दासगुप्त-सुरेन्द्रनाथः

ICPR TRANSLATION OF Indian Philosophical Classics General Editor: Sibajiban Bhattacharyya The Mahabhaṣya of Patanjali SURENDRANATH DASGUPTA ICPR TRANSLATION OF INDIAN 25122 PHILOSOPHICAL CLASSICS THE MAHABHAṢYA OF PATANJALI ICPR TRANSLATION OF INDIAN PHILOSOPHICAL CLASSICS GENERAL EDITOR Sibajiban Bhattacharyya ADVISORY BOARD Badarinath Shukla (deceased) Visva Bandhu Bhattacharyya K. Kunjunni Raja R. Balasubramanian THE MAHABHAṢYA OF PATANJALI WITH ANNOTATIONS (ĀHNIKAS I-IV) SURENDRANATH DASGUPTA Edited by SIBAJIBAN BHATTACHARYYA jor INDIAN COUNCIL OF PHILOSOPHICAL RESEARCH. NEW DELHI First Published 1991 H Indian Council of Philosophical Research 1991 All rights reserved. No part of this publication may be reproduced in any form, or by any means, without written permission of the publisher. Published by the Member Secretary for INDIAN COUNCIL OF PHILOSOPHICAL RESEARCH Rajendra Bhavan (Fourth Floor), 210 Deen Dayal Upadhaya Marg New Delhi 110002 Distributed by MUNSHIRAM MANOHARLAL PUBLISHERS PVT. LTD. 54 Rani Jhansi Road, New Delhi 110055 ISBN: 81-215-0530-3 Typeset by Neographics Trade Centre, 2/7 Ansari Road, Darya Ganj, New Delhi 110 002 Printed at Rajkamal Electric Press B-35/9 G.T. Karnal Road, Delhi 110 033 CONTENTS Editor’s Preface FIRST AHNIKA (Paspaśā) vii 1 SECOND AHNIKA 65 THIRD AHNIKA 135 FOURTH AHNIKA 213 Index 251 EDITOR’S PREFACE Dr (Mrs) Surama Dasgupta donated a collection of unpublished works of the late Dr Surendranath Dasgupta to the library of the Indian Council of Philosophical Research, Lucknow. The present work belongs to that collection. The work as published here is incomplete in two ways. Firstly, it covers only the first four Ahnikas of Prathamâdhyaya of the Mahabhaṣya. Secondly, even of the four Ahnikas, the second and fourth Ahnikas are incomplete. There were two possible courses open here. First, to get the incomplete Ahnikas completed and to append them to the main work. This way was not followed because of the following reason. The value of the present work lies not so much in the translation of the text but in the very thorough discussions on all topics in the annotations. It was felt that a comparable treatment of the portions left out by Dr Dasgupta could be given only by someone having the width and depth of knowledge of the subject which Dr Dasgupta had. As such a scholar was not readily available, on the advice of Mrs Dasgupta the work is published in the same form in which Dr Dasgupta left it. As Dr. Dasgupta dictated different parts of the manuscript to different scholars apparently of unequal competence, the manuscript contained many errors of different sorts. I thank the editorial reader, Pandit Satkari Mukhopadhyaya of Sahitya Akademi, New Delhi, for carefully going through the manuscript, correcting mis- takes which remained after my editing, and for preparing the index. We have, however, corrected only obvious mistakes in spelling and in grammar. It is hoped that this great work, incomplete though it is, will attract attention not merely of scholars of Sanskrit grammar, but also of philos- ophers of language because of the important philosophical discussions on various topics in the annotations. April, 1991 SIBAJIBAN BHATTACHARYYAFIRST AHNIKA (Paspaśā) Panini. atha sabdânuśāsanam Bhāṣya. athety ayam sabdo’ dhikārārthaḥ prayujyate. śabdanuśāsanam nama śāstram adhikrtam veditavyam. The object of grammar is directly specified here by the elaborator of grammar, Patanjali - ‘atha’ etc. Now (follows) the instruction of words. The word ‘atha’ is used to indicate the commencement of the topic. Though the topic of the instruction of words is a long process, the word ‘atha’ indicates the commencement of such a topic. The action of instruction has for its object, ’the words’. The word ‘adhikara’ means prastava or the commencement of a topic.1 Bhāṣya. keṣām sabdānām? laukikānām vaidikānām ca. tatra laukikās tāvad gaur aśvaḥ puruşo hasti sakunirmṛgo brāhmaṇa iti. vaidikāḥ khalvapi - śam no devir abhiṣṭaye; işe tvorje tva; agnim ile purohitam; agna āyāhi vitaya iti. Of what words? Of the popular and the Vedic. There the popular are: the cow, the horse, the man, the elephant, the bird, the beast, the brahmin and so forth. Also there are the Vedic (propositions as illustrated in the Atharvan, Yajus, Rc and Saman)- sanno devir abhiṣṭaye

may the

1. Dikṣita in the Sabda-kaustubha asserts that the word ‘atha’ which has the grammatical status of a nipata has only an indicatory function (dyotakatva) like the upasarga and not an explicatory function (vācakatva). The word ‘atha’ indicates that now is the commencement of the topic of instruction of words. The word ‘anusāsana’ means that the function of the instruction consists in distinguishing correct from incorrect words (vivicya bodhakatvam - vivicya asadhubhyo vibhajya bodhyante yena iti karane lyut). The word ‘instruction’ again as no reference to the instructor and for this reason the sixth case ending in sabdasya anusāsanam is by the rule kartṛ- karmanoḥ kṛti and not by the rule ubhaya-präptau karmani and hence there is no legitimate objection against compounding sabdasya with anusāsanam. 2 Mahābhāṣya of Patanjali shining good be for the sacrifices (A. V. 1.1.6); işe tvorje tva-I (cut thee) for libation and strength (Y.V.1.1.1); agnim ile purohitam- I adore Agni, the leader (R. V. 1.1.1); Agna āyāhi vitaye - Oh Agni, may thou come for drinking (the clarified butter) (S. V. 1.1.1).2 Bhāṣya. atha gaur ity atra kaḥ sabdaḥ? Now the question is what is the word in cow (gauḥ).3
2. The words about which the instruction is proposed consist of such words as are used among the people, popular words and sounds. The word ‘anuśāsana’ means ’to make known’. It also means ’explaining with discrimination’. The word ’laukika’ means known among the people (loke vidita iti laukikaḥ) or that, that which has originated among the people (loke bhava iti laukikäḥ). Though the Vaidika words are known among the people, yet they are separately mentioned for their importance. The Pradipa suggests that laukika is separately mentioned on account of the fact that it is spoken. In giving example of the laukika words, words are separately mentioned at random. But in giving the example of the Vaidika words, the definite order in which they occur cannot be overridden; and it is for that reason that the illustrations of Vaidika words are given in the form of sentences in which they occur. The Uddyota comments that the Vaidika illustrations are given in the form of sentences in order to show their accents.
