Rama Nath SharmaThis volume of the Astadhyayi of Panini contains English translation of adhyayas four and five. It includes Sanskrit text of individual siitras with their translitera- tion, word-boundaries, indication of inflectional endings and formation of compounds. Each siitra is also furnished with anuurtti, urtti, examples, detailed explanatory notes, and complete deriva- tional history of examples cited by the Kasikavrtti. Explanatory notes are presented to facilitate proper understanding of individual siitras, especially in view of their formulation, interpretation, application and relative placement. An attempt has been made to sort out and explain pertinent issues, as raised and discussed by major commentaries. The appendix contains nearly three hundred examples with complete derivational history. Over five hundred derivations have been discussed in notes under individual siitras, especially from the point of view of their relevance to the Paninian grammatical system. This volume is the fourth of a now planned six-volume study. The fifth volume of this series will cover the sitras of adhyaya six. The sixth and the last volume will include the remaining adhyayas, seven and eight.THE ASTADHYAYI OF PANINI ENGLISH TRANSLATION OF ADHYAYAS FOUR AND FIVE WITH SANSKRIT TEXT, TRANSLITERATION, WORD- BOUNDARY, ANUVRTTI, VRTTI, EXPLANATORY NOTES, DERIVATIONAL HISTORY OF EXAMPLES, AND INDICES RAMA NATH SHARMA University of Hawaii Munshiram Manoharlal Publishers Pvt. Ltd.ISBN 81-215-0051-6 (for the set) ISBN 81-215-0747-2 (Vol. IV) First published 1999 © 1999, Sharma, Rama Nath [All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form, or by any ‘means, electronic, mechanical, photocopying, recording or otherwise, without the written ‘permission of the publisher. Typeset, printed and published by Munshiram Manoharlal Publishers Pvt. Ltd., Post Box 5715, 54 Rani Jhansi Road, New Delhi 110 055.Contents Preface Abbreviations Adhyaya Four Adhyaya Five APPENDIX Derivational History of Examples Bibliography Index of Derivations Index of Terms Index of Siutras 763 791 793 796 800Preface This present is the fourth of a planned previously five but now six-volume study on the Astédhyayi of Panini. It also constitutes the third of the four translation volumes of the sitrapatha of Panini, with Sanskrit text, translit eration, word-boundary, anuvrtti, vrtti, English translation, examples, explana- tory notes, derivations and appendices. This volume contains the English translation of adhyayas four and five of the Astadhyayt. The fifth and sixth volumes of this series will soon be published with the English translation of adhyayas six, seven and eight. T have, as usual, relied very heavily on commentarial sources for sorting out issues relative to rule-formulation, interpretation, application and or- der. Three prakriya texts, the Riipavatarah of Dharmakirti, the Prakriyé- kaumudi of Ramacandra and the Siddhantakaumudi of Bhattoji Diksita have served as primary sources for derivational details. I have benefited most from the Vyakaranamahabhasya of Patafijali, the Vyakaranacandrodaya of Charudeva Shastri, the Kasikéuytti, with the Nyasa of Jinendrabuddhi and Padamafjari of Haradatta (with Hindi translation of Jayashankar Lal Tripathi), the Astadhyayibhdsya of Prajita Devi, the Hindi translation of the Laghukaumudi of Varadaraja made by Bhimsen Shastri and the Siddhantakaumudi of Bhattoji Diksita with the Tattvabodhini of Jiianendra Sarasvati and the Balamanorama of Vasudeva Diksita. I have mostly followed the Kasikavrtti with the Nyasa and Padamaiijari for explanatory notes. Here again, under individual rules, Thave taken detours to accommodate remarks of cultural importance. Ihave presented some general derivational processes at the beginning of the appendix so that repetition can be minimized. I have still offered major derivational details to illustrate issues, both in the text as well as in the ap- pendix. Frequent references to derivational details of earlier volumes have proved very helpful. I fully realize the extremely complex nature of the form and content of the Astadhyayi. I am also aware of my own limitations due to which I may have made many mistakes. I realize that there are many typographical er- rors in the earlier volumes. I have tried not to repeat them in this volume. As I get to the end of my work on the last adhyaya I feel that I should have better completed the last volume first. This would have offered some better insights into arrangements of discussions. I do intend to edit the earlier volumes of the siitrapatha and additionally, present a more detailed version of volume one. I shall greatly appreciate my reader's comments in this re- gard. I must express my deep sense of indebtedness to authors of works I have regularly consulted. My brother Narendra and Pandit Ramaprasad Tripathi,x Preface both of the Sampurnanand Sanskrit University at Varanasi, have always been generous to me in giving of their time. I have, as usual benefited greatly from the writings of Professor George Cardona. The University of Hawaii, especially through its Research Council and Research Relations, has been very generous in supporting my research endeavors with grants. The American Institute of Indian Studies still remains the major funding agency to support this entire project. Iam grateful to the University of Hawaii and the American Institute of Indian Studies for their continued support. I must express my special thanks to Chris Bopp without whose programing skills I would have been lost in the software jungle. These volumes owe a great deal to him. Finally, these volumes owe the most to my wife Kamala who has always protected me from many of my duties as a house- holder. I could not do this project without her support. Rama NATH SHARMA Honolulu, Hawaii (USA) 25 December 1998pac Abbreviations ablative accusative Aitareya Brikmana bahworihi caturthi-tatpurusa Dhatupatha dative dual dvandva dvitiya-tatpurusa idem itaretara-dvandva feminine Ganapatha genitive instrumental internal Kathaka Samhita Kagikavytti of VamanaJayaditya with Padamafjari (PM) of Haradatta and the Nyasa of Jinendrabuddhi locative Laghusabdendusekhara of Nagesa Vyakaranamahabhasya of Pataijali with the Pradipa of Kaiyata and the Udyota of Nageéa masculine nan-tatpurusa neuter nominative Padamaiijari of Haradatta as Kasikaurtti of Vamana Jayaditya Paribhasendusekhara of Nagesa Paippalada Samhita paribhasa paicami plural Rgveda Vaiyakaranasiddhéntakaumudi of Bhattoji Diksita with the Balamanoramé of Vasudeva Diksita, and the Tattvabodhini of Jianendra Sarasvati SittrapathaAbbreviations samahéra singular Sivasiitra Satapatha Brahmana Taittirtya Brahmana tatpurusa trliya Vakyapadiya of Bhartrhari with the commentary Ambakartri of Pandita Raghunatha Sarma varttika Vajasaneyi Samhita vocativeAdhyaya Four PApa ONE 4.1.1 Suromferafeenr, nydppratipadikat /iiyap-pratipadikat 5/1 = ni ca dp ca pratipadikam ca = hyappratipadikam (sam. dv.), tasmat/ (pratyayah #3.1.1, paras ca #3.1.2) adhikaro' yam; yad ita itrdhvam anukramisyéma apaiicamadhyayapari- samapleh nyappratipadikad ity evam tad veditavyam ‘An affix (henceforth, through the end of adhydya five) occurs after that which either ends in affixes marked with Niand 4P, or is a nomi- nal stem (pratipadika) . EXAMPLES: Look under subsequent rule(s). 