoughton, 1976.
Sanskrit primer that provides additional discussion and examples for
some of the material covered in our Part Three, along with many
helpful observations on other grammatical points (see the table of
contents). A brief but excellent introduction to commentarial tech-
niques may be found in the discussion of Mallinatha's commentary
on Kumiirasambhava 6.84-95 on pp. 258-267.
II. Kale, Moreshwar Ramchandra. A Higher Sanskrit Grammar. Rpt.
Delhi: Motilal Banarsidass, 1972.
The only grammar of Sanskrit in English that gives a full presen-
tation of the PaI,linian system of grammar. Unfortunately it has no
index.
10. JhalakTIcar, Mahamahopadhyaya BhImacarya. Nyiiyakosa or Dic-
tionary of Technical Terms of Indian Philosophy. Rev. by Maha-
mahopadhyaya Vasudev Shastri Abhyankar:4th ed. Bombay San-
skrit and Prakrit Series 49. Pune: Bhandarkar Oriental Research
Institute, 1978.
Dictionary of technical terms of Indian philosophy, with special em-
phasis on Nyaya terminology. A difficult work, but useful for stu-
dents who have some experience in reading Sanskrit philosophical
texts. In Sanskrit.
269
Useful Reference Works
1
I
I
J
Appendix 268
7. Devasthali, G. V. Anubandhas of Piil}ini. Publications of the Centre
of Advanced Study in Sanskrit, Class B, No.2. Pune: University of
Poona, 1967.
Explanations of the meanings of the indicatory letters used in the
technical terms of PaI,linian grammar, together with indices of gram-
matical elements both with and without the indicatory letters.
12. Katre, Sumitra M. of Piil}ini. Texas Linguistics Series.
Austin: University of Texas Press, 1987.
A complete English translation of PilJ)ini's grammar, with useful in-
dices. The text is presented not in its traditional form but in marked-
up Roman transliteration with sandhi dissolved and hyphens and ac-
cents added. '
8. Hartmann, Peter. Nominale Ausdrucksformen im Wissenschaftlichen
Sanskrit. Heidelberg: Carl Winter Universitatsverlag, 1955.
Detailed and scholarly treatment of nominal constructions in the
style of Sanskrit prose.
13. Renou, Louis. Histoire de la langue Sanskrite. Les Langues du
Monde 10. Lyon: Editions lAC, 1956.
A short history of Sanskrit commentaries and a good general intro-
duction to the style of Sanskrit prose may be found on pp.
133-146.
9. Jacobi, Hermann. "Dber den nominalen Stil des wissenschaftlichen
Sanskrits." Indogermanische Forschungen 14 (1903): 236-251. Rpt.
in: Hermann Jacobi. Kleine Schriften. Wiesbaden: Franz Steiner
Verlag GmbH, 1970. I: 6-21.
Good general introduction to the style of Sanskrit prose; in-
cludes an exemplary passage, with translation, from the Vediinta-

