Anda di halaman 1dari 102

Paribhs: Metarules in Sanskrit Grammar

Jo Brill

Dissertation, Master of Studies in Oriental Studies


University of Oxford

Submitted 17 May 2013


Corrected October 2013
Contents

First words ................................................................................................................................................................... 1

Introduction ................................................................................................................................................................. 1

Paribhs and paribh collections....................................................................................................................... 2

K. V. Abhyankars achievement ............................................................................................................................... 5

The importance of Sradeva ...................................................................................................................................... 9

Sradevas place in history ................................................................................................................................... 11

The text of Sradevas Paribhvtti ................................................................................................................. 13

Sradevas commentators ..................................................................................................................................... 13

Prior to the beginning .............................................................................................................................................. 15

First paribhs ......................................................................................................................................................... 17

arthavathow the older collections begin........................................................................................................ 19

vykhynatahow the more recent collections begin.................................................................................. 23

Contents ..................................................................................................................................................................... 28

What rules go into the table? ............................................................................................................................... 31

Sradevas innovations with respect to content ................................................................................................ 33

Order .......................................................................................................................................................................... 34

Pinis paribhs ............................................................................................................................................... 35

Repetition ............................................................................................................................................................... 35

Related rulessometimes together, sometimes apart..................................................................................... 38

Portmanteau rules ................................................................................................................................................ 40

Sradevas new order ............................................................................................................................................ 42

Contradictory statements paired ........................................................................................................................ 44

i
Limitations of jpaka.......................................................................................................................................... 47

Endings ...................................................................................................................................................................... 49

Conclusions ............................................................................................................................................................... 52

Directions for further research ................................................................................................................................ 53

AppendixSradeva on vykhynata, paribh 1 .......................................................................................... 56

AppendixDerivations for Sradevas examples under paribh 1 ............................................................... 69

AppendixSradeva on arthavat, paribh 9..................................................................................................... 70

AppendixAbhyankars table ............................................................................................................................... 76

AppendixManuscript sources ............................................................................................................................. 91

References .................................................................................................................................................................. 93

ii
First words

So much ingenuity and energy has been spent on the paribhs that [the doctrine of paribhs]

has become, for the Pinya student, the hardest nut to crack. Shripad Krishna Belvarkar.

It is true that Sradevas Bhatparibhvtti is a hard nut to crack, as he has evinced his deep

study of the subject and presented learned discussions which are difficult to follow unless one

has mastered the subject beforehand. K. V. Abhyankar.

Introduction

This paper is a first attempt to explore the conventions of the paribh treatises in vykaraa, Sanskrit

grammar. It combines a close look at two paribh rules especially as discussed by a renowned author,

Sradeva, with a broad survey of the way that paribh collections are put together: how they begin (in

most cases, with one of the two rules just mentioned), what they contain, how rules are ordered within

the work, and how they end. The survey relies heavily on a groundbreaking work by K. V. Abhyankar,

and generates many interesting questions that cannot be fully answered here.

Throughout we try to keep in mind the motivations that have been recognized in the vykaraa tradition.

First, from the very early stages, possibly even starting before Pinis Adhy took the form we know

today, grammarians and patrons have wanted to create a description that was easy to understand,

something accessible. In tension with this goal, they also sought technical completeness and perfection on

the part of the system in general and the foundational texts in particular. Finally, authoritativeness

perhaps prized most of allcan flow from this technical perfection, but religious authority is also

commonly sought.

We first undertake a few preliminariesabout paribhs in general, about K. V. Abhyankar, and about

Sradevaand then consider in turn the beginnings, content, order, and endings of paribh texts. A

brief conclusion and outline of directions for further research follow. Translation of excerpts from

Sradevas work on paribhs, and notes on various technical matters, are given in appendices.

1
Paribhs and paribh collections

The word paribh is analyzed as coming from the root bh, to speakalso the root of bh, language,

and bhya, commentary. The traditional definition of bh, vyakty vci, gives the sense of clear and

distinct or manifested utterance. The prefix pari gives the sense of about, and the verb paribh means

to lay down a convention; paribh has come to mean technical phraseology, or a general rule that is

applicable throughout the work with which it is associated.1

In the sense of principles or conventions, paribhs are a natural adjunct to any kind of technical

writing, certainly including stra in general. As Devasthali says, these are the general rules concerning

technical terms, used by writers to explain abbreviations and other peculiarities of vocabulary,

construction, syntax and style.2 Examples are easy to think of: Cookbooks often give guidelines on

measuring and handling ingredients, and dictionaries may explain up front the elements of a definition,

or speak directly to certain assumptions that users should hold in mind.3

Works of stra generally, and stra works particularly, are associated with such interpretive principles.

As Haraprasd Shstre puts it, Wherever there are stras, there must be rules for the interpretation of

stras.4 Devasthali points to the recognition of paribhs especially in the prtikhyas and the

rautastras; they are also found in works on yurveda and jyotia.5 The term paribh is also used in

reference to works on mms and nyya. But in vykaraa, the ancillary quality of paribhs emerges

especially clearly: they serve to clarify and interpret something else. Wujastyk calls them auxiliary

1
Apte, pp. 970, 981, 1196.
2
Devasthali, p. 1.
3
One latter-day paribh, from Abhyankars Dictionary of Sanskrit Grammar, echoes a cultural
preoccupation with meaning: As the senses given in such dictionaries are more of less technical or
conventional, the literal or the usual sense of the words is given only when it is allied to the conventional
sense. Abhyankar gives this guideline as one of his hints for the use of this dictionary.
4
Haraprasd Shstr, p. xxxiii-xxxiv.
5
Devasthali, p. 8 n38.

2
hypotheses. As Benson explains, paribhs exist only to serve other rules, namely, those which

actually teach operations.6

What other special attributes do paribhs have, when the topic is vykaraa, the description of

language? Pini included definitions and interpretive rules among the stras, straightforward

instructions for the correct reading of his grammar.7 The word paribh does not occur in the

Adhy, but is common in the Mahbhya and also occurs in the vrttikas.8 Later grammarians

continued to find or formulate new rules to iron out the difficulties that can arise from application of the

stras of Pini (or other foundational texts). As Wujastyk notes, it is natural to try to improve the

theory,9 and indeed this desire for technical completeness and perfection has been noted. One last point:

the formulation of a paribh is just one way to improve a theory; another is to edit the foundational

textwhich may well in turn require taking on questions of authority.

Modern vykaraa authors report on the refinements of what can be called paribh theory; while we

review the theoretical distinctions, we should also recognize that they may sit rather uneasily on the work

of early grammarians.10 Wujastyk points to the four components that paribhs are said to have:

vaiyartham, irthajpakam, sve critrthyam, and anyatra phalam.11 Abhyankar categorizes

paribhs according to their purpose: they serve to bring about (1) the correct interpretation of the rules

of grammar, or (2) the removal of conflict between two rules which occur simultaneously in the process

of the formation of words or (3) the formation of correct words.12 Devasthali lays out three additional

ways to classify paribhs. Paribhs may be acceptable (svkrya) or notand if not, for various

6
Benson 2002, p. 3.
7
Wujastyk 1993, v. II, p. xii. Abhyankar (1967, p. 3 n1) counts 48 paribhs in the Adhy. As Cardona
(1976, p. 324) notes, his list includes some rules that could be considered definitions or heading stras.
8
For example, it is found in the vrttikas on 1.1.69 and 1.3.11.
9
Wujastyk 1993, v. II, p. xi.
10
It is unfortunately beyond the scope of this paper to treat historically the development of these ideas.
11
Wujastyk (1993, v. II, pp. xii-xiii) explains these terms.
12
Abhyankar 1986, p. 240.

3
reasons;13 paribhs may apply always (in which case they are nitya) or not always (anitya), or always,

except under certain specified conditions (vyavasthita anitya); and, paribhs come from various

sources.14

These classifications can be seen in the questions that Kielhorn in his translation of the

Paribhenduekhara asks about each paribh: Is it stated, directly or indirectly, in the Mahbhya?

Has it been adopted by Ngea, and if so, is it universally valid? And is it necessary for proper

application of Pinis rules? K. V. Abhyankar, in substantial essays prefacing several works published in

the 1960s, presents and discusses additional questions with respect to paribhs: the order of their

mention, and their classification according to the subject-matter.15

As we have seen, paribhs may be presented as part of the work that they explain, but in vykaraa

they also are commonly collected into standalone texts defined in relation to a grammar (that is, a set of

operational rules).16 These collections of paribhs are called accessory treatises by Belvarkar; he uses

the same term also to refer to works on dhtus, gaas, udi affixes, genders or accents.

Abhyankar suggests that paribh collections constitute a special kind of commentary. He notes that

works in the stra genre generally require explanation, hence vtti, bhya, vrttika and ka, and to

these, works on Paribhs could be added as forming a fifth class in the Vykaraa stra.17 This

brings up the question, acknowledged in passing by Wujastyk as well as Abhyankar, of how the

paribh literature should be positioned in relation to the prakriy literature on derivations.

13
That is, they may be pratykhyta, apahita, bhyaviruddha, or anvaaka. Devasthali 1969, p. 8.
14
They may come from a direct statement (vcanik) in the stra, vrttika, or bhya; from an indirect
statement (jpaka) in any of the same sources; or from a widely-accepted maxim (nyya). Devasthali
1969, p. 8.
15
Abhyankar 1960, p. 3.
16
See for example pp. 42. Haraprasd Shstr (p. xxx) lists seven texts associated with the Pinian
school and the Adhygaapha, dhtupha, lignusana, ik, and uni and phit stras as
well as paribhpha. Other schools too have such associated texts.
17
Abhyankar 1967, pp. 1-2.

4
K. V. Abhyankars achievement

This paper relies heavily on the Paribhsagraha, a collection of paribh works compiled by a

prominent twentieth-century grammarian, K. V. Abhyankar. Abhyankars work on this book was

informed especially by two projects undertaken during his time at the Bhandarkar Oriental Research

Institute: he oversaw a second edition of Kielhorns translation of the Paribhenduekhara, published in

1960, and he produced an edition of the text of the Paribhenduekhara with the contemporary

commentary of his father, V. S. Abhyankar, published in 1962, for which he supplied a substantial

forward.18

As Wujastyk says, the publication of the Paribhsagraha as a whole, was a major event for

vykaraa studies; it highlighted the importance, extent and continuity of the paribh literature, and

made several texts available for the first time.19 Abhyankar chose 18 texts (some simple lists of paribh

rules, some with commentary as well) that spanned hundreds of years; he arranged them in what he

judged to be chronological order, and studied the various topics that the authors took up and the ways in

which the collections differed over time. He presented several new editions, including one of a twelfth-

century work of rules and commentary by Sradeva.

Abhyankar does not include all known paribh works. He mentions that he omits, on account of

incompleteness, two works in the later Mahbhya tradition: the Paribhpraka, an independent

treatise by Viubhaa, also known as Viuea, of the famous ea family, and the

Paribhpradprchis of Udayakara Phaka (also known as Nn Phaka), a scholarly independent

18
Abhyankars early work included work on Bhartharis Dpika. He also edited Kielhorns edition of the
Mahbhya, and worked on the publication of his fathers translation of the Mahbhya into Marh,
providing for that work a seventh volume with background material and very useful indices. The first
edition of his Dictionary of Sanskrit Grammar came out in 1961. In later years, he worked closely with his
student J. M. Shukla; this collaboration gave rise to an English translation of and notes on the first three
hnikas of the Mahbhya. Abhyankar died in tragic circumstances in 1976.
19
Wujastyk 1993, v. I, p. xv.

5
treatise of the mid-eighteenth century, apparently following the arrangement of Sradeva.20 He omits

without comment the Paribhrthasagraha of Vaidyantha strin, son of Ratnagiri Dkit and

nephew of Rmabhadramakhin; this work was a treatise that deals with 125 paribhs, about 1700

CE.21

Abhyankar also omits, but without comment, several paribh works associated with later non-Pinian

grammars: the Sakiptasraparibhstrak of Goycandra (or Gopcandra), associated with the

Jaumara school of Kramadvara, estimated to date from the thirteenth to the fourteenth century;22 the

Paribhvtti of Padmanbhadatta, associated with the Saupadma school, in perhaps last quarter of

fourteenth century;23 and the Paribhvtti of Rmacandravidybhaa, a work from the late 1680s,

associated with the Mugdhabodha school founded by Bopadeva in the thirteenth century.24

Still, Abhyankars collection contains many treasures. One important first edition is the work on

paribh rules with commentary ascribed to Vyi, later edited again by Wujastyk using additional

manuscript evidence. In another case, that of the Jainendra school, Abhyankar goes beyond the role of the

editor and makes an original contribution; he collects the paribhs that are quoted in a work of that

school, Abhayanandins Mahvtti, organizes them into four sections according to purpose and topic, and

furnishes them with a short gloss,25an achievement that showed its authors great mastery of

vykaraa.26

20
Abhyankar 1967, p. 33; Abhyankar 1986, p. 241 (entries for both titles). Kielhorn (1960, p. xxiv) also
mentions the Paribhpradprchis.
21
Abhyankar 1986, p. 241; NCC v. XI, p. 222; date given by Yudhiira Mmsaka as savat 1750.
22
Belvarkar, p. 92; Abhyankar 1986, p. 145; Coward and Kunjunni Raja 1990, p. 485.
23
Belvarkar, pp. 93-94; Abhyankar 1986, p. 237; Coward and Kunjunni Raja 1990, p. 484.
24
Belvarkar, pp. 88-90.
25
Abhyankar 1967, pp. 24-25.
26
Wujastyk 1993, v. I, p. xv.

6
With respect to Abhyankars editorial approach, Wujastyk makes some important points: In several

places [in the first of the two texts ascribed to Vyi] Abhyankar preferred to produce a text that made

good sense to him, rather than a text which was derivable from the manuscripts. This preference, though

wholly understandable, contravenes basic principles of textual criticism. 27

Abhyankar prefaces the Paribhsagraha with a 60-page essay that stands alone as a review of

paribh literature. He provides a brief prcis of the each school and its texts, but goes much farther,

framing the topics treated in the paribh literature and, within each topic, reviewing the various

individual rules that bear on it.

The final pages of the book are equally impressive, containing a grand table in which Abhyankar presents

all 555 rules that he has identified across the 18 collections that he considers. The table contains a row for

each rule and a column for each collection. If a rule does not occur in a collection, the corresponding entry

in the table is blank; if it does occur, the entry gives the number of that rule within that collection.

This grand table has forebears. Abhyankar gives an earlier version, for just the rules selected by Ngea,

in the 1962 edition of his fathers commentary on the Paribhenduekhara, showing the occurrence of

these rules in a half dozen collections.28 Before that, Abhyankar edited an index first put together by

Kielhorn: a bare list of Sradevas rules, cross-referenced to Ngeas. This list is given in an appendix on

pp. 529-537 of Abhyankars 1960 edition of Kielhorns translation, which, although printed from newly

set plates, is substantially the same as the appendix given on pp. 529-537 of Kielhorn 1874.29 Finally,

27
Wujastyk 1993, v. I, p. xv.
28
The table is given twice, once in Ngeas order (anukramaik, pp. 41-46 of the front matter), and once
with the rules in alphabetical order (paribhm akrdy anukramea sc, pp. 201-206). The
collections included are: the first ascribed to Vyi, both of those ascribed to Puruottamadeva, and the
collections of Sradeva, Nlakaha, and Ngea.
29
Abhyankar in his list has stated in full each of Sradevas metarules in Devangar; Kielhorn included
some such statements, but for most, he listed only the number. Abhyankar has also updated page
numbers and changed certain details. (For example, in the notation of long vowels, he has changed

7
Bhattacharya, editing Puruottamadevas Paribhvtti (collection 12) in 1946, compiled an index of

those rules cross-referenced to Sradevas.30

In fact, it is not a simple matter to decide how exactly a rule is entered in such a table. Beyond deciding

when a phrase is a rule, it involves deciding when two differently-worded rules are the same. For now,

we will just keep in mind Abhyankars cautious description of his method: In editing the index of

Paribhs, minor differences in readings are ignored. If the wording is different, the Paribhs are given

differently although they are identical in sense.31

The table inevitably contains some errorsmuch easier to find now that mechanical sorting tools are

readily available. (A new version of the table, with proposed adjustments highlighted, is given in the

fourth appendix on p. 76) There are two extra paribhsrules listed in the table, but not marked as

belonging to any collection: #202 taddhiteu nvayavrth and #454 vivakta samsapravtti.32

There are also instances of missing paribhsrules that appear in a collection, but not in the table. For

example, rules 36 gaaktam anityam and 37 na nirdiam anityam from the Klpa collection cannot

be found in the table. In fact the two rules are substantially identical to included rules: Abhyankars #164

gaakryam anityam and #238 naghaitam anityam, respectively.33 Thus there are two ways to address

Kielhorns circumflexes to macrons.) Abhyankar left untouched a couple of anomalies in Kielhorns list:
the omission of paribh 123, agaty hi khalu paribh cryante, and the scrambling of the
numbering for Ngeas rules 96, 102, and 103 (Sradevas 113, 114, 115, resp.). These differences,
combined with differences in commentary (as given for example on p. 537, note 1), seem to show that
Kielhorn was working from sources different than Abhyankars and Dubes.
30
Bhattacharya 1946, pp. 1-4. (Bhattacharya restarts his page numbering several times. This list comes
immediately after his 36-page introduction.)
31
Abhyankar 1967, p. 33.
32
Numbers preceded by a hash mark (#) refer to lines in Abhyankars table.
33
Abhyankars table is marked 67 in column 8 of row #238. This is a typo; rule 67 in the Klpa collection
in fact is a variant of #338 pratyayalope pratyayalakaam (see Abhyankar 1967, p. 69). Row #238 should

8
the problem. Adding rows to the table would emphasize the particular phrasing of each instance. Adding

entries in existing rows would emphasize the shared meaning and import of the various instances of the

rule in different collections.

There are examples of missing rules that are harder to account for. In the Ktantra-related groups, one

rule appears, with small variations, in collections 6, 7, and 8 (rule 48 in 6, prvoktaparoktayo

paroktavidhir balavn; rule 70 in 7, prvaparayo paravidhir balavn; and rule 50 in 8, prvaparayo

paro balavn ). This rule cannot be found in the table. It seems clearly related to #450 vipratiedhe para

kryambut the latter is itself is given as rule 69 in collection 7 and rule 97 in the collection 8, implying

that the respective authors thought that the two rules were different from each other. (Or perhaps, at least

in collection 7 where the two rules are adjacent, the author considered one a gloss of the other.)

Also not to be found in the table: rules 21, 65, and 101 from Nlakahas collection;34 the final rule from

the Jainendra collection, rule 108; and rule 137 from Vyis second collection.35

The importance of Sradeva

Associated with the eastern or Bengal school of Pini, Sradeva is recognized as the author of an

important collection of paribhs (called the Bhatparibhvtti, or just the Paribhvtti). Abhyankar

speaks very highly of this work, calling it magnificent, very scholarly and critical, and a masterly

independent treatise among the recognized works.36 Bhattacharya calls Sradeva an all-India celebrity,

being quoted by Mdhava and Bhaoji.37 Belvarkar (who notably does not mention Puruottamadevas

paribh or jpaka works, and like Abhyankar believes that Vyis work predates the Mahbhya)

be either left blank or marked 37, and if the latter, then one could argue that #164 should also be marked
36. In the appendix table, row #238 is marked 37.
34
Nlakahas rules 21 and 101 are closely related to rules 20 and 100. There may have been a question
about whether to number them separately.
35
This statement is marked as partially unreadable; see Abhyankar 1967, p. 43.
36
Abhyankar 1967, p. 29; Abhyankar 1962, p. 6; Abhyankar 1986, p. 430.
37
Bhattacharya, p. 21; Belvarkar, p. 45, also notes that Sradeva was quoted in the Mdhavya-Dhtuvtti.

9
calls it perhaps the earliest known regular treatise specially dealing with paribhs.38 Abhyankar

places Sradeva in the twelfth century, not many years after Puruottamadeva; like Puruottamadeva,

he was supposed to have been a Buddhist. 39

In addition to his very learned analysis and explanation, Sradeva made an important innovation with

respect to the order in which the metarules were presented; in his arrangement of the paribhs, he did

not follow the method of any of his predecessors,40 instead arranging them according to the order in

which their jpakas appear in the Mahbhya.

According to Abhyankar, the peculiarity of [Sradevas] work is that it bristles with quotations from old

works on grammar; this practice not only makes clear Sradevas great learning, but also gives valuable

information about works of other authorsperhaps most especially Maitreyarakitas Tantrapradpa, a

subcommentary on Jinendrabuddhis commentary on the Kik.41 Abhyankar notes the importance of

the Tantrapradpa and describes it as a work that is almost lost at present, which is not printed so far,

and the rare manuscripts of which are fast being worn out.42 Bhattacharya says that Sradeva is the only

author known to quote Puruottamadevas Paribhvtti, anddespite Sradevas discussion being far

more advancedthat he has borrowed from him on almost every page.43

38
Belvarkar, p. 45.
39
Abhyankar 1967, p. 29.
40
Abhyankar 1962, p. 7.
41
Abhyankar (1967, p. 29-39) also notes that Sradeva often quotes the same author by different names.
According to Bhattacharya (p. 14, note 1), Sradeva refers to Maitreyarakita 55 times as Rakita and 15
times as Maitreya.
42
See Abhyankar 1967, pp. 29-30. Two editions have recently been published, both incomplete. Kanjalils
2007 edition, based on three manuscripts, includes 1.2-1.4, 2.1-4, 3.1, 7.3, 8.1. Hind Kesaris 1991 edition is
of 1.2, 2.1 and 3.3.
43
Bhattacharya, p. 29.

10
Sradevas place in history

Sradeva was predated, by several hundred years, by what Belvarkar calls the traditional elaboration of

the system of jpakas and Paribhs, which he places somewhere between 470 (the date of

Candragomin) and 650. Belvarkar ascribes the flowering of this aspect of grammar to the desire of the

more orthodox grammarians to defend against any charges of defects in Pinya grammar.44

Despite these efforts, two eleventh-century grammariansBhoja and Hemacandracrafted compelling

new foundational texts, of special interest to us because despite having the opportunity to start fresh,

they still used external or secondary paribhs to complete the system. Bhoja integrated a list of 118 so-

called paribhstras into the second pda of the first adhyya of his Sarasvatkahabharaa, and

Hemacandra (long before the Haima list was put in its current form) included a bare list of 57

Paribhs at the end of his Bhadvtti (an autocommentary on the abdnusanam), which appear to

be a sort of selection from the Paribhstras of Bhoja.

Puruottamadeva, working in the twelfth century, is according to Abhyankar the author of two paribh

works. His Paribhvtti draws not from Bhojas and Hemacandras work but harks back to Vyi in

choice and arrangement of rules. His other work, a bare list or paribhpha, is different and bespeaks

significant innovation. Abhyankar suggests that the Paribhvtti came relatively early in

Puruottamadevas career, and that the paribhpha was appended to his Bhvtti, a mature work

commissioned by his then patron. Bhattacharya is much less definite about the ultimate source of the

paribhpha; nor does he assume that Puruottamadeva compiled the pha second.45

After Sradeva, vykaraa literature took a new direction. As Abhyankar puts it (speaking of the interval

between Sradeva and Ngea): only a few writers wrote on Paribhs. The attention of the people was

44
Belvarkar, p. 29.
45
Bhattacharya, pp. 28-29. The Bhvtti has been printed both with and without the paribhpha. An
example of the first is Dwarikadas Shastris 1971 edition; of the second, Srish Chadra Chakravartis 1918
edition, where the paribhs appear in the fourth appendix (after the gaapha, dhtupha, and
lignusanam).

11
more devoted to the study of the Stras of Pini arranged in different topics,46 that is, with the

formulation of a new genre, prakriy, or what Belvarkar calls recasts of the Adhy. Scharfe linked

the movement to teaching: classroom needs called for a more practical introduction into Sanskrit.47

Haraprasd Shstr sums up: What with Brahmanic and what with Buddhistic commentaries the study

of Pini became very cumbersome, and so, it was thought desirable to recast the Adhy and give it a

practical shape, that is, to evolve out of this vast literature, school books large and small.48 Thus in the

quest for accessibility, the grammarians moved away from traditional modes of presenting their work.

The prakriy literature grew to include the following works, the first of which probably predates

Sradeva: the Rpvatra of Dharmakrti,49 the Rpaml of Vimalasarasvat,50 and the Prakriykaumud

of Rmacandraea51 (with commentaries including the Prakriyprasda of Vialea and the

Prakriypraka of r Ka). This literature culminated with the Siddhntakaumud of Bhaoji Dkit;

Bhaojis very choice to embrace this new mode of expression is interesting, positioned as he is firmly in

the trimuni camp. Haraprasd Shstr notes that compared to Rmacandraea, Bhaoji Dkit was not

so liberal. He would not tolerate any opinion expounded or any new facts of the language noted by the

46
Abhyankar 1962, p. 7.
47
Scharfe 1977, p. 174.
48
Haraprasd Shstr, p. xcv.
49
Haraprasd Shstr (p. xcv) estimates the date as the beginning of the eleventh century, Scharfe (1977,
p. 174) as the tenth century.
50
Belvarkar (p. 36) posits not later than 1350 AD, Haraprasd Shstr (p. xcv) the beginning of the
eleventh century.
51
Haraprasd Shstr (p. xcvii), Belvarkar (p. 37), and Abhyankar 1986 (p. 260) put this in the first half of
the fifteenth century.

12
Buddhist commentators.52 At roughly the same time,53 Nryaa Bhaatiri, also known as Melputtr,

composed the Prakriysarvasva; Nryaa also wrote the Apinyaprmyasdhana.54

The text of Sradevas Paribhvtti

Sradevas Paribhvtti was previously edited, in 1887, by Harintha Dube as part of the Benares

Sanskrit Series. Abhyankar criticizes this editionthe paribhs are numbered differently and wrongly

also in the different copies, and the text is also corrupt at several places55but in fact (aside from the

vagaries in numbering), the two editions are quite similar.