3. Now when one says that the cow is white, no distinction is made between the word and the object among the people in general. One has therefore to distinguish the place in and status of the word in the awareness of the object. The awareness consists of the words, the universal and the individual. As a quality and a substance often appear as unified, so the word appears as unified with object. When the word ‘cow’ is pronounced, the universal and the particular may both appear in consciousness. The quality (guna) and the action (kriya) being indistinguishable from those that possess them, also appear in consciousness along with them. There is a view that a substance is nothing but a conglomeration of qualities, and in such a view, the parts may be regarded as qualities of which the whole is a conglomeration. In the awareness of a cow its qualities are apparent in a generic manner in the cow as a whole. When we see mango, we have a general knowledge of its juiciness but still we may enquire about the specific nature of the juice. This problem that the enquiry about specific particulars is preceded by generic awareness has been more fully discussed in the Mañjūṣā. When the word ‘cow’ (gauh) is pronounced, its etymological aspect does not come immediately before awareness. The word ‘cow’ (gauh) immediately stands before awareness as a significant something referring to its significate. Just as the cow expresses its generic character which is inseparably inherent in it, so is it associated with the quality of signification. It has been said by Bhartṛhari that all objects are penetrated through and through with words. All knowledge is inter-penetrated with words. An object seen by our eyes can at the same time remind us of the word which denotes it. Just as a ray of light works in a double manner in contributing to vision as First Ahnika 3 well as to the illumination of the object, so the word also signifies the object and itself is signified by the object. (Vakyapadiya, 1.124 and 1.55). When by looking into an object one is reminded of a word, that word must have been brought to the mind by that object; so the object also may have a reverse function by virtue of which the sight of the object reminds one of the word. So, just as the genus cow is identical with the particular cow, so the word also, by virtue of its association, is identical with the object; and so also are the qualities of the cow. Under such a theory a natural question is-how to distinguish the word? Is the word then identical with the substance of the object and its qualities etc.? How can the word be at once identical with the cow and different from it? The reply of Patanjali is that negation and position become contradictory when they refer to the same specified unit. But otherwise they are not so. Thus qualities which are identical with being may be affirmed of it with impunity. So we may affirm pleasure and pain, heaven and hell, of being and no contradiction is felt. So the identity is affirmed of words with objects by virtue of their association and hence no contradiction can be anticipated. A point may arise that if grammar consists in imparting instruction on words why should it try to analyse them in constitutive parts, such as the stem, the suffix, the prefix and the like. The answer is that the words are so vast that to exhaust them by simple enumeration would be an impossibility. The analytical method serves to classify the stems, the suffixes and the prefixes and impose a number of relation for associating them together and thereby can bring the description and words within reasonable limit. This method of analysis is merely a means by which the young learners may be introduced to the infinite realm of words. Being loyal to the method of analysis, different grammarians may use different schemes. There are some thinkers who think that when we listen to a word we are reminded of its fictitious constituent parts and these are indicative of particular meanings. This view is wrong. The introduction of fictitious parts as describing the constitution of an actually used word has no locus standi and are in themselves false, though they may, in the end, lead to the comprehension of the word and its meaning just as mere lines forming a script, though in themselves meaningless and in that way false, may indicate real objects. In reality, the signification of words is due to traditional usage. A suffix can only have an import so long as it is used with words. It does not matter, if in different grammars the suffixes have different forms. A suffix, however, can only be significant, when it is used in relation to words. A verbal form is supposed to be capable of indicating its meaning not by the memory of the conventionally artificial form of its primal root (as du pacas), but in the form in which it is practically used (e.g. pacati). In such usages as pacati Devadattaḥ (Devadatta cooks) and pacantam Devadattam pasya, the suffixes tip and satr if regarded as reminding one of the lat as signifying in one case (i.e. as in pacati Devadattaḥ) the dynamicity of the verb as being qualitatively associated in the nominative Devadatta as its locus and in the other case (pacantam devadattam pasya) when it has to be supposed that the nominative is adjectival to the dynamicity of the verbal part we have a contradiction of principles as held by the Naiyayikas. The Mimamsakas also fall into the same difficulty when they deny the verbal action on the part of the nominative in one adhikaraṇa, and in another adhikarana attributes meaning of satṛ and sanac as modifying the nominative. Their 4 Bhāṣya. Mahābhāṣya of Patanjali kim yat tat sāsnā-langula-kakuda-khura- viṣāṇy-artha-rupam sa sabdaḥ? To the previous question ‘what is the sabdaḥ in gauḥ’?, a further question is raised. Does śabda denote the object which is a combination of the dewlap, the tail, the hump, the hoof and the horn? In reply in denying the supposition that the sabda of the word denotes that, Patanjali says: Bhāṣya. netyāha; dravyam nāma tat. “No,” he says, “that is called dravya or substance”. The enquirer again is supposed to ask: Bhāṣya. yat tarhi tad ingitam cestitam nimiṣitamiti saḥ sabdaḥ? Does sabdaḥ then mean the gesturing, the moving or the winking? The final reply is that Bhāṣya. netyāha; kriyā nāma sā. “No”, he says, “that is called action.” A further question is again asked: assertion that the part denoted by the suffix sanac refers the action to the nominative cannot stand a deeper scrutiny. The principle that ought to guide the problem as to which ought to play the substantive part depends upon what may be the natural complement for the completion of the sense. In the case of Devadattaḥ pacati the nominative ‘Devadatta’ is the verbal complement for the completion of the verbal action and for this reason, it may be possible to regard the verb-suffix denoted part as having adjectival to the nominative ‘Devadatta’. In that case, the meaning would be the verbal action as qualified by the nominative as modified by the suffix-denoted part. But the Naiyayikas deny it; they think that the verb as a whole- the effort included in the action is adjectival to the nominative ‘Devadatta’. But in the case of Satṛ and sanac these suffixes standing for the original verb-suffix cannot but be regarded as adjectival to the nominative which, in its turn, is adjectival to the action denoted by the verb. Thus the Naiyayikas have to take to two different contradictory theories in explaining the relation between the subject and the action. Following the view of the Nirukta, the grammar school holds that the verbs are dominated by the idea of action and names are dominated by the sense of being. The suffixes only denote the specific relations. But it is obvious that such analytic consideration of denotation is an over-adumbrage for Patañjali is a believer in the Sphota doctrine which regards all analytic considerations as being meaningless. The word or words have a complete sense and that sense cannot be so analysed that we may be able to point out parts of each meaning as being denoted by certain analysed parts of the words. The Sabda-kaustubha goes on in this connection into special Bhāṣya. First Ahnika 5 yat tarhi tacchuklo nilah kapilaḥ kapota iti sa sabdaḥ? Is it then the white, the blue, the brown, or the gray that is to be called the sabda or word? The reply given is: Bhāṣya. netyāha; guņo nāma saḥ. “No”, he says, “this is guna.” Again, if the gunas or rather those which have different colours are not to be regarded as words, then what are the words? The Bhāṣya again proposes another reply: Bhāṣya. yat tarhi tad bhinneṣv abhinnam chinneṣv acchinnam sāmānyabhūtam saḥ śabdaḥ? Should that which being broken remains unbroken, being torn remains untorn, the being, the highest genus of all things, be called sabda?" The conclusive affirmation rebutting the previous alternatives is given by Patanjali as: Bhāṣya. netyāha; ākṛtir nāma sā. “No”, he says, “it is called ākṛti.“5 The next question is: Bhāṣya. kastarhi sabdaḥ? What is then the word or sabda? The answer, as given by Patañjali, is: elaboration of the Sphota theory which seems to us to be quite irrelevant for our present purposes.
4. The phrase ‘bhinnesy abhinnam’ signifies that the genus or the class-notion is one. The phrase ‘chinnesv acchinnam’ means that the class-notion is eternal. The word ‘samanya-bhūtam’ means like the class-notion samanyam iva samanya-bhūtam. The word bhūta is used in the sense of comparison. But Nageśa offers a different interpretation to the word ‘samanya-bhūta’. Nāgeśa says that the word samanya here need not mean the highest genus or class-notion; for in that case, it would not be possible to denote the class-notion of a smaller genus, the cow. The word ‘sāmanya’ simply means that which has an aspect of generic character. According to Nāgeśa there is no idea of comparison here. The word ‘samanya-bhūta’ means that entity which has a generic aspect.
5. Akṛti means jāti or class concept and also the build of the body (samsthāna). The akṛti on the one hand displays a definite structure and build and at the same time reminds one of the class-notion. 6 Bhāṣya. Mahābhāṣya of Patanjali yenoccaritena sāsnā-langula-kakuda-khura- viṣāṇinām sampratyayo bhavati sa sabdaḥ. That which being pronounced leads to the comprehension of the animal possessing dewlap, tail, hump, hoof and horn is called ‘sabda’ or word.6
6. The word ‘uccaritena’ has been interpreted by Kaiyata not as ‘pronounced’ but as ’expressed’ (prakāśitena). According to the grammarians, the denoting entity is something different from the alphabets constituting a word (pada) or a sentence (väkya). They argue that if the letter-sounds individually denote the object then the pronouncement of the first letter-sound would be enough and the pronouncement of the other letter-sounds constituting the word would be unnecessary. If each letter- sound has a meaning then they jointly cannot produce the unified meaning-the cow. If it is held that the manifestation (abhivyakti) of the meaning takes place in succession, then as the letter-sounds vanish each moment that they are produced, it has to be admitted that the combination of the different letter-sounds contributing to the final manifestation takes place on the basis of memory; on such an alternative the word ‘sara’ will have the same manifestation of meaning as ‘rasa’ for the constituent letter-sounds as held in memory are the same. It is for this reason that the grammarians admit the existence of a separate transcendent whole called the sphota. Here on this view the pada is a transcendent whole different from the varņas. We have already shown why the letter-sounds or the varnas can neither separately nor jointly manifest a meaning. The argument that the memory of the letter-sounds may integrate the letter-sounds in the same temporal order in which they were pronounced is objected to by the grammarians who hold that the sameness of order cannot be guaranteed in the case of memory. Nageśa in explaining the situation takes to the analogy of a painting. On a canvas something may be painted by various colours in a particular order, but the finished painting produces a sense of wholeness which cannot be deduced from the manner and order in which the strokes of the brush were applied. Bhattoji Dikṣita in his