14. Speijer, J. S. Sanskrit Syntax. 1886. Delhi: Motilal Banarsidass,
1973.
Comprehensive treatment of Sanskrit syntax that provides additional
discussion and examples for some of the material covered in our Part
Three (see the table of contents and the Sanskrit index). Speijer often
refers to PaI,lini on specific grammatical points.
270
Appendix
271
Commentarial Passages Quoted
15. Whitney, William Dwight. Sanskrit Grammar. 2nd ed. 1889. Cam-
bridge, MA: Harvard University Press, 1971.
Comprehensive treatment of Sanskrit and Vedic morphology from a
Western viewpoint; very little discussion of Pi!I)inian grammar or of
Sanskrit syntax.
GBh 1.1 188, 263
GBh 2.10 189,208
GBh 2.14 209
GBh 2.16 190
GBh2.18 245
GBh 2.20 181
GBh 2.21 228
GBh 2.23 225
GBh 2.48 225
GBh 4.20 263
GBh 8.19 263
Kiisikii 1.2.9 164
Kasikii 1.2.26 164
Kiisika3.1.7 44
Kasika3 .1.8 46
Kasikii 3.1.10 47
Kiisika3.1.11 47
Kasika3.1.12 , .47
Kiisika3.1.15 .48
Kasikii 3.1.17 48
Kasika3.1.18 48
Kasika3.1.19 48
Kiisika 3.1.20 .48
Kasikii 3.1.21 47
Kiisika3.1.22 45
Kasika3.1.23 45
Kasika3.1.24 .45
KUikii3.1.25 a
Kiisika3 .1.26 46
Kasikii 3.2.83 64
Kasika3.3.11 51
Kasikii3.3.12 51
Kiisikii 3.3.13 51
Kasikii 5.1.5 , 67
Kiisikii 5.1.10 67
Kasika 8.1.54 164
Kir.M. 13.69 71
Kum.C. 1.10 , 142
Kum.C. 1.46 116
Kum.C. 7.53 52
Kum.M. 1.10 61
Kum.M. 1.23 58
Kum.M. 1.50 61
Kum.M. 1.1 114, 129
Kum.M. 1.2 32, 103, 146
Kum.M. 1.4 15,80,121
Kum.M. 1.5 98, 165
Kum.M. 1.6 129
Kum.M. 1.7 55
Kum.M. 1.10..... . . 142
Kum.M. 1.11 . 23, 126, 133; 141
Kum.M. 1.13 107
Kum.M. 1.15 24, 27
Kum.M. 1.16 109, 139
Kum.M. 1.18 24
Kum.M. 1.20 59
Kum.M. 1.23 45
Kum.M. 1.24 28
Kum.M. 1.25 129
Kum.M. 1.26 71
Kum.M. 1.27 81,145
Kum.M. 1.28 122
Kum.M. 1.32 167
Kum.M. 1.35 16
Kum.M. 1.39 25
Kum.M. 1.41 77
Kum.M. 1.43 23
272 Commentarial Passages Quoted
,
273
Kum.M. 1.44 100
Kum.M.1.46 23,81.116
Kum.M. 1,47 140
Kum.M. 1,48 32, 95
Kum.M. 1.51 98
Kum.M. 1.53 129
Kum.M. 1.55 80
Kum.M. 1.58 , 56, 100
Kum.M. 1.60 95
Kum.M. 2.1 30. 77
Kum.M. 2.2 72
Kum.M. 2,4 54
Kum.M. 2.7 44
Kum.M. 2.10 160
Kum.M. 2.11. 75, 164
Kum.M. 2.17 110
Kum.M. 2.20 129
Kum.M. 2.21 31
Kum.M. 2.22 129
Kum.M. 2.23 98
Kum.M. 2.25 100
Kum.M. 2.27 27
Kum.M. 2.28 117
Kum.M. 2.29 26
Kum.M. 2.30 99
Kum.M. 2.32 103
Kum.M. 2.35 76
Kum.M. 2.41 125
Kum.M. 2.45 99. 100
Kum.M. 2.46 56
Kum.M. 2.47 30, 107, 136
Kum.M. 2.51 61.167
Kum.M. 2.56 163
Kum.M. 2.57 73
Kum.M. 2.59 31
Kum.M. 2.64 30
Kum.M. 3.1 135
Kum.M. 3.5 25
Kum.M. 3.10 133
Kum.M. 3.14 44
Kum.M. 3.19 166
Kum.M. 3.23 73
Kum.M. 3.26 31
Kum.M. 3.27 105
Kum.M. 3.29 107
Kum.M. 3.30 108
Kum.M. 3.36 104, 105
Kum.M. 3.37 61
Kum.M. 3.39 108
Kum.M. 3.40 54
Kum.M. 3.41 98
Kum.M. 3.44 17
Kum.M. 3.50 118
Kum.M. 3.53 105
Kum.M. 3.54 134
Kum.M. 3.57 66
Kum.M. 3.59 31
Kum.M. 3.64 71
Kum.M. 3.65 80
Kum.M. 3.67 24
Kum.M. 3.68 24, 83
Kum.M. 3.72 136
Kum.M. 3.74 166
Kum.M.3.75 131
Kum.M. 3.76 79
Kum.M. 4.3 134
Kum.M. 4.9 29
Kum.M. 4.10 24, 100
Kum.M. 4.13 83
Kum.M. 4.14 104
Kum.M. 4.17 65
Kum.M. 4.18 77
Kum.M. 4.21 89
Kum.M. 4.26 51
Kum.M. 4.27 78
Kum.M. 4.42-4.43 162
Kum.M. 4.45 114
Kum.M. 5.1 95
Kum.M. 5.36 167
Kum.M. 5.55 14
Kum.M. 5.73 25
Kum.M. 5.86 24
Kum.M. 6.93 112
Kum.V. 1.1 114
Kum.V. 1.4 75
Kum.V. 1.10 62, 118, 142
Kum.V. 1.14 124
Kum.V. 1.15 26,53
Kum.V. 1.19 62
Kum.V. 1.25 15
Kum.V. 1.27 27
Kum.V. 1.27 [M 28] 98
Kum. V. 1.37 27
Kum. V. 1.37 [M38] 99
Kum.V. 1,45 [M 46] 116
Kum. V. 1.56 32
Kum.V. 1.56 [M 57] 98, 124
Kum.V. 1.58 15
Kum.V. 1.59 50
Kum.V. 1.60 99
Kum.V. 2.1 77
Kum.V. 2.3 72
Kum.V. 2.5 126
Kum.V. 2.7 126
Kum.V. 2.9 160
Kum.V. 2.17 78, III
Kum.V. 2.18 136
Kum.V. 2.26 17
Kum.V. 2.28 17
Kum.V. 2.30 III
Kum.V.2.33 78.82
Kum.V. 2.42 16
Kum. V. 2.44 82
Kum.Y. ~ 7 30
Kum.V. 2.51 61
Kum.Y. 2.57 74
Kum.V. 3.8 100
Kum.V. 3.10 27
Kum.V. 3.18 66
Kum.V. 3.27 104
Kum.V. 3.44 17
Kum.V. 3.46 100
Kum.V. 3.52 82
Kum.V. 3.56 101
Kum.V. 3.63 162
Kum.V. 3.69 80
Kum.V. 4.1 131
Kum.V. 4.9 76
Kum.V. 4.13 83
Kum.V. 4.35 82
Kum. V. 4.36 17
Kum.V. 5.10 80
Kum.V. 5.14 82
Kum.V. 5.22 :.. 60
Kum.V. 5.29 [M 30] 71
Kum.V. 5.44 53
Kum.V. 5.63 14
Kum.V. 6.3 62
Kum.Y. 6.43 55,82
Kum.V. 6.87 [6.88] 54
Kum.V. 6.92 78
Kum.V. 7.53 52
Kum.V. 7.57 76
Kum V. 1.49 [1.50] 57
Megh.M. 1 32, 106
Megh.M. 4 114
PPp.49 254
275
PP p. 55 241
PP p. 73 248
PP p. 84 241
PPp.90 253
PP p. 107 257
PP p. 123 252
PP p. 128 , 248
PP p. 135 225
PPp.137 227,260
PP p. 140 256
PP p. 171 241
PP p. 199 243
PP p. 211 , 259
PP p. 248 242
PP p. 249 254
PP p. 257 253
PP p. 317 261
PP p. 333 241
PP pp. 100-101 239
PP p. 30 220
PP p. 33 189
PP p. 79 196
PPp.163 217
PP p. 183 198
PP p. 258 209
PP p. 356 199
Ragh.H. 2.10 42
Ragh.H. 2.3 58
Ragh.H. 3.55 [M 54] 101
Ragh.H. 5.58 [M 5.57] 78
Ragh.M. 2.10 40,42
Ragh.M. 4.72 61
Ragh.M. 5.72 61
Ragh.M. 1.1 15,20,74,125,145
Ragh.M. 1.2 . 15,20,44,55, 113
Ragh.M. 1.3 63, 65
Ragh.M. 104 16, 20, 29,140
Ragh.M. 1.5 93, 160
Ragh.M. 1.5-1.9 162
Ragh.M. 1.6 94