As mentioned above (p. 7) Kielhorn in an appendix to his translation of the Paribhenduekhara

provides a bare list of Sradevas rules.

There are two manuscripts held at the Bodleian Library at Oxford (see the last appendix); the scope of this

paper allowed for only a cursory examination. The New Catalogus Catalogorum (NCC) points to

relatively extensive manuscript resources.

Sradevas commentators

After pointing out the difficulty and importance of preserving Sradevas work, Abhyankar notes that

there are a few commentary works on Sradevas Bhatparibhvtti which are in the form of

manuscripts.56 The New Catalogus Catalogorum lists three:57 the Vijay of r Mnaarman, the

52
Haraprasd Shstr p. cvii. He adds that Bhaoji accepted the arrangement of Bopadeva.
53
Houben 2012b, p. 13.
54
There has been substantial recent work on Nryaa; see for example Houben 2012a, Houben 2012b,
Candotti 2012a, and Candotti 2012b.
55
Abhyankar 1967, p. 29 n1.
56
Abhyankar 1967, p. 30.
57
Volume XI (Pa), pp. 223-224.

13
Vykhy of Rmabhadra Dkita,58 and the k of Govindamira. Abhyankar includes r Mnaarmans

commentary in the Paribhsagraha; as of 1983, when Volume XI of the NCC was published, this was

its first printing and it is the only one of the three to have been printed.

r Mnaarman, the son of Lakmpati, was from Campha;59 he wrote the Paribhvttiippa, a

very brief commentary on the Paribhvtti of Sradeva.60 This work is included in the

Paribhsagraha, where (as in the NCC) it also bears the title Vijay. Bhattacharya gives his dates as

the mid to late fifteenth century and reviews his other works and influences.61

One of the other two commentators, Rmabhadra Dkita, is well known as a scholar in hajis court and

part of the newly founded academic community of Tiruviainallrin.62 A student of Nlakaha Dkita

among others, Rmabhadra in his Paribhvttivykhy singles out another teacher, Cokkantha Makh,

for special thanks.63 He wrote several other works, including a mahkvyam on the life of Patajali.64

Before concluding this review of Sradevas work, two clarifications should be made with respect to

Abhyankars 1962 description of it.65 First, the description of Sradevas numbering (Abhyankar 1962, pp.

58
This work goes by several names; NCC practice is followed here. Thiruvengadanathan 2002 calls it the
Paribhvttivykhyna; see especially pp. 137-140. Abhyankar sometimes calls it the
Paribhvttivykhy (Abhyankar 1986, p. 324) and sometimes the Paribhvttik (Abhyankar 1986,
p. 242).
59
NCC v. XI, p. 224, from Yudhihira Mmsakas history of Sanskrit grammar.
60
Abhyankar 1986, p. 242.
61
Bhattacharya, pp. 16-18.
62
See Thiruvengadanathan 2002, pp. 5-10, and Wujastyk 2011, pp. 4-6. (The phrase academic
community is Wujastyks.) Thiruvengadanathan and Wujastyk differ in their descriptions of
Rmabhadras father and children.
63
Thiruvengadanathan 2002, p. 11.
64
Belvarkar (p. 28, n3) mentions this work as giving a fanciful explanation of the fact that some of
Pinis stras are not to be found in the Mahbhya.
65
The two misleading statements listed here are unfortunately repeated in Devasthali 1985, p. 13.

14
6-7) does not match the collection as presented, which is clearly divided into two big sections (rules 1-99

and rules 100-130), the latter consisting of paribhs that are nyyamla, the former subdivided into

adhyya and pda according to where their jpakas occur in the Mahbhya. A correct description is

given by Thiruvengadanathan.66

Second, Abhyankar says of Sradeva, he has arranged the Paribhs according to the place of their first

mention in the Mahbhya.67 Here, place of mention cannot mean a point at which Patajali refers to

the metarule by direct reference. Those sites of mention are in fact given in Abhyankars own index to his

fathers translation of the Mahbhya into Marh.68 This very useful index lists paribh rules

(including but not limited to rules in Ngeas Paribhenduekhara) and refers to the places they are

mentioned in the Mahbhya. With respect to Sradevas ordering, what Abhyankar actually meant is

that Sradeva has arranged the paribhs according to the place in which their jpakas are found.

Prior to the beginning

Before I speak, I have something important to say. Groucho Marx.

A recent collection of papers, strrambha, powerfully demonstrates that the elements placed at the

beginning of a Sanskrit work are among the first tools used by authors to establish the authoritativeness

of the work. Minkowski in the lead essayin addition to laying out the ways that magalcaraa has

been deployed at different time in different disciplines by authors from different backgroundsargues

that the magalcaraa is a rich paratextual element not to be ignored.69

In the Paribhsagraha, Abhyankar presents introductory materialmost often a verse, sometimes a

simple invocation or a prose passagefor all the collections that include commentary as well as rules.

Conversely, the collections that consist only of rules are presented without introductory material (with

66
Thiruvengadanathan 2002, p. 138.
67
Abhyankar 1962, p. 6.
68
The index appears as an appendix, pp. 13-24 of the back matter, in Abhyankar 1954.
69
Minkowski 2008.

15
one exception, the Haima work). Of course, some of the pha texts, as edited elsewhere, do have

magala expressions. For example, V. S. Abhyankars and Kielhorns editions of the

Paribhenduekhara begin with a line of homage, a magalaloka, and a brief two-sentence prose

introduction. Others were originally appendices to larger works that start auspiciously; for example,

Abhyankar also tells us that Puruottamadevas second collection is a mere list (1962, p. 5) appended to

his Bhvtti, and that the latter has a magalalokaone showing, incidentally, both that

Puruottamadeva was Buddhist, and that his work was composed in accordance with the statements by

the three great sages only.70

Sradeva begins with an introductory verse71 in the beautiful mlin meter that may not meet

Minkowskis definition of a magalcaraa, for while the verse clearly comprises more than mere

mention of the subject matter to follow, it neither pays homage to nor makes requests of a deity. On the

other hand, one can tell at once that the verse is fully poetic,72 even without searching for potentially

70
Abhyankar 1967, pp. 28-29. The Bhvtti has been printed both with and without the paribhpha.
An example of the first is Dwarikadas Shastris 1971 edition; of the second, Srish Chadra Chakravartis
1918 edition, where the paribhs appear in the fourth appendix (after the gaapha, dhtupha, and
lignusanam). Both these editions start with the magalaloka quoted by Abhyankar; Dinesh Chandra
Bhattacharya (p. 23, n1) points out, on the other hand, that though a Buddhist by his profession in the
introductory verse, as well as by the numerous Buddhist illustrations in the body of the book
[Puruottamadeva] is claimed as a Brahmanical writer by all his commentators.
71
Abhyankar 1967, p. 161. The verse is given in just the same way in Dubes edition. A preliminary
examination of the verse included in Manuscript A showed that it is almost identical, with a few akaras
differing. For example, abhinava in pda a appears as atinava. The verse can be translated thus
In which delight is newly springing forth, confident, like the sun,
By which faults are quieted perforce, and by which the society of good people is expanded,
An examination of which is valuable for what has been taught
For the sake of dispelling heaps of dullness,
Let this commentary on paribhs be ardently undertaken.
72
This phrase is Minkowskis, p. 3.

16
subtle figures like lea, upam, or rpaka, for Sradeva unmistakably deploys abdlakra, figures of

sound. He incorporates plenty of anuprsa (with fully 60 percent of the guru syllables ending in long ),

and a little yamaka (repetition) to lead off pda c.

In sum, Sradeva uses the something very like a magalcaraa, to make a general imperative statement,

presumably with sense of benediction, that the following work be enthusiastically undertaken, by the

author or reader or both. It notes that the works purpose is to dispel dullness. Notably, it contains no

expression of humility or self-deprecation; on the contrary, Sradeva by choosing words like pragalbha

conveys a sense of confidence and even pride.73

First paribhs

What paribh does an author put in first place, and why? Is he claiming a kind of authority, plunging

into a topic, or setting up some structure?

Particularly in the absence of a magalcaraa (staying for the moment with the notion of propitious

beginnings), the first words of the body of a work are fair game for retrofitting with magala. Fixing

an auspicious word like atha or vddhi in the mind is one of at least three ways that a magala

statement, though not necessarily in verse, has been recognized as appropriately setting the stage.74 In

vykaraa especially, there is a long history of contending that initial words are chosen to convey

auspiciousness; in the Mahbhya, Patajali makes just that argument with respect to siddha, which he

holds to be the first word of the vrttika work.75 Patajalis own beginning, atha abdnusanam, is cast

73
Deshpande 2012 (pp. 179-180) finds examples of strrambha in vykaraastra that illustrate
humility on the one hand and pride on the other: respectively, Kaiyaas humble introductory verse for
the Pradpa, and Bhaoji Dkits supremely confident introductory verse for the abdakaustubha.
74
Minkowski 2008, pp. 14, 21. Retrofitting is his word.
75
Having first argued that siddha means nitya, Patajali argues that as a starting word it is preferable to
nitya, on account of its other attributes: magalrtham. mgalika crya mahata straughasya
magalrtham siddhaabdam dita prayukte. magaldni hi stri prathante vrapuruaki ca
bhavanti yumatpuruaki ca. adhyetra ca siddhrth yath syu iti. Abhyankar and Shukla (pp.

17
in the familiar formula used such in foundational works as the mmsstra and the dharmastra;

Abhyankar and Shukla say in their translation of the Mahbhya that the word atha, apart from its

sense, is looked upon as supremely auspicious just as the word om.76

Two of the paribhs collected by Abhyankar are notable for the fact that they each head six collections.

The paribh arthavadgrahae nnarthakasya stands at the head of some of the oldest collections, and

may be able to claim auspiciousness on account of starting with a, 77 and perhaps also because artha, in its

several meanings, is generally a good thing to have.78 By placing this paribh first, an author plunges

immediately into a major topic in the paribh literaturewhether a given word is to be taken as

26-27) translate: to serve the purpose of a benedictory beginning. The revered author of the Vrttikas,
inclined to recite benedictory words at the beginning of his great stream of Vrttikas, uses the
benedictory word siddha at the beginning. Works marked with benedictory words at the beginning gain
reputation, and people reading them become valiant and long-lived. In fact, scholars who study these
works are able to achieve all their objects.
76
Abhyankar and Shukla 1975, p. 4. No doubt they had in mind the krik quoted on p. 13, Minkowski
2008.
77
Similar reasoning has been applied to explain the fact that the first verse of the Kumrasabhavam
does not seem to display the customary attributes of a magalcaraa; the argument is that these
purposes are served by the fact that the verse begins with asti (see for example Amer 2013, pp. 4-5) or
indeed just by its beginning with the letter a. In fact, Thiruvengadanathan argues (p. 28) for a similar
interpretation of the first wordastiof the Patajalicarita, a poem in the mahkvya style by
Rmabhadra Dkit, one of Sradevas three known commentators.
78
Those who find this basis plausible could use it to dispute Minkowskis observation that akaras
commentary on the Brahmastra does not in practice make any concession to the need for a magala
beginning (Minkowski 2008, p. 13), for this work too begins with artha. In any case, the existence of the
argument serves to underscore the felt need for a sign that some effort for auspicious beginning has been
made.

18
meaning the thing signified, or as a string of phonemes, and how that may change when the word has a

technical meaning.79

The paribh vykhynato vieapratipattir na hi sandehd alakaam is the first listed for the six most

recent collections presented in the Paribhsagraha. Does Patajalis oft-quoted definition80 of

vykhyna impart or reflect the degree of auspiciousness conveyed by such words as atha or vddhi?

Perhaps not: in contrast to atha, vykhyna is not used across stra, and unlike vddhi, the word appears

in the expected prose order, rather than one that by being unusual signals emphasis of some kind. Still, as

this paribh itself says, vykhyna is the source of authoritativeness, the settling of all doubts that may

arise in the study of grammar. Whence vykhyna is of course the next question.

arthavathow the older collections begin

arthavadgrahae nnarthakasya (#55)

As noted above, this paribh leads off six of the older collections in the Paribhsagraha. Abhyankar

does not go so far as to say that the early authors deliberately chose to lead off with the arthavatparibh

in order to convey the emphasis that it deserves. But he does infer from its placement that the early

authors particularly attached a remarkable importance to the sense-element in grammar.81

With respect to this topic, the association of words with sense, recall that Pinis rule 1.1.68, sva rpa

abdasyabdasaj (#543), sets aside the usual assumptionthat when we use a word, we are talking

about what it signifiesfor the ground rule within the Adhy: when we use a word, we mean its

79
sva rpa abdasyabdasaj, Pinis 1.1.68, which occurs in two Ktantra-related collections
(#543), opens the same topic.
80
MB, I,11.20-23. Katre, p. xlvi, provides a nice translation, concluding that examples, counter-examples,
filling in ellipsis together constitute vykhynawhat I have called traditional teaching. Also see
Abhyankar and Shukla 1975, p. 41. This passage is often quoted by those studying the examples used in
vykaraa; see for example Candotti 2012a, p. 103.
81
Abhyankar 1967, p. 33.

19
form, the set of phonemes that it comprises. In Abhyankars words: Pini in his rule [1.1.68] has laid

down a definite principle, that as far as grammar rules are concerned, words in Stras [other than

technical terms] are to be taken strictly as collections of letters in a specific order and no attention is to be

paid to their sense except in the case of technical terms.82

The present rule, arthavat, appears in all 18 collections, and further refines sva rpa83by

reassociating what Abhyankar calls the sense element with words undergoing grammatical operations,

that is, with kryins. Benson translates it thus: When a meaningful form is mentioned it does not denote

a meaningless form.84

A look at the rules incidence within the various collections shows that it does often (but not always)

occur in tandem with one or more of three rules: The rule #175 gauamukhyayor mukhye

kryasampratyaya in a way restates arthavat : of the secondary (attributed sense) and primary (natural

sense), the settling is to be upon the primary. The rule #246 na varagrahaeu sets out an exception in

the case of single letters.85 The rule #23 aninasmangrahanyarthavat cnarthakena ca tadantavidhi

prayojayanti gives another exception to the arthavatparibh: when a wording possessed of sense is

available, the same, devoid of sense, should not be taken except in the case of as, in, as, and man.86

Concerning the interaction of arthavat with #23, aninasman, Kielhorn proposes just the sort of structural

explanation that is satisfying to any student of Pini. He says that Ngeas placement of aninasman is

such that it should be read as an exception to arthavat, despite being separated from it by the

82
Abhyankar 1967, p. 14.
83
For the rest of the paper, rules will frequently be referred to by their first few words.
84
Benson 1990, p. 109.
85
This paribh, which is directly stated in the Mahbhya, is not given its own number by Ngea, but
just mentioned in the commentary. na varagrahaeu also serves as an exception to
pratyaypratyayayo pratyayasya grahaam (Ngeas rule 102), and Ngea mentions it in his
commentary there as well. This double duty may be one reason that Ngea decided not to number it.
86
Abhyankar 1967, p. 34. Also see Abhyankar 1962, p. 18.

20
interposition of another statement, since that statement, gauamukhyayo, is really just an elaboration of

arthavat. 87
Kielhorns observation holds with respect to the relative placement of these rules in

Puruottamadevas Paribhpha, but cannot be sustained for other collections, including

Puruottamadevas Paribhvtti and Sradevas, in which aninasman is taught at a distance from

arthavat.

Seven collections altogether follow the arthavatparibh with na varagrahaeufour of the six

collections that begin with arthavat (both of Vyis, kayanas, and Puruottamadevas first), plus

those of Sradeva, Nlakaha and Haribhskara. The next most common successor is gauamukhyayor

mukhye kryasapratyaya, which follows arthavat in four collections, including Puruottamadevas

second and Ngeas.

Four early collections do not start with arthavat, but do get to it quickly. Of the Ktantra collections, two

(5 and 7) number this rule 4, and follow it with na varagrahaeu; collection 8 numbers it 7. (On the

other hand, sva rpa, which is also in all these collections, is nowhere near these paribhs despite

their common topic, connection of word with sense.) In the Jainendra collection, arthavat is numbered 3,

and is followed by aninasman and then gauamukhyayo.

Sradeva on arthavat

Sradeva sets out to show that when a stra applies to a word, that word qualifies as a kryin only if it

bears meaning in that very form. (What follows is a summary of the argument. For details, see the

translation and notes thereon in the third appendix, starting on page 70.) The stra would not apply to a

word that is just part of another word, the larger word being the bearer of meaning. The first example

is the word van and a stra 4.1.7 that replaces the final of van with r. Sradeva establishes that the stra

will apply to van in words made with the meaningful kvanip/vanip pratyayas, but not to van in words

(such as ativan) that simply happen to contain the string v-a-n. Similarly, 6.4.12 and 6.4.13, which are

stated to apply to han, do not affect the word plhan, which happens to contain h-a-n.

87
Kielhorn 1960, p. 96 n3.

21
Sradeva then discusses svarpa in this context, noting that happenstantial strings like v-a-n and h-a-n,

mere aggregates of letters, do not themselves have their own form as their object. They only contribute to

the svarpa of a bigger word.

Sradeva next takes up a trickier example involving the string iwhich may be, of course, either the

augment i as in se/ani, or the first person singular tmanepada ti affix i. He refers in passing to

several stras for which it is clear which is meant, and then alludes to Maitreyarakitas question of

whether both might be meant in 3.4.106; he argues by analogy to marker letter that one anubandha does

not mean two things.

He then returns to the argument that operations are not performed on parts of words that happen to have

meaning in other contexts: d and dh are certainly verbal roots on their own, but operations that pertain

to them do not pertain to longer roots such as d and dhv. Likewise, operations that pertain to the

Vedic dual noun ending e do not operate on the meaningless string of letters -e in the words ke and

kue.

Noting that that the arthavat in this paribh occurs also in the definition of prtipadika (P. 1.2.45),

Sradeva discusses the verbal compounding of pra, etc.particles that despite being meaningless, are

ultimately considered prtipadikas. The resolution is indicated by the condition anarthaka later in the

section on particles, because it implies the possibility of meaningfulness. A particle like pra can thus

receive a sup ending in preparation for having that ending elided in the making of a compound.

Finally Sradeva discusses how the rule is known, from sva rpa abdasyabdasaj 1.1.68 (upon

which he relies for the purpose of determining where to place this rule) and also from vraca-bhrasja-sja-

mja-yaja-rja-bhrja-ccha- a 8.2.36. For the first argument, he notes that a words svam, what it

has, could potentially refer to either form or meaningbut that it must in fact be form, because of the

three properties that form has: the property that it is permanently connected to word, the property that it

cannot be let go, and the property that its association is closer [than that of meaning]. The fact that both

svam and rpam are mentioned in this rule thus indicates that there is more than just svam, and an

interpretation with meaning is to be preferred to one without.

22
The second indicator is supported by a simple argument that if not for the present paribh, there would

be no need to state both rj and bhrj in 8.2.36, for bhrj would be implied just on account of containing

rj. Ben-Dor points out that this argument for the jpaka, and the jpaka itself, are rejected by Vyi on

the grounds that bhrj occurs twice in the Dhtupha.88 However, Sradeva does not express similar

reservations.

vykhynatahow the more recent collections begin

vykhynato vieapratipattir na hi sandehd (#464)

In Abhyankars words, this rule emphatically states that every rule of grammar must be clearly

interpreted and not a single rule should be discarded as ambiguous or doubtful. Anyone who doubts a

rule simply shows his own weaknessthat is, his failure to grasp the explanation given by the ancient

grammarians.89

This paribh occurs in almost all collections, all but Bhvamiras (6) and Jainendra (9). Moreover, it

stands at the beginning of the six most recent collectionsthat is, starting from Puruottamadevas

Paribhpha (12a). Abhyankar ascribes this placement to the authors having in mind the making-

adequate quality of paribhs (aniyame niyamakri paribh ).90 Classing vykhynata as a rule that

helps with right interpretation of the rules of grammar, he also concludes that later grammarians

attach more importance to right interpretation,91 and indeed, that this is one the defining characteristics

of what he calls the third school.92

88
Ben-Dor 2009, pp. 92-93. Also see Wujastyk 1993, vol. II, pp. 5-6; after translating and explaining
Vyis argument, Wujastyk concludes that it is unsatisfactory.
89
Abhyankar 1962, p. 12.
90
Abhyankar 1962, p. 3.
91
Abhyankar 1967, p. 33.
92
Abhyankar 1962, p. 8.

23
Kielhorn emphasizes this paribhs special nature in his introduction to the Paribhenduekhara:

There are a few paribhs which assist us neither in interpreting nor applying the rules of grammar,

but which rather lay down some of the principles followed by Pini himself in the formation of his

rules. Here he first mentions Ngeas last rule, 122, which extols the virtue of brevity, and continues:

[Ngeas first rule] refers us for the solution of any doubt or the removal of any ambiguity which

cannot be solved or removed otherwise, to the interpretation of the learned.93

Scharfe explicitly calls out the connection between vykhynata and authoritativeness: From this

authoritative character of Pinis and Ktyyanas work follows the first rule of interpretation (and

indirectly all others): The specific sense [of an expression in a stra] is obtained from interpretation; for

there cannot be [assumed to be given a definition which is a] non-definition [just] because there is a doubt

[as to its specific sense]. 94 In fact, this rule both boldly claims authoritativeness, and underscores the

quest at the heart of the entire paribh projectto fix up the system and ultimately to make it perfect.

Scharfe names only two of the munis, and Patajali by his very discussion of this principle seems to

imply that he too looks to the ancients for the ultimate word. He states the rule, and also what vykhyna

entails, up front in the paspahnika.95 The second time that Patajali mentions this rule, under la, the

sixth of the ivastras, he says: etaj jpayaty cryo bhavaty e paribhvykhynato vieapratipattir

na hi sandehd alakaam iti. Abhyankar and Shukla translate: the Strakra by use of the same letter

twice, indicates the regulatory dictum that the decision in cases of doubt is to be taken from the

explanations of ancient scholars, evidently because the interpretation of a rule is never to be kept

doubtful.96 (Abhyankar and Shukla use indicates for Patajali's word jpayati; the causative of j

93
Kielhorn, p. i.
94
Scharfe 1977, p. 158.
95
Abhyankar and Shukla 1975, pp. 26, 41.
96
Abhyankar and Shukla 1975, p. 123. Elsewhere (1962, p. 3), Abhyankar seems to contradict this, saying
that while Patajali states and uses this paribh, it is left to other authors to provide jpakas.

24
takes on formal significance in the paribh literature.) Here, Patajali says that the rule vykhynata

must be true, precisely because we need it to make sense of the second use of akra.

Who are the learned ancient grammarians doing the explaining, vykhyna? Is Patajali included,

despite his too referring to vykhyna? Bronkhorst posits that whoever it may be, Patajali is the

mouthpiece: The desired rule is the rule desired by Pini. How do we know which is the rule desired

by Pini? From tradition. And how do we know which is the traditional position with regard to these

specific derivations? By asking Patajali.97 Kielhorn says the group includes the author of the Bhya

and others.98

Finally, beyond grammar, this rule may help establish principles shared by other vedgas. Kahrs makes

the point that vykhynata may have a wider scope than just the rules of the Adhy, having argued

earlier thatgiven the prevalence of the substitution operation in organizing knowledge in Indian

thoughtthe substitutive genitive may have wider application than just vykaraa.99

Sradeva on vykhynata

In discussing this paribh, Sradeva first gives a very quick gloss of pratipatti (avagama) and

vykhyna (cryapramparyopdea). He then turns immediately to one place that the paribh is

mentioned in the Mahbhyaunder 1.1.49; we need this paribh to understand Pinis rule for the

special use of the sixth case in the grammar, what could be called the genitive of place (where place is

the locus of the operation to be performed).

Sradeva now sets out to strengthen the proofs given by earlier grammarians for this rule. (What follows

is a summary of the argument. For details, see the translation and notes thereon in the first appendix,

starting on page 56.) He first reviews the argument based on the use of the anubandha in the

97
Bronkhorst 2004, p. 32.
98
Kielhorn 1960, p. 2 n2.
99
Kahrs 1998, p. 214.

25
ivastrasit appears twice, so that any pratyhra ending in has two potential values100and the

confusion that thus arises when trying to understand stras in which a or i is mentioned. Rather than

rehash the discussion in the Mahbhya, which for each case of a uses relatively simple arguments to

resolve the ambiguity, Sradeva just mentions the argument he does not accept. This argument is used to

prove, for two of the cases (6.3.111 and 7.4.13), that the earlier a is meant. It rests on the reference in

these stras to the terms hrasva and drgha, and claims that the implication is that only vowels could be

meant. The simple argument relies on reading 1.2.28, the paribh stra that follows the stra defining

hrasva, drgha, and plta, as in place of vowels only; what it says is and in place of vowels. Sradeva

finds this argument facile, and goes through the derivations of some rather peculiar wordsagka,

adhkato explore what happens when the notions of short and long are extended to h and the

semivowels, r in particular. He is able to conclude that such an interpretation adds nothing to the

derivations for these words, and so for these stras that indeed the first a is meant.

He then takes up the case of the only stra in which the second a is read, 1.1.69, another paribh stra:

audit savarasya cpratyaya. This time he is arguing not with Patajali but with Jinendrabuddhi, who

is having it two ways in his argument based on ur t 7.4.7. This stra enjoins the substitution of for and

of a causative stem, in the reduplicated aorist.101 The simple argument here is to assume that is not

savara with , and proceed to a contradiction. We see the short vowel marked by t, and that is

pointless, so we are done. But Sradeva points out that the first word in the stra, the sthnin, also is ,

and claims that Jinendrabuddhi, who mentions the case when the sthnin is drgha, has assumed one

thing for the of the dea (that it refers to the particular letter, since we have assumed is not in a), and

something else for the of the sthnin (that it refers to the entire class of letters savara with , and so we

are relying on being in a). He then tries to salvage Jinendrabuddhis argument by reading the first

word of the stra as the sixth case not of a masculine singular but of a neuter samhra dvandva of

100
That is, a-i-u, or a-i-u---e-o-ai-au-h-y-v-r-l. The set of vowels, ac, contains the first a, and is contained
in the second.
101
This translation is Vasus (v. II, p. 1459).