Sabda-kaustubha speaks of eight kinds of sphota such as varṇa-sphota, pada-sphota, vakya-sphota, akhanda-pada-sphoța, and so on. In an- other type of classification, he speaks of five kinds of vyakti-sphota and three kinds of jati-sphota. When the same consonance in different combinations yields different kinds of meanings, we have the case of varṇa-sphoța. Thus, kar, kār, kur make different meanings. In such cases as Rāma, Rāmeņa, Rāmāya, it is impossible to determine what part of it denotes the substantive character of Rama and what part of it denotes the case characters. This is called pada-sphota. Again, in dadhidam, Hare’eva, there being the same difficulty as to which of the parts denotes what, it has to be admitted that the sentences as such have crystallised meanings. This is called vakya-sphota. As a piece of pictorial art manifests its wholeness through all its constituent parts, so a pada or a vakya also manifests its character as a whole, though the manifestation is suggested or imparted by the individual constituent sounds. This appearance of wholeness is not conditioned by the parts, for in the expression of a picture also, one might regard the parts as playing the role of conditions (upādhi). A First Ahnika 7 difference must here be noted between vyañjana and upadhi. In the case of vyañjanā, something is projected which was nct present in the projecting constituents. Thus rasa or an aesthetic emotion is projected by a piece of literary art and so a picture may be projected through the diverse placings of colours and lines. But an upadhi is that which modifies in some manner or other, the object or objects which is perceived in and through the upadhi. The modification may be special, temporal or the attribution of new characters. Bhattoji maintains that the projection of the type of sphota is a case of vyañjana and not of upadhi. The next question is, under what category, can we include sphota. The reply is, that though it may not be included within the well-known categories of the philosophies like that of Vaiseṣika and Samkhya, it is a category of reality (bhava). No objection can be raised against it, on the ground that it is not included within the well-known categories, for since its existence is logically proved, it does not matter under what category it may be included. No philosopher who evolved any set of categories of existence ever put a limit to the number of individual existence. In the case of upadhis, it is always possible to set a limit to the conditional modifications. Or one may divide the categories as existent or non-existent, eternal or non-eternal. In the case of the categories of existence, it is fruitless to make any enumeration of appearances due to the operations of upadhi. In both these alterna- tives, the varnas are unnecessary. The manifesting set of sounds of a unique nature conditions the sphota which is manifested as ‘ka’ or ‘kha’. This leads to the monistic theory of sphota, such that, the one sphota may manifest itself as diverse sphotas of varnas, padas, etc. through the diverse conglomeration of sounds co-operating together as conditions or upadhis for the manifestation of particular sphotas as ‘ka’ or ‘kha’. In support of his view Bhattoji quotes from the Tattva-bindu which is supposed to hold that the sphota of ‘ka’ is not different from the sphota of ‘ga’. In counting on Bopadeva’s support, he says that as letter-sounds may be pronounced in shrill or a tender voice, so does one sphota manifest itself in the form of diverse letter- sounds. Bhartṛhari also in his Vakyapadiya (1.73) says that just as the parts do not exist in the varnas, so the varnas do not exist in the padas and the vakyas also differ absolutely from the padas. For this reason, one has to admit the theory of akhanda- pada- sphota and akhanda-vakya-sphota. These together with the varṇa-sphota and vakya-sphota as described before give us the fivefold sphotas. There are others, however, who hold the theory of jati-sphoța. According to this school of thinkers, the letter-sounds denote the jati or the class-notion and these class-notions through a gradual widening of its limits ultimately refers to Brahman. Thus according to this view, Brahman is the ultimate significance of words but different meanings appear on account of the conditioning apparatus (upadhi) of the different letter-sounds. There is, however, another view as to which different meanings or significance are projected by words due to the diverse manifestations of the functions of avidya (avidya-vṛtti-brahmatattvameva tattadupahitam väcyam vācakañca avidya avidyāka- dharma-viseṣo vā jātir iti pakṣe tu saiva vācikā astu ityāhuḥ). Bhattoji holds that as already said that there are eight types of sphota. Both Patanjali and Kaiyața admit the varṇa-sphota, pada-sphota and vakya-sphota views. But true eternity according to them belongs only to the jäti-sphoța. Every word ultimately expresses the ultimate reality, the Brahman or rather it is the reality of the 8 Bhāṣya. Mahābhāṣya of Patanjali kāni punaḥ śabdânuśāsanasya prayojanani? What are the needs of the instruction regarding words?? Brahman that appears as words (anupadam eva brahmatattvam eva hi Sabda-rupataya pratibhāti). The other view of course is that avidya is the jāti and from that point of view in the phenomenal aspect jäti is also eternal. The analysis of grammar is all unreal just as the scripts and writings of all descriptions are all unreal. The real aspect, however, is the concrete expression of the whole as expressed in concrete wholes through words or propositions. Bhattoji holds that this view has been maintained by Patanjali and Kaiyața. Bhartṛhari also believes that the jati or the class concept is ultimately the Brahman. Thus he says it is only the Being (sattva) that appears diverse as the cow or the cat in diverse relations and this is the highest universal or jäti and all words are ultimately grounded in them. All nouns and verbs signify this reality. This is the eternal Great Self. This is signified by all suffixes like tva and tal that signify abstractness. Though a reference has been made to eight kinds of sphotas, both Patanjali and Kaiyata believe in the doctrine of vakya-sphota, ultimately leading to jati-sphota. The position may briefly be stated thus: a proposition or a combination of propositions forming a vakya or sentence signifies a unified whole and this through a hierarchy ultimately leads to the concept of the pure Being which involves within it all specific forms of Being as being impositions on it. It is only by a reverse process and for illustrative purposes that we adopt the process of grammatical analysis. Philosophers who adhere to other systems of thought start at the wrong end and believe that we must begin with the analysis of words and the relations to suffixes, prefixes and the like and then proceed upwards on the analogy of atomic and molecular construction to bigger wholes as forming the propositions. There are, however, situations in which such a view seems to have a better footing. Bhartṛhari also has said that what one school regards as primary is regarded as secondary by other schools. He further says that in the synthetic word the ‘brāhmaṇa-kambala’ (the blanket of the Brahmin) there is no separate meaning of the word ‘brāhmaṇa’; so in a proposition or a combination of propositions the individual words have in themselves no individual meaning (Vakyapadiya, 2.14). Thus the function of ’nipatas’ as suggestiveness and all the vikaranas as being without meaning and the like are analytical discussions for the comprehension of the structure of words while in a state of formation. Thus Bhartṛhari that the distinction of the upasarga and the dhatu has been made for the sake of advising as to the place that is to be taken by certain prefixes like at and the like. In reality, however, the dhatu and the upasarga together constitute the dhatu (Vakyapadiya, 2.182). Thus also the name and its meaning are identical. The supposition of the special importance of the suffixes has only value as illustrative of the constructive process of the word. These discussions regarding the various functions of prefixes, suffixes and the like, though ultimately unreal still have a significance analogous to the science of scripts such as letters and the like to the understanding of the composition. says
7. Kaiyata points out that the question is whether the study of grammar is to be regarded as an obligatory duty or an optative work. Nageśa, however, puts it First Ahnika 9 Bhāṣya. rakṣohāgamalaghvasandehaḥ prayojanam. The needs (of studying grammar) are preservation, modification, injunction, brevity and certainty.8 Bhāṣya. rakṣārtham vedānām adhyeyam vyākaraṇam. lopāgama-varṇa-vikārajño hi samyag vedan paripālayiṣyati. The grammar is to be studied for the preservation of the Vedas. Only the person who knows about elisions, introductions and the change of sounds can properly follow the Vedas.9 Bhāṣya. uhaḥ khalvapi-na sarvair lingair na ca sarvabhirvibhaktibhir vede mantrā nigaditāḥ. te câvaśyam yajña-gatena puruşeņa yathā- yatham vipariņamayitavyāḥ. tānnāvaiyā- karaṇaḥ śaknoti yathāyatham vipariņa- mayitum. tasmād adhyeyam vyākaraṇam. Certainly also modification; in the Vedas the mantras are not always uttered in all genders and with all kinds of suffixes. The person engaged in sacrifice must properly modify them. No one who is not a grammarian otherwise and thinks the word prayojana means prayojaka, i.e. the question is whether any real purpose is served by the study of grammar. Bhattoji, however, says that it falls within the obligatory duties of a Brahmin to study the Vedas and as grammar is an accessory literature (arga) of the Vedas, its study is also obligatory. So the study of grammar has the same obligatoriness as the performance of the daily sandhya prayers.
8. Rakṣā or preservation means the preservation of the Vedas. There exist such forms in the Vedas as are not found in spoken Sanskrit and unless people read grammar, ignorant people might, following the usage of ordinary Sanskrit speech, consider the Vedic forms to be erroneous. The Vedic language differs from spoken language in various ways. There may be elisions, introductions, elision and introduc- tion both, change of sounds and the like.
9. Thus certain suffixes are elided only in the Vedas. Thus in lan of the verb duha, the ta of the suffix at the third person plural number is elided and we have the form deva aduhra which is different from the current Sanskrit expression aduhata. Again sometimes sounds are changed; instead of udgraha and nigraha we have the Vedic forms udgrābha and nigrābha.10 Mahābhāṣya of Patanjali can make the proper modification. For that reason one should study grammar. Bhāṣya. 10 agamaḥ khalvapi-‘brāhmaṇena niṣkāraṇo dharmaḥ ṣadango vedo’dheyo jñeyaśca’ iti. pradhanam ṣadangeṣu vyakaraṇam. pradhāne ca kṛto yatnaḥ phalavan bhavati. There is also the Vedic injunction- ‘a Brahmin should, without any objective, study and know the dharma as the Veda with its six angas.’ Grammar is the prominent of the six angas. Effort made in the prominent anga produces fruit.11 Bhāṣya. laghvartham câdhyeyam vyakaranam. ‘brāhmaṇenâvasyam sabda jñeyāḥ’ iti. na cântareṇa vyakaraṇam laghunopāyena śabdāḥ śakyā jñātum.
10. There are two kinds of sacrifices called prakṛti and vikṛti. Prakṛti is that kind of sacrifice in which all accessories are mentioned fully, such as darśa-pūrṇamāsa. Vikrti is that kind of sacrifice in which the details are not mentioned, but the details of the mantra is to be imagined by the analogy of the prakṛti sacrifice. Thus when in the prakṛti sacrifice the mantra is given; ‘agnaye tva justam nirvapāmi’; in the vikṛti sacrifice there is a caru-yaga and the vidhi is sauryam carum nirvapet brahmavarcasa- kāmaḥ, i.e. he who desires the brahminic spirit should offer the caru to Surya, but the exact mantra of the offering of caru is not given; here the mantra is to be formulated after the analogy of the mantra of offering made to Agni as specified in the prakṛti. It has been noted above that the mantra of offering obtains to ‘agnaye tvā justam nirvapāmi’. Now the mantra offering oblations to Surya must be formulated in analogy to this mantra by changing agnaye tvā to sūryāya tvā. Such a change cannot be made if the sacrificer has no knowledge of grammar.