Ragh.M. 1.7 23,28,29,44
Ragh.M. 1.8 10, 26, 75,133
Ragh.M. 1.9 99
Ragh.M. 1.11 122
Ragh.M. 1.12 80,82,99, 107
Ragh.M. 1.13 .. 77,84, 106, 133
Ragh.M. 1.14 22, 161
Ragh.M. 1.15 129
Ragh.M. 1.16 22, 103
Ragh.M. 1.17 14,21,124
Ragh.M. 1.18 .. 32,91, 129, 161,
167
Ragh.M. 1.20 21, 54, 72
Ragh.M. 1.21. 21
Ragh.M. 1.23 62, 136
Ragh.M. 1.24. 11, 16,21,22,99
Ragh.M. 1.25 22, 25, 70
Ragh.M. 1.26 21
Ragh.M. 1.27 20
Ragh.M. 1.28 20
Ragh.M. 1.29 20
Ragh.M. 1.30 83, 130
Ragh.M. 1.31 63
Ragh.M. 1.32 165, 166
Ragh.M. 1.35 46,123,163
Ragh.M. 1.36 20, 73, 160
Ragh.M. 1.38 53
Ragh.M. 1040 145, 167
Ragh.M. 1.43 63
Ragh.M. 1.44 29,56,78, 95
Ragh.M. 1.49 64, 123
Ragh.M. 1.50 124
Ragh.M. 1.51 82, 140
Ragh.M. 1.53 80
Ragh.M. 1.54 11
Ragh.M. 1.55 73, 129
Ragh.M. 1.58 105, 139
Ragh.M. 1.59 26, 31
Ragh.M. 1.60 72
Ragh.M. 1.61 59
Ragh.M. 1.63 84
Ragh.M. 1.66 81
Ragh.M. 1.67 113
Ragh.M. 1.69 165
Ragh.M. 1.71 65
Ragh.M. 1.72 21
Ragh.M. 1.78 167
Ragh.M. 1.81 ., 84, 131
Ragh.M. 1.83 103, 107, 161
Ragh.M. 1.84 113
Ragh.M. 1.85 60, 63
Ragh.M. 1.86 65
Ragh.M. 1.88 73
Ragh.M. 1.89 166
Ragh.M. 1.90 16
Ragh.M. 1.91 80,95
Ragh.M. 2.3 59
Ragh.M. 204 54
Ragh.M. 2.9 115
Ragh.M. 2.10 99
Ragh.M. 2.17 47
Ragh.M. 2.20 76
Ragh.M. 2.22 130
Ragh.M. 2.23 131
Ragh.M. 2.29 76
Ragh.M. 2.30 , 74
Ragh.M. 2.54 51
Ragh.M. 2.55 72
Ragh.M. 2.61 .47
Ragh.M. 2.65 54
Ragh.M. 2.70 73
Ragh.M. 2.71 60
RaghM. 2.72 53,58
Ragh.M. 2.73 28
Ragh.M. 3.5 65, 75
Ragh.M. 3.10 62
Ragh.M. 3.19 121
Ragh.M. 3.35 47
Ragh.M. 3.41 115
Ragh.M. 3045 III
Ragh.M. 3.51 51
Ragh.M. 3.54 101
Ragh.M. 3.55 134
Ragh.M. 3.57 122
Ragh.M. 3.59 122
Ragh.M. 3.62 62
Ragh.M. 4.1 76
Ragh.M. 404 71
Ragh.M. 4.5 100, 105
Ragh.M. 4.6 65
RaghM. 4.12 " 53
Ragh.M. 4.31 71
Ragh.M. 4.33 : 28
Ragh.M. 4.35 77
Ragh.M. 4040 76
Ragh.M. 4042 53
Ragh.M. 4045 74
Ragh.M. 4.61 108
Ragh.M. 4.74 74
Ragh.M. 4.87 , 75
Ragh.M. 5.2 56, 76, 78
Ragh.M. 5.3 79
Ragh.M. 5.9 72
Ragh.M. 5.15 73,124
Ragh.M. 5.19 68
Ragh.M. 5.25 135
Ragh.M. 5.32 57
277
Ragh.M. 5.36 83
Ragh.M. 5.40 55
Ragh.M. 5.45 50
Ragh.M. 5.52 80
Ragh.M. 5.57 54
Ragh.M. 5.61 50
Ragh.M. 5.62 51
Ragh.M. 5.69 63
Ragh.M. 5.73 , . 47, 108
Ragh.M. 5.75 77
Ragh.M. 11.8 66
Ragh.M. 11.10 65
Ragh.M. 11.11 160
Ragh.M. 12.68 53
Ragh.M. 12.71 , 163
Ragh.M. 12.73 115
Ragh.M. 12.74 64
Ragh.M. 12.75 65
Ragh.M 4.4 64
Ragh.M 4.85 58
SBh 1.1.4 182
SBh 1.1.1 177,180-182,189,
193,197,199,204,206,
210-212,216,219,220,
226,227,233,236,239,
241,242,246,252,255-
258,260,263
SBh 1.1.2 ... 180, 182, 190,203,
205,207,210,214,215,
224,255,256
SBh 1.l.3 ... 187, 200, 226, 242,
249
SBh 1.1.4 181,
196,197,199,204,205,
207-209,211,214,217,
219,220,227-230,233-
236,238,241-243,245-
248,252,254,255,261
SBh 1.1.5 ... 198,207,211,214,
215,228,238,239,257,
260
SBh 1.1.6 209, 244
SBh 1.1.7 181, 204, 217, 219
SBh 1.1.11 217
SBh 1.1.12 .. 214,219,224,233,
243,247,259
SBh 1.1.14 208, 217
SBh 1.1.19 .. 190,231,245,246,
248,257
SBh 1.1.20 224
SBh 1.1.22 .. 206, 223, 246, 253,
254
SBh 1.1.23 .. 206, 252, 254, 262
SBh 1.1.24 248,249,259
SBh 1.1.25 219,248
SBh 1.1.26 224
SBh 1.1.28 245
SBh 1.l.31 226, 245
SBh 1.2.1 216,224,253
SBh 2.1.1 233
SBh 2.1.4 177, 198,236
SBh 2.1.6 226, 243
SBh 2.1.7 233
SBh 2.1.24 232
SBh 2.2.17 176
SBh 4.1.1 251
SOS p. 2 235
SOS p. 3 256
SOS p. 6 235, 263
SOS p. 7 245, 247
SOS p. 8 244
SOS p. 25 260
SOS p. 31 259
SOS p. 2 181
SOS p. 3 180, 206
Sisu.M. 1.41 71
Sisu.M. 14.4 71
278
Grammatical Citations 279
Grammatical Citations
Coulson 46-49 195
Coulson 49-50 231
Coulson 61-63 251
Coulson 76 251
Coulson 107-11 . . . . . . . . . .. 185
Coulson 140-42 175
Coulson 154-58 175
Coulson 170-71 204
Coulson 171-77 251
Coulson 191-95 251
Coulson 195 231
Coulson 203-05 213
Coulson 207-14 251
Coulson 214-15 231
Coulson 230-35 213
Pa. 1.2.64-1.2.73 145
Pa. 1.2.68 81
Pa. 1.4.58 113
Pa. 1.4.60 116
Pa. 1.4.61-79 116
Pa. 2.1.5-2.1.20 92
Pa. 2.1.6 95
pa. 2.1.7 94
Pa. 2.1.13 93
Pa. 2.1.22-2.1.48 96
Pa. 2.1.22-2.2.22 96
Pa. 2.1.24 98
pa. 2.1.34-35 100
Pa. 2.1.37 99
Pa. 2.1.47 103
Pa. 2.1.49 104
Pa. 2.1.49-2.1.72 102
Pa. 2.1.50 108, 109
Pa. 2.1.51 109-112
Pa. 2.1.52 108
Pa. 2.1.55 106, 107
Pa. 2.1.56 107
Pa. 2.1.57 ff. . . . . . . . . . . . . . . 102
Pa. 2.1.60 105
Pa. 2.1.61 103
Pa. 2.1.64 113
Pa. 2.1.68 103
pa. 2.1.69 104
Pa. 2.1.72 123
pa. 2.2.1-2 101
pa. 2.2.3-5 101
Pa. 2.2.4 123
Pa. 2.2.18 113,114,116
pa. 2.2.19 119
Pa. 2.2.23-2.2.28 127
Pa. 2.2.25 134, 135
Pa. 2.2.26 135
Pa. 2.2.27 135
Pa. 2.2.29 125
Pa. 2.2.31 101
Pa. 2.3.25 180
Pa. 2.4.1 111
Pa. 2.4.2 126
pa. 2.4.29 112
Pa. 3.1.5-3.1.32 .43
Pa. 3.1.6--3.1.7 43
pa. 3.1.7 44
Pa. 3.1.8 46
Pa. 3.1.8-3.1.9 46
pa. 3.1.10 47
Pa. 3.1.11 47
Pa. 3.1.12-3.1.13 47
pa. 3.1.15 48
Pa. 3.1.17 48
Pa. 3.1.18 48
Pa. 3.1.21 47
Pa. 3.1.22-3.1.24 45
Pa. 3.1.25 43
Pa. 3.1.26 45
Pa. 3.1.91-3.4.117 49
Pa.3.1.I01 66
Pa. 3.1.108 59
Pa. 3.1.115 66
pa. 3.1.124 79
pa. 3.1.134 114
pa. 3.1.136 42
pa. 3.1.145-147 57
pa. 3.2.78 64