1. This governing rule (adhikéra) specifies bases after which affixes enu- merated in its domain can be introduced. The expression *ydp refers to a form which ends in feminine affixes marked with Ni and 4P. These indi- vidual terms can each refer to three feminine affixes: Ni§, NiP and NiN, on the one hand, and CaP, TéP and DaP, on the other. That Niand 4P each refer to a class of three affixes each is made possible by their specification by means of Nand 4P, used with no additional it-elements: N/(NiN), $/ (NS), P/(NiP); and C/(CaP), T/(TaP), D/ (DaP). Rule 4.1.2 svaujasmaut . . . constitutesa subdomain, whereby a set of twenty- one nominal endings termed vibhakti (1.4.104 vibhaktis ca) are introduced. These nominal endings are generally referred to by the abbreviatory term sUP (1.4.103 supah). Rule 4.1.3 striyam heads another subdomain under which feminine affixes specified by riyap are introduced. These two subdomains, headed by 4.1.2 svaujasmaut... and 4.1.3 striyam, both require a nominal stem as their input. Additionally, an output of the subdomain headed by 4.1.3 striyam, i.e., a form which ends in feminine affixes characterized as ‘iydp, can also form an input to the subdomain of 4.1.2 svaujasmaut. .. . Thats, a nominal stem which takes the option of 4.1.3 striyam, to derive a feminine form ending in ‘yap, re-enters the domain of 4.1.1 to avail the option of 4.1.2 svaujas. . .. This re-entry of a feminine form to the subdomain of 4.1.2 is obligatory. A nominal stem which, upon entering the domain of 4.1.1, opts for the option of 4.1.2 may also get re-entry to the domain of 4.1.1, especially for derivational options of its subdomain headed by 4.1.76 taddhitah.2 The Astadhyayt of Panini 411 2. Note that rule 4.1.1 does not have any predication (vidheya). Conse- quently, the purpose of its formulation should be accepted as facilitating the application of other rules (pardirtha). It is also a governing rule, and hence, has a valid scope of application prior to the domain of 6.4.1 agasya. Kasika explains that riyappratipadikam specifies two types of forms: (i) nominals ending in feminine affixes marked with Niand 4P; and nominal stems covered by rules 1.2.45 arthavad.... and 1.2.46 kyttaddhita. ... This specification is made possible by the samahara type of dvandva compound in ‘yappratipadikam. That this specification is made via samaharadvandvais clear from the singular (ekavacana) in nyappritipadikam. An itaretarayoga type for- mation of a dvandva compound would have required its specification via plural, as... . pratipadikebhyah. This samahara specification also rules out the possibility of any other compound -formation. Ifsuch a possibility is not ruled out then riyappratipadikam may also be interpreted as a genitive tatpurusa compound, parallel to xiyapoh pratipadikam, meaning: a nominal stem termed compound with constituents terminating in Ni and aP. Consequently, sU, etc., could be introduced only after compound bases such as rajakumari, etc., which end in Ni, and not after kumar, etc., which could not qualify as compounds. One can argue here that ryapoh pratipadikam refers to a nomi- nal stem which ends in Niand 4P, and thus could refer to both a compound as well as a non-compound base. One can thus facilitate the introduction of sUP also after forms such as kumdri. This, however, will create difficulty of interpretation. For, riyapoh pritipadikam will have pratipadikamas part of predi- cation. The pratipadika of 4.1.1 is part of the subject (uddesya). Consequently, the sense of one cannot be equated with the sense of the other, especially relative to the introduction of sUP (4.1.2 svaujas.. ..), the predicate (vidheya). 3. It is argued that to state that affixes denoted by the abbreviatory term sUP occur after ‘that which ends in Nand 4P, or else, is a nominal stem’ is not necessary. For, an affix, in view of 3.1.2 paras ca, will have to occur after something, anyway. Ruie 4.1.1 makes a specification with regard to that af- ter which sUP can be introduced. One can even say that affixes denoted by the abbreviatory term sUP can be introduced only after bases covered by specifications made by nyéppratipadikat. If one does not accept this then there could be problems. Padamaiijari, ad Kas., outlines four types of bases: (i) verbal roots (dhatu), (ii) items ending in affixes denoted by the abbreviatory term iN (tiranta), (iii) a sentence (vakya) and (iv) items which end in affixes denoted by the abbreviatory term sUP. If a particular specifica- tion in terms of riyappratipadikat is not made then sUP, TaP, etc., feminine affixes, aN, and the taddhita affixes commonly known as the suarthikas (5.3.96 ive pratikytau, etc.), will occur after these four types of bases in accord with the order of their enumeration (yathasamkhya).. Kasika claims that, based upon deduction, one can still find those same bases forming left context of an affixal operation, as one would otherwise41.1 Adhyaya Four: Pada One 3 have by means of a specification such as riyappratipadikam. However, it is in view of specifications such as vrddha ‘that among whose vowels the first is a urddhi (1.1.73 vrddhir yasyacam adis tad. . .)', avyddha‘that which is not vrddha’, @-varna ‘the vowel a’, svara ‘accent’ and duyac ‘that which contains two vow- els’, that a specification such as riyappratipadikam becomes necessary: vrddhévrddhévarnasvaradvyajlaksanapratyayavidhau tatsampratyayartham ryappratipadikagrahanam kartavyam ‘to employ riyappratipadikamis necessary for facilitating the understand- ing of an affixal operation characterized by specifics such as vrddha, avyddha, avarna, svara, and duyae For, vrddha, etc., are given as qualifiers (vigesana) to nyappratipadikam. It will indeed be a confusing situation to find modifiers used without their corresponding modifieds. Thus, if iyappratipadikam is not specified then operational scopes, for example of rules such as 4.1.157 urddhad agotrat, 4.1.160 pracdm avrddhat phin, 4.1.95 ata if, 4.2.44 anudattader af and 4.4.7 nau duyacas than, would run into interpretational confusion. For, these rules must be supplied with ryappratipadikam as a qualified (viéesya) for modifiers such as urddha, avrddha, adanta, etc. A question is raised against including syap in the specification of this tule, especially when pratipadikat itself can account for it. For, a pratipadika also carries reference to its counterparts in other genders (Paribhasa (72): pritipadikagrahane lingavisist.