26
and , that is, .102 But this wont give us the sixth case form u that is found in the stra, because isnt a

vowel: according to our original assumption, it is not savara with , nor is it mentioned in the ivastras.

Asking how the compound could be considered short, despite having two mtras, he recalls that

hrasvo napusake prtipadikasya 1.2.47 teaches shortness in the neuter for the final vowel of a

prtipadika. Having applied the stra, the prtipadika ends in and hence can be declined in the sixth

case as u (the more familiar form, a, being optional as it results from an augment). The first word

now is well-formed, the contradiction arises from marker t in the second word, and the use of the longer

a in 1.1.69 is proved.

Sradeva concludes this section of the commentary by noting that Patajali says that [the strakra]

indicates the existence of this paribh. He is quoting the penultimate statement of Patajalis

commentary on la.

The second half of Sradevas commentary on vykhynata is a deep look at the compound

vieapratipatti. This involves a review of the conditions under which the sixth case can be assigned to a

word in relation to a word formed with a kt suffix (such as pratipatti, that is), and an examination of the

circumstances under which two such words, one in a sixth-case relation to the other, can be compounded.

The first question has to do with whether the both the karman and agent are present (if so, the sixth case

can be used only for the karman) and the second question has to do with whether the words in the

prospective compound are themselves the karman or the agent of the verbal action. Here, of course,

viea is the karman, and Sradeva mentions the iydistudent, etc.as the agent. After an intricate

argument having to do with explicit mention versus implicit understanding, Sradeva concludes that

stras 2.2.16 and 2.2.14 on the one hand, and 2.3.65 and 2.3.66 on the other, end up mutually determining

each others scope of action. I do not fully understand the flow of the argument, but it seems that

Sradeva succeeds in drastically narrowing the application of 2.2.14 (which otherwise would rule out

sixth-case compounds like vieapratipatti for which the first word is the karman), by establishing it as a

niyama in relation to the vidhi 2.2.16, just as 2.3.66 is a niyama in relation to the vidhi 2.3.65.

102
By aka savare drgha 6.1.102 [sahitym 6.1.72, aci 6.1.77, eka prvaparayo 6.1.84].

27
Contents

Abhyankar finds and lists 555 paribh rules; the collections, as he numbers them, range in length from

62 to 140 rules. Which rules do we find in which collections, and why? To get a feel for the choices made

by the paribh authors, we can look at the overlapor lack of overlapbetween collections.

A first observation: Abhyankars list includes a couple of dozen of Pinis stras, including many from

the second half of pda 1.1. These do occur, just as one would expect, in the Ktantra-related group, and

especially in Bhojas work, but not at all in the later, explicitly Pinian collections.103

For example, of the rules that pertain the substitution operation, three rules that provide for the

metalinguistic use of case endings#474 ahy nirdie svasya, #496 saptamy nirdie prvasya, and

#286 pacamy nirdie parasyaare clearly equivalent to Pinis stras 1.1.49, 1.1.66, and 1.1.67. In

addition to giving a very basic demonstration of how the contents of various paribh collections may

differ, this appropriation into the non-Pinian grammars of ground rules for the substitution operation

incidentally shows that these grammars too take for granted what Kahrs describes as a framework of

archetype-ectype or prakti-vikti.104

Sometimes the wording in the non-Pinian paribh works is identical to the stra, or nearly so, as is

the case for familiar rules like aco hrasvadrghaplut (1.2.28, #9), yathsakhyam anudea samnm

(1.3.10, #386), and vipratiedhe para kryam (1.4.2, #450). We also see rules about it-letters and the

placement of augments. The Pini rule seen most in the paribh collections, dyantavad ekasmin

(1.1.21, #83), occurs in seven works, including the Jainendra and Haima collections as well as Bhojas and

the Ktantra-related ones.

Can we get a sense of how far apart two collections might be? It seems easy to get a quick measure of

which collections are more, or less, like the others. For example, neither Puruottamadevas

Paribhpha nor Sradevas work contain any rules that are not also in one or more other collections.

103
The non-Pinian versions of these rules are sometimes phrased differently.
104
Kahrs 1998, p 184. Kahrs also discusses samnya-viea (p. 183) and utsarga-apavda (p. 185).

28
On the other end of the spectrum, Candragomins and Bhojas collections both surpass 25 percent in

terms of the number of rules in the collection that occur only in that collection, and for both the Haima

and Jainendra collections, this measure is over 40 percent. This is hardly surprising, at least for the

Jainendra collection; giving new names to all the technical terms necessarily changes the rules.

But as we have seen, two rules that are worded differently may yet be substantially the same, and in this

case, our quick measure will understate likeness. As noted above (p. 8) Abhyankar has said that he errs

on the side of listing rules twice. Thus one rule may occur in two different rows; for example, #253 na

strpratyaye cnupasarjane and #537 strpratyaye cnupasarjane na. Two more examples: #21 anityam

tideikam and #78 tideikam anityam; #20 gamastram anityam and #73 anityam gamasanam.105

And there is another phenomenon that should be understood before adopting a measure of difference. In

some cases in some collections, a statement that is numbered as a rule seems to be merely a gloss on the

preceding or subsequent rule. For example, in the second collection ascribed to Vyi, rule 43 (#38

antarage krye bahiragam asiddha syt) appears to be nothing more than an explanation of rule 42

(#70 asiddha bahiragam antarage). Similarly, as mentioned above in passing, Kielhorn maintains that

Ngeas 15 (#175 gauamukhyayor mukhye kryasampratyaya) is really only a part of his rule 16

(#55 arthavadgrahae nnarthakasya).106 Does counting such rules as different paribhs make sense?

Conversely, in some collections, certain rules are not given their own numbers, but are just mentioned

under other rules. Four examples are given here. First is #246 na varagrahaeu (which Ngea

mentions under his rule 14, the by now familiar arthavat rule); this principle does get paribh status in

every other collection, but (see note 85 above) not the Paribhenduekhara. Second, Bhvamira in

collection 6, at least as edited by Abhyankar, does not separately number the rule parasyde where it

appears between rules 36 and 37. This rule, which has to do with the placement of conditions on the

105
Some rules that may at first appear to be simple rephrasings may in fact be different: asiddha
bahiragam antarage is not the same as antaragabahiragayor antarago vidhir balyn.
106
Kielhorn 1960, p. 96 n3.

29
substitution operation (and takes the place of P. 1.1.66 and 1.1.67) is given paribh status by Bhoja in

collection 10.

For the third and fourth examples of rules that sometimes are not accorded paribh status, we turn to a

set of rules highlighted by Abhyankar, who counts four rules individually numbered by

Puruottamadeva and other authors, but not numbered by Ngea (who however does mention them).107

Two of these illustrate the present point. Ngea gives the phrase ktrimktrimayo ktrime

kryasapratyaya (#139, of the conventional or technical sense of a word, and the literal sense, the

settling is to be upon the conventional108) in his introduction to his rule 9, ubhayagatir iha bhavati) but

does not number it separately; as a paribh, ktrimktrimayo occurs in 11 collections altogether.

Ngea gives (but does not separately number) #414 lakye lakaasya sakd eva pravtti109 in his

commentary under parjanyaval lakaapravtti, which is listed by Abhyankar (and Kielhorn) as

Ngeas rule 111,110 and occurs in 12 collections altogether.

The remaining two examples given by Abhyankar illustrate editorial differences of opinion not between

the paribh authors, as for the four examples just reviewed, but rather between V. S. Abhyankar and

Kielhorn. In both cases, the dispute is about which of two related statements gets numbered: Abhyankar

highlights as a paribh a statement treated by Kielhorn as introductory to the actual rule which follows.

(Abhyankar surely had his reasons, but Kielhorns presentation is the one consistent with Ngeas near-

ubiquitous practice of leading into each rule with a short introductory statement culminating in ha.)

Ngea gives #137 ktktaprasagi nityam to characterize nitya (one rule is nitya with respect to

107
Abhyankar 1962, p. 8.
108
Compare gauamukhyayo, discussed on p. 20.
109
This wording is Nlakahas. In Puruottamadevas second collection, the rule appears as lakye
lakaa sakd eva pravartate.
110
Abhyankar 1976, p. 377; Abhyankar 1962, p. 190.

30
another if it is undertaken whether or not the other has been applied) after the phrase #297 parn nitya

balavat; Kielhorn lists the former as Ngeas rule 42, Abhyankar the latter.111

And Ngea gives #401 yena nprpte yo vidhir rabhyate sa tasya bdhako bhavati (what rule applies

when there is not nonobtainment by another, that is the blocker of the other) after the phrase #35

antaragd apy apavdo balyn; Kielhorn lists the former as Ngeas rule 57, Abhyankar the latter.112

What rules go into the table?

A close look at Abhyankars grand table of rules makes it clear that Abhyankar grappled with these

issues, and was not always consistent in his treatment. Two rules that have different statements may yet

be identical or close in sense, and that rules with identical meanings are not always treated the same. We

also see that Abhyankars treatment changes over time, between 1960-1962, when he edited the

Mahbhya and Kielhorns translation, and 1967, when the Paribhsagraha was published. Three

observations follow, backed up with examples.

Two rules that are worded differently may nonetheless be listed in the same row of the table. That is,

despite his statement to the contrary (q.v., page 8 above), Abhyankar does sometimes allow differences in

wording. For example, in row #186 we find both jpakalakaa na sarvatra (in Vyis second

collection) and jpakasiddha na sarvatra (in Puruottamadevas two collections, the collections in the

Sradeva group, and the Paribhenduekhara). In row #5 we find Sradevas 123, agaty hi khalu

paribh cryante, and 119 from Puruottamadevas first, agaty nu paribhryante.

111
Abhyankar 1967, p. 376; Abhyankar 1962, p. 73. Four other authors include just ktktaprasagi
nityam, and one other just parn nitya balavat. Collection 8 includes both.
112
Abhyankar 1967, p. 376; Abhyankar 1962, p. 134. See Kielhorn 1960, p. 320-321ff; antaragd
apyapavdo balyn is presented as introductory text to the actual paribh: yena nprpte yo vidhir
rabhyate sa tasya bdhako bhavati. Nine other authors include yena nprpte, but no one else numbers
antaragd apyapavdo balyn.

31
Two rules, although listed in different rows in Abhyankar 1967, may be identified as identical in the

tables in Abhyankar 1962. For example, although (as noted above on page 29) Sradevas 26 na

strpratyaye cnupasarjane and Ngeas 26 strpratyaye cnupasarjane na are listed on two different

rows in the big table of 1967 ( #253 & #537, resp.), they are collapsed into one row in the earlier work.

Two rules, although listed in different rows in both Abhyankars 1967 and 1962 tables, may be identified

with one another by Kielhorn 1960 in his table of correspondences, pp. 529-537. For example, #100

upajtanimitto 'pyutsarga upajaniyamanimittenpavdena bdhyate (Sradevas 93) is shown to be

substantially the same as #102 upasajaniyamanimitto py apavda upasajtanimittam apy utsarga

bdhata iti (Ngeas 64). Another example: #300 parihtypavdaviayam utsargo 'bhiniviate

(Sradevas 94) is the same as #319 prakalpya vpavdaviaya tata utsargo 'bhiniviate , Ngeas 63.113

Note also that Abhyankar in his big table sometimes gives, for a particular collection, the same number to

more than one row. One can take this (if not a typo) as a sign showing that he may consider the rules in

these rows essentially the same. This is the case, for example, for tideikam anityam and gamastram

anityam. Rows #21 and #78 are both marked as 95 in the second collection ascribed to Vyi, and rows

#20 and #73 are both marked as 84 in Puruottamadevas first collection, 98 in Sradevas, and 100 in

Haribhskaras.

Likewise, rules #149kvacit svrthik praktito ligavatannyativartante and #552 svrthik pratyay

praktito ligavacannyativartante are both given the same number within Sradevas and two more

collections (Nlakahas and the Jainendra) despite the fact that the rules actually in the collections differ

slightly from #149 and #552 (and each other). Puruottamadevas second collection, and Ngeas and

edris collections include #149; three more collections (2, 12, and 15) include slight variations on #552.

113
The phrasing of version #319 goes at least as far back as Jainendras rule. See Benson 1990, pp. 135-136,
for a description of Ngeas calling on Sradevas rule to discuss the principle: having determined the
range of the exceptions, the general rule is applied [elsewhere].

32
Another example, from the Jainendra collection, may betray some hesitation on Abhyankars part. He

marks #232 dhto svarpagrahae tatpratyaye kryavijnam with a 31 in column 9,114 but also includes

row #234 dho svarpagrahae tattvavijnam. The latter is nothing but #232 shortened up Jainendra-

style (note that dhu is the Jainendra technical term for dhtu115). There is no need to include the row #234

if 31 is entered in #232. Perhaps Abhyankar decided against the extra row, but forgot to take it outor

perhaps he decided the extra row was warranted, but forgot to move the entry 31 to the new row.116

Similarly, Abhyankar marks #490 sakdgatau vipratiedhe yad bdhita tad bdhitam eva (which occurs

in all the collections) with a 60 in column 4, but also includes row #451 vidhipratiedhena yad bdhita

tad bdhitam evaand marks it with a 60 in column 4 (the only collection to include this rule). The two

statements are not quite the same: the former is more specific. Perhaps Abhyankar decided to retain the

distinction but neglected to remove the marking from the more common rule, #490.

Sradevas innovations with respect to content

As noted above, all of Sradevas paribhs appear in other collectionsif only because Haribhskara

follows him so closely.117 Still, as many as nine of his rules do not appear in earlier collections and can be

regarded as his innovations.118 The table below contains the rules from the big table that have no entries

for collections before Sradeva and so may have been first formulated by him.119 Caution is in order, for

114
This marking has been altered in the table attached.
115
Abhyankar 1967, p. 104.
116
The latter adjustment has been made for the appendix table.
117
As we saw above, Sradevas paribh 26, which appears at first to be unique in row #253, is the same
as strpratyaye cnupasarjane na.`
118
As Abhyankar points out, Ngea also innovates (1962, p. 10).
119
According to Bhattacharya, Sradeva added fourteen rules to Puruottamadevas Paribhvtti. (He
also asserts that Sradeva omitted three of Puruottamadevas paribhs, numbers 60, 64, and 108
(Abhyankars #67, #526, and #282). In fact, Puruottamadevas 64 does appear in Sradevas collection as
rule 106.

33
some of these candidates on closer examination turn out to be closely related or even identical in impact

to other rules used by earlier authors.

A couple such show up right away: as we just saw, Sradevas 94 is identical to #319, that is, Jainendras

91.120 And Sradevas 54 is identical to #363, which is seen in Puruottamadevas Paribhvtti as rule 59.

Also, Sradevas 61 is closely related to the previous rule in the collection, #105 ubhayata raye

nntdivat, which occurs in seven other collections in various time periods.

Sradeva Abhyankar Ngea MB?


54 145 kriyviean karmatva napusakaligat ca
61 435 6.1.85 varraye vidhau nntdivattvam
74 231 88 7.2.114 dhto kryamucyamna tatpratyaye bhavati
89 449 93.10 vidhau paribhopatihate nnuvde
93 100 64 upajtanimitto 'pyutsarga upajaniyamanimittenpavdena bdhyate
94 300 63 parihtypavdaviayamutsargo 'bhiniviate
108 305 3.1.3 pitsvarccitsvaro balyn
122 303 115 paryyaabdn gurulghavacarc ndriyate
127 399 2.4.62 yugapadadhikaraavacano dvandva
128 309 prva dhtu sdhanena yujyate pacd upasargena
130 499 samsakttaddhiteu sabandhbhidhnam anyatra rhyabhinna-
rpvyabhicaritasambandhebhya

The rules are quite diverse and cover a variety of topics, including balbala rules, rules clarifying the

nature of anuvtti, compounding rules, and more.

Order

What motivated the paribh authors decisions about order of presentation? A grammarian may wish

to: use a historical ordering, as Puruottamadeva did for his first collection; divide into sections by type,

as the Ktantra authors and Bhoja did; organize by topic (possibly with pedagogical considerations in

mind), as Puruottamadeva began to do in the first section of his second collection; or organize for brevity

or elegance with devices such as juxtaposition and anuvtti.

120
It is important to keep in mind that Jainendras collection is not historical; rather, it was compiled by
Abhyankar and first published as part of the Paribhsagraha.

34
Sradeva used an organizing principle different from all of these. Before examining his innovation and

where it leads him, we first turn to three matters. First, we look to Pini for lessons about organizing for

brevity or elegance. Second, we notice a phenomenon that looks disorganizedthe fact that some

paribhs are repeated in some collections. Third, we use Abhyankars table to identify a few clusters of

related rules, and last we see how some authors have put together several rulesof a similar type or

functioninto one.

Pinis paribhs

Of course the Adhy itself is the epitome of works put together for brevity and eleganceand devices

such as juxtaposition and anuvtti are found no less when Pini is presenting paribhs than in the rest

of the work. Four sets of stras show the care with which he arranges the rules.

Stras 1.1.46 through 1.1.47 pertain to augments and their placement, and are placed just before 1.1.49 as

exceptions.

Stras 1.1.52 through 1.1.55 pertain to substitutes, and determine whether a substitute is to take the

place of the original (sthnin) as a whole, or only part of it; and in the latter case whether the final or the

initial is to be substituted. ah from 1.1.49 is read with each of these rules; ala from 1.1.52 is read with

1.1.53 and 1.1.54, and antyasya from 1.1.52 is read with 1.1.53.

Stras 1.1.60 through 1.1.63 define and distinguish between the various types of elisionlopa, which is to

be considered sthnivat , and luk, lu, and lup, which are not. In this cluster of rules, 1.1.63 is a restriction

of the scope of 1.1.62, and makes sense only when 1.1.62 is read with it.

Stras 1.1.68 through 1.1.69 pertain to the association of sense with word. sva rpa from 1.1.68 is read

with 1.1.69, which restricts the first stras application.

Repetition

In several cases, a paribh is repeated within a collectionwith the second instance at a distance from

the first. Of course, Pini himself repeats some rules in the Adhy, also at a distance. (The most

common of these, bahula chandasi, is a general acknowledgment of Vedic variants.) On the other hand,

35
the tradition holds that some rules are read twice, in the same place, with different implications;

halantyam is a well-known example.121

Why would a paribh author repeat a rule? Are the readings, or supporting arguments, different? We

have two instances of repetition in Vyis work, four or five in Ktantra-related collections, and one in

Nlakathas collection.

In the first collection ascribed to Vyi, rules 79 and 81 are exact repetitions of rules 36 and 35. (Both are

paribhs of precedence, and both find a place in all 18 collections presented by Abhyankartwo of five

rules that do so.) With respect to 36/79 (#70 asidda bahiragam antarage), on the basis of the

examples that Vyi gives in the two places, Abhyankar posits different readings.122 Wujastyks

translation shows that Vyis arguments supporting the two instances of the rule are different123 and

notes that the rule is frequently used by Ktyyana.124 The second occurrence of asidda bahiragam

antarage is followed by an exception to it, #257 njnantarye bahivaprakpti.

The second rule that appears twice, 35/81, is #490 sakdgatau vipratiedhe yad bdhita tad bdhitam

eva; again, the arguments are different in the two places.125 The second occurrence of sakdgatau

vipratiedhe yad bdhita tad bdhitam eva is followed by a directly contradictory rule, #306

punaprasagavijnt siddham; more about this later.

121
The treatment of halantyam in the Siddhntakaumud is discussed by Bronkhorst, p. 1-2.
122
Abhyankar 1967, p. 59. Abhyankar suggests that the first occurrence is ex ante, to resolve the
simultaneous occurrence of antaraga and bahiraga operations and that the second is ex post,
concerning a bahiraga operation which has taken place and which comes in the way of other
operations and hence has to be made invalid, and so, that the two statements of the rule may involve
two kinds of bahiragsiddhatva.
123
Wujastyk 1993, v. II, pp. 144-148, 240-242. (Wujastyk numbers the rules 32 and 73.)
124
Wujastyk 1993, v. II, p. 144-148, note 246.
125
Wujastyk 1993, v. II, pp. 141-143, 246-247. (Wujastyk numbers the rules so they are 31 and 75.)

36
The rule #76 gamdeayor gamo vidhir balavn is repeated in collections 7 (as 43 and 78) and 8 (as 45

and 94).126 In collection 7, its first occurrence is just after rule #44 antybhve antyasadeasypi grahaam

and just before #75 gamt sarvdeo vidhir balavn, and the second is with a group of balbala rules. In

collection 8, which does not display the same division into sections by purpose, the first occurrence of

#76is after a couple of balbala rules and just before gamt sarvdeo vidhir balavn, and the second is

immediately before antybhve antyasadeasypi grahaam.

In collection 7, #98 utsargpavdayor apavdavidhir balavn occurs twice, as 66, which is the

penultimate rule of the general interpretation section (vykhynata is the last), and as 96, which is last of

the collection as well as last of the balbala section.

In collection 8, #272 nityd antaraga balya occurs as 52 (with some nitya rules) and 117 (flanked by

two other balbala rules, at the end of the collection).

Also in collection 8, rule 87 is #175 gauamukhyayor mukhye kryasampratyaya, which comes just

before vykhynata, and just after a quirky rule that occurs only in this collection and may be a line of

commentary introducing gauamukhyayo. As rule 112, gauamukhyayo follows #525 smnyavidhir

utsargo vieavidhir apavda and precedes #541 sthitnm anvkhyna hi vykaraam; these two

flanking rules occur (in this form) only in the Ktantra-related collections.

In collection 14, #289 padgdhikre tasya ca tadantasya ca, a tadantavidhi rule, occurs first as 36, here

following #142 kdgrahae gatikrakaprvasypi grahaam, a rule that prohibits tadantavidhi; it is

followed in turn by a balbala rule regarding the application of case endings. As rule 108,

padgdhikre appears in a small group of rules about blocking.

126
Although Abhyankar lists, for both collections, both numbers in row #78, in fact the phrasing of rule 94
of collection 7 and of rule 45 of collection 8 slightly differs in that gamavidhi is given as a compound.

37
Related rulessometimes together, sometimes apart

The paribhs listed below are presented separately in some collections, but in others are either

concatenated and given one number, or given two numbers but commented on together. In yet other

cases, two rules may be integrated using a connective ca or v.

Rules #379 and #395 are closely linked: ya vidhi pratyupadeo 'narthaka sa vidhirbdhyate and

yasya vidher nimitta nsau bdhyate. They appear in sequence in collections 1, 2, 4, 5, and 11. In 1, they

are commented on separately; in 5, together. (Collections 2, 4, and 11 are bare lists.) In kayanas list,

#379 is given together with another clause that does not appear in the table. The two rules are

concatenated, and together given one number, in two of the Ktantra-related collections and in Bhojas,

and for these, Abhyankar lists that rule number in both rows in the table.

Rules #406 (yogavibhgd iasiddhi; accomplishment of the desired outcome is from splitting the rule)

and #463 (vyavasthitavibhaypi kryi kriyante; operations are done by means of a fixed option) are

combined into one rule in the Cndrapha: vyavasthitavibhay yogavibhgd v iasiddhi. (This

combined rule does not appear in the table; rather, the rule number in the collection, 55, is given in both

rows #406 and #463.) In other collections, the two rules are presented separately, close to one another in

Sradevas and Haribhskaras texts (in their nyyamla sections), and at a distance in collections 2, 3, 12,

14, and 16. Vyis first collection is the only one to include #406 but not #463; no collection includes #463

without #406. One more point of interest: #406, together with three more rules (#303 paryyaabdn

lghavagauravacarc ndriyate, #24 anirdirth pratyay svrthe, and #186 jpakasiddha na

sarvatra) is part of a subset of rules that Sradeva, Nlakaha, Haribhskara and Ngea give together.127

Sradeva and Haribhskara also include #463 with this subset of rules.

127
Kielhorn asserts that the rules in this subset (Ngeas 113-116) are among the 16 rules for which it
would be superfluous to ask whether they have been considered universally valid or not, because the
principles laid down in them are such that if it be considered at all necessary to adopt them, they must be
adopted once for all. See p. xviii.

38
Rules #512 and #514, sarvavidhibhya ividhir balavn and sarvavidhibhyo lopavidhir balavn, are listed

separately but sequentially by Vyi, kayana, Puruottamadeva in his first collection, and

Nlakaha. Vyi in his second collection expands the sequence by a third rule, #513 sarvavidhibhyo

lugvidhir balavn. In Puruottamadevas Paribhpha and the Paribhenduekhara, the rules are

combinedsarvavidhibhyo lopavidhir ividhi ca balavnand commented on together, but also given

two sequential numbers.128 Just the Ktantra-related collections include one rule (#514) without the other.

Only Sradeva and Haribhskara treat the two rules at a distance from each other. (Note that #512 is

considered to have a jpaka, 7.2.12.129 Rule #514 occurs in these authors nyyamla sections; however,

Kielhorn presents the rule, Ngeas rule 93.7, as having a jpaka, 1.1.62.130)

One final example will be useful later. Rules #490 and #306, sakdgatau vipratiedhe yad bdhita tad

bdhitam eva131 and punaprasagavijnt siddham, are represented widely; #490 is given in every

collection and #306 in all but the Ktantra-related and Haima collections. Only kayana presents the

two at a distance from one another; otherwise, when both rules are present, #306 immediately follows

#490.132 Puruottamadeva in his first collection, Sradeva, Ngea, and edri comment on both rules

together; the other authors, separately.

128
See discussion in Abhyankar 1962, pp. 9-10.
129
See Kielhorn 1960 on rule 93.8, pp. ix, 445-446. Note that Abhyankars edition of Sradevas text has a
typo; see Abhyankar 1967, p. 242, lines 8-9); sani grahaguhoca is 7.1.12, not 7.1.10. The same error
appears in Dubes edition.
130
Kielhorn 1960 on rule 93.7, pp. ix, 445-446.
131
This paribh, marked by Abhyankar as appearing in every collection, is given variously as
sakdbdhito vidhir bdhita eva (collections 5, 6, and 8), sakdgate pariaye bdhito bdhita eva (9), and
sakdgate spharddhe yad bdhita tad bdhitam eva (3 and 11).
132
Bhoja gives the two as rules 104 and 105oddly, given Abhyankars characterization of the group 105-
118 as a section specially pulled together as maxims for the guidance of scholars in the matter of general
interpretation and original composition in Sanskrit. Abhyankar 1967, p. 26. Surely both are in this final
section, or (less likely) neither is.