pa. 3.2.79 64
pa. 3.2.83 64
pa. 3.2.84-3.2.122 56
pa. 3.2.99 11
pa. 3.2.134-3.2.179 57
pa. 3.3.1-2, 3.4.75 49
pa. 3.3.126 113
Pa.3.4.68 66
pa. 4.1.15 72,74
pa. 4.1.76--5.4.160 67
pa. 4.1.92 74
pa. 4.1.137 75
pa. 4.1.170 74,76
pa. 4.2.49 76
pa. 4.2.129 73
pa. 4.3.11 73
pa. 4.3.17 73
pa. 4.3.23 72
pa. 4.3.74 71,72
pa. 4.3.79 71
pa. 4.3.81 124
Pa. 4.3.82 124
Pa. 4.3.143 ff 124
Pa. 4.3.148 ff 124
Pa. 4.4.1 74
Pa. 4.4.45 74
Pa. 4.4.76 74
Pa. 4.4.77 11
Pa. 4.4.92 72
Pa. 4.4.98 74
Pa. 4.4.99 73
Pa. 4.4.105 73
pa. 5.1.5 67
Pa. 5.1.9 71
Pa. 5.1.10 71
Pa. 5.1.36 70
Pa. 5.1.42 76
Pa. 5.1.105 77
Pa. 5.1.115 70
Pa. 5.1.115-116 234
Pa. 5.1.117 78
Pa. 5.1.119 .'175
Pa. 5.1.125 75
Pa. 5.1.128 :. 76
Pa. 5.1.129 10, 75
Pa. 5.1.130 75
Pa.5.2.36 77
Pa. 5.2.37 124, 194
Pa. 5.2.42 78
Pa. 5.2.121 80
Pa. 5.2.124 80
Pa. 5.2.134 68
Pa. 5.3.7 , 81
Pa. 5.3.57 ,: 82
Pa. 5.3.67 83, 191
Pa. 5.3.103 82
Pa. 5.4.25 78
Pa. 5.4.38 82
280 Grammatical Citations
. "'!'!.! .
281
Pa. 5.4.50 83, 117
Pa. 5.4.77 .. , 84, 126
pa. 5.4.78 84
pa. 5.4.91. 112, 114
pa. 5.4.151 84
pa. 5.4.153 84, 131
pa. 5.4.154 188
Pa.6.1.94 122
pa. 6.3.73-77 , 117
pa. 6.3.101 ff.. 112
pa. 6.3.107 113
pa. 6.3.109 121
pa. 6.4.174 76
pa. 7.4.32 83
Speijer 16-17 195
Speijer31 195
Speijer 193.3 185
Speijer 197 185
Speijer 213 b 185
Speijer 228-31 185
Speijer 229.1 231
Speijer 235-39 175
Speijer 241 231
Speijer 261-90 251
Speijer 365-68 204
Speijer383-90 213
Speijer 394-490 251
Speijer 430 231
Speijer 470 a 231
Speijer 491-99 213
Whitney 303 204
Whitney 607 .43
Whitney 775 43
Whitney 996 ff .43
Whitney 1000--1025 45
Whitney 1026-1040 44
Whitney 1040 43
Whitney 1041-1052 45
Whitney 1041 b 43
Whitney 1053-1068 46
Whitney 1056 43
Whitney 1090 ff 116
Whitney 1107 231
Whitney 1143-1201 , .. .49
Whitney 1202-1245 67
Whitney 1222 h 185
Whitney 1237 175
Whitney 1239 175
Whitney 1252-1258 125
Whitney 1253.1 125
Whitney 1253.2 126
Whitney 1262-1291 96
Whitney 1264-1278 96
Whitney 1279-1291 102
Whitney 1280 102
Whitney 1281 ff 112
Whitney 1286-1287 119
Whitney 1288.1. 117
Whitney 1292 ff 115
Whitney 1293-1308 127
Whitney 1310 . . . . . . . . . . . . . 115
Whitney 1312 . . . . . . . . . . . .. 108
Whitney 1313 92
Sanskrit Index
-a, 10, 11,71,74-76,83,206
-aka, 73
agre,229
aC,84
aN, 10,75
72, 74-77,82
ata eva, 256
atal), 209, 256, 258
atra,253, 258, 259
atra ke cid ahul,1, 243, 249
atrocyate, 244
atrabhidhiyate, 244
ailia, 240--242, 251, 252
ailiava, 249
adal),262
adhikaraJ.le, 55
-ana, 38, 39, 72
anavadyam, 218
anubandha,xxvi,16,37
anuvartate, 164
anuvJ;lla, 163
anuv!"lti, 163, 164
anekapada-bahuvnlhi, 132
anena,25
-anta,231
-antara, 191
anya, 191,261
anyac ca, 254
anyailia, 246
anye,262
anvaya, 154, 157, 159,217
anvaya-mukhena, 149
anvayamukhi, 147-149, 151, 153-
156, 159
apara aha, 243, 249
aparalJ1 ca, 254
apare,262
apadane,55
api, 198,210,257,258
apica, 254, 257
a-prak!ta, 224
abhava, 206, 207
abhiita-tadbhave, 83
abhyupagatam, 218
abhyupagama, 207
abhyupetam,218
ayam, 69, 261, 264
-artha, 191
arthal,1,217
arh,219
aiaukika-vigraha, 34
avataraJ.la, 151, 224
avasitam,218
avidya-hetukatvam, 179
avyayibhava, 86, 87, 90, 92, 93,
95, 185
asiddhi, 206
astu,248
asmin,259
asya vidyate, 79
asyasti,79
a-sva-pada-vigraha, 90
a, 16
a1c?ipati, 27
243
aksepasamadhana, 5, 171
-iikhya,30
282 Sanskrit Index
283
an,16
atma-mane, 64
-adi, 22, 186, 190,231-233,235
-adika, 232
-adya, 232
apatti, 207, 246
apad,246
apadyate, 246
amna,229
amnata, 229
MaIik,242
asarikyate, 242
aha, 225, 243
-ika, 73, 74
ieehati,44
it, 37
itaJ;1, 256
itaC,78
itaretara dvandva, 145
itaretara-dvandva, 125
iti, 15,52,56,57,61-63,79,173,
179,195,198-200,203,
207,208,213-218,220,
221,223,228,240,244,
245,254,261,263
iti iqtva, 208
iti gamyate, 26, 218
iti eet, 242
iti efuvam-adi, 234
iti tasmat, 208
iti tatparyaJ;1, 25
iti tena, 208
iti 218, 229
iti 229
iti dhvanyate, 26
iti niseitam, 217, 218
iti nisclyate, 217
iti nyayat, 229
iti prasiddham, 218, 230
iti prasidhyate, 230
iti prapte, 243
iti badhitam, 218
iti bhavitum arhati, 219
iti 218
iti bhavaJ;1, 23-25
iti matva, 214
iti yavat, 23, 25
iti yuktam, 218
iti viruddham, 217, 218
iti virudhyate, 217
iti virodhaJ;1, 217
iti 25
iti 26
iti 160, 163
iti srilyate, 230
iti sambhavyate, 218
iti siddham, 218
iti sugamam, 218
iti siieitaJ;1, 26
iti siieyate, 26
iti 218
iti smaryate, 230
itivat, 213, 235
ittham,236
ity ataJ;1, 208
ity atra, 254
ity anavadyam, 217
ity anvaYaJ;1' 159, 217
ity abhiprayaJ;1, 25
ity abhyupagatam, 217
ity abhyupetam, 218
ity arthaJ;1, 23-26, 217
ity avasitam, 218
ity amnayate, 229
ity asankya, 242
ity MaIikyaha, 225
ity asayaJ;1, 25
ity uktva, 214
ity upapannam, 218
ity eval)l prapte, 214
ity-adi, 213, 233, 234
ity-adi ea, 234
ityadivat, 213, 235
ity-evam-adi, 234
idantii, 182
idam, 52, 54, 55, 77, 182, 223,
253,258,262-264
-in, 64, 68
inl, 68,80
indrasya, 179
iva,64,89,231,234-236
44
iha, 253,254,258,259