11. The word karana in niskāraṇa means ’end’ or ‘object’ (phala). It is suggested by Nageśa that a study of grammar is obligatory for a Brahmin and optative for others. The fruitfulness of the study of grammar consists in this alone that only by a study of grammar one can understand the meanings of the Vedas and the other angas which consist of speech. Nageśa in commenting on the word ‘dharma’ says: ‘dharmatvañca vedasya purusa-yatna-sadhyataya dharmatvenâbhimatâdhyayana-jñāna-karmatvena- upacārikam’, i.e. though dharma cannot really be the object of study or of knowledge, yet since the Vedas are the source of dharma, a study of the Vedas and its knowledge are regarded as dharma in a secondary sense. First Ahnika 11 Grammar is to be studied for the sake of brevity. ‘A Brahmin must know words.’ There is no other simple method of knowing words than grammar, 12 Bhāṣya. asandehartham câdhyeyam vyakaraṇam. yājñikāḥ paṭhanti-‘sthula-pṛṣatīm āgni- vāruṇīm anaḍvāhīm ālabheta’ iti. tasyām sandehaḥ-sthūlā câsau prṣati ca sthula- pṛṣati, sthūlāni vā pṛṣanti yasyāḥ seyam sthūlapṛṣatīti? tām nâvaiyakaraṇaḥ svarato’- dhyavasyati. yadi pūrva-pada-prakṛti- svaratvam tato bahuvrihiḥ, atha samāsāntodāttatvam tatas tatpuruşa iti. Grammar should be studied also for attaining certainty. The sacrific- ers read as follows: sthula, etc. Now there is a doubt whether the word ‘sthula-pṛṣati’ is a tatpuruşa or bahuvrihi compound. In the first case it would mean that which is fat and has spots on it or in the other case it would mean that has large spots on it. No one but a grammarian would distinguish the difference between the two (in the first case meaning a cow which is fat and has spots on its back and in the second case meaning having large spots on its back)-describing the cow which is to be sacrificed in honour of the gods, Agni and Varuna (sthula-prṣatim agni- vāruṇīm anaḍvāhīm ālabheta) leading to the accent of the first pada in the case of bahuvrihi or leading to an udatta accent at the end of the compound word in the case of the tatpurusa.13
12. A brahmin earns his living by teaching and no teaching is possible without words. Again, words being limitless they can only be studied through grammar which systematizes them under general principles and exceptions and it is this method that makes grammar the simplest method of learning a language.
13. The word ‘asandeha’ involves a negation of the pragabhava type, i.e. the grammarian has no doubt in his mind which is later on destroyed. In his case no doubt arises. In the case in question the samāsa to be accepted is bahuvrihi and the accent is to be given on the first member of the compound (purva-pada-prakṛti-svara). From the manner of the accent of such Vedic passages, which the sacrificer learnt from the teacher, he knows that the cow he has to sacrifice is such that would have large spots on it and this he could not know unless he was a grammarian. A grammarian knows that a word which has accent on the first member is a bahuvrihi compound. So by the manner of his own reading he can clearly comprehend the meaning of the compound as being bahuvrihi which is of practical importance for him for selecting the sort of the cow he requires. 12 Bhāṣya. Mahābhāṣya of Patanjali imāni ca bhūyaḥ śabdânuśāsanasya aniwprayojanani-’te ‘suraḥ’, ‘duṣṭaḥ sabdaḥ’; ‘yad adhitam’; ‘yastu prayunkte’; ‘avi- dvāmsaḥ’; ‘vibhaktim kurvanti’; ‘yo vā imam’; ‘catvāri’; ‘uta tvaḥ’; ‘saktumiva’; ‘sarasvatim’; ‘dasamyām putrasya’; ‘sudevo’si varuna’ iti. There are further the following objects for the instruction of words: “Those asuras’; ‘wrongly pronounced word’; ‘whatever is studied’; ‘whoever uses’; ’the ignorant’; ’they make the suffixes’; ‘he who (reads) these’; ’the four’; ‘and what one does’; ’like the powdered pulse’; ‘be-’ longing to Sarasvati’; ‘of the son, on the tenth lunar day’; ‘o Varuna, you are highly shining’.14 Bhāṣya. te’surāḥ- ’te’sura helayo helaya iti kurvan- taḥ parābabhūvus tasmād brāhmaṇena na mlecchitavai nâpabhāṣitavai. mleccho ha vā eşa yadapaśabdaḥ’. mleccha mã bhūma’ity adhyeyam vyākaraṇam.-te’surāḥ. Te’surah (those asuras): those asuras were defeated because they pronounced ‘helayah’, ‘helayah’ (i.e. instead of pronouncing the word ‘arayah’ the enemy, as preceded by the vocative term ‘he’ with a long pluta accent, the asuras on account of their defective speech pronounced it as ‘helaya’; therefore the brahmin should not pronounce as a mleccha does and should not use vulgar words: it is the vulgar word that is called a mleccha word. Grammar ought to be studied in order that we may not be mlecchas.15
14. These cryptic phrases which have been interpreted by Patanjali in the imme- diately succeeding passages carry no meaning to us in themselves and it is not at all intelligible why Patañjali should enumerate them as well-known cryptic phrases unless in his time these have already attained popular celebrity. The commentators are practically all silent about the significance of the enumeration of these cryptic phrases. These objections are regarded as being accessory (anuṣangika) to the items of usefulness of studying grammar which has just been enumerated.
15. Kaiyata points out a different reading in as the use of “haihayo” instead of ‘haihe’ by two asuras. The mistake in this case consists in dropping the use of the vocative pluta accent. In the reading accepted above, the mistake consists in Bhāṣya. First Ahnika duṣṭaḥ śabdaḥ- duṣṭaḥ sabdaḥ svarato varṇato vā, mithya-prayukto na tamartham aha/ sa vāg-vajro yajamānam hinasti, yathendraśatruḥ svarato’ parādhāt // iti duṣṭāñchabdān mã prayunkṣmahity- adhyeyam vyākaraṇam.- duṣṭaḥ sabdaḥ. 13 A word wrongly accented or sounded, is a confounded use and cannot convey the desired meaning. That speech is like a thunder that injures the sacrificer just as the (word) ‘Indra-śatru’ (did) on account of its wrong accent. We ought to read grammar so that we may not pronounce words wrongly. Just as the homa performed by adhvaryu produces virtue for the sacrificer, so the mispronounciation of the adhvaryu injures the sacrificer (yajamāna). The story of the mispronunciation of the word ‘Indra-satru’. When Viśvarupa, the son of Tvastr was killed by Indra, Tvastr began a sacrifice for producing a son who would be able to kill Indra. In doing this sacrifice he uttered the mantra Indra-śatru (slayer of Indra) vardhasva (grow)’. Here the object of sacrifice is to make that one grow who would be the slayer of Indra, and therefore according to grammar the udatta accent should be on the last member, but the priest had by mistake put the udatta accent on the first member thus completely changing the meaning of the compound and producing a son who would be killed by Indra (indraḥ śatruḥ ghatako yasya). pronouncing ’la’ instead of ‘ra’ and in conjoining two padas in the vakya ‘helayaḥ’ instead of pronouncing two padas ‘he’ and ‘arayah’ separately. This prohibitory injunction is of the nature of arthavada. The arthavada consists generally in the narrative of a story in strengthening of the emphasis of an obligatory or a prohibitory injunction. It is supposed that there is a prohibition in the śruti against the vulgariza- tion of words in pronouncing as leading to defect and the present story of the asuras is an arthavada for strengthening the force of the prohibition. It is suggested that this prohibitory arthavada may be treated as having the force of a prohibitory injunction. The word apabhāṣa is used to refer to such words which through error or confusion slightly differ from the usage sanctioned in grammar (apaśabdatvam vyākaraṇânu- gatasabdasyeṣad bhramsana eva prasiddham-Nageśa). The ‘mleccha’ here does not mean either particular races of people or country. It only means speech not sanctioned by grammar and thus only means ‘unacceptable’ (nindya). Bhattoji says that in the Satapathabrāhmaṇa (Madhyandina) the reading is ‘helavo helavah’ instead of helayo helayah as given here. The present reading does not seem to be traceable. 14 Mahābhāṣya of Patanjali Bhāṣya. yad adhitam- yadhitamavijñātam nigadenaiva śabdyate/ anagnaviva śuṣkaidho na tajjvalati karhicit // tasmād anarthakam mādhigiṣmahitya- dhyeyam vyakaranam. -yadadhitam. That which is only studied and merely pronounced but not properly understood does not ever burn like dry wood (in a hearth) without fire. So grammar should be studied so that our reading may not be fruitless.16 yastu prayunkte- Bhāṣya. yas tu prayunkte kuśalo visese sabdan yathavad vyavahara-kale/ so nantam apnoti jayam paratra vagyogavid duşyati câpaśabdaiḥ // Kaḥ? vagyogavid eva. kuta etat? yo hi sabdan jānāty apaśabdan apуasau jānāti. yathaiva hi sabda-jñāne dharmaḥ, evam apaśabda-jñāne’ pyadharmaḥ. atha vā bhūyan adharmaḥ prāpnoti. bhūyāmso’ pa- śabdāḥ, alpiyāmsaḥ śabda iti. ekaikasya hi sabdasya bahavo’pabhramśaḥ. tad yathā gaur ity asya sabdasya gāvi goņi gotā gopotaliketyādayo bahavo’ pabhramśāḥ. atha yo’vagyogavid ajñānam tasya saranam. Who uses an adept in special peculiarities (i.e. which word is correct in which sense and the like) who uses words properly at the time of usage attains victory in the other world. (The person) who knows the proper conjunction of words is corrupted by vulgar words. (Is corrupted by vulgar words)-
16. The above verse is found in the Nirukta (only there is the reading grhitam instead of adhitam). This is preceded by the following passage: sthānurayam bhāraharaḥ kilābhūd adhitya vedam na vijānāti yo’rtham/ yo’rthajña it sakalam bhadramaśnute nākameti jñāna-vidhūta-pāpmā // First Ahnika 15 Who? Only he who knows the conjunction of words. Why so? He who knows the correct words also knows the incorrect or vulgar words. As there is merit in knowing correct words there is demerit in knowing vulgar words.orok ze Or, rather there is chance of there being more demerit. There are more vulgar words and less correct words. For each correct word there are many vulgar usages. Thus the correct word ‘gauh’ (cow) has many vulgar usages-gavi’, ‘goni’, ‘gotā’, ‘gopotalika’, and many other vulgar forms. He who does not know the proper conjunction of words is only to thank (his) ignorance.17 Bhāṣya. viṣama upanyāsaḥ-nātyantāyâjñānam saraṇam bhavitumarhati. yo hyajānan vai brāhmaṇam hanyāt surām vā pibet so’pi manye patitaḥ syāt. The thesis is contradictory. Ignorance cannot be ultimate (lit. endless) support. One who unknowingly kills a brahmin or drinks liquor may well be supposed to have fallen low.18 Bhāṣya. evam tarhi- ‘so’nantam apnoti jayam paratra/ vagyogavid-dusyati câpaśabdaiḥ // kaḥ? avagyogavid eva. atha yo vāgyogavid vijñānam tasya saranam.