-yapi grahanam). The inclusion of yap is made to ensure that a taddhita affix can also be introduced after a syntactically related nominal stem ending in affixes denoting feminine (SK:. . . ryabgra- hanam wyabantat taddhitotpattir yatha syan ryabhyam prank ma bhiid ity evam artham). Consider the derivation of aryaka from arya + TaP + kaN, where haN, a taddhita affix, denotes svartha, the meaning of its nominal base. If affix kaNis introduced after arya, prior to TaP, then the result will be dryaké, Rule 7.3.44 pratyayasthat . .. will then apply on dryaké to replace its a, which precedes k, with i. The result will be: aryika. This derivate will be unique since the éreplacement for a is obligatory (nitya). That is, aryaka, a parallel form to aryika, can never be derived. If TaPis introduced prior to the intro- duction of kaN, then, given arya + ka, the final 4 of arya will be replaced with its short counterpart by rule 7.4.18 ke nah. This wilhgive us dryaka. One can now derive dryika, from dryaka + TaP where the short a before k will be treated as the @ it replaced. This will then facilitate the optional application of rule 7.3.46 udicdm tah . .., whereby the a before & will be optionally re- placed with i. We will thus get ayiha, parallel to dryaka. Kasika recognizes that ‘a specification by means of #yap also facilitates introducing taddhita affixes after bases which end in yap (tadantat taddhitavidhanértham nyabgrahanam). Consider the derivation of kalitard ‘a female relatively more black . . .', where the comparative suffix taraP is in-4 The Astadhyayi of Panini 412 troduced after kali (5.3.57 duivacanavibhajyopapade. ..), a derivate ending in NiSof 4.1.42 janapadakunda . .., to denote relative prominence (prakarsa). Commentators state that kalitard cannot be derived unless the specification of its base is made by 1iyap. For, denotation of feminine by NiS, and that of relative prominence by taraP, will create a conflicting situation. Thatis, the scope elsewhere (savakdsatua) of a feminine affix must be one where ‘femi- nine’ alone is denoted. The scope elsewhere of prakarsa ‘relative promi- nence’ must be one where prakarsa alone is denoted. Thus, kaliand kalatarah illustrate instances where both meanings have their independent scope (savakasatua). An example such as kdlitard, presents a situation of conflict where both meanings find their scope. This conflict can be resolved in ac- cord with 1.4.2 vipratisedhe param karyam. Consequently, only prakarsa, and no feminine can be expressed, mostly because taraP is subsequent. If, how- ever, the bases are specified with yap then kalitard can be derived with no problem. 4.1.2 eireicemaifienaiarstinat aera etegy svaujasmautchastabhyambhishebhyambhyastiasibhyambhyashasosaintiyossup /svaujas. .. sup 1/1 = su ca au ca jas ca... 0§ ca sup ca = svaujas.... sup (sam. dv.)/ (pratyayah #8.1.1, paras ca #3.1.2, fiyappratipadikat #1) “su, au, jas, am, aut, Sas, ta, bhydm, Dhis, te, bhydm, bhyas, rasi, bhyam, bhyas, has, os, dm, ti, 0s, sup’ ity ete pratyaya nydppratipadikad bhavanti Affixes sU, au, Jas; am, auT, Sas; ta, bhydam, bhis; Ne, bhydm, bhyas; Nasl, bhydm, bhyas; Nas, os, am and Ni, os, suP occur after that which either ends in affixes marked by Niand 4P, or is termed a nominal stem. EXAMPLES: kumar ‘girl’ gauri ‘a female of fair complexion; Shiva’s consort’ Sarigaravi ‘a female name’ f ‘a city with many kings’ havisagandhya ‘a granddaughter of Karisagandhi? drsad ‘nominative singular of drsad ‘pebble, rock”’ 1. Refer to the appendix for derivational details of these examples. The first, kumari, constitutes an example for both riyappratipadikatand NiP (4.1.20 vayasi prathame). The next two, gauri and sarigaravi, represent affixes MS (4.1.41 sidgauradibhyas ca) and MiN (4.1.73 Sarigaravadyanio ...). The next three examples, khatud, bakuraja and karisagandhya, represent derivates of TaP (4.1.4 ajadyatas ¢ap), DaP (4.1.13 dab ubhabhyam ...) and CaP (4.1.74 yanas cap), respectively.4.1.2 Adhyaya Four: Pada One 5 2. The twenty-one suffixes enumerated by this rule have been classified into seven triads of prathama ‘first’, dvifya ‘second’, trliyd ‘third’, caturthi ‘fourth’, paficami ‘fifth’, sasthi ‘sixth’, and saptami ‘seventh’ (1.4.101 tinas- trini .. .). Individual members of each triad are further assigned the terms ehavacana ‘singular’, dvivacana ‘dual’ and bahuvacana ‘plural’, respectively (1.4.102 tany ekavacana.. . .). The following is a tabular listing for reference: singular dual Plural prathama sU au Jas dvitiya am auT Sas brtiya Ta bhyam bhis caturthi Ne bhyam bhyas paricami Nast bhyam bhyas sasthi Nas os am Saptaini Ni os suP The vowels which have been used as jit in these elements either facilitate articulation, or distinguish endings from each. others (visesandrtha). The T, as an itin Ta, is used to facilitate formation of the abbreviatory term sUT. The final Pof suPis brought as an itso that these nominal endings all can be referred to by the single abbreviatory term sUP. The denotata of these nominal endings have been outlined as number (samkhya), nominal stem notion (pratipadikartha) and karman ‘object’ etc., by rules enumerated elsewhere (2.3.1 anabhihite, etc.; Kas.: samkhyaharmadayas ca suddinam arthah sastrantarena vihitas tena sahdsyaikavakyata). This rule, for its proper interpretation, must share the ‘single-sentenceness’ relationship (ckavakyata) with them. Thus, rule 2.3.2 karmani dvifiya, read with 2.3.1 anabhihite and rule 4.1.2 svaujas. . ., can be interpreted as a single sentence as follows: ‘affixes termed dvifiyd, particularly its members termed ekavacana‘sin- gular’, dvivacana ‘dual’ and bahuvacana ‘plural’, occur after an item which either ends in a feminine affix denoted by Nrand 4P, or is termed a pratipadika, provided an object (harman) not expressed otherwise, is to be expressed” Recovering thé context of a rule, via ‘single-sentenceness’ (chavakyata), often entails an elaborate network of related rules. My interpretation of 2.3.2 karmani dvitiya shows the involvement of only a few important rules. Note that accepting thie signification of affixes denoted by sUP as samkhyaharmadi ‘number and object, etc.’ is also problematic. For example, if samkhyakarmadi is the denotatum of a sUP then these affixes cannot be introduced after indeclinable bases (avyaya). Consequently, Panini’s provi- sion of deleting a sUPby means of LUKafter an indeclinable (2.4.82 avyayadThe Astadhyayi of Panini 413 ap supah) becomes meaningless (vyartha). But since Panini provides for de- leting a sUP after an indeclinable, introduction of the same after an inde- clinable cannot be challenged. That is, a sUP will be introduced after an indeclinable even when samkhyakarmadi are not denoted. It is a different story that nominative (prathamé) singular (ekavacana) alone could then be introduced. 4.1.3 foray, striyam /striyam7/1/ (pratyayah #3.1.1, paras ca #3.1.2, nyappratipadikat #4.1.1) yad ita trdhvam anukramisydmas striyam ity evam veditavyam The following affixes occur after a nominal stem when feminine is to be denoted. EXAMPLES: Refer to subsequent rules. 