iha tu, 254
iha lake, 254
ihapi,254
-1,72,74, 84
1-,83
-Ina, 71, 73, 74
IyasUN,82
ukta, 226--228
ukta-nyayena, 228
ueyate, 225, 244
ul),49
ul)-adi,49
uttarapada-dvigu, 110
uttaram aha, 225
upapatti, 206, 207
upapada, 37, 42, 57-64, 70, 119,
154
upapannam, 218
upamana, 234,235
upamanapiirvapada, 106
upamita-samasa, 107
upameya, 236
upada, 223
upadana, 223
upadhi, 57, 73
-ii, 84
81,145,145,146
35
eke,262
-el)ya,73
etat, 214, 223, 262-264
etena,25
-eya,71
eVal)l prapte, 214
eVal)l sati, 211
evam, 211, 214,234,236
evam api, 258
-ka, 82, 83, 179, 185, 188
Ka, 37, 38
ka, 188
kathal)l punaJ;1, 241
kathal)l-bhiita, 148
kathal)lbhiitinI, 147, 148, 150, 151,
154-156, 159, 163
katham, 241
kaP,84
karal)e, 54, 62
karal)e LyuT, 39
kartari,52-54
284 Sanskrit Index
285
kartari!qt, 49, 60, 61, 64
kartary upamane, 64
kannaI)i, 52, 54, 56
!qt, 61
kannadharaya,36,86-89,91,102,
106, 186, 187, 189
kaHipaka, 161
-kaIpa, 191
kalpaP, 83, 191
kasmal,259
kamam, 248
-kara, 15
karaka, 52, 60, 61
karaI)a, 208
kfuika, 1
kavya,2
kil)l ca, 251, 254
kil)l tarhi, 255
kil)lIU, 255
148
kim, 185, 190,241,258
ku-, 112
kUla!), 258, 259
kUlra, 258
ku1aka, 161, 162
lq, 51, 83
!qt, 11, 35, 38, 42, 45, 46, 49, 50,
52-M
lqd-anla, 49
lqd-V!1li, 35
ke cil, 262
kevaIa-samasa, 88, 89
kr!,61
kva, 241, 258
KviP,38
kha,71
khaN,73
gauravam, 176
-gala, 192
116
gam, 192
gamyate, 26, 218
-garbha, 192
gUI)a!), 176
guru, 176
gurula, 176
gurulvam, 176
guru-bhava!), 176
g,hIla, 223
gmin!,80
grah,223
Nas, 34
N-IP, 72, 74
ca, 88,205, 206
cikilsali, 43
cinlayali, 43
cur-adi,43
cel, 252
corayali, 43
cvi, 47, 83
-jana, 192
-jala, 193
jugupsale, 43
!Ika, 147, 149
!yuL,72
!haK,74
!haN,73
Qa,11
QUlqN,39
QUlqN karaI)e, 39
<;lhaK, 11
<;lhaN,71
71
I)amu1, 51
!)ic,45
I)ij-anta-dhatu, 45
74
79
-1-, 72
tac-c!ll'a, 57, 58
lal, 67, 77, 164, 185, 189, 190,
209,228,245,252,256-
259,262-264
tal-, 63
lallalha, 40
lallatha-bhiltam, 40
lallath6klam, 40
lala!), 209, 256, 258, 259
58-60, 68, 70, 83, 86,
87,90,92,96,121, 154,
157,176,186,187,194
latta,211,252,253,258,259
latta bhavam, 72, 73
latta sadhul), 73
lattilival)l sali, 2Il
lal-sila, 57
176, 178, 179, 185, 186,
188, 189, 191-193
latha, 167,211,236,260,261
latha ca, 254
latha sali, 211
latha hi, 231, 255
lathapi, 258
lad ayuklam, 245
lad asal, 245
lad idam, 263
lad elal, 263
lad eva, 257
lad yatha, 231, 255
lad-artham, 78
tad-arham, 78
taddhila, 10,34,35,39,67, 68,
74,83, 191, 194
laddhila-v!1li, 35
laddhilartha-dvigu, 109
laddhitanla, 67
tadvat, 236
tan na, 245
Ian rna bhiid ili, 245
layaP,78
-laya, 180
layo!), 259
laraP,82
larhi,248
-las, 81, 258
lasIL, 81
tasmai hilam, 67
lasmat, 209, 256, 259
tasmad eva, 256
tasmin, 211, 259
lasya, 179
tasya kanna, 75
lasya bhava!), 75
tasya raja, 76
tasya vikilra!), 76
tasya samuha!), 76
tasyedam, 77
tasyapalyam,74
-ta, 175-183, 188,204,206
tacclu'ya, 63
286 Sanskrit Index
287
Uicchilye, 64
tiiraklldi, 78
tavat,253
tavata,253
tavata,209
-ti, 39, 206
tin, 49
43
tu,255
tena, 209, 256
253
madhye, 253
-tra,81,258
-tva, 175-183, 188,204,206
-tvat, 180
-tvena, 180
darsana, 207
-darsanat, 229
226
135
dig-bahuvnlli, 135
dldaJ1lsate, 43
dfS,229
dfsyate, 229
218, 229
dosa,246
85-91, 125, 176, 178,
185, 186, 189
145
dvigu,108
dhatu-lopa, 114
dhvanyate, 26
na,16,244
na ca, 199,244
nacacodanlyam,244
na ca mantavyam, 244
na ca vacyam, 244
na 246
na 246
nanu, 240
146
nahi, 199
nan, 16
36
-nama, 30
nama, 28, 30
niima-dhiilU, 46
nitya,91
nitya-samiisa, 90, 91
nipiita,68
nipiitana, 65, 68
nipatita, 36, 65, 68
nimitta-saptami, 210
niscitam, 217, 218
niSclyate, 217
nivflta, 164
nivflti, 164
nyaya, 229
nilitad evam, 245
246
naYaf!l 246
27, 238
pa!hati, 16
padaccheda, 4, 9, 13
padapa!ha, 13
padiirthokti, 4, 9
paraJ1l tu, 255
parigrah, 223
vibhakti-vigraha, 41
pa!ha, 16
puna\.!, 255
puna\.! puna\.!, 45
purohitadi, 76
-piirva, 193
238
220, 237, 238, 240,
243,244
238, 239, 244
239
121
prakfla, 224
prajnftdi, 82
pratIka, 150
prathamanta-vigraha,41
-prabhfli, 232
prabhfti, 232
prasanga, 207, 246
prasajyate, 246
prasafij,246
prasiddham, 218, 230
prasiddhi, 207
-prasiddhel), 230
prasidh, 230
prasidhyate, 230
113, 114
pradi-bahuvnlli, 130
pradhanya, 86
prapta,243
prapte,214
bahuvn1ll,30,40,41, 77, 79,83-
89, 91, 127, 128, 130,
132,134,136,157,178,
179,185-188,190-193,
231-233,235
ba<)ham, 248
badhitam,218
bThhatsate,43
briima\.!, 244
bhagavan/224
57, 218
bhavet,247
-bhava\.!, 176
bhave,50,51,59,65
bhave kfl, 59, 60
171, 172
224
bhii,83,219
-bhiita, 165, 166, 189, 198
bhiita, 198
bhiite,56
bhfsam,45
-bheda, 194
matUP, 80, 81
mate, 230
matvartblya, 68, 79
matva,214
madhyamapada1opa-bahuvnIii, 42
-maya, 124,208
mayaT, 124, 208
mayiiravYaf!lsakadi, 123
mahakavya, xxvi
mahii-balatvat, 179
mahakavya, 2
mahakiivya, 2
-matra, 193, 194
matraC, 78, 82, 194
-matra, 193
mlmilJ11sate, 43 .
miila, 147-151, 154, 156, 157,223-
226
miilakiira, 223-228
-ya, 11,70-76
288 Sanskrit Index