17. The above represents the opponents’ point of view. The opponent is supposed to object to the accruing of merit through the study of grammar. He is supposed to hold that the grammarian in knowing the correct words naturally knows also the limitless incorrect words. Therefore if he happens to attain any merit by the knowledge of correct words, he attains a much greater measure of demerit through the knowledge of incorrect words. In such a situation, it is the grammarian who is in a grave situation for through his knowledge of incorrect words he is being corrupted in a limitless manner (väg-yogavid eva dusyati câpaśabdaih). The non-grammarian, however, the objector is supposed to think, is obviously in a better position, for since he does not know the correct word, he cannot know such a variety of incorrect words as a grammarian would do. He would probably know one of the incorrect forms and be quite ignorant of the variety of other forms. His ignorance therefore saves him from much mischief which would come to the learned man through his knowledge of many incorrect forms naturally associted with the knowledge of correct forms.
18. The word atyantaya is an obscure avyaya word having the same meaning as atyantam. The purport here is that there is a difference between men and beasts. The beasts are absolutely shrouded in ignorance and are therefore below good and evil, 16 Mahābhāṣya of Patanjali So then he attains endless victory in the other world; he who knows the conjunction of words is corrupted by (his knowledge of) vulgar words. Who? Only he who is not conversant with the conjunction of words. But he who has knowledge of the conjunction of words has knowledge for his support.19 Bhāṣya. kva punar idam pathitam? bhrājā nāma ślokāḥ. Where has it been read? These are verses known as bhrājā slokas (i.e. verses composed by Katyāyana as Nāgeśa points out).. Bhāṣya. kiñca bhoḥ ślokā api pramāņam? Oh! Are verses also to be regarded authoritative? Bhāṣya. kiñcâtaḥ? Why such a question? Bhāṣya. bivago yadi ślokā api pramāņam ayam api ślokaḥ pramāṇam bhavitum arhati- “yadudumbara-varṇānām ghaṭīnām mandalam mahat/ pitam na gamayet svargam kim tat kratugatam nayet //” iti. pramatta-gita eşa tatra-bhavatah. yas- tvapramatta-gitastat pramāņam.-yastu prayunkte. If verses are to be regarded as authoritative, then the following verse is also authoritative (says the objector): “If (the contents) of a collection of copper-coloured jugs (i.e. liquor) do not, when drunk, lead (one) to heaven, will that be attained through sacrifice?” but man who has intelligence and who has the right to know good and evil is expected to make an effort towards a knowledge of them. He cannot offer ignorance as a plea.
19. The idea is that mere knowledge of vulgar words does not cause any demerit. It is the use of vulgar words that causes demerit. He who knows the science of words knows both the correct and incorrect words, but he uses only the correct words. Me First Ahnika 17 This is a mad man’s song that you have quoted. That which is not a mad song may be authoritative.20 Bhāṣya. avidvāmsaḥ- “avidvāmsaḥ pratyabhivāde namno ye na plutim viduḥ/ kāmam teṣu tu viprosya striṣvivāyamaham vadet //” abhivāde strivan ma bhūmety adhyeyam vyākaraṇam. -avidvāmsaḥ. The ignorant-“Those ignorant persons who at the time of returning a greeting do not know that the pluta is to the name (of the person first greeting) may be addressed at pleasure after returning from a distant place merely as ‘I am here’ as one might do with regard to women.” Grammar should be studied so that at the time of greeting we may not be like women.21 Bhāṣya. vibhaktim kurvanti- yājñikāḥ paṭhanti-‘prayājāḥ savibhaktikāḥ kāryaḥ’ iti. na cântarena vyakaraṇam prayājāḥ savibhaktikāḥ śakyāḥ kartum. -vibhaktim kurvanti.
20. The idea is that the sloka of Katyayana quoted above is authoritative because it has been written by a wise person and Kaiyața observes that this has a support of the Vedas also; cf. ekaḥ śabdaḥ sujñātaḥ susthu prayuktaḥ, svarge loke kamadhuk bhavati, i.e. even a single word when properly known and properly used may produce all the fruits of a man’s desire in the world of heaven.
21. It is the rule that a surperior person when returning a greeting made by an inferior person should call the latter by his name with the pluta vowel. It is also the rule that when an inferior person addresses a superior person he does it as follows– ‘I greet you, I am Devadatta’ - but in the case of offering a greeting to women one should not mention one’s name but say merely (in the form) ‘I great you’ (ayam aham abhivadaye). But as already said, in the other case one should say abhivadaye Devadatto’ ham. But with reference to the person who does not know that at the time of returning the greeting of an inferior person he should return the greeting by accosting him by his name with the pluta at the end, may behave as he might do to a woman, i.e the inferior person may greet such a superior person without mentioning his own name as he would do in the case of greeting a woman. 18 Mahābhāṣya of Patanjali The sacrificial priests read thus-“The prayāja mantras should be (suitably) associated with suffixes”, but it is not possible to do so without studying grammar.22 Bhāṣya. yo vā imām And who this “yo va imām padaśaḥ svaraśo’ kṣaraśaś ca vācam vidadhāti sa ārtvijino bhavati. 99 ārtvijīnāḥ syāmetyadheyam vyākaraṇam. -yo vā imām. “And who properly articulates this speech according to accent and according to consonant and vowel sounds becomes a priest.” Grammar ought to be studied so that we may be priests. Bhāṣya. catvāri- “catvari śriga trayo asya pādāḥ dve širṣe sapta hastāso asya / tridhā baddho vṛṣabho roraviti maho devo martya āviveśa //” Magm ‘catvāri śrigāṇi’-catvāri padajātāni nāmākhyātopasarganipātāś ca. ’trayo asya Sievi ir pādāḥ’—trayaḥ kālā bhūtabhaviṣyad- — vartamānāḥ. ‘dve sirşe’- dvau śabdātmānau nityaḥ kāryaś ca. ‘sapta hastāso asya’ sapta vibhaktayah. ’tridha baddhaḥ’-trişu sthāneṣu baddha urasi kanthe śirasīti. ‘vṛṣabho’ varṣaṇāt. roraviti-sabdam karoti. kuta etat? rautiḥ śabda-karmā. ‘maho devo martya āviveśeti’- mahān devaḥ sabdaḥ. martya maraṇadharmāņo manuṣyās tān āviveśa. mahatā devena naḥ samyam yathā syad ityadhyeyam vyākaraṇam.
22. The prayāja sacrifice mantras are generally directed with proper suffixes, but if after the commencement of the sacrifice the priest has a stomach-ache or is otherwise ill, then a separate compensatory sacrifice is to be performed. In that connection it has been directed that the prayājas should be associated with proper First Ahnika M 19 Four-“It has four horns, three legs, two heads, seven hands, bound in three-wise, the bull bellows, the great deity has entered the dying ones.” (R. V. 4.58.3). Four horns are the four kinds of padas or nouns, verbs, prefixes and the particles. It has three feet (which means) three times-past, present, and future. The two heads (are the) two types of words as eternal and as produced. Its seven hands-the seven declensions. ‘Bound in three ways’ -bound in three places, the heart, the throat and the head. “The bull’- because it showers. ‘Bellows’- because it makes sounds. Why so? The root ru indicates a verb having its object, sound, i.e the making of sound. The great deity entered the dying ones- the great deity is the word. The dying ones-men who are mortal; (it) entered them. Grammar ought to be read so that we may have similarity with the great deity.23 Bhāṣya. apara aha- “catvāri vāk-parimitā padāni

23. The word is here symbolised as a bull. The showering which is the etymolo- gical sense of vṛṣabha from the word vṛs, to shower, indicates that just as the cloud showers water, so on account of the qualities that the word possesses it satisfies all20 Mahābhāṣya of Patanjali Words are only fourfold. The four kinds of padas are nouns, verbs, prefixes and particles. These are known only by scholarly Brahmins. He who collects (within him) the mind is called wise (manisin). The three hidden in the cavern do not twinkle or move. It is only the fourth part of the speech that belongs to men. Turiya means the fourth.24 Bhāṣya. uta tvaḥ- “uta tvaḥ pasyan na dadarśa vācam uta tvaḥ śrnvan na śṛnotyenām / uto tvasmai tanvam vi sasre jayeva patya usati suvāsāḥ //” ‘uta tvaḥ’-api khalvekaḥ pasyannapi na paśyati vācam. api khalvekaḥ śṛṇvannapi na śṛṇotyenām iti. avidvāmsam āhârdham ‘uto tvasmai tanvam visasre’ tanum vivṛnute. ‘jayeva patya usati suvāsāḥ’-tad yathā jāyā inpatye kamayamānā suvāsāḥ svam ātmānam vivṛṇute, evam vāg vägvide svātmānam vivṛṇute. vanno vivṛṇuyad ātmānam domainityadhyeyam vyākaraṇam.-uta tvaḥ. “There are others who though perceiving cannot see speech; others again though listening cannot hear it(i.e. speech). There are others again to whom she reveals herself as a well-dressed wife desiring her husband would do."(R. V. 10.71.4) our desires if it can be properly used. It is said that the word entered into the mortals. This means that it revealed itself as identical with it. It is the object of the science of grammar to reveal the great deity, the word as Brahman. It may therefore be supposed that he who knows grammar becomes inspired with it. The aspirate ha is supposed to have its origin in the heart and some varņas such r, the ta group, ra and șa are supposed to have sprung from the head. The eternal form of sabda represents the overlord, the omnipotent, the Great who is symbolised as the word ‘bull’ and the man who knows grammar by using words with full knowledge can free himself from all sins and can ultimately demolish his egoism.