1. That this rule is formulated to facilitate application of other rules (parartha) is clear from the specification of striyam in locative (saptami), es- pecially with absence of any predicate (vidheya). It offers jati ‘class’, particu- larly stritua ‘femininity’, as scope for application of rules it governs. A ques- tion is raised as to what stritvais. It is explained that, like the general class of gotva ‘cowness’, stritva also finds varying loci (dsryaya), both general (samdnya) and particular (visesa). In the absence of a well defined particular locus, it may, on occasions, be comprehended in a manner similar to brémanatua “Brahmananess’ based on upadeéa. Kasikd, therefore, explains this jatiof stritoa as samanyavises d bahuprakarah ‘general and particular of varying types’ (Kai siimanyavisesah stritvadayo gotuadaya iva bahuprakaré kvacid asrayavisesabhdvad upadesavyanya eva bhavanti yatha brahmanatvadayah). Sanskrit grammarians do not clearly define what gender is. Pataijali rec- ognizes prominence of ‘breasts’ and ‘hair’ as distinguishing marks for ‘fe- males’ and ‘males’, respectively. A thing which is different from two is char- acterized as ‘neuter’: stanakesavati stri syal lomasah purusah smytah/ ubhayor antaram yac ca tadabhave napwrsakam/ / But this could only characterize biological gender. It cannot apply to a host of things such as khatud ‘cot’, mala ‘garland’, ghata ‘pov’, pata ‘cloth’, etc. Grammarians rely solely upon usages for determining gender (Jokasrayatval liigasya). What about the belief that gender is already included within the denotatum of a nominal stem (pratipadikagrahane lirigavisistasyapi grahanam)? If the notion of femininity (séritva) is included within the sense ofa pratipadika4.14 Adhyaya Four: Pada One 7 then what is the purpose of introducing a feminine affix. It is argued that feminine affixes are employed to manifest this notion of gender inhered by nominal stems. They thus reinforce its comprehension by way of definitizing it. As far as comprehension of gender, per se, is concerned, one may, in many instances, comprehend it even without the use of a feminine suffix. 2. Note that this rule forms a subdomain of 4.1.1 ryappratipadikat. The anuurtti (from rule 4.1.1) of pritipadikat alone is valid here since a form which ends in a feminine affix (xyabanta) cannot form an input to the subdomain which introduces that affix. 4.1.4 SATEEN, ajadyatas tap /ajadyatah 5/1 = aja tap \/1/ (pratyayah #3.1.1, paras ca #3.1.2, pratipadikat #41, striyam #3) Gjadibhyah pratipadikebhyo' akérantac ca pratipadikat striyam tap pratyayo bhavati Affix Tap occurs after a nominal stem which is abstracted from the group headed by aja ‘she-goat’, or after a stem which ends ip -a7 (1.1.70 taparas tatkélasya), when feminine is denoted. ca at ca = ajadyat; tasmat = ajadyatah (sam. dv.); EXAMPLES: aja'she-goat’ edaka ‘ewe’ calaka ‘sparrow’ asva ‘mare’ 1. The P of TaP is given as an it for the general low pitch (anudaitta) accent of rule 3.1.4 anudattau suppitau. The Tis used as an it to distinguish T@P from affixes such as CaP and DaP. The T of aT'is, of course, part of taparakarana (1.1.70 taparas tatkalasya), whereby stems ending in a (of short duration) alone can qualify. This provision of TaP, especially in view of its specification with ajadi‘that which begins with aja’, is made against affixes NiS and NP, which may other- wise find their scope of application. Note that the listing of the ajadi group is not made with the enumeration of nominal stems aja, etc. It is, instead, made with aja, etc., which are items derived with affix TaP already intro- duced (kytalappratyaya). Thus, TaP will be introduced after stems specific to those of aja, etc. Affixes NiSand NeP, as against TaP, could be conditioned by factors such as: (i) signification of jati ‘class’, (ii) marked association of a female with a corresponding male (pumyoga-laksana), (iii) a following word with the denotatum of flower of fruits (puspaphalottarpada), or (iv) age (vaya), etc.8 The Astadhyayi of Panini 4.14 Consider, for example, the derivation of sdidra ‘a female of the class charac terized as Sidra’. This provision of TaPis made against denotation of a ‘wife’, in relation to the ‘husband’. Thus, the wife of a siidra is not called siadrd. Instead, saidra denotes a class of females (jati). For denoting a ‘female (wife)" corresponding to a ‘male (husband)', one has to introduce affix Ni$ to de- rive Sidri, parallel to siidrasya bharya ‘wife of a Sidra’. This ‘association with a male (pumyoga)' does not work with forms which have mahat as their first constituent. Thus, the wife of an abhira, a class synonymous with mah: will not be called mahasiidri. Instead, TaP will be used to derive mahi This exception is made possible because of tadantavidhi, interpreting a form as ending in that by means of which a specification is made (1.1.73 yena vidhis tadantasya). Incidentally, an dbhirais characterized as the offspring of a brahmana father and an ugra mother. An ugré, in turn, is a female born of a Sidra mother and a ksatriya father. Kasika states that the following paribhasa is not valid in this connection: grahanavaté pratipadikena tadantavidhir nasti ‘a specification made by means of a particular nominal stem is not interpreted as referring also to a form which may end in that nominal stem’ Consequently, we get forms such as atidhivari‘an excellent fisherwoman’, atipivari‘a very fat woman’, atibhavati‘a very respected woman’ and atimahati ‘avery great woman’, Examples such as ajé ‘she-goat’, aiva‘mare’ and cataka ‘sparrow’ denote jati ‘class’; bala ‘maiden’, pak ‘girl’, vatsa ‘female calf’, and manda ‘feeble woman’, etc., denote age (vaya). One can similarly get other examples in other categories where T@Pis introduced to block NiS or MP. 2. Why should ajadi be explicitly mentioned in this rule especially when, since all bases of ajadiend in a, specification made with atah can account for TaP? For, all the ajadi bases end in a. A special mention of ajadi is made here to block provisions of NiS (4.1.63 jater astrivisayad ...) and NiP (4.1.20 vayasi prathame) 3. Bhattoji Diksita interprets ajadyatah as ending in genitive (sasthyanta). A paricamyanta ‘ending in a fifth triplet of sUP’ interpretation is problem- atic, For example, given paiicaja‘a group of five goats’, TaP will yield a wrong form: *paficaja. A genitive interpretation will yield the correct form paiicaji, with MiP of 4.1.21 dvigoh. Actually, these two interpretations relate to ques- tions of whether denotation of feminine should be accepted as meaning of a base, or of an affix. If affix, then a paficami interpretation of ajadyatah, read in view of TaP, will yield the following meaning: ‘affix TaP occurs after a nominal stem which ends in a, or after a nomi- nal which is inferred from the group headed by aja, when feminine is denoted’415 Adhydya Four: Pada One 9 This interpretation will accept ‘feminine’ as the meaning of the affix (pratyayartha). A genitive interpretation will accept ‘feminine’ as included within the meaning of the base. What would be the function of a feminine affix under this interpretation? It will simply manifest the notion of gender inhered by the base (dyotayati). Thus, a genitive interpretation of ajadyatas sap will yield the following meaning: ‘affix TP occurs after a nominal stem which ends in a, or is inferred from the group headed by aja, when the sense of feminine inhered by the base is made manifest” Bhattoji Diksita accepts the genitive interpretation, along with ‘revela- tion or manifestation’ as the meaning of feminine affixes. The tradition has discussed both views relative to the