f
1
289
yaK, 76
yan,45
yan-anta-dhiitu, 45
yaT, 11,34,70-72,74-76,78,82
yat, 164, 185, 190,208,228,245,
258,259,261,262
yat tu, 228
yat ll1ktam, 228
yata!), 208, 258
yatra,258,259
yatra yatra, 259
yathii, 167, 231, 235, 236, 260,
261
yathii tatha, 167
yathokta, 228
yadi,252
yady api, 248, 258
yadva,249
yasmlit,208
yasmin,259
yasya,230
yasya mate, 230
yavata,208
yu,38
yuktam, 218, 219
yugma, 161
yugmaka, 161, 162
yuj,218
yuv1idi,75
yena,208,259
ya ya!), 259
yaga,207
yajya, 160
rupakasamasa, 108
laukika-vigraha, 34
36
27
LyuT, 38, 39
vaktavyam, 200
226
226
-vat, 231, 234, 235
vat!, 71, 79
-vaot,234
vanate, 164
vanamane, 56
vakyayajaoa, 5, 9
vakya-vrtti, 35, 187
vigraha, 4, 9, 21, 33, 35, 86, 91
vinI,80
87,186,189
vibhakti-sarpjnaka, 81
vibhaktisarpjnaka, 68
viruddham, 217, 218
virudhyate,217
virodha!),217
226
25
26
30, 194
161
vuN,73
vrtti,35
Vfddhi, 10,49, 176
sak,218
sakyam,219
sakya,219
-sabda, 15
sastra, 1,2
sasttiya-vigraha,34
-Sila, 58, 63
Silam asya, 63
sISarpsate, 43
sukla, 176
suklatvam, 176
sravaI)a, 207
-sravaI)at, 230
sru,230
sruti,230
sruyate,230
-sa, 43
saqIkhya-bahuvnni, 134
-sarpjna, 30
sarpdiinitaka, 161
sarpbadhyate, 160
sarpbandha, 160
sarpbandhanIya, 160
sarpbandhin, 77
sat, 165, 166, 198
sali, 166,210,211
salI, 165
satyam, 248
satya-vadI, 57
sao, 35,43,44, 165
sao-adi,49
sao-ady-aota-dhatu, 43
sao-ady-anta-dhatu-vrtti,35
saot, 198, 211
sann-aota-dhatu, 44
samasiinta, 81
samasa-vrtti, 35, 187
samasiinta, 83
samahara, 90
samahara-dvaodva, 36, 126
samahara-dvigu, 111
sampradiine, 54
-sambandhin, 78
sambhava,206, 207
sambhiivyafe,218
sarvathilpi, 248
saha-bahuvnni, 131
siddham,218
siddhiinta, 237, 238, 244
siddhiintin, 238-240, 244
siddhi,207
sU,34
su-, 112
sugamam,218
sup, 67
sup-sup-samasa, 89
sueita!), 26
sueyate, 26
sutra, 1,3, 163, 164,238
224,226,227, 242
stli,67
218
smara!)a, 207
-smaraI)at, 230
srnr,230
syat,247
syad etat, 247
syad etad evam, 247
svarthika, 81
svarthika, 68
hi, 203, 211, 251
hita,67
he, 167
290 English Index
291
English Index
ablative and instrumental phrases,
206
abstract nouns, 177-185
causal, 182, 183
common constructions, 179-
182
fonnation, 177, 178
in ablative case, 182, 183
in accusative case, 183, 185
in compound, 180
in dative case, 183
in genitive case, 184
in instrumental case, 182-185
in locative case, 184
in nominative case, 183
predicative instrumental, 184,
185
significance, i 78, 179
with subjective genitive, 179-
182
with two genitives, 182
with verbs of motion, 185
abstract suffixes, 177
accompanying words, see upapada
compounds
adhikaralJe krt, see primary deriva-
tion, in kiiraka meanings
iidi in glosses, 22
adverbs, 169
pronominal, 260-264
Almikiiraca,tiimalJi of Hemacandra,
3
alternative interpretations, 251
citing others, 25 I
giving second views, 25 I
alternatives given by commenta-
tors, 156
of Amarasirylha, 10
Amarasirylha, 10, 29
analysis, 4, 5, 12, 19, 33-42
and inflection of the original,
41,42
general principles, 35-39
glossing within, 40
in ordinary language, 34, 37,
153
in technical language, 34, 37-
39, 153
insertion of, 21, 39-42
interruption of, 41, 42
obscuration of, 37, 40
types of, 34
Annambhana, 2, 161
antecendent of pronoun
supplying, 32
anubandhas, see indicatory phonemes
anuvrtti, see words carried over
anya
repeated, 263, 264
apiidiine krt, see primary deriva-
apposition, 20
arranging the words of the text,
149-160
anvayamukhf approach, ISO,
lSI
insertion of explanatory ma-
terial, 151-156
katha,!,bhiitinfapproach,149,
ISO
overview, 149
variations on the two main ap-
proaches, 156-160
Arul)agirinatha, 159
atha in objections, 243
attibutive noun, 199
attributive adjectives, 167
attributive nouns
with bhiita, 200
with sant, 200
attributive substantives, 167, 168
author
references to, see references
authorities
autocommentaries, 2, 3
auto-commentary, 229
avyayfbhiiva compounds, see com-
pounds, avyayfbhiiva
bahuvrfhi compounds, see com-
pounds, bahuvrfhi
Bhagavadgftii, 168
Bhamaha,2
bhiirya style, 173-266
introduction, 173-175
Bhani,2
Bha!!ikiivya of Bha!!i, 2
bhiive krt, see primary derivation
krt, see primary deriva-
tionwith temporal restric-
tions
-bhiit(l
marking a karmadhiiraya, 191,
192
bhiite krt, see primary derivation-
with temporal restrictions
Biihtlingk; O. von, 38
bold type, xxvi, II, 14, lSI
bound fonns, 92, 119-121, see
upapada compounds
of Sailkara,
xxvi
ca
in series of causes, 207, 208
carat, see circumflex accent
Cardona, George, 38
Ciritravardhana, 158
causal sentences, 205-214
with ca in causal series, 207,
208
with hi, 213, 214
with iti clauses, 209, 210
with ablative and instrumen-
tal phrases, 206
with locative absolute, 211-
213
with relatives, 210, 211
with series of causes, 206, 207
with verbal nouns, 208, 209
causative verbs, 45, 46
circumflex accent, xxvi
citation of words, 14-16
using iti, IS
using -sabda or -kiira, IS
using indicatory phonemes, 16
using inflected indeclinable,
IS, 16
colon, xxvii
commentaries
abundance of, 1-3
English Index
continuative
in -Qln, 51
continuatlves, 51
correlatives, 261-264
multiple, 262
repeated, 261
Coulson, Michael, xxv, xxvii
cvi formations, 47, 83, 84
293
definition
marking as, 29
definitions, 28-30
offered by commentator, 28,
29
quoted from dictionaries, 29,
30
demonstratives, 264-266
level of proximity, 264, 265
referring to earlier and later
passages. 265
denominative verbs, 46-48
"becoming something", 47
"behavior like", 47 .
"behavior toward", 47
"desire to have", 46
"doing or making something",
48
"feeling something". 48
"making something so", 47
"making use of something",
48
derivation
primary. see primary deriva-
tion
secondary, see secondary deriva-
tion
derivative verbs, 43-48
additional meaning, 89, 90
aggregation, 90
clarification of connection,
90
comparison, 89
basic categories, 86-89
containing pronouns, 192
ending in -antara, 193
ending in -arrha, 193
ending in -garbha, 194
ending in -gata, 194
ending in -jana, 194
ending in -jiita, 195
ending in -kalpa, 193
ending in -tniitra, 195, 196
ending in -purva, 195
ending in 196
ending in special terms, 193-
196
irregular, 92
longer, 137-145
analyzing, 187-189
choice of options, 142-145
special techniques, 141, 142
obligatory, 90-92
asvapadavigraha, 90, 91
bound forms, 92,119-121
conventional meanings, 91
optionality of, 90
relation to outside words, 189,
190
connective particles, 253-258
construction
indicating, see indicating the
construction
1
I
I
of reciprocal combat, 136
special, 135, 136
dvandva, 87, 88, 125-127
itaretara, 126
satniihiira, 126, 127
marked by tat, 191
dvigu, 109-112
satniihiira, Ill, 112
taddhitiJrtha, 110
uttarapada, 110, III
see
with inseparable first mem-
ber
kartnadhiiraya,87,102-106
marked by -bhuta, 191, 192
of comparison, 106-108
kevala-satniisa, 88, 89
see
ber
87, 96-125
dvigu, see dvigu
galf, see with in-
separable first member
kartnadhiiraya, see kartnadhiiraya
tnayuravyaTf'lsakiidi, 123, 124
priidi, see with
inseparable first member
121-123
upapada, 119-121
case, see depen-
dent
dependent, 96-102
irregular, 121-125
negative, 118, 119
ber, 112-118
as vehicles of originality, 2
assistance they offer, 1,3-5
five services of, 3, 9, 10
importance of, 1-3
reasons for, 1-3
upon commentaries, 2, 3
commentary
references to in commentary,
see references
comparisons, 233-238
usind itivat and ityiidivat, 237
using iva, 238
using vat, 236, 237
using yathii, 238
complex formations, 4, 5, 12, 19,
33-42
modes of, see modes of com-
plex formations
compound nouns, see compounds
compounds, 4, 14,85-145, 187-
196
avyayfbhiiva, 86, 87,92-95
irregular, 95
obligatory, 93-95
optional, 93
bahuvrfhi, 88, 127-137
anekapada, 132, 133
dig, 136
pradi, 131
saha, 131, 132
saTf'lkhyii, 135, 136
appositional, 128-133
in adi, 233-236
irregular, 136, 137
marked by -ka, 190
negative, 130
nonappositional, 133, 134
292
English Index 294
types of, 43, 44
desiderative verbs, 44
Devasthali, G. V., 38
Dhi'itupii!ha appendix to PaI)ini's
grammar, 39
dialectic style, 239-242
example of, 241, 242
dictionaries
formats of, 29
in Sanskrit, 10, 12,29,30
Sanskrit-to-English,4
diminutive suffixes, see secondary
derivation, svarthika suf-
fixes
discussion, 239-251
direct address in, 241
double meanings, see puns
dvandva compounds, see compounds,
dvandva
dvigu compounds, see compounds,
dvigu
formations, 147, 148
male and female, 147
neuter, 148
ellipsis, xxvi
emphatic particles, 258-260
enclitics
glossing of, 20, 21
euphonic combination, see sandhi
examples, 233-238
using -iidi, 233-236