24. The above verse is of similar import as the one referred to before, which is quoted by Patanjali in support of the previous one. The phrase ‘Parimită padani’ is irregular, being a Vedic form. In the ordinary form it should be parimitāni as adjective to padāni, which means measured or limited. Nägeśa says that the word ‘manişin’ First Ahnika 21 ‘And others’. There is again some one who though perceiving cannot see speech and there is again some one who though auditive cannot hear her. This half-rk-verse speaks about the ignorant. ‘Again to him reveal herself’. Reveals the body. ‘Like a wife well- dressed desiring the husband’, i.e. just as the wife well-dressed and desiring the husband reveals her own body, so to one who knows speech (she, i.e. speech) reveals her own self. Grammar should be studied so that the speech may reveal herself to us.25 Bhāṣya. palug boot saktumiva “saktum iva titaünā punanto yatra dhira manasā vācamakrata/ atrā sakhāyaḥ sakhyāni jānate bhadraiṣām lakṣmir nihitādhivāci//” dat ‘saktuḥ’-sacater durdhāvo bhavati, kasater vā viparītād vikasito bhavati. ’titaü’ pari- denotes a person who has risen high by being able to control their mind through gradual purification. All the four kinds of words such as nouns, verbs, etc. have four stages-the para, pasyanti, madhyama and vaikhari. Of these the para represents the Brahman stage of words. The pasyanti is that stage of words in which it is in the form of incipient ideation, and therefore has not the capacity of being used as language. The madhyama is the stage wherein the effort for constructive word-formation is being noticed in the heart and vaikhari is that while it attains the audible stage. Of the four classes of words, each one has these four stages. Of these the first three remain in the darkness of ignorance for the ordinary men. The grammarians, however, by virtue of their wisdom of the sastras can break open the darkness of ignorance and know the word in all its stages. The ordinary man merely speaks. He does not know the mystery of speech or its origin. Grammar is therefore to be studied for attaining such wisdom as has been specified above. The para stage of speech has been declared by Bhartṛhari as being like internal eternal light and by its true intuition a man attains salvation. [Bharthari is usually understood as not having any para stage of speech. Ed].
25. The word uta means api or also, and the word tvaḥ means others (tva-sabdaḥ anyavāci). There are many students who though they are alike in body and intelligent cannot understand the proper meaning of speech and such persons cannot compre- hend the import of speech though they may listen to it. The first half of the verse explains the condition of the ignorant. But to some grammarians the speech reveals her import and to such a person does speech reveal herself like a wife who desiring her husband takes off the garment through love. The comparison is supposed to point out that the speech reveals her import by revealing the significance of all suffix, prefix, etc. which are compared with the garments of the wife. 22 Mahābhāṣya of Patanjali pavanam bhavati, tatavad vā tunnavad vā. ‘dhira’-dhyānavantaḥ. ‘manasa’ prajñānena. ‘vācam akrata’-vācam akṛṣata. atrā sakhāyaḥ sakhyāni jānate’- atra sakhāyaḥ santaḥ sakhyāni jānate. kva? ya eşa durgo marga ekagamyo vägviṣayaḥ, ke punaste? vaiyakaraṇāḥ, kuta etat? ‘bhadraiṣām lakṣmir nihitādhivāci’- eṣām vāci bhadra lakṣmir nihita bhavati. lakṣmir lakṣaṇād bhāsanāt parivṛḍhā bhavati.-saktumiva. “Like the powdered pulse - just as one cleanses the powdered pulse with sieve so the meditative persons by their wisdom purify the speech. It is here that men of equal knowledge know each other. In their speech the auspicious beauty is inherent.” (R. V. 10.71.2) The word saktu (powdered pulse) is from the root ‘sacati ‘(to sprinkle) and it is not easily purifiable; or the word may have been formed from the root ‘kasati’ (which means movement and control) and by a transposition of consonants it becomes saktu in the sense of flowering. The word ’titaü’ means to purify (i.e. by removing the husk and other impurities). The word tataü either means spread out or interspersed with holes (i.e. a sieve). ‘Dhirah’ means meditative persons. The word ‘manasa’ means through superior knowledge. The word ‘vācam akrata’ means ‘purified the speech’. Here the friends attain friendship. Where? In this what is the impassable path attainable only by words. Again who are they? The grammarians. Why so? In their speech the noble Lakṣmi is well hidden. The word ’lakṣmi’ comes from ’lakṣaṇa’, i.e. revelation, by virtue of which the speech becomes all-encompassing, i.e. can express all things, even the finest shades of thought.26
26. The word ‘sakhāyaḥ’ has been explained by Kaiyața as samanakhyātayaḥ as having the knowledge of equality, i.e. those with whom the knowledge of plurality has vanished and the wisdom of oneness has appeared, that is those who have knowledge of Brahman as one and who regard all things as one. The word ‘sakhyani’ has been explained by Kaiyața as sayujyani or absorption into the divine essence; and ‘jānate’ means ‘attain’. It should be noted that we have translated the words sakha and sakhya in the ordinary sense of ‘friend’ and ‘friendship’. We think that Kaiyata’s interpretation of sakhāyaḥ as samanakhyātayaḥ (in the sense of those who have attained unity) and sakhya as sayujya or absorption is rather far-fetched. In our Bhāṣya. First Ahnika sarasvatim- yājñikāḥ paṭhanti-“ahitagnir apaśabdam prayujya prāyaścittiyām sārasvatim istim nirvaped” iti. prāyaścittiyā mā bhūmety adhyeyam vyākaraṇam.-sārasvatim. 23 The sacrificial priests read: “Those who lay sacrificial fire in the house should have to perform the sarasvati sacrifice by way of expiation if they use vulgar words.” We ought to read grammar so that we may not be laid under the obligation of performing an expiation. Bhāṣya. feine dasamyām putrasya- yājñikāḥ paṭhanti-“daśamyuttarakālam putrasya jātasya nama vidadhyād, ghoṣavad- ädyantarantaḥstham avṛddham tripuruṣānūkam anaripratisthitam. tad dhi pratisthitatamam bhavati, dvyakṣaram caturakṣaram vā nāma kṛtam kuryan na taddhitam” iti. na cântarena vyākaraṇam kṛtas taddhitā vā sakyā vijñātum.- daśamyām putrasya. Of the son of the tenth lunar day- The sacrificial priests read: “One should, after the tenth day, initiate the born baby with names and such interpretation the meaning is that those who speak the same type of correct speech easily become friends. Friendship can naturally grow among people of the same level of high culture of speech. So also our interpretation of the last line ‘bhadraiṣām’ etc. is different from that of Kaiyața and Nageśa. It is too far-fetched to think that by the knowledge of grammar or of speech one can attain Brahmahood. In our interpretation the sentence simply means that in the speech of the cultured people there lies deeply hidden unique grace and beauty. It is pity that most of the interpreters of the Bhāṣya of Patanjali belong to the monistic Vedanta faith. We have still to discover whether Patanjali had anticipated the interpretation of the Upanisads as given by Sankara. The word ‘adhinihita’ means deeply laid. Referring to our interpretation of the line ‘bhadraiṣām’ etc., we must say that the meaning of the word ‘parivṛdha’ as ajñāna- nivartaka or ’the dispeller of ignorance’ as given by Nagesa seems to be doubtful. The word means ‘all-encompassing’, and it does not seem that it had any special Vedic sense. The unique grace of one well-versed in language consists in its all encompas- sing character as being able to express all shades of thought. We cannot agree with Nägeŝa that the passage has any reference to the four stages of words ultimately leading to Brahman and thereby leading also to pure consciousness as the only reality and the vapourisation of all plurality of the world before us. 24 Mahābhāṣya of Patanjali names shall begin with the type of alphabet sounds called ghoṣavat or hoarse-sounding such as ha, ya, va, ra, la and the medial aspirate and the nasals of each of the five groups; and such names should have in between the alphabet sounds ya, ra, la, va, and should refer to the three generations of the name of the father and these names should not be associated to one’s enemy. It is only then that such names become well-established. It should consist of two or four syllables ending in a kṛt suffix and not have a taddhita. Now without grammar no one can distinguish the kṛt from the taddhita. Bhāṣya. sudevo asi- “sudevo asi varuņa yasya te sapta sindhavaḥ/ anukṣaranti kākudam surmyam suṣirām iva //” ‘sudevo asi varuna’- satya-devo’si. ‘yasya te sapta sindhavaḥ’ sapta vibhaktayaḥ. ‘anukṣaranti kākudam’- kādudam tālu, kākur jihvā sā’sminnudyata iti kākudam. ‘sūrmyam suṣirām iva’. tad yatha- sobhanām ūrmim suṣirām agnir antaḥ praviśya dahati, evam te sapta sindhavaḥ sapta vibhaktayas tālvanukṣaranti. tenâsi satyadevah. satyadevāḥ syāmetyadhyeyam vyākaraṇam.-sudevo’si “O Varuna, you are a true deity. Your palate showers the seven rivers just as an image interspersed with holes” (R. V. 8.69.12). O Varuna, you are a good shining one, i.e. you are a true deity. Of you there are the seven rivers the seven declensions of forms. Your kakuda or the palate- ‘kāku’ means tongue, that in which it is struck upwards is called kākuda. Like an iron image full of holes, the word ‘surmi’ means beautiful image. Just as fire burnt inside a beautiful image full of holes so the seven rivers like the seven declensions of forms are being showered by the palate. For that reason you are the truly shining one. We ought to study grammar so that we may be the truly shining one.27
27. Varuna is here spoken of as the truly shining deity. From this Varuna flow the seven rivers which are but the seven declensional forms from the palate. The word bas Bhāṣya. First Ahnika kim punar idam vyākaraṇam eva adhijigāmsamānebhyaḥ prayojanam anvākhyāyate, na punar anyad api kiñcit? om ity uktvā vṛttantaśaḥ ‘sam’ityevamādin sabdan pathanti? 25 So far the reason why grammar ought to be read has been described; is there nothing else to say? O yes, saying this they would begin with the Vedas as sanno devir abhisṭaye’ etc. Bhāṣya. 28 purakalpa etad āsīt-samskārottara-kālam brāhmaṇā vyākaraṇam smādhiyate. tebhyas tattat-sthāna-karaṇa-nādānupradana-jñebhyo vaidikāḥ śabdā upadisyante. tad adyatve na tathā. vedam adhitya tvarita vaktāro bhavanti- ‘vedanno vaidikāḥ śabdāḥ siddhā lokācca laukikāḥ, anarthakam vyākaraṇam’ iti. tebhya evam vipratipanna- buddhibhyo’dhyetṛbhyah suhrd bhūtvā ācārya idam śāstram anvācaṣṭe-‘imāni prayojanani, adhyeyam vyākaraṇam’ iti. uktaḥ śabdaḥ. svarupam apy uktam. prayojananyapyuktāni. In ancient times it so happened that the Brahmins after receiving their initiations used to study grammar. To such boys who knew the proper kākuda is formed thus: kākur jihvā, udayate utkṣipyate asmin iti kākudam tālu. The word ‘udayate’ is from the root vad to throw upwards (rather a peculiar meaning of the root)- kāku + vad + ghañ = kāku + uda. There is a shortening of the vowel as it is supposed to come under the sakandhvādi list. The word ‘surmya’ is a Vedic form of ‘surmi’. Just as fire entering an iron image full of holes burns all the impurities so showering the seven declensions of forms are supposed to purify us.