39
Portmanteau rules

In what could be called portmanteau rules, three or more rules on a similar topic are gathered together

and combined into one rule. Examples from three topic areas follow.

First, anubandha rules: A number of paribhs have to do with the function of indicatory letters in

defining the nature of the Kryin.133 Three rules in particular (#261, #262, and #263) declare that words

ought not be viewed as having more than one letter, or not ending in e/o/ai/au, or being different in

form, just on account of anubandhasthat is, nnubandhaktam anekltvam, nnubandhaktam

anejamtatvam, and nnubandhaktam asrpyam. Candragomin, Puruottamadeva (both collections),

Nlakaha, and Ngea present these rules in sequence, as does Vyi (both collections) for the two that

he includes.134 Sradeva (and Haribhskara) include all three rules but not in sequence; the topical

grouping is set aside so that the rules can be grouped according to their jpakas. Bhoja and Hemahasa

do something different: they combine the three rules into one rule covering all three circumstances:

nnubandhaktny asrpynejantatvnekltvni (#264).135

Second, anitya rules: In his very next rule, Bhoja again compresses several rules into one, this time

various kinds of rules whose application is voluntary.136 Each of the six kinds picked up by Bhoja has a

standalone version (and perhaps more than one) used by other authors. There are rules that teach the

addition of samsnta affixes,137 that is, affixes added to the ends of compounds without changing their

sense138 (#504); rules that teach the addition of an augment139 (#20, #72); rules using technical terms to

133
Abhyankar 1967, p. 42.
134
Sradeva and Haribhskara include all three rules but not in sequence.
135
Devasthali 1967, pp. 31-32, describes the nnubandhaktam paribhs and Bhojas compression of
them into one rule.
136
Abhyankar 1986, p. 20. anitya rules are discussed in Abhyankar 1967, p. 53.
137
See Kielhorn under Ngeas paribh 84, p. 407.
138
Vasu under 5.4.69, v. 2, p. 1008.
139
See Kielhorn under Ngeas paribh 93.2, p. 441.

40
denote that which is taught,140 sajprvaka (#477); rules taught by indication (#185 jpakajpit

vidhayo hy anity);141 rules teaching operations that affect a whole class of words, gaakryam (#164);

and rules that contain the negative particle na (#238).142 The portmanteau rule as given by Bhoja and

Hemahasa is #505 samsntgamasajjpakagaananirdini anityni.

A few more observations: Ngea too collects anitya rules, but (as edited by Kielhorn) he numbers them

separately. Ngea's sequence of rules 93.1 through 93.6 include a couple that Bhoja omitted: #28

anudttettvalakaam tmanepadam anityam, (his rule 93.4, first presented by Puruottamadeva in his

Paribhpha), and #78 tideikam anityam (his rule 93.6, which harks back to Vyis second

collection).143 Another anitya rule, #125 eknubandhaktam anityam, occurs only in Vyis collections

and Jainendras. Other authors gather some anitya rules and scatter others; for example, the anitya rules

in collection 8 are numbered 21 and 36-39.

Third, balbala rules: The rule that Ngea chooses to headline his second sectionof rules deciding the

priority of application in case of conflictalso collects a number of rules given individually in other

collections (and indeed within the Paribhenduekhara): #316 prvaparanityntaragpavdnm

uttarottara balya. This rule comes from Puruottamadevas Paribhpha and appears in slightly

altered form (without the leading prva, #292) in edris collection. One of the Ktantra collections

contains a rule that means the same thing: #296 parn nitya nityd antaraga antaragc

cnavaka balya.

140
Abhyankar 1986, p. 404; Kielhorn under Ngeas paribh 93.1, pp. 440-441.
141
This rule occurs in the Ktantra-related collections. As noted above (p. 31) Puruottamadeva and the
later authors phrase the rule without using the word anitya: #186 jpakasiddha na sarvatra.
142
Ngea note that this rule serves to solve a problem with 1.1.4 but argues (as he does for the anitya
rules numbered 93 by Kielhorn) that it ought not be accepted; after all, bhye darant, it is not given in
the Mahbhya. See Kielhorn under Ngeas 93.5, p. 442-444.
143
Ngeas sequence of rules 43-49his first sustained deviation from the content and order presented
in Puruottamadevas paribhphause the word anitya but have to do primarily with order of
operations.

41
There are more than 70 simpler rules that can be considered balbala paribhs. Such paribhs may

decide precedence between two rules, one with characteristic A, one without Afor example, #32

antaragabahiragayor antaraga balavat, or #98 utsargpavdayor apavdvidhir balavn. Or, they

may decide precedence between two rules, one with characteristic A and one with characteristic Bfor

example, #35 antaragdapyapavdo balyn, or #272 nityd antaraga balya.

This particular portmanteau rule, containing as it does the catchphrase uttarottaram, shows not only the

gathering together of rules but also their placement in a hierarchy. In this (although not in content) it

resembles the mms principle rutiligavkyaprakaraasthnasamkhyn samavye

pradaurbalyam arthaviprakart.144 Lubin shows that the device of establishing a hierarchy of criteria

goes back to the ritual codes; these may help us think about precisely what paribhs apply toperhaps,

that is, not other rules, but rather operations enjoined.145

Sradevas new order

Before Sradeva, who proposed new ways to order collections of paribhs? And what were they

thinking?

Vyi was an early paribh author who writes in reaction to the Adhy; Abhyankar attributes two

collections to him. What accounts for the second work being fifty percent longer, and organized anew?

Abhyankar and Belvarkar agree that Vyis work predated the Mahbhya. Wujastyk says it was the

other way round (with reason, says Cardona146) and ultimately proposes a much later date for Vyi.

Here, it is enough to observe that dating Vyi prior to the Mahbhya neatly integrates the whole idea

of external paribh works into the trimuni tradition, given Patajalis unquestionable authoritativeness.

Puruottamadevathe other author to whom two collections are ascribedfor his Paribhvtti goes

back to Vyis contents and order. His Paribhpha, on the other hand, displays a new order where

144
Mmsstra 3.3.14.
145
Lubin 2013, pp. 2-3.
146
Cardona 1999, p. 145.

42
the paribhs appear to have been arranged into three groups, and it is to this arrangement that Ngea

returns 500 years later.147

Sradeva first groups his paribhs according to their basisthat is, into two groups: those that are

indicated by a statement or part of a statement in the Adhy, the vrttikas, or the Mahbhya, and

those that are evident from accepted maxims.

Sradevas arrangement of rules in the first group is an innovation not seen in any previous paribh

collection. He places the rules in this group according to the order in which their jpakas occur in the

trimuni texts. For each rule, he typically concludes with a brief statement of its authority. In explaining

the basis of the vykhynata and arthavat, for example, Sradeva uses the expressions bhyakrea

jpit and bhye jpit. Some other paribhs are justified using words such as pahita.

Bhattacharya suggests that this new order was a product not of Sradevas imagination alone but instead

was his adaptation of ideas already in playindeed that Sradeva was taking up the cue evidently from

the Jpakasamuccaya of Puruottamadeva.148 Abhyankar surmises that Sradeva intended to help the

studentthat he chose the jpaka order possibly with a view to facilitating the finding out of the

references in the Mahbhya.149

But Sradevas order is not as simple as it seems. Looking at the jpakas that Sradeva mentions,

grouped as they are by adhyya and pda, we can see that he carves out at least a couple of exceptions to

jpaka order. In particular, a rule that influences the application of another rule may be listed with that

rule even if its jpaka would put it somewhere else. Two examples follow.

Sradevas rules 9 and 10 (#55 arthavadgrahae nnarthakasya and #246 na varagrahaeu): The jpaka

for rule 10 is 1.1.17,150 but rule 10 is presented directly after rule 9, which is pointed to by 1.1.68.151 (As

147
Devasthali 1985, p. 12.
148
Bhattacharya, p. 29.
149
Abhyankar 1967, p. 29.
150
Abhyankar 1967, p. 173, lines 9-11.

43
discussed above (pp. 19-21), these rules have to do with the sense connection of words; na

varagrahaeu is a narrowing of arthavats application.) A glance at the other collections shows that the

second rule occurs in just seven collections, whereas arthavat appears in every one; when the second rule

is present, it always directly follows the first.

Sradevas rules 69 and 70 (#70 asiddha bahiragam antarage and #257 njnantarye

bahivaprakpti): Although the jpaka for rule 70 is 6.1.86,152 it is instead given directly after rule 69,

which is pointed to by 6.4.132.153 (These rules have to do with the priority of antaraga rules; 70 is an

exception to 69.) All the collections from Puruottamadeva on give these two rules in sequence; in the

earlier collections, the rules may be close, or far, or the second rule may not appear at all.

It is also true, however, that Sradeva sometimes follows jpaka order even when it breaks up two

closely related rules.

Contradictory statements paired

Juxtaposition of two rules that are mutually contradictory and have the same scope is a device used

several times by paribh authors: a true option is provided by presenting both possibilities together.

Wujastyk mentions the device in explaining Vyis rules 81 and 82 (#490 and #306): these contradictory

rules are applied as appropriate to produce the desired form.154 As Abhyankar puts it, option is

defined as presence along with absence, and this method of presentation is reminiscent of the dictum of

the Mmsakas as observed in the case of two orders of an opposite nature; he alludes to the phrase

atirtre oainam ghti, ntirtre oainam ghti.155

151
Abhyankar 1967, p. 171, line 23ff. Also see Kielhorn 1960, pp. xiii, 84-85 (note that #9 is Ngeas 14).
152
Abhyankar 1967, p. 235, line 22. Also see Kielhorn 1960, p. viii, 84-85 (note that #70 is Ngeas 51).
153
Abhyankar 1967, p. 234, line 6. Also see Kielhorn 1960, p. viii, 84-85 (note that #69 is Ngeas 50).
154
Wujastyk 1993, v. II, p. 248.
155
Abhyankar 1967, pp. 53-54. This example is discussed below.

44
Kielhorn lists several optional principles that appear in the Paribhenduekhara, including at least five

that are expressed as pairs of mutually contradictory rules.

Sradeva Abhyankar Ngea


115 387 2 yathoddea saj paribham
116 131 3 kryakla saj paribham
14 119 17 ekayoganirdin saha v pravtti saha v nivtti
8 155 18 kvacid ekadeo 'pyanuvartate
21 306 39 punaprasagavijnt siddham
20 490 40 sakdgatau vipratiedhe yad bdhita tad bdhitam eva
47 373 60 madhye 'pavd prvn vidhn bdhante, nottarn
48 17 61 anantarasya vidhir v bhavati pratiedho veti
78 343 107 pratyeka vkyaparisampti
79 148 108 kvacit samudye 'pi
128 309 prva dhtu sdhanena yujyate pacd upasargena
129 308 prva dhtur upasargena yujyate pact sdhanavcin pratyayena

In introducing his list, Kielhorn notes that the very idea of optionality in application renders the system

less than fully deterministic. In his words, a determination of universal validity is impossible in regard

to those Paribhs which, while they allow a choice between several alternatives, leave the decision as to

what course should be followed in individual cases, either to the interpretation of the learned, or to our

own judgment guided as it must be by the particular forms of the language which it may be desirable to

form, or for the formation of which we may wish to account, by the rules of Pinis grammar.156 In the

tradition, the concept of vyavasthita as applied to paribhs captures the fact that it is difficult or

impossible to state definitively the conditions under which some rules apply.

In each of the listed pairs, the two rules share the same scope. If their scope differed, then one rule would

be interpreted as an exception to the other, at least for some range of application. Benson makes this clear

in his edition and translation of the Mmmsnyyasamgraha, 10.8.3: atrobhayor vieaviayatvena

paryudsatvsabhavd vikalpa.157 That is, in the phrase [atirtre oainam ghti, ntirtre

oainam ghti] there is an option, on account of the impossibility of being an exception on the part of

both (statements) because of their (shared) particular scope; as Benson puts it, because both (clauses)

156
Kielhorn 1960, pp. xviii-xx.
157
Benson 2010, p. 282.

45
have a particular range there cannot be an exclusion, and therefore there is an option; in contrast to a

situation in which the first clause has a general range, the second clause a particular one.158

Sradeva as well as Ngea includes these contradictory pairs in his collection. Four of the pairs fall in

Sradevas first section of jpita rules; notably, Sradeva too lists the two rules of each pair in sequence.

Does this, like the utsarga/apavda structure seen above, require Sradeva to temporarily set aside his

organizing principle?

It seems not, in three of the cases: rules 20 and 21 (sakdgatau vipratiedhe yad bdhita tad bdhitam

eva and punaprasagavijnt siddham),159 rules 47 and 48 (madhye 'pavd prvn vidhn bdhante,

nottarn and anantarasya vidhir v bhavati pratiedho veti), and rules 78 and 79 (pratyeka

vkyaparisampti and kvacit samudye 'pi). Rather, for these rules, the jpakas proposed by Sradeva

seem to be the same for both rules of the pair, or are said to come from adjacent stras. For example, as

discussed above (page 39), Sradeva comments on 20 and 21 together,160 to the effect that the two rules are

taught by two interpretations of 1.4.2, vipratiedhe para kryam.161

In at least one case, however, jpaka order is preserved even though it breaks up the pair: Sradevas

rule 8 (#155 kvacid ekadeo 'pyanuvartate) has as jpaka the itis in P. 1.1.66 and 67,162 and his rule 14

(#119 ekayoganirdin saha v pravtti saha v nivtti) has jpaka 1.2.14.163

158
Benson 2010, pp. 724-725, and note 707.
159
These rules teach both that the rule set aside remains always set aside, and that a rule may apply even
after the correct form is reached.
160
Abhyankar 1967, pp. 184-186.
161
Abhyankar 1967, p. 184, lines 7-8 especially. Also see Kielhorn 1960, p. xv (note that Sradevas 20 and
21 are Ngeas 39 and 40).
162
Abhyankar 1967, p. 170, lines 8-9.
163
Abhyankar 1967, p. 176, line 12. Kielhorn (1960, p. 100, n4) points out that Ngea has a different
opinion, basing this rule on a lokanyya.

46
Sradeva also preserves the juxtaposition of the two rules that famously govern the site at which

paribhs are combined with the stras with which they are read: #387 yathoddea sajparibham

and #131 kryakla sajparibham.164 Here, there is no question of running afoul of jpaka order,

for Sradeva puts these rules as numbers 116 and 117 into the second, nyyamla section of his collection.

There are pairs of opposites that do not occur in the Paribhenduekhara and so are not listed by

Kielhorn. An example is Sradevas 128 and 129 (#309 prva dhtu sdhanena yujyate pacd

upasargena and #308 prva dhtur upasargena yujyate pact sdhanavcin pratyayena). In

Abhyankars words, Sradeva presents both views, viz., that [upasargas] could be prefixed before the

conjugational affixes are added to roots, or after.165 In this case, Sradeva deviates from the presentation

we have seen above, by framing the second rule with anye tu hu . It is possible here that Sradeva

was intending not to present an option but rather to dismiss others opinions. On the other hand,

Nlakaha, who also includes both rules, presents them without such framing. Two other authors

present only one side of the argument: Jainendra #308, and Haribhskara #309.

One last example serves to underline the requirement for shared scope. Puruottamadeva and Ngea

present together two rules that prescribe and prohibit, respectively, tadntavidhi,166 that is to say, #337

pratyayagrahae yasmtsa vihitas taddes tadantasya grahaam and #336 pratyayagrahae

cpacamy.167 (Sradeva gives only one of these: #337 is his rule 25.) These two rules do not present an

option; rather, the validity of #337 is restricted by #336.168

Limitations of jpaka

For at least four rules, Ngea as reported by Kielhorn offers jpakas that are inconsistent with

Sradevas ordering.

164
Joshi 1965 and especially Benson 2012 provide detailed analysis of these rules.
165
Abhyankar 1967, p. 55.
166
Abhyankar 1967, pp. 46-47.
167
This pair of rules is discussed by Benson (1990, p. 48-53).
168
Kielhorn 1960, p. xix.

47
Abhy # Sradeva # S jpaka Ngea # N jpaka rule
95 15 1.2 (45) 22 8.3.46 (vii) udayo 'vyutpannni prtipadikni
326 18 1.3 (18) 36 6.4.59, 8.2.46 (viii) praktivad anukaraa bhavati
115 40 3.2 (56) 37 4.1.83 (xi) ekadeaviktamananyavat
547 53 4.3 (140) 79 6.2.142 (viii) svaravidhau vyajanam avidyamnavat

How can we explain this divergence? It may be simply that Ngea is using jpaka in a narrower sense

than Sradeva. A broad interpretation of jpaka is simply that which teaches something.169 The tradition

also developed a narrow interpretation requiring that a jpaka consist of four parts: vaiyarthya, jpana,

svasmicritrthya, and anyatraphala. Kielhorn illustrates the difference between the two interpretations

using vykhnata as an example: the use of in la, the sixth ivastra, indeed teaches the paribh,

but does not also meet the third of the formal conditions (that it is also absolutely necessary and serves a

purpose when that paribh has been adopted) since some other marker letter could have been chosen

for the sixth ivastra and so this use of in la is not considered by Ngea to be a jpaka for it.170 As

seen in the translation of vykhnata (see translation on page 61), Sradeva does not mention or require

any such condition.

Or, there may in fact exist more than one jpaka. In Kielhorns words, there is no reason why the

existence of one and the same Paribh should not be indicated by the two or even more Jpakas; nor

can there be any cause why the existence of a Paribh the validity of which may be established

otherwise than by a Jpaka, should not at the same time be indicated by a Jpaka.171 The next question

is how discretionary then were Sradevas choices, or Ngeas for that matter. Sradeva for some rules

gives more than one jpaka; in fact, this is the case for arthavat (see translation on pages 74-75).

169
Aussant 2011, p. 204, points out that this broader sense of jpaka is also found in the Cndravtti, a
commentary on Candragomins grammar. Aussant contrasts this use of jpaka with cra as it appears
in the Kikwhich evokes, more frankly, the weight of the tradition or of a spiritual heritage, while
jpaka seems more neutral.
170
Kielhorn 1960, p. v-vi, n3.
171
Kielhorn 1960, p. vii.

48
Even given some discretion, Sradevas organizing principle crowds out other goals. Paradoxically,

returning to the framework dictated by the Mahbhyaitself organized according to the Adhy

effectively rules out the use of most stra devices.

Endings

How do the paribh collections end? Is the finish notable in some way, or is it just a matter of having

got through the material? Let us note first that endings may be especially vulnerable to alteration during

transmissionrules may get added on, or dropped.172 And a thorough treatment of the endings would

require attention to the colophons; of course, this would require a careful look at the manuscripts. With

these caveats in mind: in some cases, collections seem to end with a bang; in others, it is hard to tell.

For the works that are divided into sections, some have general rules of interpretation first, others last.

Abhyankar notes for example the novel feature of Bhojas collection, a group of maxims (numbered

105-118) that deal with the matter of general interpretation and original composition in Sanskrit,

positioned at the end of the work.173 Notably, this group contains the paribhs that Ngea lists first,

second and third in his collection. Ngeas final section comes after general rules for interpretation, and

balbala rules, have already been dealt with; these are the rules that are more narrow in application. In

fact the Ktantra collections were the first to divide into sections according to function, giving first the

rules of interpretation and then the balbala rules. Sradevas Paribhvtti also falls into two sections, of

course, but according to source or basis rather than function; his final section comprises 31 rules that are

nyyamla rather than jpita. Necessarily it comprises a diverse group.

As far as the individual rules involved, vykhynatawhich we have seen leading off a half dozen

collectionscan also contribute to a big finish, perhaps not surprising, since the rule firmly underscores

the centrality of traditional teaching. We find this rule at the end of Durgasihas collection, at the end of

the first section of the Ktantra-related collection 7, penultimate in the second section of the Haima

172
For example, manuscript A is missing Sradevas final rule, rule 130. See the fifth appendix, p. 91.
173
Abhyankar 1967, p. 34.

49
collection, third from the end of Bhojas collection, and at the very end of Puruottamadevas first

collection, the Paribhvtti.

Although Ngea generally takes Puruottamadevas second collection, the paribhpha, as a model,

he ends his collection the same way that Puruottamadeva ends his first collection, the Paribhvtti

with #58, ardhamtrlghavena putrotsava manyante vaiykara. This statement of brevity as a

desideratum is an interesting choice, enunciating a principle embraced by the strakra but not worked

as much as might be expected by the paribh authors.

How does Sradeva end? We look at the last eight rules of his collection. His rule 123, #5 agaty hi khalu

paribhcryante (paribhs are relied upon when there is no (other) way to proceed), is among

the rules that give hints of the application of Paribhs;174 it is the penultimate rule of

Puruottamadevas Paribhvtti.

Sradevas 124, #463 vyavasthitavibhaypi kryi kriyante, sets out the regulated option (the phrase

is Abhyankars) for a stra to apply only in a few specified cases.175 It occurs in seven other collections

from various time periods, though not necessarily near their respective ends.

Sradevas rules 125 and 126 also occur among the last dozen of the Paribhenduekhara. Rule 125, #24

anirdirth pratyay svrthe, concerns the application of affixes. Rule 126 seems a good candidate for

a capstone: #186 jpakasiddha na sarvatra, and it does occur at the end of Vyis second collection.

This rule provides that a rule that is indicated is not always performedthereby, in Sradevas case,

calling into question a good three quarters of his work.

With a single exceptionJainendras prior employment of Sradevas 129Sradevas last four rules are

all innovations that are later picked up to some extent by Nlakaha and Haribhskara but not at all by

Ngea or edri. His 127, #399 yugapadadhikaraavacano dvandva, provides that a single member of

174
Abhyankar 1967, p. 41.
175
Abhyankar 1967, p. 54.

50
a dvandva compound can denote both members.176 As noted above (p. 47), Sradeva presents with his

rules 128 and 129 the two different views of whether upasargas are applied to dhtus before or after

inflectional endings.

Sradevas final rule, 130, states that in words which are derived from nouns and roots, as also in

compound words, the element of relationship is an additional feature which is due to the composite

expression.177 Having set out the formation of composite wordings as one of the half dozen major types

of grammatical operation that the kryin may undergo,178 Abhyankar classifies this rule with others

(including Sradevas 127) that concern composite expressions and limitations on their use.

We would be remiss in leaving the discussion of endings without pointing out that the final rule in the

Jainendra collection, K. V. Abhyankars own compilation and commentary, auspiciously bears the

number 108.

176
Abhyankar 1967, p. 60-61.
177
#499 samsakttaddhiteu sabandhbhidhnam anyatra rhyabhinnarpvyabhicarita-
sambandhebhya. The translation is Abhyankars, p. 61.
178
Abhyankar 1967, p. 50-51.

51
Conclusions

1. The decision of what constitutes a paribh is non-trivial. Wording can differ; some authors

combine things that others leave separate; learned people can disagree about what portion of a

vtti text really is the paribh; some authors label as a paribh a statement that looks like a

gloss, and some authors treat as gloss something that looks like a paribh; sometimes it takes

two statements to express one principle.

2. Despite the difficulty of comparing collections formally, we can see that the collections really do

fall into groups: Ktantra-related; Vyi 2 & Puruottamadeva 1; Puruottamadeva 2 and

Ngea; Sradeva and Haribhskara.

3. Although Sradeva was known for his innovative arrangement, he at times set jpaka order

aside, showing that certain pairs or small sets of rules simply do not make sense unless presented

together and in order.

4. Use of order and juxtaposition certainly matters in the sense that small groups of rules mutually

determine their appropriate scopes of action. This kind of artful arrangement appears in the

paribh literature, butpossibly because the works are relatively smallnot nearly on the

scale seen in the Adhy. We do not see the kind of close interplay and subtle relationship that

Sradeva lays out for the stras he analyzes whilst commenting on vykhynata.

5. Sradeva was perhaps the last enthusiastic participant in the jpaka strand of grammatical

literature. However, the limitations of the jpaka model were recognized by Sradeva and other

paribh authors, as evidenced by individual paribhs pointing to the variability of rules

derived from jpakas.

6. In terms of content and arrangement, Ngea had three roughly contemporaneous (11th-12th

century) models to work from just within the trimuni tradition: Puruottamadevas two works,

and Sradevas. He rejected Sradevas orderdespite the deep bow that it makes to the authority

of the Mahbhyachoosing instead the model that organized paribhs first by type and then

by topic.

7. From the beginning, the paribh literature was the product of a tradition that vested authority

in the three big works. Despite the also-strong desire for accessibility, that assumption of

52
authoritativeness delayed the emergence of recasts, and one reason the idea of jpaka gained

currency was that it could reiterate and underscore the authority of the traditional texts.

8. Even for the authors of new systems, however, paribh texts sometimes remained ancillary to

the main text, perhaps because of their metalinguistic nature.

9. Wujastyks claim that paribhs are falsifiable179 certainly holds for some rules (for example, the

balbala paribhs), but for many it does not. First, vykhynata seems to be in direct conflict

with falsifiability, certainly to the extent that interpretation is the best current understanding of

a living, changing tradition. Only if vykhynata can be tied back to some well-defined and

complete source (Bronkhorst suggests the Mahbhya) could it be consistent with falsifiability.

Moreover, the existence of formal devices by which to present options in some cases requires a

judgment call, thus rendering untestable the hypothesis corresponding to either option. The

paribh literature is fascinating precisely because it skates on the thin ice near the limits of the

formal system.

Directions for further research

Much more could be done along the lines pursued in this paper. For example, a look at the use of

particles would be usefulca, v, na, apiperhaps taking as a model Clooneys examination of the

significance of frequently used particles in the Mmsstra.180 And, systematically charting the use of

examples across collections is a first step toward understanding the arguments used.