using vat, 236, 237
extension
meaning by, 27
five services, 3, 9,10
footnotes, 3
formulas of analysis, see analysis
future participle, 57
genitive case
in reference to views. 232
gerund
in English, 50
in Sanskrit, see continuatives
gerundive, 65
glosses
location of, 20, 21
glosses, expansions of, 22, 23
glossing, 4, 5, 19-32
by equivalent forms, 21
types of, 10-12
with participles, 21
with periphrastic constructions,
21
habitual action, see primary deriva-
tion
halfpenny, 10I
Hemacandra, 3
Hemadri, 37, 42
hi
in causal sentences, 213, 214
honorifics, 82
implication, see extension
implications
explicit statement of, 22
implied contrasts
explicit statement of, 22, 23
indeclinable
inflected, 15, 16
indeclinables
glossing of, 31
1
i
indicating the construction, 5, 149-
169
adjectives and adverbs, 163
arranging the words of the text,
see arranging the words
of the text
connections between statements,
164
connections between verses,
163
direct statement of connec-
tions, 161-165
marking features of syntax,
166-169
attributive adjectives, 167
attributive substantives, 167,
168
locative absolute, 168, 169
subject and predicate, 166,
167
with sa,!,bandha, etc., 162
with 163
with yojya, 163
words carried over, see words
carried over
words supplied, see words sup-
plied
indicatory phonemes, xxvi, 16, 37-
39
infinitives, 51
Ingalls, Daniel H. H., xxv
instrumental case
of present participle, 228
intensive verbs, 45
interpretations
alternative, see alternative in-
terpretations
295
introduction to section of commen-
tary, 153
iti
at end of passage, 222, 223
common constructions, 215,
216
in analysis of krt formations,
52,56,57
in comment, 218-220
in explanation, 218
in quotation, 216-218
scope of, 200, 201
uses of, 215-223
with infinitive, 220-222
iti cet, 244
ityiidi in examples, 235, 236
-ka
marking a bahuvrfhi, 190
Kale, M. R., 38
Kalidasa,2, 10, 150
kiirkake krt, see primary deriVa-
tion
kara'.'e krt, see primary derivation,
in kiiraka meanings
kiirikii works, I
karmadhiiraya compounds, see com-
pounds,kannadhiiraya
karma'.'i krt, see primary deriva-
kartari krt, see primary derivation,
in kiiraka meanings
Kiisikiivrtti of Jayaditya and Vamana,
46
kiivya poems, 2
Kiivyiilankiira of Bhamaha, 2
296 English Index
297
Kiivyii"usiisana of Hemacandra,
3
Kesavasvamin, 29
kevalasamiisa compounds, see com-
pounds, kevala-samiisa
Kiparsky, Paul, 91
Kira"iivalr of Udayana, 2
krt suffixes, see primary deriva-
tion
Kumiirasambhava of Kalidasa, 138,
141, 142, 150, 152, 157
locative absolute, 168, 169, 184
abbreviated, 213
concessive, 213
expressing a cause, 212
expressing a condition, 212
in causal sentences, 211-213
locative case
in dictionaries, 29
in giving the meaning of a
root, 39
in giving the meaning of suf-
fixes, 39
in glossing particles, 31
Magha,153
mahiikiivya poems, xxvi, 2, 154
Mallinatha, xxvi, 10, 11, 37, 40,
42, 123, 139, 142-145,
151,155-157,167,168
memorization, 2
modes of complex formations, 35
mula (root text), 149-153, 156-
159
na ca in objections, 246
lJamul gerund, see continuative in
-am
nanu in objections, 243
159, 161
nominal compounds, see compounds
nominative case
in analyses, 41, 56, 62, 64
in definitions, 10
in dictionaries, 29
nouns
abstract, see abstract nouns
null suffix, 37, 38
Nyiiyakosa ofBhlmacarya Jha1akiKar,
3
objections, 242-246
replies to, 246-250
answering,S
answering of, 173
using iikipta, 245
using iti eet, 244
using ityiismiya, 244
using na ca, 246
using nanu or atha, 243
using priipta, 245
using interrogatives, 243
using verbs of speaking, 245,
246
omission of words from the root
text, 156
oral argument, 1
oral debate,S, 239, 240
oral instruction,S
original text, see mula (root text)
Padapii!ha text of the Veda, 13
Padiirthasmigraha ofPrasastapada,
2
PaJ:.linian grammar, xxi, xxii, xxvi,
5,9, 10, 12, 16, 19,33-
39, 43, 46, 49, 50, 52,
56-59, 64, 67, 68, 73,
74, 77, 78, 83, 88, 89,
91, 94, 123, 152, 161,
193, 196, 236
paraphrases
more removed, 23-27
paraphrasing, 4, 19-32
with anena, 25, 26
with iti yiivat, 25
with ity arthai}, 23, 24
with ity bhiivai}, 24
with other tags, 25
Parasarapuriirza, 3
parentheses, xxvii
participles, 51, 64, 65
in references, 230
particles, 253-260
connecting sentences, 253-258
giving the mearting of, 31
of emphasis, 258-260
partitive genitive, 101
past active participle, 56, 62
past passive participle, 56, 63, 65
patronymics, 74
pejoratives, 82
phrase mode, 35
pick-up pronoun, 28, 41, 42
plural
in references, 229
possessive suffixes, see secondary
derivation
Prasastapada, 2
predicate nominative, 197-199
predicative accusative, 198
predicative instrumental, see ab-
stract nouns
preliminary' view, see
present participle, 65
present passive participle, 56
primary derivation, II, 42, 49-66
habitual action, 57, 58
in kiiraka meanings, 52-55
adhikara"e, 55
apiidiine, 55
kara"e, 54
karma"i, 52, 54
kartari, 52, 53
sampradiine, 54
in the sense of the action, 50,
51,59,60,65
irregular forms, 65, 66
other restrictions of meaning,
57,58
verbal participles, see partici-
ples
with accompanying words, see
upapada compounds
with temporal restrictions, 55-
57
future, 57
past, 56
present, 56, 57
primary nominal derivation, see
primary derivation
pronoun
pick-up, see pick-up pronoun
relative, see relative pronouns
pronouns, 260-266
adverbs used as, 260, 261
correlative, 261-264
demonstrative, 264--266
298 English Index
299
glossing of, 32
in compounds
containing pronouns, 192
pronouns subjects, 198
proper names
glossing of, 30
prose word order, 150, 156
puns
glossing of, 27, 28
purvapalqa, 240
quotation marks, 15
Raghuva,!,sa of Kalidasa, 10
Rama Govinda Siddh1intavaglsa,
158
references, 225-232
to authorities, 231, 232
to commentary, 229-231
to earlier topic, 231
to text or author, 225-229
to views, 232
relative pronouns, 151
in analysis of krt formations,
52
in analysis of taddhita forma-
tions, 80
in references, 230
repetition of text, 14, 15
replies, 246-250
using etc., 248, 249
using prasanga, etc., 248
using tarhi, 250
using concessives, 250
using denials, 247
using optatives, 249
using verbs of speaking, 246,
247
root text, see mula (root text)
root-forming suffixes, 43, 49
Samilsacakra, 36, 150
sampradane krt, see primary deriva-
tion, in kilraka meanings
sandhi, xxvi, 4, 13
Satlkara, xxvi
Sanskrit texts
modem editions, 13
Sanskrit tradition
fondness for commentaries,
1,2
Indian curriculum, 2
peculiar features, I
pedagogical concerns, 2, 3
richness of commentaries, I
sastra works, I, 2
secondary conjugations, see deriva-
tive verbs
secondary derivation, 10, 35, 67-
84
case-substitute suffixes, 8I
compound formulas, 78, 79
evi, see cvi formations
dependent formulas, 68-74.
ablative, 71, 72
accusative, 70
dative, 71
instrumental, 71
locative, 72-74
genitive formulas, 74-77
connection in general, 77
tasya bhilva/:!, 75
tasyiipatyam, 75
nominative formulas, 77, 78
possessive suffixes, 79-81
straightforward possession,
80
with additional meaning, 80,
81
sviirthika suffixes, 82-84
diminutives, 82
samasiinta suffixes, 84
secondary nominal derivation, see
secondary derivation
secondary usage, see extension
semicolon, xxvii
sentences
causal, see causal sentences
long, 201-203
separation of words, see word di-
VISIOn
series of causes, 206, 207
siddhiinta, 240
Sisupalavadha of Magha, 144, 145,
153
skeleton sentence, 149, 150, 161
special items
glossing of, 30-32
species names
glossing of, 31
speech
referring to, 26-28
Speijer, J. S., xxv, xxvii
square brackets, xxvi, xxvii, II
standard of comparison, 236, 237
students
beginning, 4, 168, 173, 174
frustrations of, 4
in traditional curriculum, 2,
36
memory of, 30
of the.acred lore, 26, 68
ponies for, 13
problems of, 3
tedious life of, 5
subcommentaries,2,3
subject and predicate, 166, 167
subject at hand, 226
subjective genitive, 179-182
suffixes
PaQini's treatment of, 37-39
suffix-formations resembling com-
pounds, 125, 193, 196
suggestions
conveying of, 26
siitra works, 1
siitra works, 1,3, 165, 166
sviirthika suffixes, see secondary
derivation
synonyms
simple, 20, 40
taddhita suffixes, see secondlrry
derivation
Tarkadfpika of AnnaI)lbhana, 2,
161
Tarkasmigraha of AnnaI)lbhana,
2, 161
tat
marking a dvandva, 191
compounds, see com-
pounds,
technical terms
glossing of, 28
temporal restrictions, see primary
derivation
tenth-class verbs, 43
300 English Index
text
text being commented on, see mula
(root text)
"therefore"
words for, 258
thesaurus, 29
topic of discussion, 226
Udayana,2