28. Kaiyața notes as follows: In ancient times people used to study grammar first and due to their tender age (balyat) they were unable to ask anything. So for them it was not necessary to state the reasons for the study of grammar. But in these days when life is short people first study the Vedas and are therefore quite grown up when they read grammar and hence able to ask the reason for studying grammar and for that purpose it is now necessary to state the reasons for studying grammar. 26 Mahābhāṣya of Patanjali places of pronunciation, the inner efforts and the outer efforts, instructions were given about Vedic words. But in our days it is not so. By studying the Vedas, they soon become loquacious, (they think) that from the Vedas we know the Vedic words and from popular usage we know the spoken language; grammar is thus useless. To such deluded students a teacher (Patanjali), as a friend, instructs the sastra-that these are the objects for studying grammar. So far words have been described, their nature has also been told as also the necessity of studying grammar. Bhāṣya. śabdânuśāsanam idānim kartavyam. tat katham kartavyam? kim śabdopadeśaḥ kar- tavyaḥ, āhosvid apaśabdopadeśaḥ, āhosvid ubhayopadeśa iti? anyataropadesena kṛtam syāt. tad yatha bhakṣya-niyamena abhakṣya-pratiṣedho gamyate- “pañca pañca-nakha bhakṣyāḥ” ityukte gamyata etad ‘ato’nye abhakṣyāḥ’ iti. abhakṣya-pratiṣedhena va bhakṣya- niyamaḥ. tad yathā “abhakṣyo gramyakukkutaḥ,” “abhakṣyo gramya- sukaraḥ” ityukte gamyata etad ‘araṇyo bhakṣyah’ iti. evam ihāpi. yadi tavac chab- dopadeśaḥ kriyate, gaurity etasminnupadişte gamyata etad ‘gāvyādayo’paśabdāḥ’ iti. athâpy apaśabdopadeśaḥ kriyate, gāvyādiṣu upadişteşu gamyata etad ‘gaur ityeşa sabdaḥ’ - iti. Now instructions have to be made regarding (correct) words. How is that to be done? Have we to make instructions about correct words or about incorrect words or about both? (An onesided opponent is supposed to answer as follows): Any one of these may do. Thus, by restricting the edibles, the prohibition of the non-edibles is implied, when it is said that the five five-nailed animals are edible it is implied that others are non- edibles. Or, rather by prohibiting the non-edibles we may find the rule about the edibles. Thus when it is said that the village fowl is non-edible, the village pig is non-edible it is implied that the forest ones are edible. So First Ahnika 27 here also when instruction is given about words that ‘gauḥ’ is the proper word instructed, it is implied that words like ‘gavi’ etc. are vulgar words. But on the other hand, instruction can be given about vulgar words; when words like ‘gāvi’ etc. are instructed it is implied that the word gauh is the proper word. Bhāṣya. kim punar atra jyāyaḥ? laghutvāc chabdopadeśaḥ. laghiyān chabdopadeśo gariyan apaśabdopadeśaḥ. ekaikasya sabdasya bahavo’ pabhramsāḥ. tad yatha-gaur ity asya sabdasya gāvi-goņi- gota-gopotalikety evam adayo’apabhramśāḥ. iṣṭānvākhyanam khalvapi bhavati. Which is again better here? It is becasue of simplicity that the instruction of correct words is better. The instruction of correct words is simpler and the instruction of vulgar words is cumbrous. For each correct word there are many vulgar equivalents. Thus for the word ‘gauh’ there are many vulgar equivalents such as gāvi, goņi, gotā, gopotalikā, etc. (Moreover by instructing the correct words) one attains what is desirable. Kaiyața notes that by the instruction of correct words one may attain merit or there is a direct comprehension.29 Bhāṣya. athaitasmin sabdopadeśe sati kim śabdānām pratipattau pratipada-pāṭhaḥ kartavyah? gaur aśvaḥ puruso hasti sakunir mrgo brāhmaṇa ity
29. Bhattoji in the Sabda-kaustubha makes a long discussion as to what may be the criterion of correct words. The discussion is long and it is not of much use to give here an account of the whole of it in detail. The substance of the discussion is that only those words about which no traditional origin or commencement can be pointed out should be called sadhu or correct words, as such it is indispensable that some one should instruct what words have such characteristic and can be called correct. Deviation of pronunciation through indolence, inattention and the like leading to the use of words which differ from standard words are called incorrect. It thus appears that there should be some teacher who should advise about correct words. The first advice so far as we now have is that given by Panini. Patanjali supplements the instructions of Panini and hence he says that his duty is to make an anusasana of correct words. Sistasya punaḥ sasanam anuśāsanam. Anusāsana means the supplementary instruc- tion on that on which instruction has already been given. 28 Mahābhāṣya of Patanjali evamādayaḥ śabdāḥ pathitavyāḥ? nety aha. anabhyupaya eṣa śabdānām pra- tip.ttau pratipada-pathaḥ. evam hi śruyate- ‘bṛhaspatir indraya divyam varṣa-sahasram pratipadoktānām sabdānām sabda-pārā- yaṇam provāca, nantam jagāma. bṛhaspa- tiśca pravaktā, indraścâdhyetā, divyam varṣa-sahasram adhyayana-kālo, na cântam jagāma. kim punar adyatve? yaḥ sarvathā ciram jivati varsa-śatam jivati. caturbhiśca prakārair vidyopajuktā bhavati-agama- kālena, svādhyāya-kālena, pravacana-kālena, vyāvahāra-kāleneti. tatra câsyagama- kālenaiva ayuḥ krtsnam paryupayuktam syāt; tasmād anabhyupayaḥ śabdānām prati- pattau pratipada-pāṭhaḥ. Now in the matter of instruction of correct words, should one enumer- ate off the correct words for their proper comprehension? Should the separate words such as ’the cow’, ’the horse’, ’the man’, ’the elephant’, ’the bird’, ’the beast’, ’the Brahmin’ etc. be instructed. No. This (process) of separate enumeration of words is not the proper means for their comprehension. So this we hear: Bṛhaspati, for one thou- sand heavenly years, instructed Indra the separate words for reaching the other shore of words but could not exhaust it. Bṛhaspati the teacher and Indra the pupil, the time of study is one thousand heavenly years, yet he could not exhuast it. What about these days? He who lives long, lives a hundred years. In four ways is learning acquired-the time of instruction, the time of preparation, the time of teaching, and the time of its use (in sacrifices etc.). So by the time, the teaching is finished, the life would be wholly spent up. Therefore, this method of severally enumerating the words is not the means for their comprehension. Bhāṣya. katham tarhi ime sabdaḥ pratipattavyāḥ? kiñcit sāmānya-viseṣavallakṣaṇam pravartyam yenâlpena yatnena mahato ma- hataḥ śabdaughān pratipadyeran. kim punas First Ahnika tat? utsargāpavadau. kaścid utsargaḥ kar- tavyaḥ, kaścid apavādaḥ. kathamjātīya-kaḥ punar utsargaḥ kartavyaḥ, 29 kathamjātiyako’pavādah? sāmānyena utsargaḥ kartavyaḥ. tad yatha “karmanyan” (Pa. 3.2.1), tasya viseṣenâpavadaḥ; tad yathā “ato’nupasarge kah” (Pa. 3.2.3). How are these words to be comprehended? Some rule has to be introduced involving the use of universal and special principles so that with little labour, the great collection of words may be comprehended. What is that? General rule and exceptions. In the sutras there should be some universal laws and some exceptions. There should be some times general principles and their exceptions. Of what class should again be the general principles be made and what class the exceptions? The general principles should be made in universal lines such as “karmanyan” (the suffix an is enjoined when a verb carries an accusative case as its pre- word). A limitation of that is an exception (apavāda). “ato’nupasarge kaḥ” (the ka should be suffixed to verbs ending in a and without a pre-particle- anupasarga). Bhāṣya. kim punar ākṛtiḥ padarthaḥ, āhosvid dra- vyam? ‘ubhayam’ ity aha. katham jñāyate? ubhayatha hy acaryeņa sūtrāņi paṭhitāni, ākṛtim padartham matvā -“jātyākhyāyām ekasmin bahuvacanam anyatarasyām (Pa. 1.2.58)” ity ucyate. dravyam padartham matvā “sarūpāṇām” (Pa. 1.2.64) - ity ekaseṣa ārabhyate. A new topic now comes in, regarding the nature of that which is signified by a word. Does the word signify the class-notion (akṛti) of the individual or the individual substance? Both- he says. How is it known? The teacher (Panini) has read the sutra both the ways taking the class notion for the meaning of the pada, he has formed the sutra-“jatyākhyāyām ekasmin bahuvacanam anyatarasyām”. In the sense of jāti or a class notion a word may optionally be used in the plural. Thinking individual substance as the30 Mahābhāṣya of Patanjali significance of words, he has formulated the sutra-“sarūpāṇām ekaseṣa ekavibhaktau”–when the words having the same form and having the same declensional suffix, one may be left (other may be dropped).30 Bhāṣya. kim punar nityaḥ śabdaḥ, āhosvit kāryaḥ? samgraha etat pradhanyena parikṣitam- nityo vā syāt kāryo veti. tatroktā doṣāḥ, prayojanany apy uktāni, tatra tveṣa nirnayaḥ -‘yadyeva nityo’thāpi kāryaḥ, ubhayathā’- pi lakṣaṇam pravartyam’ iti.