denotation of gender. Those who accept gender as part of the meaning of a nominal stem believe that an affix is introduced to ‘manifest or reveal’ this meaning (striyam yat pratipadikam vartate tasmat tabadayah). This view accepts the base after which a feminine affix is to be introduced as qualified with ‘feminineness’ (stritvavisista). Adherents of the other view believe that an affix is introduced after a base when gender is to be denoted (striyam abhidheyayam tabadayo bhavanti). This view will give prominence to affixal meaning: prakrtipratyayau saharthabrittas tayoh pratyayérthasyaiva pradhényam ‘affixal meaning is principal when a specifica- tion is made concurrently by means of a base as well as an affix’. But this will create problems. Since ‘feminine’ is to be expressed as affixal meaning (pratyayartha) qualified by meaning of the base (prakytyartha), singular (ekavacana) alone can be used. What is the use of dual and plural forms? If feminine is to be accepted as a notion inhered by the base then this prob- Jem would not arise. Commentators have discussed und resolved this, and host of other problems concerning the two views, in relation to the fivefold (paricaka) meaning of bases (pratipadika) and the notions of jati‘class’, vyakti ‘individual’, padartha ‘substance’ and satta ‘existence’. | omit any further discussion of them for lack of interest. My translations and discussions re- flect acceptance of both the views. 415 FET ST ranebhyo nip /ranebhyah 5/3 = rt can ca = ynn (sam. dv.); tasmat; ip 1/1/ (pratyayah #3.1.1, paras ca #3.1.2, prétipadikat #1 striydm #3) rhardintebhyo nakarantebhyas ca priitipadikebhyah striyam nip pratyayo bhavati Affix NiP occurs to denote feminine after nominal stems which end in r (rT, cf. 1.1.70 taparas tatkdlasya) and n. EXAMPLES: karbi ‘she who does .10 The Astadhyayi of Panini 4.16 harti ‘she who carries . . .’ dandini ‘she (a female ascetic) who carries a shaft’ chatrini ‘she who carries an umbrella’ 1. Note that N, as an itin NiP, is intended for reference to affixes iP, Ni and NiN in general (sémanyagrahana). That is, in its absence, a reference with N of Nyap in rule 4.1.1 riyappratipadikat, will have to be limited to just two affixes. What is the purpose of Pas an if It is used so that NiP can be distinguished (visesanarthah) from NiN and NiS. This P, as has already been stated, also accounts for anudatta. 2. Note that rnnebhyahis interpreted as: ‘after that which ends in rand n’. Such an interpretation is possible only when one treats mnebhyah as a quali- fier (visesana) to pratipadihat (here read as pratipadikebhyah), in accord with tadantavidhi ‘weating a qualified (visesya) as ending in that which qualifies (visesana)’. 4.1.6 Braga ugitas ca /ugitah 5/1 = ug it yasya (bv.), tasmat; cao/ (pratyayah #3.1.1, paras ca#3.1.2, pratipadikat #1, striyam #3) ugid yatra sambhavati yathakathafcit tad ugic chabdaritpam, tadantat striyain ‘ip pratyayo bhavati Affix NiPoccurs to denote feminine also after a form which ends in an it denoted by the abbreviatory term uK (u, 7, fy cf, Sé. 1-2). EXAMPLES: bhavati ‘you (feminine polite of bhavat)’ atibhavati ‘you (feminine extremely polite)’ pacanti ‘she who is cooking’ yajanti ‘she who is performing a ritual sacrifice’ 1. Note that uK denotes 1, 7, /. The form ugitis interpreted as a bahuvrihi compound parallel to ug it yasya ‘that which has sounds denoted by uKas an it’. We know that the meaning of a bahuwihi falls outside of its constituent meanings (anyapadartha) . Can pratipadika of pratipadikat be accepted as the denotatum of the bahuvihi compound ugif? That is, can ugit be accepted as a modifier (visesana) to pratipadika? To accept pratipadikat as a qualified (visesya) for ugit will facilitate the introduction of NiPin bhavafiand mahati, where nominal stems bhavatU/and mahatU/have Uas an it. But this interpre- tation will block MiP in examples such as gomafi ‘a river; she who is rich in cows’ where it is affix matUP, and not the nominal stem go, which has uK as an it, Examples such as arvati ‘mare’ will also be excluded since what is an ugit there is tR, a replacement of n of arvan, itself an ugit. It is therefore suggested that ugit be interpreted not as a nominal stem but as a qualifier.4.7 Adhyaya Four: Pada One u This qualifier status of ugit, when read with pratipadikat, will facilitate tadantavidhi and yield the following interpretation: ‘affix NiPoccurs after a nominal stem which ends in a form containing a sound denoted by the abbreviatory term uK as an if 2. A varttika proposes that roots with uKas an itshould be excluded from the scope of this rule. This will exclude roots samsUand dhvamsU of ukhdsrat ‘fallen from a pot’ and parnadhvat ‘fallen from a leaf’. An additional vérttika suggests that forms ending in affix afclJ should be included to enable the derivation of prac ‘east’ and pratici ‘west’ with NiP. The Mahabhasya rejects this varttika, and accepts yogavibhaga, whereby anyatarasyam of rule 4.1.8 pado nyatarasyam will be treated as a rule by itself. It will be accepted asa residual (sesa) rule to 4.1.6, 4.1.7 and 4.1.8. This rule will then be treated as wyavasthitavibhasa, whereby a dhatu which qualifies as ugit will receive NiP only when it ends in avicU. 4LTTHTS vano ra ca /vanah 6/1 ra (deleted 1/1) cad/ (pratyayah #3.1.1, paras ca #3.1.2, pratipadikat #1, striyam #3) vannantad pratipadikat striyam nip pratyayo bhavati rephas cantadesah Affix Nip occurs to denote feminine after a nominal stem which termi- nates in van, with an additional provision that r comes in place of the final sound of that which ends in van. EXAMPLES: dhivari ‘fisherman's wife’ pivari‘a fat woman’ Sarvari ‘night; Rati, Kamas wife’ paralokadrsvari‘a female; capable of seeing through the world beyond’ 1. Commentators state that 4.1.5 ynnebhyo ... could easily account for derivates of this rule. For, stems which end in vanalso qualify as those which end in n. A separate formulation of this rule is needed so that r can be concurrently be introduced as a replacement. Note that vanis a cover term for affixes KvanIP, vanlPand NvanIP. These affixes enable their bases to qualify as those ending in n. Thatis, van of the rule means: that (a nominal stem) which endsin van (vann-antat pratipadikal). Commentators state that van also refers to a nominal stem which ends in a form ending in affix van. That is, van means: a nominal stem which ends in a nominal stem ending in van (vann-antat pratipadikantat pratipadikat) . This second interpretation is gotten via Paribhasa (24): pratyayagrahane yasmal sa vihitas tadades tadantasya grahanam ‘that after which an affix is introduced12 The Astidhyayi of Panini 4.18 refers not only to a form which ends in that affix, but also to that which ends in the form ending in that affix’. ‘The word-form van cannot be here interpreted as denoting verbal root vanU''to beg’, because ‘a specification relatable to both an affix (pratyaya) as well as a non-affix (apratyaya), is interpreted as referring to an affix alone’ (Paribhasa (111): pratyayapratyayayoh pratyayasyaiva grahanam). Besides, in- terpreting van to refer to verbal root vanU would not make any sense, A varttika states that this rule cannot introduce NiP after a form which, prior to introduction of an affix denoted by van, ended in a sound segment denoted by the abbreviatory term hS (Ss. 5-10; vano na hasah). Consider Sarvari where a van affix has been introduced after sy which does not end in 4S. A form, such as sar which ends in ris produced through application of guna before van. Now consider sahayudhva brahmani where sahayudh, which ends in dh (a hS), receives kuanIP (3.2.94 drseh kuanip). Affix NiP, and rasa replacement for nas well, are thus blocked. 