Wujastyk makes the point that the literature on prakriy (by which he means the literature on

derivation, including the paribh literature) is one way to string along discrete gobbets of

information, and compares the task to that faced by those who compiled and presented subhita

texts.181 Looking at the way that paribh authors move from one rule to the next will shed light on the

rules structural relationship. Ngea is especially methodical in his introduction of each paribh and it

179
Wujastyk 1982.
180
Clooney, p. 44
181
Wujastyk 1993, v. I, p. x.

53
would be good to look carefully at whether Puruottamadeva or Sradevaor any of the non-Pinian

authorswas a model in this regard.

These taskslooking at particles, examples, and connections between rulesall require a granular

approach that is best supported by a systematic look at available manuscripts.

Regarding a paribh collection as one way to package and deliver information encourages us to take a

step back, as Trimble does in his examination of stra literature, to assess what genre we are

discussing, and to distinguish between that and the mode, or manner of enunciation. Here, the paribh

collection is one mode, and the prakriy work another, used to present information about a particular

kind of thematic content: the application of operational rules of various kinds.182

Candotti has looked at how Pinis paribhs were arranged by the prakriy authors;183 surely they

must also take recourse to some extra-Adhy rules. Tracing the way that balbala rules, for example,

show up in early prakriy literature, might help contribute to understanding the choices made by

grammarians in deciding what to work on and how.

There is a distinct jpaka literature, including at least two published works: the Jpakasamuccaya

ascribed to Puruottamadeva (there is extensive manuscript evidence for this work as well) and the

Jpakasagraha, attributed by Tatacharya to Ngea. These works are presumably in the trimuni

tradition; a natural question to ask is how jpakas work in the non-Pinian grammars. It may well be

that attempts to treat jpakas systematically reveal more about their limitations than their usefulness,

but that remains to be seen.

If we want to draw insights from comparing collections, then more should be done to look at their

provenance and actual place in history. For example, where did the Cndra paribhpha come from?

According to Belvarkar, no work on Paribhs in connection with the Chndra school has come down

182
Trimble, pp. 6-7.
183
Candotti 2012b.

54
to us;184 Haraprasd Shstr (while mentioning the work of Bruno Liebich with respect to this school in

general) makes the same observation.185 Abhyankar presents a list, noting only that he has given

paribhs of this school as they are found printed at the end of the Cndrastras edited by Liebisch

(sic).186 Where did Liebich get them?

Another example: many modern authors remark on the uncanny way that Puruottamadevas second

collectionlike the Cndra collection, a bare listpresages Ngeas Paribhenduekhara. Might the

transmission of Puruottamadevas pha have been affected by the Paribhenduekhara?

A thorough historicization would involve looking at how individual ideas in the paribh literature fare

over time. How does the phrasing of the rule change? Do examples and arguments also change? Is it

possible to tease out how specific topics wax and wane?

184
Belvarkar, p. 50.
185
Haraprasd Shstr, pp. li-lii.
186
Abhyankar 1967, Prefatory Note, p. 2. Abhyankar here is referring to Bruno Liebichs Konkordanz
Panini-Candra, pp. 49-52.

55
AppendixSradeva on vykhynata, paribh 1

[p. 161, line 9] There is understanding of particulars from traditional explanation, for there is no lack of

clear instruction arising from doubt. 1

There is comprehension187 of particulars, even when there is a doubt, on account of traditional

explanationthat is, on account of the teaching by the lineage of teachers. There is no lack of clear

instructionby this, even if 1.1.49188 were rejected [as being redundant] then in the places that sixth case

is taught, despite the possibility of there being many relations,189 [we would still know that] only that

[genitive] which is connected to location comes about. What is said [in the Mahbhy]: Given [this

paribh, viz.,] vykhynato vieapratipattirna hi sadehd alakaam, we will explain

sthneyog in the traditional manner.190

And so, there are [in the Adhyy] four mentions of a with the earlier akra191 and one [mention of

a], in 1.1.69,192 with the latter [akra]; mentions of i are with the latter [akra] only.193

187
Sradeva glosses pratipatti with avagama.
188
ah sthneyog 1.1.49.
189
The several relations between two nouns that can be expressed by the sixth case include sva-svmi
(possession), avayava (part-whole), nirdhraa (specification) and many more. Sradeva is asserting that,
should confusion arise about which interpretation is appropriate, traditional teaching (no less than 1.1.49)
specifies sthneyogathe genitive connected to location, that is, the so-called substitutive genitive.
190
This quotation from the Mahbhya is near the end of Patajalis remarks on 1.1.49, under the fourth
and final vrttika on the rule. The wording of the final clause, given by Sradeva as sthneyogeti
vykhysyma, is slightly different in the Mahbhya: sthna iti vykhysyma.
191
a refers to a pratyhra read from the ivastrasthat is, one of two possible pratyhras, since the
marker letter is used twice. The prior a is a, i and u. The latter includes not only a, i, and u, but also ,
, e, o, ai, au, h, y, r, v, and l. As discussed in the Mahbhya under la, the four stras are that mention
the prior a are: hralope prvasya drghoa 6.3.111; ke a 7.4.13 [agasya 6.4.1, hrasva 7.4.12];
ao praghyasya anunsika 8.4.57 [sahitym 8.2.108, vvasne 8.4.56]; and ur a rapara 1.1.51
[sthne 1.1.49]. (Words carried over from previous rules by anuvtti are put in square brackets.)

56
[p. 161, line 16] And in this matter, the reasons are explained as far as possible. Among the reasons: in the

stras 7.4.13194 and 6.3.111,195 the very mention of a has the power to show196 that it is determined as

being the former one. If it were the latter [a ], hrasva and drgha necessarily replace vowels only; [but]

this is established just by the paribh 1.2.28.197 Therefore, why mention a ?198

[But] this line of reasoning is liable to mislead.199 The fact that he [presumably the strakra] makes

mention of a, when short and long are effected by 1.2.28 alone, means that a has precisely this

purpose: in order that long and short would occur in place of [a phoneme belonging to] a despite [that

phonemes] not being a vowel. [For example:] agka, adhka, valika, madhulika, varh,

varhum.200 And thus what was said by Jinendrabuddhi201 under labecause a-consonants are not

possible when ka follows, and likewise, and when 1.2.28 is brought into play by the hearing of

drgha, length must occur only in place of a vowelthat is dubious.

192
auditsavarasya cpratyaya 1.1.69 [sva rpa 1.1.68].
193
The i pratyhra is i, u, , , e, o, ai, au, h, y, r, v, and l.
194
ke a 7.4.13 [agasya 6.4.1, hrasva 7.4.12].
195
hralope prvasya drgho a 6.3.111.
196
That is, the smarthyam.
197
aca ca 1.2.28 [aj hrasvadrghapluta 1.1.27].
198
Sradeva is arguing that if the latter a were meant, there would be no point in mentioning it at all,
since as a superset of the vowels, to which this operation is already limited, it would have no restrictive
force. This is the first rejection of the argument that the latter a is referenced in 7.4.13 and 6.3.111. He
now turns to considering what would happen absent the conventional implication of 1.2.28that is, if
hrasva or drgha could replace a consonant.
199
vyabhicrayitumto lead astray.
200
The first two of these examples are obscure words that may mean something like poor little mute
one and poor little harmless one. Proposed derivations are provided in the second appendix, p. 69.
201
Author of the prominent commentary on the Kik known as the Nysa.

57
[p. 161, line 25] In this matter, some suppose: In such examples as agka, once the visarga substitutions,

etc., are done,202 the h [of lih and of vh] and the r [of gir and dhur]which [for the sake of argument] are

ano longer exist. And there is no state of being asiddha203on account of [asiddhas] prohibition by

means of splitting 8.2.3 into two.204 Nor can one say inability to express the meaning desired205on

account of agka, etc., being traditionally explained206 in the Nysa at 8.3.16.207

And with respect to the rule teaching long (that is, 6.3.111) in examples such as varh, varhum,208

and due to understanding the pratyhra with the latter akra, [it could be] stated that there is length in

place of repha despite its not being a vowelbut this is not so. If this [making r long] were the purpose,

then one could make the pratyhra with the letter of [the ivastra] hayavara. (That is to say,) when

there is lopa of h or r, there is length in place of the preceding a phoneme.209 Or rather, in such

202
kharavasnayor visarjanya 8.3.15 [padasya 8.1.16, sahitym 8.2.108, ra 8.3.14] provides for the
substitution of visarga for word-final r when a voiceless stop or sibilant follows, or when a pause follows.
ho ha 8.2.31 [padasya 8.1.16, jhali 8.2.26, ante ca 8.2.29] provides for the substitution of h for word-
final h in the derivation of the last four of the six example words given above.
203
That is, one cannot say that the aforementioned operations are not effected, or are as if not effected.
204
na mu ne 8.2.3 [asiddham 8.2.1.] The proposed split is: na 8.2.3a, mu ne 8.2.3b. 8.2.3a then negates 8.2.1
as necessary, according to the vrttikas on this rule.
205
This is Abhyankars definition of anabhidhna, a reason conventionally given in grammar to reject the
formation of words not found in current use recognized by learned persons or scholars. Abhyankar
1986, p. 17.
206
Sradeva uses the word vykhytatva for these examples.
207
ro supi 8.3.16 [sahitym 8.2.108, ra 8.3.14, visarjanya 8.3.15.] The point is that merely on account
of their having been mentioned by an authoritative grammarian, these words are known to be valid.
208
These forms are made with the affixes tc and tumun after the dhtu vh. Note that for these last two
examples, while (as has been mentioned) h has been replaced and thus is no longer there, r (as a result of
guation) remains, and is thus (under the supposition that r is in a ) liable to lengthening by 6.3.111.
209
This is identical to 6.3.111 except for a in place of a.

58
examples as varh, varhum, let there be kra that is indeed long.210 Even so, once that is done, then by

6.1.87211 with gua in place of both akra and kra, and once [the substituted gua] akra is followed by

r,212 precisely thus the [desired] form varh, varhum. And so, [these] having been accomplished by

1.2.28, the mention of a is for the purpose of [making] the pratyhra [formed] with the prior akra.213

[p. 162, line 6] But nor is it the case, given the understanding of the pratyhra with the latter akra, that

a blocking of 6.1.87 should be assumed due to the pointlessness of lengthening for repha.214 For thus a

two-fold capacity would come about.215 Marking with t after [ ] in 7.4.7216 is an indicator, showing that

there is understanding of the pratyhra with the latter akra in the savara-rule.217 If, even there [in

1.1.69], it were with the former [akra ], and thus when the kra that is expressly mentioned in the stra

[7.4.7] is being taught in place of the [single] measure expressly mentioned in the stra, then on account

of [ and ] not being savara, there is indeed no attainment [of a sound] of two measures; consequently

the marking with t would be pointless.218

210
According to 6.3.111, by substitution for consonant r. This works for both versions of 6.3.111as in the
Adhyy, or with a in place of a, since r is a member of both pratyhras.
211
d gua 6.1.87 [sahitym 6.1.72, aci 6.1.77, eka prvaparayo 6.1.84.]
212
By ur a rapara 1.1.51 [sthne 1.1.49.]
213
We can make this conclusion because the outcome is not dependent on whether or not the
lengthening (replacement by ) of r (which is there by guation) takes place; either way, the forms
varh, varhum come about. So there is no point to supposing that it is possible and appropriate to
lengthen r. This is the second rejection of the argument that the later a is referenced in 6.3.111.
214
As Sradeva is about to demonstrate, 6.1.87 is not inconsistent with reading the latter a, although the
stras usefulness in the above argument might lead one to think so.
215
That is, the rule 6.3.111 would teach two thingsnot only length for repha, but also that gua is
blocked. Such a double capacity goes against the principle of efficiency in teaching.
216
ur t 7.4.7 [agasya 6.4.1, au cai upadhyy 7.4.1, v 7.4.6.]
217
auditsavarasya cpratyaya 1.1.69 [sva rpa 1.1.68.]
218
Sradeva has provided proof that the longer a is referenced in the savara ruletaking care as he
does so not to contradict the arguments he made for the shorter a in 6.3.111.

59
However, here under la, [this is] what is said by Jinendrabuddhieven in place of drgha, only hrasva

must be heard; consequently the marking with t afterward would be useless since there is nothing to be

excludedin that statement, the claim that the sthnin is a sound with two mtras is inconsistent.219

Because there is nonexistence also of a sthnin that has two mtrasjust as for the dea. If [one were to

claim] that the sound with two mtrasin the [universal /] categoryis a sthnin, we reply, not so.

Thus also with respect to the dea, an unwanted conclusion results from the property of having two

mtras.

[p. 162, line 14] Now, the universal for the sthnin, [but] the particular for the deathis is not

convincing.220

Thus then by the singularity, as a samhra dvandva, and by the shortness, as a neuter [noun], of both

kra and kra [compounded together],221 when 7.4.7 is taught, a long sound also could be the sthnin.222

And the absence of num is because the teaching of augments is not obligatory.223

This too is incorrect.224 On the contrary,225 is not [considered] a vowel, because of its noninclusion in the

pratyahra a, and because there is not the understanding of as being savara with .226

219
Sradeva is attacking Jinendrabuddhis argument for its assumption of particularity on the part of the
dea (that is, that it can only be short ), but of universality on the part of the sthnin, which can be any
of the sounds that are savara with .
220
This concludes Sradevas criticism of Jinendrabuddhi: although Jinendrabuddhi reaches the correct
conclusion (that marking with t indicates the understanding of the latter akra,) his argument is not
sound. Sradeva now considers other ways that the sthnin could be long.
221
Since the final vowel of a prtipadika in the neuter is replaced by a short vowel: hrasvo napusake
prtipadikasya 1.2.47 [ .]
222
He first proposes that, rather than referring to a sound in the universal / category, the u in the stra
directly denotes the two phonemes and in a (singular) samhra dvandva compound.
223
Neuter nouns ending in vowels optionally get augment n before endings that begin with vowels, from
iko ci vibhaktau 7.1.73 [agasya 6.4.1, num 7.1.58, napusakasya 7.1.72.] With this option, more often
taken than not, the dvanda in the sixth case would be a.

60
[p. 162, line 18] From this, the vowel being short is being taught; how [then] could it be [in place] of ?

We reply: By 1.2.47,227 the vowel being short is taught. And does exist in the pratyhra ac [satisfying

the conditions of the stra]; that being so, from the reliance on universal -ness by the letter , there must

be shortness of the letter despite its having two mtras, so [an expression] such as in place of the long

sthnin is proper.228

[p. 162, line 22] This paribh is pointed out (jpit) by Patajali under la. He said: Is this akra a

cause for doubt, from the fact that despite being attached already,229 it is nonetheless attached again as

though there were no other letters? And by that there is a doubt, when there are mentions of a and i, as

to whether the former or latter [is meant]. So the teacher [Pini] points out [here], this (vykhynata)

paribh exists.230

224
That is, reading the sthnin u as a samhra dvandva is no more convincing than the immediately
previous argument, which treats the sthnin as general but the dea as particular.
225
Amending vyaktireke to vyatireke. Manuscript A gives this reading (back of folio 2, line 2, starting
from the 33rd akara). This scribe is sometimes careless about compound consonants, but he does write kt
properly twice before on that page (vyaktire starting from the 13th akara on line 1, and yukta, the
word just previous to vyatireke. On the other hand, the Benares edition, as well as Abhyankars, gives
vyaktireke.
226
And so, not being a vowel, its sixth case singular form cannot be u.
227
hrasvo napusake prtipdikasya 1.2.47 [aca ca 1.2.28].
228
Having applied the stra, the prtipadika ends in and hence can be declined in the sixth case as u
(the more familiar form, a, being optional as it results from an augment). The first word now is well-
formed, the contradiction arises from marker t in the second word of 7.4.7, and so the savara rule must
indeed refer to the latter a.
229
That is, used as an anubandhahere, to facilitate the formation of pratyharas.
230
These statements attributed to Patajali are drawn from the end of his discussion under la: kim punar
varotsattv iva (aya?) akro dvir anubadhyate. etaj jpayaty cryo bhavaty e paribh
vykhynato vieapratipattir na hi sadehd alakaam iti. auditsavara parihya
prvegrahaa parea igrahaam iti vykhysyma. Sradeva paraphrases: ki punar aya

61
[p. 162, line 26]: But how do we have a sixth case compound here in the term vieapratipatti? Because

there must occur a prohibition by 2.2.14.231 And it is not the case that the compound can be made by

kdyoga-laka ah samasyate [2.2.8 v 1],232 since that prohibition, 2.2.14, is introduced precisely

within its [the vrttikas] scope.

[page 163, line 1] To this we say: The prohibition of a compound is taught, by 2.2.14, for just the sixth case

that occurs by means of the restrictive rule (niyama) ubhayaprptau karmai ca 2.3.66.233 And in this

varotsattv iva akra sadehahetur anubaddho pi punar anubadhyate yasmd anengrahaev


igrahaeu ca sadeha kim prvea parea veti. eva tarhi jpayaty cryo bhavaty e paribh
vykhynata itydi.
231
karmai ca 2.2.14 [samsa 2.1.3, saha sup 2.1.4, sup 2.1.9, tatpurua 2.1.22, ah 2.2.8, na
2.2.10]. While 2.3.65 and (under the condition ubhayaprptau ) 2.3.66 allow the sixth case for the karman
of the verbal action, 2.2.14 teaches that in this situationsixth case being used for the karmanthe two
words cannot be compounded. An overview of the context: 2.2.8 teaches sixth case compounds; its first
vrttika effectively imposes the condition that the second, principal word be made with a kt affix. 2.2.14
rules out sixth case compounds when the first word is the karman of the verbal action. (Of course this
sweeping exception must be narrowed in order to admit common compounds like fuel-cutter, or, more
importantly for our purposes, vieapratipatti.) 2.2.15 rules out sixth case compounds when the first
word has the sense of agent and the second word ends with tc or aka (capturing the idea of instigated
agent). 2.2.16 rules out sixth case compounds when the second word ends with tc or aka and has the
sense of agent.
232
ah 2.2.8 [samsa 2.1.3, saha sup 2.1.4, sup 2.1.9, vibh 2.1.11, tatpurua 2.1.22]. The first
vrttika on this rule, kdyog ca ah samasyata iti vaktavyam, provides only that a word taking the
sixth case by virtue of its relationship with a word ending in a kt affix (for example, pratipatti) can be
compounded with the latter. Its import is that words taking the genitive for another reason cannot be
compounded. The discussions of this vrttika in the Mahbhya and Kik both begin with the
sentence kdyog ca ah samasyata iti vaktavyam.
233
ubhayaprptau karmai ca 2.3.66 [anabhihite 2.3.1, ah ee 2.3.50, kti 2.3.65]. Again for context:
This stra restricts the application of the previous stra, kartkarmao kti 2.3.65 [anabhihite 2.3.1,
ah ee 2.3.50], which provides that sixth case can be used in the sense of agent or karman. Sradeva is

62
matter, the stra kartari ca 2.2.16234 is the means of valid knowledge. Otherwise235 for such examples as

creator of waters, eater of rice,236 kartari ca237 would be useless, because the prohibition is

accomplished just by karmai ca. But it is not the case, in [the phrases taught in] 1.4.53,238 the one who

causes taking on the part of the instigated agent [and] the one who causes doing on the part of the

instigated agent, that this [2.2.16] is for the sake of prohibiting sixth case compounds [when the first

word is] in the sense of an agent. Because that [prohibition] is accomplished by 2.2.15239 alone.

[p. 163, line 7] And moreover, what is stated in the Kik at 2.2.15, viz., that the mention of tc is for the

sake of the subsequent [stra]that presumes that that stra, kartari ca, exists.240 Some say: For the sake

of the subsequent is expressed in consideration of [a form] not ending in ic [that is, a noncausative

form]. And, the sixth case [produced for the karman] by the restriction 2.3.66 [that is, ubhayaprptau]

referring to the argument for a narrow interpretation of the prohibition in 2.2.14: not all karmans in the
sixth case are prevented from compounding, just those that take the sixth case by 2.3.66 because the agent
is also present.
234
kartari ca 2.2.16 [samsa 2.1.3, saha sup 2.1.4, sup 2.1.9, tatpurua 2.1.22, ah 2.2.8, na 2.2.10,
tjakbhym 2.2.15].
235
That is, if kartari ca 2.2.16 did not exist, or if it were not authoritative.
236
These examples, ap sra (with tc ) and odanasya bhojaka (with aka ), are given in the Kik
under 2.2.16. See discussion in Sharma, vol. III, pp. 83-84.
237
Recall that kartari ca continues the prohibition on ahtatpurua compoundswhen the second word
ends with tc or aka in the sense of agent.
238
hkor anyatarasym 1.4.53 [krake 1.4.23, karma 1.4.49, aikart sa au 1.4.52]. This stra provides
that an instigated agent can optionally be called karman in relation to the causative of the verbs h and k.
239
tjakbhym kartari 2.2.15 [samsa 2.1.3, saha sup 2.1.4, sup 2.1.9, tatpurua 2.1.22, ah 2.2.8,
na 2.2.10, karmai 2.2.14].
240
r Mnaarman in discussing this statement mentions (besides instigated agent) another construction
for which sixth case is used in the sense of agent: the Brhmans sleeping, brahmaasya ayanam.

63
would occur only where there is the explicit mention (updna) that is obtained by signifying power of

both agent and karman, on account of the continuation of understanding prayoga from 2.3.26.241

As for example, the milking of cows by one who is not a cowherd is astonishing.242 For in this way

astonishing is explainable, and thuswhen the explicit mention (updna) is obtained by the innate

signifying capacity of the karmans, cows that are difficult to milk, and of the agent who is not a well-

taught cowherdsixth occurs, not otherwise.

In vieapratipatti [p.163, line 12], because the agentthe student, etc.is not doing anything, an explicit

mention (updna) of the agent is not obtained by innate capacity. And this being so, third case for yena

in 1.4.28243 is not proper. Because 2.3.65 enjoins sixth with respect to the agent.244

[p. 163, line 15] In response to that, we say: Precisely on account of [Pinis] usage, third case is thus not

wrong. Someone else explains: [Even] when karman is merely understood (gamyamna), the term

241
ah hetuprayoge 2.3.26 [hetau 2.3.23]. Sradeva is claiming here (in order to discredit the
prvapakin) a continuation by anuvtti that is not recognized by the tradition; prayoge in fact continues
only to 2.3.27.
242
Note the syntax in the Sanskrit text: karman in sixth case, agent in third: caryo gav doho
goplakeneti. The activity, milking, is a word ending with a kt affix, namely gha, so the conditions of
2.3.65 are fulfilled; despite this, the agent is not in the sixth case, because of the restriction provided by
2.3.66.
243
antardhau yendaranam icchati 1.4.28 [krake 1.4.23, apdnam 1.4.24]: He by whom one wishes not
to be seen is termed apdna when hiding is connoted. (Translation adapted from Sharma, volume II, p.
238.) Here, Sradeva is drawing on the Adhy itself for examples of correct usage. In this stra, yena
refers to the agent of not-seeing. What is not seenthe karmanis not mentioned. (Note, however, that
the other verbal activity in the sentence, wishing, also has an implicit agent, possibly the same individual
that is the karman of not-seeing.)
244
Since the karman (although the object of not-seeing, and the agent of wishing) is not mentioned, in so
many words, in 1.4.28.

64
ubhayaprptau is indeed a restrictive rule (niyama).245 And in this way: Under 1.4.28 by the nysakt

Jinendrabuddhi, and under 2.3.66 by Maitreyarakita, the full implication of [Pinis] usagethrough

the absence of sixth in 1.4.28 [for the agent of not-seeing] due to the restrictive rule (niyama) [operating]

even when the karman is merely understood (gamyamna)246is stated.247

[p. 163, line 20] But thus, what is said in the Nysa248 with respect to tasya grahaam bhavatyekena, [in

the Kik] under aiu, viz. tasyeti kartari ah,249 is not correct. For there as well, at the point where

the karman to be grasped is being mentioned, the sixth case [on the part of the agent, in this case the

anubandha akra] must not occur on account of the restrictive rule (niyama) ubhayaprptau. And this is

traditionally explained (vykhyta) upon the introduction of grammar-teaching by Jinendrabuddhi.250

245
That is to say, the restriction will run even when the karman is merely understood, with the result that
agent cant be expressed in the sixth case.
246
gamyamna, being understood, is in contrast to updna, explicit mention.
247
In other words, the fact that the sixth case is not used in 1.4.28 means that the niyama must have
operated, and given also that the karman is understood, we must therefore not need explicit mention of
that karman in order for the niyama to run.
248
In fact, the Nysakt is just stating what another says.
249
That is, it is in sixth case expressing the agent. Here, as Sradeva notes, Jinendra is analyzing tasya
grahaam bhavatyekena, a phrase drawn from the Kiks explanation of the formulation of the
pratyhra a from the ivastras using the anubandha therein. (The antecedent of tasya is akra.)
Having already considered the possibility that tasyas sixth case expresses karman (with respect the
verbal action of grahaa) he proposes that it rather express the agent (while the karman is presumably
the letters in the middle, that is, between a and in aiu) and that ekenas third case expresses going-
with.
250
Grammar-teaching here is likely a reference to the term abdnusanam itselfa ahtatpurua
compound with its second word ending in a kt affix, rather than a reference to the Adhyy or any
other work on grammar. This term abdnusanam is analyzed by Jinendrabuddhi right at the
introduction of the whole system of grammar, just after discussing the Kiks three introductory
kriks. He concludes that the underlying sixth case can indeed express karman, 2.2.14 notwithstanding,
since the agent is not also mentionedas for idhmapravracana, a cutter of fuel (a hatchet). These are

65
And there what the subcommentator [Maitreyarakita] explained (vyjahra), out of excessive pride in

his own greatness (prgalbhya251):

[p. 163, line 23] Whenever there is explicit mention of both, obtained by innate signifying capacitythat

is the scope of the stra 2.3.66. Otherwise [to continue what Maitreya mistakenly says] there would be no

difference in scope of this rule 2.3.65 and [the scope] of a rule such as 2.3.66. And, on account of the

maxim there is no karman without an agent.252 That would be wrong, because [in fact] the restrictive

rule (niyama) does not result in a contradiction when the agent is explicitly mentioned (abhihita) in such

examples as Devadatta, destroyer of cities.253

[p. 163, line 27] In this matter we reply: Just as, not having accepted the understanding (grahaam) of

prayoga from 2.3.26 despite its continuing, 2.3.66 is claimed to be a restriction (niyama), when karman is

understood (gamyamna), by all of youjust so, not having accepted the anabhihite adhikra, the

absence of difference [in scope] is also claimed by all of us.254 And this is not correct.255 A prohibition,

when there is aka- or a-pratyaya (akkrayo),256 is to be stated; by that, where the restriction (niyama)

[2.3.66] is blocked [by its vrttika], it is because the purpose of the rule (vidhi) [2.3.65] has been fulfilled

both lyu, and (as pointed out in the Kik), in the opinion of some are optionally excluded from the
restriction 2.3.66. From the Kik: abdnm anusanam cryea cryasyeti v.
251
Recall Sradevas use of the word pragalbha in his introductory verse. Here, the implication is that the
pride is unwarranted.
252
With this sentence, Maitreyarakita is concluding that explicit mention of the agentbeyond the
existence of an agent implied by the existence of a verbal actionis what serves to differentiate the two
stras 2.3.65 and 2.3.66.
253
pur bhett devadatta.
254
Traditionally, prayoge is considered to continue only to 2.3.27 from 2.3.26. anabhihite is considered to
continue to 2.3.73. The argument may be: You reject prayoge, and say that 2.3.66 provides a restriction
so we reject not explicitly mentioned and say that it does not. (If the scope of 2.3.65 and 2.3.66 were the
same, 2.3.66 would present not a restriction but an option.)
255
Sradeva has used Maitreyarakitas reasoning to come up with a contradiction.
256
akkrayo str pratyayayo prayoge neti vaktyam, v. 1 on 2.3.66.