uncommon words
glossing of, 28
1
upapada compounds, 58-64
with kartari krt, 60, 61
with krt in another karaka mean
ing, 62, 63
with other conditions of mean
ing, 63, 64
with passive krt formation, 62
with temporal restriction, 62,
63
uppercase letters, xxvi
v, 225, 226
of 2
Vallabhadeva, 142-145, 150, 154,
157, 159
variant readings, 16, 17
vartamane krt, see primary deriva-
tionwith temporal restric
tions
vat in comparisons, 236-238
Vedanta, xxvi
Vedantasiltra of BadarayaQa, 1
verbal endings, 49
verbal nouns
in series of causes, 208, 209
verbal participles, see participles
verbs of saying, 228
verbs of speaking
in objections, 245, 246
in replies, 246, 247
verses in syntactic connection, 163
vibhaktisa,!,jiiaka suffixes, see sec
ondary derivation, case-
substitute suffixes
views
vigraha, see analysis
Viveka of Hemacandra, 3
vocatives, 151, 169
vrddhi derivatives, 178
Whitney, W. D., xxvii
word division, 4, 13-17
specification of, 14
word order, 197-203
scope of iti, 200, 201
with attributive noun, 199
with predicate nominative, 197-
199
with pronouns subjects, 198
words
citation of, see citation of words
repetition of, see repetition of
text
words carried over, 165, 166
words implied, xxvii
words supplied, xxvi, 165, 166
yat as conjunction, 263

Scholastic Sanskrit

Uploaded by

Copyright:

Available Formats

Scholastic Sanskrit

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Scholastic Sanskrit

Uploaded by

Copyright:

Available Formats

TREASURY OF THE INDIC SCIENCES series

Editor-in-Chief: Robert A.F. Thurman

aisvarye. [nagfidhiriija iti] [yadi va] adhikas casau raja

You might also like