4.1.8 WASAATAT, pado' nyatarasyam /padah 5/1 anyatarasyam'7/1/ (pratyayah #3.1.1, paras ca #3.1.2, pratipadikat #1, striyam #3) padantat pratipadikad anyatarasyam striydm nip pratyayo bhavati Affix MiP optionally occurs to denote feminine after a nominal stem which ends in pad ‘foot’. EXAMPLES: dvipat ‘that which has two feet’ dvipadi ‘id’. catuspat that which has four feet" catuspadi ‘id’. sahayudhué brahmani ‘a Brahmana woman who was a fighting compan- ion 1. This rule specifies padah, an ablative of pad, where the samasanta affix TaC, of pada, has already been deleted. That is, pad, with deleted TaC, oc- curs at the end of a compound. Thus, dvipad is a bahuvrihi compound origi- nally with the form dvipada. Rule 5.4.140 samkhya . .. deletes its final a, sub- sequent to which pads replaced with pad by rule 6.4.130 padah pat. We will get dvipad if the option of NiPis not accepted. Recall that 6.4.130 padah pat is contained within the subdomain of 6.4.129 dhasya. Our string dvipad can- not be termed bha unless affix NiPis introduced. Lack of assignment of the term bha blocks the replacement of pad by pad. The d of dvipad is optionally replaced with ¢ (cartva) of 8.4.55 vd’ vasane.4.1.10 Adhyaya Four: Pada One 13 4.1.9 erafa tab rei [tap i/\ ri 7/1/ (pratyayah #3.1.1, paras ca #3.1.2, pratipadikat #1, striyam #3, padah #8) 1c vacyayam padantat pratipadikat striyam tap pratyayo bhavati Affix TaP occurs after a nominal stem which ends in pad, to denote feminine, when a hymn (red) is signified. EXaMPLEs: dvipadé rk ‘a hymn with two quarters’ tripada rk ‘a hymn with three quarters’ catuspada 7k ‘a hymn with four quarters’ 1. Note that our preceding rule makes a provision for NiP. This rule over- rides that provision by introducing TaP under the special meaning condi- tion of red. The sense of locative in raiis that of denotation (abhidheya). That is, this locative does not denote scope (visaya). For, a scope interpretation will restrict this rule to the Rgveda. Accepting the denotational meaning of locative will let this rule apply everywhere (Nyasa: yadi hi visaya-nirdesah syat. . . regrantha eva syat, nanyatra; abhidheya-nirdese ti sarvatra bhavatiti bhavah). 4.1.10 7 Weer: na satsvasradibhyah /na $ satsvasradibhyah 5/3 = suas adir yesam te svasradayali; sat ca svasridayas ca = satsvasradayas, tebhyah (bv. with int. itar. dv.)/ (pratyayah #3.1.1, paras ca #3.1.2, pratipadikat #1, striyam #3, ajadyatas tap #4, mnebhyo nip #5) satsamjnakebhyah svasradibhyas ca pratipadikebhyah stripratyayo na bhavati An affix which denotes feminine does not occur after a nominal stem which is termed sat (1.1.24 snanta sat), or is listed in the group headed by svasr ‘sister’. EXampLes: patica brahmanyak ‘five Brahmana women’ sapta brahmanyah ‘seven Brahmana women’ svasé ‘sister’ duhita ‘daughter’ nananda ‘husband’s sister” yata ‘husband's brother's wife” catasrah ‘four . ..” 1. Note that this rule negates all feminine affixes which have been made available thus far (Kaé.: yo yatah prapnoti sa sarvah pratisidhyate). It should not14 The Astadhyayi of Panini 4.1.10 be interpreted simply as a negation of NiP, since NiP constitutes the most immediate context (anantaratvat). Our rule also negates TaP. Incidentally, MP still forms the most immediate context since a provision for TaP of 4.1.10 tab rciis made against it. Besides, this TéP is not intended to be car- ried via anuvytti (asvaritavat). It is argued that items termed sat are number words (samkhya), and as such, they simply facilitate comprehending quantitative differences of things. They do not facilitate any understanding of gender per se. Comprehension of gender takes place via syntactic coordination (sdmanadhikaranya) of number words with things which they distinguish (Nyasa: samkhyas ca bhedamatram dravyasya pratyayayanti, na strituam; ya tu tasya sbitvapratitih sa Sabdantarena sémanadhikaranyéd bhavati). A negation of feminine in relation to sat, when the same is not made available, is vacuous (vyartha). Negation (pratisedha) makes sense only when a positive provision is available. It was made clear in connection with 4.1.3 striyam that a qualifier, such as sukla ‘white’, qualifies for a feminine affix in accord with its qualified. This same is also true of items termed sat. 2. Note that paiica (of patica brahmanyah) and catasr represent two sets of derivates where feminine affixes are negated. The first word is blocked from receiving NaP (of 4.1.5 rnnebhyo itp) which was available in view of the final n of paiican ‘five’. Recall that 1.1.24 snanta sat assigns the term safto paficanon account of its final n. Thus, given parican + Jas we get paiican + (Jas) = paiica (n>), where 7.1.22 sadbhyo luk and 8.2.7 nalopah pratipadikantasya are responsible for deleting Jas and n, respectively. The second example catasr becomes an exception to 4.1.5 because catasr is a replacement for catur (7.2.99 tricaturos striyam ...). This negation of MiP and TaP, condi- tioned by sat, is applicable also items which end in a sat. For example, con- sider priyapaiica of priyaparica draupadi ‘Draupadi for whom five are dea ‘A question is asked as to why an item termed sat cannot qualify for TaP after deletion of its n has taken effect. Why patica, at the strength of now ending in a, cannot get TaP (satsamjiidndm ante lupte tab utpattih kasman na syat)?. One cannot argue here that this negation then would become vacu- ous (vyartha). This negation would have served its purpose by blocking the next immediate (anantara) NiP anyway (Paribhasa (62): anantarasya vidhir bhavati pratisedho va). It is stated that TaP can be blocked by accepting the deleted n of paiican as suspended (asiddha; 8.2.2 nalopah supsvarasamjna- tugvidhisu). But how could n-deletion be treated as asiddha when operations relative to TéPare not included within the scope of 8.2.2. This scope of TaP comes well within the scope of sub-vidhz of 8.2.2 when one accepts the for- mation of the abbreviatory term sUP, formed with sUof 4.1.2 svaujas.. . and the Pof 4.1.74 yanaé cap. That is, TaP cannot be introduced because dele- tion of nwill become asiddha. But accepting the formation of the abbreviatory term (pratyahara) sUP with the P of 4.1.74 yanas cap, to include operations41.11 Adhyaya Four: Pada One 15 relative to TaP, will create problems elsewhere. The Mahabhdsya discusses this, and other aspects of this problem, in detail. Patanjali rejects this notion of asiddhatva, especially in the derivational context of bahucarmika nagar’ ‘a city with hide in abundance’, where ndeletion of bahucarman, if treated as asiddha, will block the i (itva) of 7.3.44 pratyayasthat hat. ... That is, given bahucarmaka + TaP, the a of aka cannot be found immediately before the k. For, the deleted n of bakucarman will then be there through asiddhatua. This will then block itva. Consequently, ‘striyam’ ity arthamatram apeksya tatra yad uktam bhavati tan na bhavati‘that (NiPand TaP) which is provided there in view of denotation of meaning specified by striyam does not apply here’. 