66
(and thus has no further purpose). The suns desire to make gleams of light.257And likewisein

Devadattas straw mat ought to be made,258 on account of the unwanted conclusion implied by the

sixth when the karman is explicitly specified [for then, Devadatta must not be the agent].

[p. 164, line 1] To this we say: Of this rule 2.3.65 and of 2.3.66, etc.these two are not co-referential sixth

cases, but rather have different referents. That means this: The rule kartari ca that is the relatum of 2.3.65,

and so also the rule karmai ca (that is the relatum) of 2.3.66, with respect to both of these, there should

be no splitting of the scope, this is the import. If even when the agent is merely being understood

(gamyamna), [then] given 2.3.66, there would be a restriction with respect to karmanat that point, for

examples such as idhmavracana,259 karmai ca would be a block [for sixth case], when lyu follows [the

word that governs the karman] in the sense of instrument.260 Even when agential lyu261 follows

257
This phrase, cikr raver loknm, which is marked as a problematic reading by Abhyankar, is a
little difficult to render in English with an objective genitive. cikr is an abstract feminine noun made
with pratyaya a from the desiderative stem according to 3.3.102. Compare the examples given here
(under the vrttika on 2.3.66, which blocks the stras restriction on sixth for the agent) in the Kik
cikr devadattasya kaasyaand in the Mahbhyacikr viumitrasya kaasya.
258
krya kao devadattasya. In this phrase it is not clear whether or not Devadatta himself is making the
mat. If so, the agent has been explicitly mentioned. In regard to technicalities, krya is made with the kt
suffix yat, which does not fit the conditions of the comment akkrayor strpratyayor prayoge
(pratiedho) neti vaktavyam under the vrttika on 2.3.66. By 3.3.113, it as a ktya has senses other than
those directly taught. (3.3.113 also applies to lyu.)
259
This example appears in the Mahbhya at 2.2.8, 2.2.14, and 2.3.22. The word for hatchet or saw
means fuel-cutting thing.
260
This is the case in which the agent is (merely) understood, and lyu is in sense of instrument. The
following sentence takes up the case in which the agent is explicitly mentioned, and lyu is in sense of
agent. In both cases, the sixth case is in fact assigned (and the compound is made), apparently in
agreement with the some who say, according to the Kik, that other kt affixes, not just feminine
ones, are optionally excepted from 2.3.66.
261
That is, analyzing kartlyu as kartari lyuthat is, in contrast to the karae lyu in the previous
sentence.

67
idhmavracana devadatta, Devadatta the fuel-cutterand in this [case] by the word Devadatta itself, on

account of his being understood (gamyamna) as the agent [of cutting], the occurrence of both is indeed

non-contradicted.

And thus, whenever sixth case is in the sense of karman, for all these [sixth cases], given that a compound

is prohibited by karmai ca, kartari ca ought not be introduced. And [what is] introduced: From that, it is

correct to say that there should not be a splitting of the scope. And not this: him himself, when this

agent is understood (gamyamna), from the absence of proof or authority.

[p. 164, line 10] Thus then, just as there is no restriction (niyama), on account of the agents being

specified (abhihita), in examples such as Devadatta, the destroyer of the cities, so also in grammar-

teaching [there is no restriction] even on account of nonusage (aprayoga) of the agent.262 In that regard,

although the restriction is stated when the agent is understood (gamyamna), nevertheless, having stated

the restriction, the splitting of the scope is stated, even when the agent is specified (abhihita).

And thus, there is no splitting of the scope precisely with respect to the two statements, of injunction

[2.3.65] and restriction [2.3.66]. And this is stated through not taking refuge in the injunctions purpose

being fulfilled upon the usage (prayoga) of aka and akra [in the sense of agent].

262
Sradeva is noting that the two examplesalthough they differ as to whether compounding can
occurboth admit the use of sixth case. The particular example of grammar-teaching is given, in the
Kik as well as the Mahbhya, as an example in which both agent and karman can take sixth, and the
compound is allowed: abdnm anusanam cryasya cryena v.

68
AppendixDerivations for Sradevas examples under paribh 1

to expandvh udyamane
vh + tumuN
vh + tum 1.3.2, 1.3.8
varh + tum 7.3.86, 1.1.51 pugantalaghpadhasya ca, ur a rapara
varh + tum 8.2.31 (taking ani option) ho ha
varh + dhum 8.2.40 jhaas tathor dho dha
varh + hum 8.4.41 un u
varh + hum 8.3.13 ho he lopa
Further analysis (if r were included in a) gets back to the same form:
va + hum 6.3.111
var + hum 6.1.87

poor little speechless oneg abde


g + KviP 3.3.108, v. 9 sapaddibhya kvip
na + g + KviP + ka 2.2.6; 5.3.70 (for example)
na + gir + ka 7.1.100, 1.1.51 ta iddhto, ur a rapara
a + gr + ka 8.2.76 rvor upadhy drgha ika
a + g + ka 8.3.15 kharavasnayor visarjanya
a + g + ka 8.3.39 ia a

poor little harmless onedhurv hisym


dhurv + KviP 3.2.177
na + dhurv + KviP + ka 2.2.6; 5.3.70 (for example)
a + dhur + ka 6.4.21 rllopa
a + dhr + ka 8.2.76 rvor upadhy drgha ika
a + dh + ka 8.3.15 kharavasnayor visarjanya
a + dh + ka 8.3.39 ia a

69
AppendixSradeva on arthavat, paribh 9

[p. 170, line 13] When there is mention of an item that is meaningful, that (mention) is not of something

meaningless. 9

When there is doubt between meaningful and meaningless, the understanding [should be] of the

meaningful word, not of the meaningless. Therefore, by the stra vano ra ca [4.1.7],263 since to [the

pratyayas] kvanip and vanip have the meaning of something agential, just for two meaningful items,

there is the substitution of ra in place of [the final of] van,264 [hence] dhvar, pvar.265 But not for the

[final of] the van that is meaningless: [we have] atiun, atiyn [which do not show substitution by ra].266

Likewise in the stras starting with in-han-pa-aryamm ,267 there is the mention of han, which,

ending in kvip, is meaningful. Therefore, even in plhnau, plhna,268 there is no setting aside of

lengthening by the restriction [to i].269

But how, in the examples starting with atiun, can it be said that the van word is meaningless, in as

much as there is meaningfulness merely by means of the meaning that is [its] svarpa? For example, as

263
vano ra ca 4.1.7 [pratyaya 3.1.1, para ca 3.1.2, prtipadikt 4.1.1, striym 4.1.3].
264
Or, the two vans.
265
The examples are the feminine forms of dhvan (artisan), pvan (stout person). kvanip and vanip are
taught in 3.2.74 (also see 94), anip in 3.2.103.
266
From atisvan and atiyuvan, for which 6.4.133 teaches the vocalization of v before non-taddhita affixes,
including feminine forms.
267
inhanpryamm au 6.4.12 [drha 6.3.111, agasya 6.4.1, upadhy 6.4.7, asambuddhau 6.4.8],
and sau ca 6.4.13 [drha 6.3.111, agasya 6.4.1, upadhy 6.4.7, asambuddhau 6.4.8,
inhanpryamm au 6.4.12].
268
Dual and plural nominative forms of masculine plhan, spleen.
269
Sharma explains (vol. V, p. 432) that there is in the Mahbhya a proposal to split this rule, for
otherwise it constitutes a restriction (of the lengthening operation) of sarvanmasthna endings (from
6.4.8) to just i endingswhich would block these forms.

70
stated in vddhir d aic [1.1.1]. Even before the relation of the name with the named, the word vddhi is

prtipadika on account of meaningfulness through the meaning that is [its] svarpa. And so they say

When some word is pronounced, and from it some meaning is recognized,

That, they say, is the meaning of the word; there is no other definition of meaning.

[p. 170, line 23] For that which has as its ultimate object svarpa is used; the object of it is just svarpa, on

account of its having that as its highest. And it is not the case, in examples such as atiun, that [items

such as] van, etc., which are [merely] aggregates of letters, are things that have as their highest object

svarpa. Because they [the aggregated letters] are pronounced [instead] for the sake of producing longer

words such as atiun etc. Just so, given the stra arthavad adhtur apratyaya prtipadikam [1.2.45], in

[a word such as] vanam or dhanam, the delimited part ending with n270 is meaningless. But when, in

atiun etc., van and so forth, which are made up of sounds, [also] have svarpa as their highest, only

then are they meaningful.

And in the stra to lopa ii ca [6.4.64],271 there is mention of both the i-gama, which does not have

sense, and of the i-pratyaya, which does have sense. In the stra io t [3.4.106],272 given that the letter a,

etc., which has the purpose of distinguishing,273 is to be done, and in the [stra] 6.4.64, the letter a of i

that is done, has this usageso the understanding [of a] is a general one.274 And this is explained

(vykhyta) in the Mahbhya at the stra 6.4.64 and the stra ita tmanepadn ere [3.4.79].

270
That is, the word-parts van and dhan.
271
to lopa ii ca 6.4.64 [agasya 6.4.1, rdhadhtuke 6.4.46, aci kiti 6.4.63]. Sradeva is undertaking a
comparison of the two i s.
272
io t 3.4.106 [lia 3.4.102].
273
According to Devasthali (p. 12): Mute in the dea i (3.4.78) is vieartha (in 3.4.106). That is, the
marker letter narrows the application of 3.4.106 to this particular ti ending. Devasthali refers the reader
to the Kik under 3.4.78 and 3.4.106.
274
3.4.106 provides for the first person singular forms of (tmanepada) li. The phrasing of the stra
admits either interpretation of i; the Kik mentions that thereforeciting this very paribhthe

71
Maitreyarakita said: But thus in 3.4.106, does the understanding of both augment and affix come about?

[No, only the affix;] in the same way, when another letter [] is to be done for the sake of the pratyhra

in ivahimahieven though the purpose of the letter could be for the sake of a it-operation. And

thus sasrasimahe (sras + li + i, we were falling down) when elision of the [penultimate] nasal

applies given the stra anidit hala upadhy kiti [6.4.24].275 Thus this is resolvable.276

[p. 171, line 4] Puruottamadeva however said, in the stra ddh ghvadp [1.1.20], there is no

understanding of the two words d and dh that are meaningless parts of the dhtus d (1U) and

dhvu (1U).277 Likewise, with respect to the stra e [1.1.13], [there is no understanding] of [e that is a]

meaningless part of a word, in ke and kue. Others do not allow this. Its like this:278 in the forms

(meaningful) pratyaya i is meant rather than the augment i. 6.4.64 shortens the vowel of dhtus ending
in , when certain ardhadhtuka affixes, including those beginning with augment i, follow. (The stra
appears in a section treating rdhadhtuka affixes and so is unlikely to refer to pratyaya i.)
275
The point is that the penultimate nasal is actually not elided, for the stra does not apply. While mahi
is indeed marked by , it is only for the purpose of making the pratyahra, not say to block gua and
vddhi by 1.1.5, nor (as in the present example) to elide the penultimate nasal.
276
Sradeva argues by analogy that just as in ivahimahi is clearly for the sake of the pratyhra, and
interpreting it otherwise would lead to an incorrect form, so too the is meant just to specify one
particular conjugational ending. He now returns to the question of which morphological items carry
meaning.
277
ghu comprises roots of the form d and dh, that is, including Ud (3U), Udh (3U), do (4P),
d (1P), de (1), and dhe (1P). As an aside, note that dP (2P) and daiP (1P) are not included;
according to Devasthali (p. 139), these two roots are marked with P precisely for this purpose. Here the
point is that although they start with d and dh, the roots d (1U) and dhvU (1U) are also not
included, as the d and dh portions of these dhtus are not meaningful. (There are several more such
roots that could have been mentioned but were not: dnA (1U, 10U), dy (1), dA (5P), and ds
(1U).)
278
In this sentence and the next, Sradeva points out that the mention of ghu in a stra listing verbal roots
does not imply that the stra should also apply to verbal roots that happen to contain the strings of letters
d- or dh-.

72
pranidati and pranidhvati, despite the presence of the name ghu [in the stra to be mentioned], there

is no substitution by [for the n in ni] by the stra nergada [8.4.17279] due to the absence of the name

upasarga for ni with [meaningless] words d and dh. 280


Likewise in the Mahbhya: there is no

association with a verbal activity with respect to these two [fragments] d and dh. And in the stra

beginning with sani mmghu [7.4.54281], on account of the mention of the ghu that is an aga, and

from not having s at the beginning,282 there is no fault associated with didiati 283 etc. Because the word

e in ke and kue is derivative, the name praghya is not [assigned]; [this] is explained (vykhyta) by

Jinendrabuddhi under e [1.1.13].284

[p. 171, line 12] And this paribh has the scope of the svarpa injunction, on account of the express

mention of arthavad in the definition of prtipadika (arthavadadhturapratyaya [1.2.45]). By that, after

the word pra (despite being meaningless) in pralambhate, by the stra avyaydpsupa [2.4.82285],

the luk-elision of sup is accomplished. The meaningless can be an avyaya, upasarga or prtipadika, on

account of the statement made in the Mahbhya [under 1.2.45, v. 12286], in place of the niptas that

are meaningless, the operations are done as on something meaningful. And the indicator, in [the rule]

adhipar anarthakau [1.4.93287], is the teaching of the name karmapravacanya for the purpose of stopping

279
ner-gada-nada-pata-pada-ghu-m-syati-hanti-yti-vti-drti-psti-vapati-vahati-myati-cinoti-
deghdiu 8.4.17 [sahitym 8.2.108, rabhy no na 8.4.1, akupvnumvyavye pi 8.4.2, upasargt
8.4.14].
280
The name upasarga is given only in association with verbal action: upasarg kriyayoge 1.4.59
[prdaya 1.4.58].
281
sani mmghurabhalabhaakapatapadm aca is 7.4.54 [agasya 6.4.1, si 7.4.46].
282
Since the desiderative pratyaya san has been augmented with i.
283
dr + san + tip. There is no substitution for the vowel of the reduplicated syllable (as there is for d +
san + tip = ditsati), since 7.4.54 does not apply.
284
e 1.1.13 [praghyam 1.1.11].
285
avyaydpsupa 2.4.82 [luk 2.4.58].
286
niptasya anarthakasya prtipadikatvam 1.2.45, v. 12.
287
adhipar anarthakau 1.4.93 [nipt 1.4.56, karmapravacany 1.4.83].

73
the name upasarga for the two adhi and pari, when meaningless, under the nipta-name adhikra. And

therefore from that, [the statement] niptasya anarthakasya prtipadikasaj vaktavythe

meaningless niptas are said to have the name prtipadikais rejected in the Mahbhya. Thus, if what

has the property of being avyaya, based on being a nipta, just on account of the signifying power of

that[then] the luk-elision of sup is accomplished. But that does not happen. Because the property of

being a nipta suitably occurs through just the means of accent. But the word pra is meaningless because

the meaning of eminence is well-known through context. Here, the first argumentabsence of

specification (particular number and case) on account of meaninglessnessis not accepted, from the

express mention of a prtipadika in 2.3.46,288 for this is its import: after only a prtipadika, despite being

meaningless, should the first [case] occur.

[p. 171, line 23] And this paribh is indicated (jpit) in the Mahbhya under sva rpam [1.1.68].

For there it is said: the express mention of rpa need not be made. Upon the svam of the word being

pronounced, although the form of the word and the meaning are possiblenevertheless, only form is to

be understood, not meaning. Because in pronouncing a word, there is the understanding of form at first;

then the understanding of meaning. Likewise, the word at the time of imitation lets go of meaning, not

form. Thus, because form has the property that it is permanently connected to word, the property that it

cant be let go, and the property that its understanding is close [closer than that of meaning], [from all

this] just the form of the word is to be understood, not the meaning, and not the other way around.

Consequently we understand: Just from the mention of svam [that is, given that both svam and rpam

are mentioned in this rule], when mention of form appears in the grammatical stra, meaning as well as

form is accepted.

[p. 171, line 29] Or rather, the indicator (jpaka) is the existence of the mention of bhrja [as well as rja],

in the stra vracdi [8.2.36].289 For if there were understanding of even the meaningless, then, upon the

288
prtipadikrthaligaparimavacanamtre prathama 2.3.46.
289
vracabhrajasjamjayajarjabhrjaccha a 8.2.36 [padasya 8.1.16, jhali 8.2.26, ante ca 8.2.29,
dhto 8.2.32].

74
express mention of even a part of bhrja, through just the mention of rja, [we would have] rja, bhrja

[and] the mention of bhrja need not be done. And so the indicator is made, by mention just of the

meaningful, not the nonmeaningful.

75
AppendixAbhyankars table

#20 11.35 is in #505, not here. #225 10.33 should be in #224; should have 11.117.
#31 8.93 should be in #40. #226 Should have 9.60.
#34 6.51 should be in #293.290 #227 1.90, 2.119, 3.91 should be in #228; should
#83 10.35 rather than 10.15. have 11.100; 12.60 is in #67, not here.
#88 4.28 is in #92, not here. #228 2.103 should be in #229.
#90 2.32 is in #210, not here. #229 12.21, 13.35, 14. 45, and 15.36 are in #69, not
#91 2.31 is in #256, not here. here.
#98 Should be 7.66 as well as 7.96. #232 9.31 should be in #234.
#106 There is no 4.96. #233 Should have 9.19.
#107 There is no 4.97. #235 Should have 9.64.
#120 5.25 is in #56, not here. #238 8.67 should be 8.37.
#136 12a.43 should be in #137. #239 Should have 10.76; 11.27 rather than 11.94.
#156 12.25 should be in #155. #243 Should have 9.68.
#165 2.138 rather than 1.138. #260 13.66 is in #442, not here.
#169 Should have 13.88, 14.139, and 15.91. #291 Should have 9.77.
#175 Should have 10.70. #312 4.24 should be in #313.
#198 Should have 3.21 rather than 3.19. #365 5.26 rather than 4.26; 6.14 rather than 5.14.
#200 Should have 2.80 rather than 2.82; 14.82 #377 10.19 is in #274, not here.
as well as (or instead of) 14.10. #401 17.49 rather than 17.59.
#204 Should have 8.73 rather than 8.63. #424 11.47 rather than 10.47.
#206 There is no 4.96; should have 9.89. #429 10.30 rather than 11.30.
#210 Should have 2.32 rather than 2.31; 3.31 #448 12.111 rather than 12.11.
rather than 3.32; 13.43 rather than13.44. #450 10.103 rather than 10.113.
#212 Should have 9.100. #470 6.37 is in #271, not here.
#213 Should have 9.59. #471 14.100 is not in the chart.
#214 Should have 9.52. #512 12.90 rather than 11.90.
#215 Should have 9.18. #513 Should have 2.71.
#220 Should have 9.46. #537 Should have 18.26.

290
Note that #293 includes #34.

76
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
1 akarmak hi api dhtava sopasarg sakarmak 8 2
2 aktakri khalvapi stram 111 2
3 aktavyh piny 65 84 69 88 104 50 49 61 51 56 48 11
4 akaravidhisthd api yogavibhgo garyam 93 77 83 4
5 agaty hi khalu paribhcryante 119 123 126 4
6 agavtte punarvttv avidhi 34 38 71 74 91 86 94 92 80 9
7 agdhikre tasya taduttarapadasya ca 85 77 79 4
8 aca parasmin prvavidhau 27 2
9 aco hrasvadrghaplut 25 2
10 ajnantarye na bahiragaparibh 46 2
11 auditsavaram eva gti na varamtram 66 1
12 atantra taranirdea 66 85 70 3
13 atulyavalayo spardho na bhavati 66 1
14 adarana lopa 51 53 66 3
15 addyanaddyor anadder eva 60 1
16 ananubandhakagrahae na snubandhakasya 48 2
17 anantarasya vidhir v bhavati pratiedho veti 16 19 20 30 18 15 19 27 53 133 17 55 48 51 50 61 52 17
18 anantyavikre 'ntyasadeasya 52 63 27 107 82 112 94 114 95 89 10
19 ananyaprv lde 55 1
20 anityam gamasanam 76 99 79 78 37 84 98 100 82 9
21 anityam tideikam 95 1
22 anityo ic curdnm 97 1
23 aninasmangrahany arthavat cnarthakena ca 129 3 4 76 77 17 65 15 67 16 16 12
24 anirdirth pratyay svrthe 90 125 114 127 113 105 6
25 anirdirth vidhaya svrthe ruhanti 29 1
26 anukaraaabd anukaryabdair arthavanta 13 2
27 anudttettvalakaam tmanepadam anityam 78 93-4 84 4
28 anuinuiayor anuiasyaiva 89 1
29 anekavara sarvasya 6 36 7 10 4
30 aneknt anubandh 4 4 4 3
31 antaraga bahiragt 93 42 2
32 antaragabahiragayor antaraga balavat 115 94 47 50 77 104 43 22 27 22 10
33 antaragavidhe pratipadavidhir balyn 89 116 90 3
34 antaragc cnavaka balya 79 105 54 3
35 antaragd apy apavdo balyn [yena nprpte 57 2
36 antaragn api vidhn bahirago lug bdhate 128 85 68 48 85 63 88 52 44 10
37 antaragn api vidhn bahirago lyab bdhate 86 79 54 47 34 62 34 54 46 10
38 antarage krye bahiragam asiddha syt 43 2
39 antarasth rephavarjit snunsik niranunsika 19 2

77
40 antybhve antyasadeasypi grahaam 56 39 6 42 93 28 15 7
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
41 anyakryt par dvirukti paroklidiu 47 1
42 anyrthe tadguasavijnam api 12 1
43 apavdaviaya parihtyotsarga pravartate 55 1
44 apavdaviaye 'py utsargasamvea 83 101 85 37 56 39 49 86 56 9
45 apavdo yady anyatra caritrthas tarhy antarage 65 55 3
46 apekto 'dhikra 48 12 2
47 aprptiprvik hi vibh taduddea dau 79 1
48 abdhakny api niptanni bhavanti 99 76 110 78 4
49 abhidhnapratyay sandisams 81 1
50 abhidhnalaka kttaddhitasams 72 90 76 117 4
51 abhidheyaval ligavacanni bhavanti 73 1
52 abhedak gu 57 80 107 83 109 101 6
53 abhysavikreu bdhyabdhakabhvo nsti 52 66 56 4
54 abhysavikrev apavd notsargnvidhn bdha 20 23 64 20 96 98 6
55 arthavadgrahae nnarthakasya 1 1 1 1 4 1 4 7 3 57 14 1 15 9 11 9 14 14 18
56 arthavad vibhaktiviparima 63 82 67 74 25 13 27 33 34 55 13 88 120 122 14
57 arthpatty stre siddhir uktiprmyt 84 1
58 ardhamtrlghavena putrotsava manyante vaiy 95 122 3
59 aligam asakhyam avyayam 98 1
60 alprayam antaragam 81 1
61 avayavaprasiddhe samudyaprasiddhir balyas 108 101 104 91 106 98 92 8
62 avayavalopinmato lopo nsti 88 114 89 43 4
63 avayave kta liga samudyam api viinii ce 107 2
64 avayavena vigraha samudya samsrtha 8 2
65 avayave 'py avayavivad upacra 40 1
66 avyakta pradhnagmi 86 1
67 avyaktaguasadehe napusakaliga prayujya 87 110 88 72 60 5
68 asati bdhe 1 1
69 asati sabhave bdhana bhavati 21 24 25 32 21 35 45 36 8
70 asiddha bahiragam antarage 36, 79 42 83 45 33 43 35 42 87 68 20 41 45 69 58 71 50 42 18
71 asti ca sabhavo yadubhaya syt 26 34 22 35 46 37 6
72 khytnkhytayor khyta balya 85 1
73 gamastram anityam 21 84 76 98 87 100 93-2 7
74 gam anudtt 105 1
75 gamt sarvdeo vidhir balavn 41 47 44 46 102 51 6
76 gamdeayor gamo vidhir balavn 40 46 43,78 45,94 4
77 gams tadgubhts tadgrahaena ghyante 7 14 32 15 22 63 67 18 4 4 11
78 tideikam anityam 95 93-6 2
79 tmanepadam anityam 94 1