4.1.11 WA: manah /manah5/1/ (pratyayah#3.1.1, paras ca¥3.1.2, pratipadikat#, striyam#8, nip#5 na #10) mannantad pratipadikan nip pratyayo na bhavati Affix NiPdoes not occur to denote feminine after a nominal stem which ends in man. EXAMPLES: dama ‘nominative singular of déman ‘rope’ damanau ‘nominative dual of daman’ damanah ‘nominative plural ...” sima ‘nominative singular of siman ‘limit, parting of hair” siménau ‘nominative dual . ..” siménah ‘nominative plural . . . atimahima ‘nominative singular of atimahiman; he who has excelled in glory’ atimahimanau ‘nominative dual atimahiménah ‘nominative plural . ..” 1. This rule negates the provision of 4.1.5 rnnebhyo nip. 2. Note that man refers here to forms which end in affixes, for example, the Unédi (IV.144) manIN, etc. Commentators explain that man must also refer to any stem which may end in man. That is man does not have to be affix man. How do we know this? From paribhasa (17): aninasmangrahanén arthavata can arthakena ca tadantavidhim prayojayanti ‘specifications made by means of an, in, as and man, entail tadantavidhi, a process whereby that by means of which a specification is made refers also to that which ends in it, whether or not an, etc., are meaningful’. This paribhasa has been read in the Mahabhasya under rule 1.1.72 yena vidhis tadantasya. Note that stman and atimakimanend in man, though manis not an affix. That is, man is not mean-16 The Astadhyayi of Panini 4.1.12 ingful. Examples such as péman ‘skin disease’ and déman ‘rope, string’, with nominative forms such as pama, pamanau, paméanak; and daméa, damanu, damanah, do terminate in affix man. This man is meaningful because of be- ing an affix. 4.1.12 rt agate: ano bahuvriheh /anah 5/1 bahworiheh 5/1/ (pratyayah #3.1.1, paras ca #3.1.2, pratipadikat #1, striyam #3, itp #5, na #10) annantad bahuvrtheh striyam nip pratyayo na bhavati Affix MiP does not occur to denote feminine after a bahuvrihi nominal stem which ends in an. EXampLes: uinative singular of suparvan; . . . whose joints are beauti- suparvénau ‘nominative dual . ..” suparvénah ‘nominative plural . ..” sucarmé ‘nominative singular of sucarman; . . . whose skin is beautiful’ sucarmanau ‘nominative dual . ..” sucarmanah ‘nominative plural . ..” 1. Note that bahuvriheh simply means ‘after a bahuvihi compound’. Our examples, though, illustrate only one kind of bahuvrihi, namely, that which ends in an but does not go through deletion of its penultimate (upadhalopa) sound. Rule 4.1.28 ana upadhalopino’ nyatarasyam will make an optional pro- vision for DaP, against NiP, in case of bahuvrihi compounds which end in an, and go through deletion of their upadha. The fact that this rule covers in- stances of bahuvthi compounds which end in an, with no deletion of their upadhé, becomes known by process of elimination (parigesyat). Also, this rule negates NiPagainst optional DaPof the next rule, and optional NiPof 4.1.28 anupadhalopino. .. . Thatis, this rule blocks NiPin toto. The DaP of the next rule is made optional to absence of a feminine affix. Why do we have to include the condition of bahuvriheh? So that atirdji, a talpurusa compound parallel to atikranta rajanam ‘she who has excelled the king’, may not qualify for MP. 4.1.13 STRATA, dab ubhabhyam anyatarasyam / dap 1/1 ubhabhyam 5/2 anyatarasyam 7/1/ (pratyayah #3.3.3, paras ca #3.1.2, pratipadikat #1, striyam #3) dap pratyayo bhavati ubhabhyém; mannantad pratipadikad annantac ca4.1.13 Adhyaya Four: Pada One 17 bahuvrther anyatarasyam Affix DP optionally occurs after both, i.e., a nominal stem which ends in man and a bahuvrihi compound which ends in an. EXAMPLES: ‘nominal singular of bahuraja, an assembly of many kings’ ‘nominative dual...” ‘nominative plural . . . bahudhivé ‘a very clever female’ bahudhivari id” iva ‘a very fat female’ bahupivari ‘id.’ 1. Note that ubhabhyam ‘after both . . .’ refers to nominal stems which (i) end in man (4.1.11 manah); or (ii) constitute a bahuvrihi compound ending in an (4.1.12 ano bahuvriheh). These two categories of stems are not allowed MiP in view of the two preceding rules. The MiP which is negated here may become available to these stems by rule 4.1.5 runebhyo nip. Our present rule makes an optional provision for DaP against the negation of MiP. The word ubhabhyam restricts optional DP to two categories of specified stems. Exam- ples, such as sima/sime/simdh, illustrate the option of DaP. Affix NiP must remain negated in cases where this optional D@Pis not accepted. This will yield forms, such as sima/ stmanau/ simanah, where no feminine affix is intro- duced. An optional DaP, in the context of a bahuvrihi compound ending in an, produces examples such as bahuraja/bahurdje/bahurajah. Not accepting this option will again lead to negation of NiP (4.1.12 ano bahuvriheh). This will produce examples such as bahuraja/bahurajanau/bahurajanah, with no choice of a feminine affix. Note that this optional provision for DaP would cause fideletion in a presuffixal base termed bha (6.4.143 teh). Not accept- ing DaP would result in retention of the base-final n. 2. What is the purpose of anyatarasyam in this rule (Kas: anyatarasyam grahanam kimartham)? 1s anyatarasyam used here to allow negation of NiPin cases where option of D@Pis not accepted? Or is it used for allowing option of NiP against its own negation, or else, against the option of DaP just pro- posed. The first interpretation is not appropriate (adyam tévad ayuktam) since D4P, and a negation of NiP as well, is already made available by the explicit statement of dab ubhabhyam (4.1.13) ‘Da@P after both ... (vaca- nasémarthyat)’. Obviously, this provision of DaP cannot be made optional to18 The Astadhyayi of Panini 4.1.14 the obligatory negation of NiP. A non-application of negation, or of optional DaP, will make both provisions meaningless. Thus, they both will have to be applied: NiP obligatorily, and DaP optionally. This interpretation is possible even without the explicit use of anyatarasyam. The second interpretation is also inappropriate (dvifiyam apy ayuktam). For, if anyatarasyam is intended to provide the option of NiP to a bahuvrihi, with no deletion of its upadha (4.1.28 ana upadhdlopino’ nyatarasyam), then negation of 4.1.12 ano bahuvriheh will become useless. If anyatarasyam is, in- stead, understood as providing the option of Da@P to bahuvrihi stems which go through deletion of their upadha but which do not take the option of iP of 4.1.28 ana upadhalopino’ nyatarasyam, then it does not make any sense. For, anyatarasyam of 4.1.28 itself will account for option of DaP. The anyatarasyém of this rule is thus not required. Kasiké states that anyatarasyam should be used here for