78
80 did aca prvasya 32 1
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
81 daabhya sakhy sakhyeye vartate na sa 90 2
82 ded gama 101 50 2
83 dyantavad ekasmin 18 20 18 21 29 35 5 7
84 bhc chstrasya kvacit siddhat 69 1
85 ikrokti savibhaktir deo 'svara 58 38 61 3
86 iko guavddh 24 1
87 itaretara kryam asadvat 54 1
88 itaretarrayy api kryi streu prakalpante 24 28 28 96 3 3 3 7
89 iato 'dhikr pravttinivtt 43 50 2
90 iha stre vyavasthitavibh bhavati 32 1
91 ukra evaiko bhvyamno 'pi savarn ghti 31 1
92 uktrthnm aprayoga 51 60 53 28 46 22 49 62 77 28 10
93 ugitkrya varakrya ca tadantd api bhavati 56 122 3
94 uccaritapradhvasino 'nubandh 11 13 13 14 55 58 71 7
95 udayo 'vyutpannni prtipadikni 103 70 23 15 20 15 22 22 8
96 uttarapadatve cpaddividhau 44 1
97 uttarapaddhikre pratyayagrahae na tadantagra 26 25 25 4
98 utsargpavdayor apavdvidhir balavn 63 52 66,96 48 90 85 55 7
99 udit savara ghti na savaramtram 55 89 17 131 17 5
100 upajtanimitto 'pyutsarga upajaniyamanimitten 93 96 3
101 upapadavibhakte krakavibhaktir balyas 69 59 74 101 97 46 48 81 103 90 105 94 88 13
102 upasajaniyamanimitto 'py apavda upasajt 66 64 3
103 upasargo na vyavadhy 69 1
104 ubhayagatir iha bhavati 36 10 10 9 4
105 ubhayata raye nntdivat 44 51 46 30 51 60 120 62 8
106 ubhayatreha stre ah na prakalpyate 96 1
107 ubhayathpi vkyasampti 34 34 106 3
108 ubhayanirdee pacamnirdeo balyn 97 63 60 23 37 23 70 60 8
109 ubhayasthnanipanno 'nyataravyapadeabhk 106 2
110 ubhayor vibhyor madhye yo vidhi sa nitya 11 78 3
111 krokti savibhaktir deo 'svara 14 1
112 krpadia kryam krasypi 71 1
113 ko 'o ralau 20 1
114 tor vddhimad vidhv avayavebhya 4 3 2
115 ekadeaviktam ananyavat 16 17 15 1 16 1 3 11 37 7 14 38 40 125 42 37 37 17
116 ekadeenyatra na vyapadeo 'pi tatkta tadabhv 16 2
117 ekapadrayatvenntaragatvam 82 1
118 ekapadrayatvenntaragnapi jagdhydividhn 83 2
119 ekayoganirdin saha v pravtti saha v niv 9 49 24 18 14 19 14 17 17 10

79
120 ekayoganirdinm apy ekadenuvartanam 23 27 27 39 50 19 8 10 14 18 10
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
121 ekavaro vidhir ante pravartate 6 92 2
122 ekaabdasysakhytva kvacit 91 1
123 ekasminn api vyapadeivad bhvo 'prtipadikagra 61 78 2
124 ekasy ktecarita prayogo dvityasys ttyasy 118 107 3
125 eknubandhaktam anityam 45 55 36 3
126 eknubandhakagrahae na dvyanubandhakasya 46 52 47 5 22 82 33 37 51 53 10
127 eknt 5 5 5 3
128 ekritam antaragam 85 109 2
129 eca i hrasvatve 60 23 2
130 aupadeikapryogikayor aupadeikasya grahaam 5 5 6 42 45 59 4 81 135 84 120-2 11
131 kryakla sajparibham 50 58 51 2 112 106 3 117 118 119 3 3 12
132 kryanirdea kryanirmittapratipattaye 81 1
133 kryam anubhavan hi kry nimittatay nryate 11 10 3
134 krynubandhd adhikrnuvtti 41 1
135 kry nimitta kryam ity ea nirdeakrama 56 28 59 72 4
136 ki hi vacannna bhavati 139 1
137 ktktaprasagi nityam 117 76 82 53 78 43 6
138 ktenyasmin dhtupratyayakrye pacd vddhis t 82 2
139 ktrimktrimayo ktrime sapratyaya 6 6 7 72 23 7 9 101 88 103 9 11
140 ktrimktrimayor ubhayagati 7 1
141 kdabhihito bhvo dravyavat prakate 111 1
142 kdgrahae gatikrakaprvasypi grahaam 93 126 65 8 46 29 62 35 64 28 28 11
143 kdv 'sarupo 'striym 89 1
144 ktalyutumunkhalartheu vsarpavidhir nsti 76 58 36 54 38 68 58 8
145 kriyviean karmatva napusakaligat 54 134 56 4
146 kvacitktktaprasagamtrepi nityat 46 1
147 kvacit prvam 106 1
148 kvacit samudye vkyaparisampti 30 79 105 81 108 5
149 kvacit svrthik praktito ligavatanny ativartan 96 68 58 73 83 71 7
150 kvacid anuvartate 52 1
151 kvacid apavdaviaye 'py utsargo 'bhiniviata iti 115 52 33 44 33 58 7
152 kvacid arthavatopi 26 1
153 kvacid ubhayagati 73 24 2
154 kvacid ubhayath vkyaparisampti 82 1
155 kvacid ekadeo 'py anuvartate 25 8 10 8 18 5
156 kvacidbhvyamnopya savarn 20 1
157 kvacid vikti prakti ghti 93 41 132 43 120-1 5
158 kvacinndeasya 13 1
159 kvipi vyajanakryam anityam 95 1

80
160 kvibant dhtutva na jahati prtipadikatva ca 132 105 3
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
161 kvibartha praktir avha 136 1
162 kvau naa na sthnivat 47 1
163 khavidhir balavn 71 1
164 gaakryam anityam 29 31 109 77 118 92 120 93-3 83 9
165 gatikrakopapadn kdbhi saha samsavaca 138 88 47 62 32 42 32 75 64 10
166 gatyarth jnrth 101 1
167 gamakatve karmayoga 57 1
168 gamayaty adhikorthdhika kryam 58 1
169 garbhavabdayo bhavanti 116 88 139 91 4
170 gmdgrahaev aviea 124 7 77 69 87 105 103 107 106 98 10
171 gukrye nivtte punar na tannimittam 41 1
172 guabdhik vddhir bhavati 60 1
173 guapradhnabhvena praktistyaca tyrtha sa 94 2
174 gor adhikre tadantasya ca 63 1
175 gauamukhyayor mukhye kryasampratyaya 4 4 5 53 2 88 87,112 5 70 22 5 16 100 14 102 15 15 17
176 granthdhikyd arthdhikyam 88 1
177 grahaavat prtipadikena tadantavidhir nsti 89 73 11 75 78 32 55 71 57 31 31 12
178 it 14 1
179 itvena kitva bdhyate 129 1
180 cakro yasmtparastatsajtyameva samuccinoti 118 2
181 cakrakev iato vyavasth 139 1
182 cavikrev apavd utsargn na bdhante 75 1
183 cnuka nottaratra 59 76 62 40 22 76 51 119 87 65 46 57 48 78 67 15
184 cnukena na yathsakhyam 120 1
185 jpakajpit vidhayo hy anity 60 63 82 3
186 jpakasiddha na sarvatra 140 118 91 126 119 129 116 7
187 jpake 'rtham anityam 77 1
188 akitv dyantau 18 1
189 alayo samnaviayatva smaryate 107 1
190 icy anit 34 1
191 icsaniyoge eva curdnm adantat 99 1
192 iyako 43 1
193 nilopo 'py anitya 98 1
194 e 'py akta bhavati 44 1
195 au kta dvitve na sthnivat 127 1
196 au kta sthnivat 21 92 33 95 97 5
197 tadanubandhakagrahae ntadanubandhakasya 54 49 38 52 70 54 82 70 9
198 tadgams tadgubhts tadgrahaena ghyante 21 25 11 4
199 taddeas tadgrahaena ghyate 66 26 2

81
200 tadekadeabhtas tadgrahaena ghyate 62 80 8 10,82 14 18 6
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
201 taddhatyo bhvapratyaya spekdapi 87 1
202 taddhiteu nvayavrth 0
203 tadbhaktas tadgrahaena ghyate 17 20 2
204 tadbhede vidher bheda 57 29 60 73 4
205 tanmadhyapatitas tadgrahaena ghyate 18 21 22 8 15 64 66 19 72 75 81 77 89 77 14
206 tasmd apy anavaka yat 89 1
207 tasya doa sayogdilopalatvaatveu 111 1
208 tcchlik apavd atcchliknutsargstcchlik 57 2
209 tcchlike e 'ktni bhavanti 66 73 79 75 87 75 6
210 tcchlikeu vsarpavidhir nsti 27 32 31 28 57 43 53 45 67 9
211 tip apnubandhena nirdia yad gaena ca . e 18 2
212 tivkkrak 100 1
213 tyagrahae ckya 59 1
214 tyagrahae yasmtsa tadder grahaam 52 1
215 tytyasabhave tyasya grahaam 18 1
216 tydivanyonya nsarpotsargavidhi 111 1
217 dnuvidhichandasi . bhulya vat .. 68 1
218 dvandvtpara pratyekam abhisabadhyate 137 1
219 dvitvabahutvayor ekatva bahutvrtha svarp 81 2
220 dvitve kartavye au kta sthnivad bhavati 46 1
221 dvitve sati prvasya bdhako na bdhaka 81 1
222 dvirbaddha subaddha bhavati 92 136 98 51 93 5
223 dvirbhve prvavidhv apavd notsargn bdha 24 2
224 dvirvacaneci 33 1
225 dvau naau praktam artha gamayata 117 1
226 dyvityadhikre tyagrahaa svarpagrahaam e 60 2
227 dhtavo 'nekrth 100 1
228 dhtuprtipadikd anubandhaprtipadika balya 90 119 91 4
229 dhtusvart pratyayasvaro balyn 103 1
230 dhtpasargayor antaraga krya bhavati 37 1
231 dhto kryam ucyamna tatpratyaye bhavati 74 76 88 76 5
232 dhto svarpagrahae tatpratyaye kryavijnam 75 26 55 80 5
233 dhugrahae dhvde samudyasya grahaam 19 2
234 dho svarpagrahae tattvavijnam 31 1
235 dhor adhikre tadantavidhir apy asti 64 1
236 na keval prakti prayoktavy 135 1
237 naivayuktam anyasaddhikarae tath hy arth 54 65 58 65 49 51 64 97 97 31 41 31 74 63 14
238 naghaitam anityam 37 93-5 85 3
239 nayuktasya tatsade 76 27 2

82
240 na nimittakrya naimittakasya 19 22 23 3
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
241 na padntadvirvacanavargntnusvraprathamat 10 40 11 16 28 6
242 na prtipadikena 9 1
243 na yoge yogo 'siddho 'pi tu prakarae prakaraam 68 2
244 na lakantarabdhanennityatvam 81 1
245 na lumat prakte 42 1
246 na varagrahaeu 2 2 2 2 10 12 10 7
247 na varninsmangrahaeu 58 1
248 na varrayapratyaytvapugatvehrasvni varv 66 2
249 na vikaraasvaro lasrvadhtukasvara bdhate 84 102 2
250 na tibanubandhagaasakhyaikasvarokteu 13 21 14 20 62 6
251 na sajvidhau (tadantavidhi) 6 1
252 na smnya abdokte 17 1
253 na strpratyaye cnupasarjane 26 1
254 na svarnantarye (bahiragam siddha) 21 1
255 na hi kry nimittatvenrayate 95 2 2 2 10 5
256 na hy ekade varagrahaeu ghyante 31 1
257 njnantarye bahivaprakpti 80 44 84 69 42 46 70 59 72 51 43 11
258 nnarthake 'lo 'ntyavidhir anabhysavikre 62 55 38 93 83 113 95 115 96 90 10
259 nnirth strapravtti 107 57 2
260 nnubandhakta halantatvam 35 1
261 nnubandhaktam anekltvam 12 14 14 75 11 6 6 6 6 6 7 11
262 nnubandhaktam anejantatvam 15 13 7 45 7 47 7 8 8
263 nnubandhaktam asrpyam 13 15 16 12 8 38 8 40 8 6 10
264 nnubandhaktny asrpynejantatvnekltvni 83 34 3
265 nnvcyamnanivttau pradhnasya 80 31 2
266 nmagrahae pryeopsargasya na grahaam 126 2
267 nmagrahae ligaviiasypi 3 16 2
268 nmavidher antaragt sarvato lopa 48 1
269 nmn vyutpattir avyavasthit 102 1
270 nsvaabdoktatvt 2 2 4 3
271 nitya sandhyakari drghni 37 1
272 nityd antaraga balya 93 52,117 80 53 4
273 nitynityayor nityo vidhir balavn 118 93 49 83 54 5
274 nityo 'vayabhv bhavati 41 1
275 nimittpye naimittikasypy apya 40 47 42 63 27 23 29 34 78 29 113 99 64 101 14
276 nimittikrya na nimittasya 29 94 2
277 niranubandhakagrahae na snubandhakasya 47 53 6 48 23 81 32 36 67 50 69 52 81 69 14
278 niranubandhagrahae smnyagrahaam 45 61 62 3
279 niravak hi vidhaya svakn vidhn bdhant 11 50 94 43 5
nirdiyamnasyde bhavanti

83
280 106 19 7 33 8 11 24 21 35 13 5 5 5 12 12 15
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
281 niedh ca balysa 113 112 104 3
282 numaciratjvad bhvebhyo nu prvavipratiedhen 108 2
283 naikamudharaa yogrambha prayojayati 133 2
284 naisargikalkaikayor naisargikasya eva graha 8 2
285 nyy stavirayaipry 140 1
286 pacamy nirdie parasya 22 24 31 12 4
287 padagauravd yogavibhgo garyn 121 1
288 padayo sadhir vivakito na samsntaragayo 62 65 84 4
289 padgdhikre tasya ca tadantasya ca 125 30 36, 108 29 29 5
290 paddhikre 'hanndibhi 6 1
291 paranityavicrae bhaven nityam 77 1
292 paranityntaragpavdnm uttarottara balya 38 2
293 paranimittdea prvasmin sa eva 44 42 47 18 4
294 parasyde 16 1
295 pardantaraga balya 130 1
296 parn nitya nityd antaraga antaragc cna 51 2
297 parn nitya balavat 51 52 42 3
298 parrthe prayujyamna abdo vatamantarepi v 116 2
299 parimagrahaena klaparima ghyate 9 1
300 parihtypavdaviayamutsargo 'bhiniviate 94 43 2
301 paroktam avirodhi cet 91 1
302 parjanyaval lakaapravtti 64 83 68 110 134 86 88 87 111 90 111 103 12
303 paryyaabdn gurulghavacarc ndriyate 122 116 124 115 5
304 pribhpika kvacid anitya 58 1
305 pitsvarccitsvaro balyn 108 130 110 3
306 punaprasagavijnt siddham 82 41 82 61 74 105 40 41 21 26 21 39 40 13
307 purastd apavd anantarn vidhn bdhante not 9 9 10 35 92 87 36 8 53 44 49 46 59 50 14
308 prva dhtur upasargena yujyate pact sdhan 98 129 28 4
309 prva dhtu sdhanena yujyate pacd upasa 128 29 131 4
310 prva prvottarapadayo krya krya pac 83 2
311 prva hy apavd abhiniviante pacd utsarg 56 62 53 4
312 prvatrsiddhe csayogdilopalatvaatveu 31 2
313 prvatrsiddhyam advitve 24 70 110 92 97 126 99 117 106 9
314 prvatrsiddhye na sthnivat 48 112 92 128 95 110 6
315 prvatrsiddhe nsti spardho 'sattvd uttarasya 67 2
316 prvaparanityntaragpavdnm uttarottara b 39 38 3
317 prvottarapadanimittakryt prvam antarage 'py 53 45 3
318 prvottarapadayo kryam antaragam apy ekd 84 2
319 prakalpya vpavdaviaya tata utsargo 'bhinivi 91 63 54 4

84
320 praktisabandhc chrutasabandho garym 97 2
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
321 praktigrahae yadhikasypi grahaam 70 79 73 90 85 93 91 79 8
322 praktigrahae yalugantasypi grahaam 60 77 64 4 12 20 13 19 20 61 17 72 80 68 76 70 93-9 86 18
323 praktigrahae ligaviiasypi grahaam 63 1
324 praktigrahae svrthikapratyayntnam api grah 58 2
325 praktipravtte punar vttv avidhir nihitasya 59 108 3
326 praktivad anukaraa bhavati 86 113 87 12 36 6 114 37 18 23 18 36 36 13
327 prakte prva prva syd antaragam 71 98 2
328 prakter antpahr lopa 56 69 2
329 praktyadhikre mahaddibhi 5 1
330 praktyritam antaragam 72 99 2
331 pratijnunsikyo piny 121 1
332 pratijsvarit piny 122 1
333 pratipadavidhnd yogavibhgo garyn 74 92 98 110 138 112 6
334 pratyaknumitayo pratyako vidhir balavn 95 2
335 pratyaya para 118 1
336 pratyayagrahae cpacamy 25 32 24 24 4
337 pratyayagrahae yasmt sa vihitas taddes tadan 7 44 24 25 31 25 23 23 9
338 pratyayalope pratyayalakaam 52 26 54 67 41 131 6
339 pratyayasya sarvpahr lopa 25 55 68 3
340 pratyaypratyayayo pratyayasya grahaam 72 59 92 114 98 116 102 94 8
341 pratyayrita bahiragam 73 100 2
342 pratysatte praktivad anukaraam 23 1
343 pratyeka vkyaparisampti 28 33 32 25 114 29 78 104 80 107 99 11
344 prathamay ca 24 26 2
345 pradhnaaktyabhidhne guaaktir abhihitavat p 62 2
346 pradhnasya spekasypi vtti 101 86 2
347 pradhnnvcayaiayo pradhnaie saprat 67 86 71 38 10 40 58 7
348 pradhnpradhnayo pradhne kryasampratya 120 102 89 104 97 91 7
349 prasiddhe sahacarite ca 71 1
350 prkaraikprkaraikayo prkaraikasyaiva 61 2
351 prgvata sakhyprvapadn tadantagraha 58 2
352 prtipadikagrahae ligaviiasypi grahaam 25 29 29 60 26 61 28 38 28 71 61 11
353 prtipadikavidhau tadantavidhir nsti 69 74 2
354 prdn kriyyogitva vieadyotitve svabhvt 80 2
355 balavan nityam anityt 92 41 2
356 bahuvrhau tadguasavijnam api 9 6 64 42 44 77 66 7
357 bahuu ca ita 56 1
358 bahvrita bahiragam 86 110 2
359 bdhakny eva niptanni 109 110 102 3

85
360 bhavati ca bahuvrhau tadguasavijnam iti 8 7 56 3
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
361 bhavati vykhynato vieapratipattir na sadeh 120 2
362 bhavati strtyaprakarae tadantavidhi 62 1
363 bhavati hi kriyviean karmatva napusa 59 2
364 bhvbhvtmako hi vikalpa 126 1
365 bhvini bhtavad upacra 26 14 28 65 33 39 9 7
366 bhvyamnena savarn grahaa na 20 19 19 3
367 bhvyamno ' savarn na ghti 30 35 31 11 16 11 6
368 bhvyamno 'py ukra savarn ghti 44 21 20 3
369 bhinnasabandhe pratiyoge pratighte cdhikr 42 2
370 bhtaprvakastadvad upacra 50 52 38 8 4
371 bhtaprvagatir iha stre sabhavati 14 17 18 82 4
372 matubvibhaktyudttatva prvanightasya 67 1
373 madhye 'pavd prvn vidhn bdhante nottar 10 11 36 93 88 37 9 54 47 50 49 60 51 14
374 midacontytpara 19 1
375 mdamdor ekdeo mdvad bhavati 17 1
376 mdgrahae na tadantavidhi 54 1
377 mdgrahae ligaviisypi grahaam 8 1
378 ya dv yasya sabhava sa tasya sanipta 33 2
379 ya vidhi pratyupadeo 'narthaka sa vidhir bd 48 56 50 37 47 50 63 90 38 9
380 yata evakras tatonyatrvadhraam 131 1
381 yatra dvau prasagv anyrthv ekasmin viaye y 68 96 3
382 yatkriyyuktsta prati gitisajako bhavati 99 1
383 yatrnekavidham ntarya tatra sthnata ntarya 14 13 13 4
384 yatrnyatkriypada na ruyate tatrstirbhavantp 122 2
385 yatropasargatva na sabhavati tatropasargaab 109 2
386 yathsakhyam anudea samnm 15 34 16 23 45 10 6
387 yathoddea sajparibham 59 52 113 105 2 116 117 118 2 2 10
388 yadgams tadgubhts tadgrahaena ghyant 12 4 11 4
389 yadupdher vibh tadupdhe pratiedha 123 2
390 yallakanupapanna tatsarva niptant siddh 31 2
391 yasmin vidhis taddv algrahae 127 9 10 75 34 12 17 12 33 33 10
392 yasya ca lakantarea nimitta vihanyate tad a 48 2
393 yasya ca lakantarea nimitta vihanyate na ta 79 47 3
394 yasya yenbhisabandho drastha sypi tena sa 124 2
395 yasya vidher nimitta nsau bdhyate 49 57 38 48 50 63 90 39 8
396 yasya sthne ya deste tadgrahaena ghyante 91 115 3
397 ydgjtyasya vipratiedho vidhir api tdgjtyasy 90 114 3
398 yvatsabhavas tvadvidhi 54 27 57 70 109 113 6
399 yugapadadhikaraavacano dvandva 127 130 2

86
400 yena dhtun yukt prdayasta pratyevopsarg 108 2
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
401 yena nprpte yo vidhir rabhyate sa tasya bdha 42 49 44 31 91 40 51 48 57 49 10
402 yena nvyavadhna tena vyavahite 'pi vacanap 39 46 41 32 70 5
403 yena nirdia tasyaiva 75 1
404 yena vidhis tadantasya 3 7 3 5 2 5
405 yena vin yanna bhavati tattasynimittasypi nim 125 2
406 yogavibhgd iasiddhi 68 87 72 55 100 121 115 123 114 9
407 yo yam rityotpadyate sa ta prati sanipta 12 12 3
408 yo hy andid aca prvastasya vidhi prati sth 11 12 17 34 67 124 69 8
409 yvd yvdraya ca 45 1
410 rhiu nvayavrth prakante 112 1
411 lakaapratipadoktayo pratipadoktasyaiva graha 3 3 4 2 75 102 6 59 15 3 86 39 102 41 105 97 16
412 lakaavidhau likhanaprakramd vidhi 74 1
413 lakantarea prpnuvan vidhir anitya 78 80 45 3
414 lakye lakaa sakdeva pravartate 89 112 2
415 laghvakara hi sajvidhnam anityam 135 1
416 ldeeu vsarpavidhir nsti 59 69 59 3
417 ligagrahae ligaviiasypi grahaam 17 3 18 25 4
418 ligam aiya lokrayatvt 73 105 2
419 ligavacanam atantram 40 1
420 lugvikaralugvikaraayor alugvikaraasyaiva gra 43 50 67 73 83 83 86 90 78 9
421 luptonubandha krya prati vieako bhavati 134 2
422 luptopyanubandha sdhyasiddhivat sadvat 24 2
423 lugantaragebhya 98 1
424 lubantaragebhya 47 1
425 lopasvardeayo svardeo vidhir balavn 35 45 37 44 4
426 lopd ajdea 100 49 2
427 vaktur vivakitaprvik hi abdrthapratipatti 61 64 83 3
428 vatsapravttyeha stre kryi sabhavanti 19 1
429 varakvilukiravaropadhtveu 30 1
430 varagrahae jtigrahaam 64 1
431 varagrahae nimittatvt 32 34 41 3
432 varagrahae savarasypi grahaam 4 1
433 varntasya vidhi 5 5 5 9 4
434 varraye nsti pratyayalakaam 96 45 52 22 7 9 7 21 21 9
435 varraye vidhau nntdivattvam 61 123 63 3
436 varaikade varagrahaena ghyante 26 30 21 65 4
437 vartsyatpravttyeha kryi kriyante 56 16 84 140 87 5
438 vasupravty iha stre krya bhavati 15 18 2
439 vgvibhakte krakavibhaktir balyas 104 1

87
440 vcyavalligavacane 71 1
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
441 vrt prkta balya 81 107 72 95 44 5
442 vrd ga balyo bhavati 39 74 49 66 60 68 55 47 8
443 vro vidhir antaragam 80 106 2
444 vikaraebhyo niyamo balyn 53 42 19 24 19 41 41 7
445 vicitr abdaaktaya 138 1
446 vidhiniyamasambhave vidhir eva jyyn 130 84 108 132 84 109 96 111 100 9
447 vidhipratiedhayo pratiedho vidhir balavn 94 118 2
448 vidhividhnavidhibhj tray sanidhne ta 111 13 18 13 5
449 vidhau paribhopatihate nnuvde 89 84 92 93-10 87 5
450 vipratiedhe para kryam 69 97 103 3
451 vidhipratiedhena yad bdhita tad bdhitam eva 60 2
452 vibhaktau ligaviigrahaam 30 27 29 39 29 72 6
453 vivakta krakvasya 68 46 11 3
454 vivakta samsapravtti 0
455 vivakvypter ivasya 67 1
456 vieaniiddhn smnyaniedhavattvam 8 2
457 vieavidhi smnyt 52 1
458 vietidia prakta na bdhate 19 54 20 28 54 5
459 vieyavalligavacanni bhavanti vieanm 106 2
460 vyapadeivad ekasmin 17 14 31 30 30 5
461 vyapadeivadbhvo 'prtipadikena 65 79 33 56 72 58 32 32 8
462 vyavadhnepy uktiprmyt 62 1
463 vyavasthitavibhaypi kryi kriyante 36 35 55 62 124 93 125 99 8
464 vykhynato vieapratipattir na hi sandehdalak 58 75 61 66 65 67 88 116 121 120 1 1 1 1 1 1 16
465 ap tipnubandhena nirdia yad gaena ca . y 40 2
466 abdntarasya prpnuvan vidhir anityo bhavati 77 100 54 44 43 5
467 abdntart prpnuvata abdntare prpnuvata 44 2
468 idanekl sarvasya 17 1
469 ldiu vydayo na bhavanti 76 110 2
470 uddhadhtnm aktri rpam 104 1
471 / yatraikjgrahaa caiva pacaitni na yaluki 94 120-3 109 3
472 rutnumitayo rutasambandho balavn 49 92 116 78 102 107 100 109 104 96 10
473 lugvikaralugvikaraayor alugvikaraa 45 1
474 ahy nirdie svasya 23 25 32 3
475 ahyntasya 13 1
476 saj na sajntarabdhik 84 1
477 sajprvako vidhir anitya 53 64 57 30 32 38 83 75 71 78 73 93-1 81 13
478 sajvidhau pratyayagrahae tadantavidhir nst 71 75 53 27 27 5
479 sajvidhau pratyayagrahae tadantagrahaa 81 28 34 27 27 6

88
480 sajottarapaddhikre pratyayagrahae pratyay 74 2
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
481 sadehe bahuvacana prayoktavyam 61 119 129 121 4
482 sadhysakarm idutau hrasvdee 43 46 95 3
483 saniptalakao vidhir animitta tadvightasya 10 12 12 13 31 12 33 40 39 65 19 10 70 64 75 66 85 73 18
484 saniyogain saha v pravtti saha v niv 14 2
485 saniyogainm anyatarpya ubhayor apy a 41 48 43 64 42 79 30 65 71 72 77 74 86 74 14
486 samprasraa tadraya ca krya balavat 96 103 59 133 61 119 108 8
487 sabhave vyabhicre ca vieaam arthavat 114 1
488 sakrpadia krya taddeasya akrasya 72 2
489 sakduktrthasya prayoga 61 1
490 sakdgatau vipratiedhe yad bdhita tad bdhita35, 81 40 38 60 36 57 38 57 73 104 80 39 40 20 25 20 40 39 18
491 satiiasvaro balyn anyatra vikaraebhya 104 1
492 satyapi sabhave tadubhaya syt 26 1
493 satyapi sabhave bdhana bhavati 22 25 33 23 86 47 89 7
494 satyavidhau na tadantavidhi 55 1
495 sada karaa anukaraa 109 1
496 saptamy nirdie prvasya 21 23 30 11 4
497 samartha padavidhi 56 1
498 samnrayamasiddha vyraya siddha bha 38 45 2
499 samsakttaddhiteu sabandhbhidhnam anya 130 132 3
500 samsataddhitn vttirvikalpena, vttiviaye ca 89 2
501 samsapratyayavidhau tadantapratiedha ugidva 80 57 121 59 5
502 samsavttistaddhitavtty bdhyate 107 1
503 samsa ca dvipadrayatvd bahiragopi lyapa 35 2
504 samsntavidhir anitya 75 94 78 39 82 69 63 74 65 84 72 11
505 samsntgamasajjpakagaananirdini 84 35 3
506 samudye 'pi vkyaparisampti 29 33 26 115 100 5
507 samudyeu abd pravtt avayavev api varta 85 86 2
508 sarva vkya svadhraa 115 1
509 sarvatra tadantavidhir anyatra chatuksvarasadh 10 2
510 sarvatrpi vieea smnya bdhyate natu sm 128 2
511 sarvanmno vttimtre puvadbhva 103 1
512 sarvavidhibhya ividhir balavn 56 69 59 90 79 82 136 85 93-8 9
513 sarvavidhibhyo lugvidhir balavn 71 1
514 sarvavidhibhyo lopavidhir balavn 57 70 60 34 44 36 43 91 79 111 137 113 93-7 13
515 sarvpahr lopa pratyayasya 53 1
516 sarvebhya lopa 99 48 2
517 sarve vidhaya chandasi vikalpyante 56 35 24 30 24 35 35 7
518 sarvo dvandvo vibhayaikavad bhavati 73 91 50 36 16 22 16 34 34 9
519 savieae vidhiniedhau vieaam upasakrm 2 2

89
520 sahacaritapratysatte taditaretarayoga 51 1
# rule 1 2 3 4 5 6 7 8 9 10 11 12 12a 13 14 15 16 17 total
521 sahacaritsahacaritayo sahacaritasyaiva graha 67 39 49 115 99 117 103 95 9
522 smpratikbhve bhtaprvagati 15 63 37 55 39 76 65 7
523 spekam asamartham 27 85 2
524 spekasypi naa savidhir bhavati 102 1
525 smnyavidhir utsargo vieavidhir apavda 53 87 111 4
526 smnytidee vienatidea 91 123 92 127 64 85 106 97 108 101 93 11
527 smnyraya siddha vyraya tvasiddam e 40 2
528 svakaniravakayor anavako vidhir balavn 95 76 103 4
529 shacaryt sadasyaiva 63 1
530 siddhaabdopaden netaretarrayadoa 85 1
531 siddhe satyrambho niyamya 59 30 62 77 1 74 25 7
532 subasupor ekdea subvad bhavati 16 1
533 susarvrdhadikabdebhyo janapadasya 57 3 2 3
534 stre ligavacanam atantram 70 61 80 47 117 30 30 73 62 9
535 strkhalan alo bdhak striy khalanau 112 1
536 strtyagrahanneyate 61 1
537 strpratyaye cnupasarjane na 45 27 33 26 26 26 6
538 sthnivad deo hy avaravidhau 20 8 39 9 12 26 6
539 sthnivadbhvapuvadbhvaikaeadvandvaikat 96 2
540 sthnentaratama 22 16 35 17 24 22 6
541 sthitnm anvkhyna hi vykaraam 89 113 2
542 sphdikhalatvaatveu nsti 49 1
543 sva rpa abdasyabdasaj 28 30 35 1 1 5
544 svagurpalakaokta ca 90 1
545 svaradrghayolopeu lopa 29 1
546 svaribhinnasya prpnuvan vidhir anityo bhavati 49 2
547 svaravidhau vyajanam avidyamnavat 33 37 37 86 32 66 53 67 55 79 10
548 svarasya hrasvadrghaplut 4 1
549 svardea paranimittaka prvavidhi prati sth 9 41 10 15 5
550 svgam avyavadhyi 68 1
551 svrthik praktiligasakhye anuvartante 95 1
552 svrthik pratyay praktito ligavacanny ativa 74 96 58 58 73 60 7
553 svrthe ijnityam 79 1
554 halsvaraprptau vyajanam avidyamnavat 68 80 68 3
555 hrasvadrghpadia krya na plutasya 73 1



90

AppendixManuscript sources

Oxfords Bodleian Library holds two manuscripts of Sradevas Paribhvtti, both in the Chandra Shum
Shere collection. In his handlist for this collection, Wujastyk includes a third manuscript of paribhs
alonethat is, with no commentaryas following Sradevas order. This manuscripts order does
closely resemble that of Sradeva, but also clearly shows the influence of Haribhskara, merely by the
presence of the rule samsa ca dvipadrayatvd bahirago pi lyapa bdhate (numbered 35 in the
manuscript, Abhyankars #503), which is unique to Haribhskaras collection.291

None of the manuscripts is dated. In the discussion that follows, the paribh numbers are given as by
Abhyankar.292 The paribhs in B are not numbered; a space between two sets of double daas
indicates the transition from the commentary on one paribh to that on the next.

A Oxford. Bodleian Library MS Chandra Shum Shere d.236 (2).293 91 folios, numbered 1-91. Size
varies, up to 27.5 cm x 10.9 cm. 8-9 lines. About 45 akaras. Paper. Incomplete.

The paper for the folios first and last folios is different than for the rest of the manuscript: lighter in color,
smoother, and slightly larger. The hand may also be different.

The first folio starts with the invocation given in the Abhyankar and Dube editions.294 There is a difficult-
to-decipher line of text in the top margin. The next to last folio ends near the end of the commentary on
paribh 107 (yath brhmaebhyo dadhi dya-).295

291
Abhyankar 1967, p. 31, notes that the only way in which Haribhskaras collection of rules differs from
Sradevas, other than this rule, is the separate numbering of two rules that Sradeva numbers together.
The ordering and numbering in this manuscript varies somewhat from that given by Abhyankar for
Sradevas and Haribhskaras collections of rules.
292
Wujastyk uses Dubes numbers to refer to individual paribhs. However, the numbering in
manuscript A matches Abhyankars. Also, the numbering in the Dube edition is anomalous in places.
293
Wujastyk 1984, No. 191 on p. 34.
294
A few of the akaras differ from the verse as edited by Abhyankar and Dube. For example, abhinava in
pda a appears as atinava.

91
The last folio begins about halfway through the commentary of paribh 128 (dya
tvadadaam).296 That folio continues through the end of the commentary on paribh 129; the
commentary on paribh 130 is not given.

Colophon: iti grman mahmahopdhyyarsradevakt paribhvtti sampta // ubham astu297


//

The commentary on paribh 1, vykhynata, begins with the first folio and runs through the end of
line 2 on the back of folio 4. The commentary on paribh 9, arthavat, begins on the last line of the front
of folio 10 and runs through the bottom line of the front of folio 12.

B Oxford. Bodleian Library MS Chandra Shum Shere d.253 (1,i).298 49 folios, numbered 41-91.299 24.3
cm x 7.8 cm. 8-15 lines. About 43-52 akaras. Paper. Incomplete.

The first folio begins in the middle of the commentary on paribh 32 (-de akandhvditvt samsas
tena sdhutvam.)300 As Wujastyk notes, there are two folios missing after folio number 54; the following
folio is numbered 57. The last folio ends about a third of the way through the commentary on paribh
115 (laukika ev-).301

C Oxford. Bodleian Library MS Chandra Shum Shere d.245 (5).302 5 folios, numbered 1-5. 24.9 cm x
13.2 cm. 10 lines. 25-30 akaras. Paper. Complete.

Colophon: iti paribhprakaraam / ubham astu // <candra bindu> // rrmya nama //

295
Abhyankar 1967, p. 262, line 15.
296
Abhyankar 1967, p. 271, four lines from the bottom.
297
This is a best guess; the last few akaras are not easily decipherable.
298
Wujastyk 1984, No. 192 on p. 34.
299
The first twelve folios also bear what look like earlier numbers 25-36, in some cases crossed out. Folios
numbered 61-77 also bear small numbers 1-12; likewise, folios numbered 85-91 bear small numbers 1-8.
300
Abhyankar 1967, p. 198, five lines from bottom.
301
Abhyankar 1967, p. 266, line 21.
302
Wujastyk 1984, No. 14 on p. 3.

92
References

Abhyankar, Kashinath Vasudev. 1967. Paribhsagragha: A collection of original works on Vykaraa


Paribhs. Poona: Bhandarkar Oriental Research Institute.

Abhyankar, Kashinath Vasudev, and Jayadev Mohanlal Shukla, ed., trans. 1975. Patajalis Vykaraa-
Mahbhya, hnikas 1-3 with English Translation and Notes, with J. M. Shukla. Poona: Bhandarkar
Oriental Research Institute.

Abhyankar, Kashinath Vasudev. 1986. A Dictionary of Sanskrit Grammar. Third edition, with J. M.
Shukla. Gaekwads Oriental Series, No. 134. Maharaja Sayajirao University of Baroda Press.

Abhyankar, Kashinath Vasudev, ed. 1962. Paribhenduekhara of Ngojibhaa, with the commentary
Tattvdara of Vasudev Shastri Abhyankar, part 1. Poona: Bhandarkar Oriental Research Institute.

Abhyakara, Vsudevastr. 1938-1954. Vykaraamahbhya : ma i Marh bhshntara,


volumes 1-7, edited by K. V. Abhyankar. Poona: Deccan Education Society.

Apte, Vaman Shivaram. 2003. The Practical Sanskrit-English Dictionary. Kyoto: Rinsen Book Co.

Aussant, Emilie. 2011. Kikvtti and Cndra vykaraa: A Comparison of the Pratyhra Section. In
Studies in the Kikvtti: The Section on Pratyhras, edited by Pascale Haag and Vincenzo Vergiani.
Anthem Press.

Ben-Dor, Sharon. 2009. The paribhs arthavadgrahae nnarthakasya, lakaapratipadoktayo


pratipadoktasyaiva grahaam and ekadeaviktam ananyavat: Studies on some Metarules in Pinian
system. Doctoral dissertation, University of Helsinki.

Benson, James W. 1990. Patajalis Remarks on aga. Delhi: Oxford University Press.

Benson, James W. 2002. How Names Work in Grammar. In Indian Linguistic Studies, edited by
Madhav M. Deshpande and Peter E. Hook. Delhi: Motilal Banarsidass.

Benson, James W. 2010. Mahdeva Vedntin: Mmmsnyyasamgraha : a compendium of the principles


of Mmms (Ethno-Indology, v. 5.) Wiesbaden : Harrassowitz.

Bhate, Saroja. 2011. The Mahbhya and the Kikvtti: A Case Study. In Studies in the Kikvtti:
The Section on Pratyhras, edited by Pascale Haag and Vincenzo Vergiani. Anthem Press.

93
Bhattacharya, Dinesh Chandra. 1946. Paribhvtti, Jpakasamuccaya, Krakacakra of
Puruottamadeva. Rajshahi, Bengal: Varendra Research Museum.

Bronkhorst, Johannes. 2003. Tradition and Argument in Classical Indian Linguistics: The Bahiraga-
Paribh in the Paribhenduekhara. Delhi: Motilal Banarsidass. First published in 1986 by D. Reidel.

Bronkhorst, Johannes. 2004. From Pini to Patajali: The Search for Linearity. Pune: Bhandarkar Oriental
Institute.

Bronner, Yigal. 2010. Extreme poetry: the South Asian movement of simultaneous narration. New York:
Columbia University Press.

Candotti, Maria Piera. 2012a. Traditional Stocks of Examples and Their Influence of the First Prakiry
Grammars. In Studies in Sanskrit Grammars: Proceedings of the Vykaraa Section of the 14th World
Sanskrit Conference, edited by George Cardona, Ashok Aklujkar and Hideyo Ogawa, pp. 103-122. New
Delhi : D.K. Printworld.

Candotti, Maria Piera. 2012b. The Role and Import of the Metalinguistic Chapters in the New Pinian
Grammars. In Saskta-sdhuta: Goodness of Sanskrit : Studies in Honour of Professor Ashok Aklujkar,
edited by Chikafumi Watanabe, Michele Desmarais, and Yoshichika Honda, pp. 86-99. New Delhi : D.K.
Printworld.

Cardona, George. 1973. Review of Pinis Metalanguage by Hartmut Scharfe, Indo-Iranian Journal, vol.
15: pp. 207-221.

Cardona, George. 1976. Pini: A Survey of Research. Delhi: Motilal Barnarsidass. Reprint: 1997.

Cardona, George. 1979. Review of Grammatical Literature (A History of Indian Literature, Part II, fascicle
2) by Hartmut Scharfe, Indo-Iranian Journal, vol. 21, no. 2: pp. 117-139.

Cardona, George. 1997. Pini: His Work and its Traditions. Vol. 1. Delhi: Motilal Barnarsidass.

Cardona, George. 1999. Recent Research in Pinian Studies. Delhi: Motilal Barnarsidass.

Chakrabarti, Samiran Chandra. 1980. The Paribhs in the rautastras. Calcutta: Sanskrit Pustak
Bhandar.

94
Chakravarti, Srish Chadra, ed. 1918. The Bhashavritti, A commentary on Paninis Grammatical
Aphorisms, by Purushottamadeva. Rajshahi: Bimala Charan Maitra.

Clooney, Francis X. 1990. Thinking Ritually: rediscovering the Prva Mms of Jaimini. Vienna: E.J.
Brill.

Coward, Harold G and K. Kunjunni Raja. 1990. The Philosophy of the Grammarians. (Volume 5 of the
Encyclopedia of Indian Philosophies series edited by Karl H. Potter.) Princeton: Princeton University
Press.

Das, Karunasindhu. 1986. Paribhs in the Pinian System of Grammar. Calcutta: Sri Abhoy Burman,
Sanskrit Book Depot Limited.

Deshpande, Madhav M. 1997. Who Inspired Pini? Reconstructing the Hindu and Buddhist Counter-
Claims. Journal of the American Oriental Society, vol. 117, no. 3: pp. 444-465.

Deshpande, Madhav M. 2011. Ultimate Source of Validation for the Sanskrit Grammatical Tradition:
Elite Usage versus Rules of Grammar. In Boundaries, dynamics and construction of traditions in South
Asia, edited by Federico Squarcini. London: Anthem.

Deshpande, Madhav M. 2012. Bhaoji Dkitas Perceptions of Intellectual History: Narrative of Fall and
Recovery of the Grammatical Authority. In Saskta-sdhuta: Goodness of Sanskrit : Studies in Honour
of Professor Ashok Aklujkar, edited by Chikafumi Watanabe, Michele Desmarais, and Yoshichika Honda,
pp. 172-196. New Delhi : D.K. Printworld.

Devasthali, G.V. 1967. Anubandhas of Pini. University of Poona.

Devasthali, G.V. 1985. Paribh: Introduction and General Survey. In Glimpses of Veda and
Vyakarana: Reflections on Some Less Familiar Topics - Professor R.N. Dandekar Felicitation Volume,
edited by G. V. Devasthali, pp. 1-13. (Originally published in 1969.) Bombay: Popular Prakashan.

Dimri, J.P. 2002. Metarules of Pini. In Aspects of Pinian Sematics, edited by C. Rajendran, pp. 55-
60. New Delhi: Sahitya Akademi.

Dube, Harintha. 1887. Paribhvtti: A Treatise on Sanskrit Grammar by Sradeva. Braj B. Das.

Gandhi, Dharmaraje Narayan; revised by R.N. Kanade. 2005 (reprint). Dhaturupakosa. Sri Satguru
Publications, Indian Book Center (Delhi).

95
Goldstcker, Theodor. 1861. Pini: His Place in Sanskrit Literature. London: N. Trbner and Co.

Haag, Pascale. 2010. Problems of Textual Transmission in Grammatical Literature: The pratyahra
Section of the Kikvtti, Vienna Journal of South Asian Studies, Bd. LII-LIII-2009-2010, pp. 45-61.

Haag, Pascale. 2011. Paratextual Elements in Indian Manuscripts: The Copyists Invocations and the
Incipit of the Kikvtti. In Studies in the Kikvtti: The Section on Pratyhras, edited by Pascale
Haag and Vincenzo Vergiani. Anthem Press.

Houben, Jan E.M. 2003. Three Myths in modern Pinian studies. (Review of Recent Research in
Pinian Studies, by George Cardona.) Asiatische Studien/tudes Asiatiques, vol. 57, no. 1: pp. 121-179.

Houben, Jan E.M. 2012a. Studies in Indias Vedic Grammarians, 1: Nryaa Bhaas Prakriy-sarvasva
and Pinis e. In Studies in Sanskrit Grammars: Proceedings of the Vykaraa Section of the 14th
World Sanskrit Conference, edited by George Cardona, Ashok Aklujkar and Hideyo Ogawa, pp. 163-194.
New Delhi : D.K. Printworld.

Houben, Jan E.M. 2012b. Grammar and Other Modes of the Mind. In Saskta-sdhuta: Goodness of
Sanskrit : Studies in Honour of Professor Ashok Aklujkar, edited by Chikafumi Watanabe, Michele
Desmarais, and Yoshichika Honda, pp. 311-329. New Delhi : D.K. Printworld.

Houston, Vyaas. 2008 (reprint). The Sanskrit Atlas. American Sanskrit Institute, reprint 2008.

Joshi, S.D. 1965. Two Methods for Interpreting Pini. Publications of the Centre of Advanced Study in
Sanskrit, Class A, no. 5. Poona: University of Poona.

Kahrs, Eivind. 1998. Indian semantic analysis: The nirvacana tradition. Cambridge: Cambridge
University Press.

Kale, M.R. 2007. A Higher Sanskrit Grammar. Motilal Banarsidass, reprint 2007.

Kanjilal, Dileep Jumar, ed. 2007. Tantrapradpa of Maitreya Rakita: A Treatise on Kik and Nyya.
Kolkata: Sanskrit Pustak Bhandar.

Katre, Sumitra M. 1987. Adhyy of Pini. University of Texas Press.

Kesari, Hind, ed. 1991. Tantrapradipah, Commentary on Vritti and Nyasa, by Maitreyarakshita. Jaipur:
Kendriya Sanskrit Vidyapeetha.

96
Kielhorn, Franz, ed. 1868. Paribhenduekhara of Ngojibhaa, Part I. Bombay: Indu-Prakash Press.
New edition 1985, Parimal, Delhi.

Kielhorn, Franz, ed., trans. 1960. Paribhenduekhara of Ngojibhaa, Part II, with a second edition by
K. V. Abhyankar, 1960. Poona: Bhandarkar Oriental Research Institute.

Liebich, Bruno. 1928. Konkordanz Panini-Candra. Indische Forschungen in zwanglosen Heften


herausgegeben von Alfred Hillebrandt, 6. Breslau: M. & H. Marcus.

Lubin, Timothy. 2013. Customary Practice in the Vedic Ritual Codes as an Emergent Dharma Principle,
paper for the 223rd meeting of the American Oriental Society, Portland, March 2013.

Minkowski, Christpher. 2008. Why Should We Read the Magala Verses? In strrambha: Inquiries
into the Preamble in Sanskrit, Abhandlugen fr die Kunde des Morgenlandes, Band LXII, pp. 1-24.
Wiesbaden: Harrassowitz Verlag.

Mira, Ramakara. 2005 (reprint). Adhyystrapha. Motilal Banarsidass.

New catalogus catalogorum: an alphabetical register of Sanskrit and allied works and authors, volumes I-
XIX, 1949-. University of Madras. Sanskrit Department.

Saghdha, Diarmuid. 2004. Object-Language and Metalanguage in Sanskrit Grammatical Texts.


MPhil Thesis, Faculty of Oriental Studies, University of Cambridge.

Pollock, Sheldon. 2011. The Languages of Science in Early Modern India, in Pollock, ed., Forms of
knowledge in early modern Asia: Explorations in the intellectual history of India and Tibet, 1500-1800,
pp. 19-48. Duke University Press.

Saini, Ranjit Singh. 1999. Post-Pinian Systems of Sanskrit Grammar. Parimal Publications.

str, J.L. Dhtupha. Dill: Motll Banrasds. Reprint, 2002.

Scharf, Peter M. 2012. Rule Selection in the Adhyy, or Is Pinis Grammar Mechanistic? In Studies
in Sanskrit Grammars: Proceedings of the Vykaraa Section of the 14th World Sanskrit Conference,
edited by George Cardona, Ashok Aklujkar and Hideyo Ogawa, pp. 319-350. New Delhi : D.K.
Printworld.

97
Scharfe, Hartmut. 1971. Pinis Metalanguage. Memoirs of the American Philosophical Society, vol. 89.
American Philosophical Society.

Scharfe, Hartmut. 1977. Grammatical Literature (A History of Indian Literature, Volume V, fascicle 2.)
Wiesbaden: Harassowitz.

Sharma, Rama Nath. 2002. The Adhyy of Pini, Vols. II, III, IV, V, VI. Second edition. Munshiram
Manoharlal.

Shstre, Haraprasd. 1931. A Descriptive Catalog of Sanskrit Manuscripts in the Government Collection,
Vol 6: Vykaraa Maunscripts. Kolkata: The Asiatic Society. Reprinted 2006.

Shastri, Bal, ed. 2001 (reprint). Patajalis Vykaraa Mahbhya with Kaiyaas Pradpa and
Ngojibhaas Uddyota and Bhaoji Dkitas abdakaustubha with the Commentary Abhinava
Rjalakm by Guruprasad Shastri. Delhi: Pratibha Prakashan.

Shastri, Dwarikas, ed. 1971. Bhvtti, a Commentary on Pinis Grammar, by Purushottamadeva.


Varanasi: Tara Publications.

Slaje, Walter, ed. 2008. strrambha: Inquiries into the Preamble in Sanskrit, Abhandlugen fr die Kunde
des Morgenlandes, Band LXII. Wiesbaden: Harrassowitz Verlag.

Shnen, Renate. On the Concept and Presentation of "yamaka" in Early Indian Poetic Theory. Bulletin
of the School of Oriental and African Studies, University of London. Vol. 58, No. 3 (1995), pp. 495-520.

Subrahmanyam, P.S. 2005. Secondary Paribhs of Pinian Grammar. Tirupati: Rashtriya Sanskrit
Vidyapeetha.

Tatacharya, N. S. Ramanuja, ed. 1972. The Jpakasagraha, attributed by Tatacharya to Ngea Bhaa
with Vivti of N. S. Ramanuja Tatacharya. Tirupati: Rashtriya Sanskrit Vidyapeetha.

Thiruvengadanathan. 2002. Rmabhadra Dkit and His Works: A Study. Chennai: Kuppuswami Sastri
Research Institute.

Trimble, W. Walker, 2005. Principles and Development of the Brhmaical Stra Genre. (Ph.D.
dissertation, University of Pennsylvania.

Tripathi, Kripa Ram. 1991. Arrangement of the Rules in Pinis Adhy. Delhi: Parimal.

98
Vasu, ra Chandra. 2003 (reprint). The Adhyy of Pini. Motilal Banarsidass.

Wujastyk, Dominik. 1982. Do Paribhs wrongly immunize Pini's theory against criticism?
Proceedings of the International Seminar on Pini : 97-103.

Wujastyk, Dominik. 1984. A Handlist of the Vykaraa Manuscripts in the Chandra Shum Shere

Collection Preserved in the Bodleian Library, Oxford. Bound manuscript, draft of February 1984.

Wujastyk, Dominik, ed., trans. 1993. Metarules of Pinian Grammar: Vyis Paribhvtti, Volumes I
and II. Groningen: Egbert Forsten.

Wujastyk, Dominik. 2011. Rmasubrahmayas Manuscripts: A Kaveri Delta Collection Then and Now.

Now, draft of 1 December 2011, retrieved from http://univie.academia.edu/DominikWujastyk/Papers

on 10 April 2012.

Yano, M. and M. Fushimi. 2004. Pancanga version 3.13, http://www.cc.kyoto-

su.ac.jp/~yanom/pancanga/.

99

Anda mungkin juga menyukai