offering DaPas an option to MiP of 4.1.7 vano ra ca, which, without this anyatarasyam, will re- main negated because of 4.1.12 ano bahuvriheh (Nyasa: nibrephav apraptau, ‘ano bahuvriheh iti pratisedhat, tasmat tadvidhanartham anyatarasyam grahanam kartavyam). Consider bahudhivaand bahudhivarias examples. The Mahabhasya proposes split-interpretation (yogavibhaga) of this rule as follows: (i) dab ubhabhyam bhavati, where both manah (4.1.11) and ano bahu- vriheh are carried; (ii) anyatarasyam, where ano bahuvriheh alone is carried. This second split-rule is intended for making affix NiP (of 4.1.5 rnnebhyo nip), and replacement in for n (of 4.1.7 vano ra ca), both optionally avail- able. 4.1.14 STATE, anupasarjanat /anupasarjanat 5/1 = na upasarjanam = anupasarjanam (naj. tat.); tasmat/ (pratyayah #3.1.1, paras ca #3.1.2, pratipadikat #1, striyam #3) yad ita irdhvam anukramisyamo nupasarjandd ity evam tad veditavyam ‘Affixes enumerated hereafter occur after a nominal stem which is not termed an upasarjana (1.2.43 prathamé nirdistam samasa upasarjanam). EXAMPLES: hurucari ‘she who wanders in the country of the Kurus’ madracari ‘she who wanders in the country of Madras’ 1. This is an adhikéra, valid through 4.1.77 yiinas tih. Note that anupasarjana, of anupasarjanat, is explained by rule 1.2.43 prathamanirdis- tam samasa upasarjanam as a compound constituent which is specified with nominative (prathamé) in a rule allowing formation of compounds.4.1.14 Adhyaya Four: Pada One 19 The word anupasarjana ‘non-secondary’, of anupasarjanat, is a negation (pratisedha) of the prasajya type yielding the interpretation: upasarjanan nibadayo na bhavanti ‘affixes NiP, etc., do not occur after an upasarjana’. A dasa interpretation will yield the interpretation: upasarjanad anyasman idayo bhavanti ‘affixes NiP, etc., occur after a form other than an upasarjana’. The first interpretation is a straight negation (nisedha). It will not permit affix NiS, in kukkutipadah, because kukkufi will be termed an upasarjana by rule 2.2.8 sasthi where sasthi, a pada ending in genitive, is specified with nominative (prathama). This argument is not valid since in- troduction of Ni$ will take place prior to compound-formation at the strength of being internally conditioned (antarariga). The formation of compound, an externally conditioned operation, will take place with kukkufi as a constituent ending in NiS. But the NaS of kukkufi will also be blocked. How? The word anupasarjana means pradhana ‘principal’; and kukkufi de- notes jati ‘class’. A pradhana is comprehended relative to an apradhana ‘sec- ondary’. We do not have anything non-principal, or secondary, relative to which kukkufa can be accepted as ‘principal’. Consequently, the paryudasa, or operational interpretation of negation, is faulty (casmad vidhi pakso' yam dustah). The negation in anupasarjanat is therefore accepted as prasajya, whereby we understand that ‘MP, etc., do not occur after that which is upasarjana ‘secondary; non-principal’.” Now consider kurucari, a tatpurusa compound (2.2.19 upapadam atin) formed with kuru + sUPand cara, where carais a derivate of Ta (3.2.16 cares Jah), a kyt affix introduced after verbal root car under the co-occurrence condition of kurz + suP. The word cara cannot be accepted as an upasarjana since it is part of a tatpurusa compound where meaning of the following constituent is principal (uttarapadapradhana). It can also not be accepted as an upasarjana because it is not specified with a nominative in the wording of the rule which allows formation of this compound. Consequently, we get affix NP. The same could not be true of bahukurucara ‘an area full of those who wander in the country of the Kurus’, a bahuurihi compound where indi- vidual constituents are all considered as upasarjana. A NiP will thus be blocked. Incidentally, the final a of kurucara and madracara will be deleted by 6.4.147 yasyeti ca, before the MiP of 4.1.15 tiddhdnaii. . ., of course, intro- duced at the strength of Tas an itin affix Ta. 2. Note that anupasarjanat qualifies pratipadikat which, when read with atah of 4.1.4 ajadyatas (ap, will yield this following interpretation: ‘an affix occurs to denote feminine after a nominal stem which ends in a, and is not an upasarjana...” This negation of anupasarjanatis not required if tadantavidhi ‘a specifica- tion made by an item also refers to that which ends in that item’ is not accepted. For, Paribhasa (32): grahanavata pratipadikena tadantavidhir nasti20 The Astadhyayi of Panini 4115 * tadantavidhi becomes inoperative when a specification happens to be made by enumerating individual nominal stems’. But anupasarjandt is required for offering negation (pratisedha) in the following rules. The negation itself cannot be accomplished without accepting the notion of tadantavidhi. What is the way out of this difficulty? Kasika states that tadantavidhi is operative in this section (asmin prakarane tadantavidhir iti) But it becomes operative only in the context of that which is ‘non-secondary’ or ‘principal’. A varttika (Sidra cémahatpitrod jatih) under 4.1.4 ajadyatas tap makes this known. If tadantavidhi isnot accepted then, given pratyayagrahane yasmat sa vihitas tadddes tadantasya ca grahanam bhavati‘a specification made with an affix refers to a form which begins with the base after which that affix is introduced, and also to the form which ends in that affix’, the aN of the next rule will refer to a form which endsin aN. We can thus derive aupagavi female descendant of Upagu’ with the introduction of NiPafter aupagava, a taddhita derivates ending in aN. But this paribhasa will create problems in deriving kumbhakéri ‘a female potter’ where karais a derivate ending in aN. The compound base kumbhakara cannot be accepted as ending in a form (kara) which, in turn, ends in aN. That is, kara alone can qualify to receive the affix. A tadantavidhi interpreta- tion will facilitate affixal introduction after kumbhakara, a form ending in hava (Nyasa: tadantavidhau tu jfdpite kavasabdantid apikarah sidhyati). Inci- dentally, aN’of next rule is interpreted as a taddhita (4.1.88 prag divyato'n) as well as a kt (3.2.1 karmany an). 4.1.15 RESO GA USA ARTA SASH SAAT: tiddhdnatidvayasajdaghnafimatractayapthakthanihaitkvarapah /tiddhanai.. . kvarapah 5/1 = ta ityasya sa tit (bv.); tic ca dhas ca... .varap ca= tiddhanah .. . kvarap (sam. dv.); tasmat/ (pratyayah #3.1.1, paras ca #3. dtipadikat #1, striyam #3) lidadibhyah pratipadikebhyah striyam tip pratyayo bhavati Affix MiP occurs to denote feminine after nominal stems which end in a (adanta) and have, as their final (antya), a non-secondary (anupasarjana) form either marked with T, or ending in affixes dha, aN, aN, dvayasaC, daghnaC, matraC, tayaP, thaK, thaN, kaN'and KvaraP. EXAMPLES: kurucari ‘she who wanders in the country of the Kurus’ madracari ‘she who wanders in the country of the Madras’ sauparneyi ‘a female descendent of Suparni’ vainateyi ‘a female descendent of Vinata’ kumbhakari ‘she who makes pots’ nagarakari ‘she who builds cities; a female architect’ autsi ‘a female descendent of Utsa’ audapani ‘a female descendent of Udapana’