30. Kaiyața reminds that the interpretation of grammar as a whole cannot be made either on the assumption that the signification of words is limited to class-notion or to individual things. He refers to two paribhāṣās - “sakṛt gatau vipratiṣedhe yad badhitam tad badhitam eva” and “punaḥ prasanga-vijñānāt siddham” (see Paribhāṣendu-sekhara, paribhāṣās 39 and 40) referred to in the Bhasya on the rule, “vipratiṣedhe param kāryam” and interprets the Bhāṣya there as holding that in the opinion of the Bhāṣya in some cases the significance of pada is to be referred to a jāti and in other cases it is to be referred to vyakti but Nägeśa in his Paribhāṣendusekhara rejects the interpretation of Kaiyața and holds that every case may be interpreted either on the supposition that the significance of words refers to class-notions or to particualrs. Kaiyața, in summarising the views of those who hold that the significance of words has reference only to class-notions, says that all words whether substantive or qualitative are permeated through and through with class-notions. Thus in denoting substantive qualities, actions and artificial names always refer to some kind of class- notions. There are others, however, who hold that words only denote particulars and the objection that such a reference would involve the necessity of extending the denoting power of words to limitless objects for otherwise if such a denoting power was limited to some particualr, others would not fall within its scope, is invalid for as in fixing the denoting power of words to class-notions denotation applies to particulars; so in giving the denoting power to particulars, the class-notion may be included within it by a similar kind of implication and as such the objections made would be invalid. The solution here is that when the denotation applies to particulars such an application necessarily applies to certain characteristic features by virtue of which a particular is a particular as distinguished from other particulars of other classes; a denotation of particulars has thus a natural reference to such characteristics as contribute to the formation of class notions; for this reason, a denotation of particulars also invloves a denotion of class-characters. Thus even though a word denotes a particualr it carries with such a denotation also a notion of class-character- istics involved in it as a constituent. Thus in both the views, in the passage from the class-notion to the particulars and vice versa an implication of some sort involving in extension of meaning or a restriction of it has to be admitted. First Ahnika 31 (A doubt naturally arises) are words eternal (i.e. uncreated) or are they produced? In the Sangraha (possibly of Vyaḍi) this (problem) has been chiefly examined-“Are they eternal or are they created?” There the defects as well as the necessity have been described. There, however, the conclusion arrived at is, whether it be eternal or produced, in each case, the definite regulations have to be made.31 Bhāṣya. 31 katham punar idam bhagavataḥ Pāṇiner ācāryasya lakṣaṇam pravṛttam? Värttika. siddhe śabdartha-sambandhe. Bhāṣya. siddhe śabde arthe sambandhe ceti. atha siddha-sabdasya kaḥ padarthaḥ? nitya-paryaya-vāci siddha-sabdaḥ. katham jñāyate? yat kūṭastheṣvavicāliṣu bhāveṣu vartate; tad yatha-siddha dyauḥ, siddhā pṛthivi, siddham ākāśam.
31. Some hold that the alphabet form of speech is eternal and is only manifested by articulated sound. Others, however, think that the articulated sounds correspond- ing to the alphabetic forms manifest the words (padas) as wholes. Others, however, hold the view of vakya-sphoța. Those who hold that the words are nothing but sound, naturally adhere to the view that the words are non-eternal and produced. The Mimamsakas think that the alphabet forms are eternal and are manifested by articulated sounds. According to this view, of course, both the pada and the vakya are of the nature of alphabet forms which are eternal in that form and manifested only by the articulated sounds. The sphota view is held by the grammarians. The more advanced thinkers of the grammar school think that in reality the whole proposition is a combination of propositions forming the vakya appear as one whole. The appearance of padas as wholes is unreal and has only a pragmatic value for demonstrating an analytic purpose. The supposition that the padas or words have separate meanings is only illusory or imaginary. The Vaiseṣika school, however, thinks that the word is nothing but the articulated sound. The articulated sounds, however, may be both of the alphabetic order or of other orders, such as noise etc. Patanjali refers to the Sangraha. But we have no knowledge as to the nature of the book or as to its author. Nageśa says that according to tradition, Vyāḍi wrote the Sangraha, a work containing a hundred thousand words. Patanjali says that so far as the writer of grammar is concerned, it is immaterial whether the words are eternal or non-eternal, for in either case the instruction of the grammatical science is necessary for the instruction of correct words. Bhattoji, however, holds that in grammar the denotation of words may well be regarded as 32 Mahābhāṣya of Patañjali How does his holiness Panini, the teacher apply his definitive instruc- tions (at this point comes the Varttika, the relation between the words and their meanings being already there to which Patanjali says as follows): The words being already there and so the object and the signifying relation being also there (a question comes-what is the meaning of the word- ‘being there’, ‘siddha’)? The word siddha means eternal. The question is how is it known? The reply is: that (is eternal) which refers to unchanging and immovable ideas or entities, such as, the heaven is there, the earth is there (siddha).32 Bhāṣya. nanu ca bhoḥ karyeṣvapi vartate. tad yatha- siddha odanaḥ, siddhaḥ supaḥ, siddhā yavāgūr iti. yāvatā kāryeṣvapi vartate, tatra kuta etan nitya-paryaya-vācino grahaṇam, na punaḥ kārye yaḥ siddha-sabda iti? sangrahe tavat karya-prati-dvandvi- bhāvān-manyāmahe ’nitya-paryaya-vācino graha-nam’ iti. ihāpi tad eva. athava santy eka padany apy avadhāraṇāni. tad yatha-abbhakṣo’, ‘vāyu-bhakṣaḥ’ iti. apa eva bhakṣayati, vāyum eva bhakṣayatīti gamyate. evam ihāpi siddha eva, na saddhya iti. applying to universals and particulars. The denotation is to be determined in accordance with contextual usage of the problematic term.
32. The import of Patanjali’s question, as to how Panini wrote his work, is according to Kaiyața whether Panini was the creator of the relation of words with objects or he merely was the compiler of the relations that existed before. Patanjali here mentions the name of Panini as he is presently introducing a Vārttika line for interpretation and previous to this, he was interpreting the intention of Panini and Katyayana together and so far that portion was wholly his own contribution. Nägeśa says here the purport of the question is-whether Panini himself coined the words or he only explained the relations that already existed between the existing words and their objects. Kaiyața further states that in both the views of jati-sphota and vyakti- sphota words are eternal. Even in the Vaiseṣika view where words are regarded as produced, they are regarded as eternal as being traditionally current from begin- ningless time. The objects themselves also are eternal in their aspect as class-notions; even if they be regarded as only particular things, there is eternity, there is their aspect First Ahnika 33 But the word ‘siddha’ is also used with reference to products (of actions), such as the rice is prepared, the soup is prepared, the ricegruel is prepared. But if that word is used in the sense of products, why it is held as synonymous with nitya or eternity and why is the other meaning in the sense of product is discarded. (The reply is) in the Sangraha, however, it is taken in the sense opposite to that of the product and therefore in the sense of eternity, here also therefore that is the meaning or rather there are words which though single and used with a determining sense such as the words ‘abbhakṣa’, ‘vayu-bhakṣa’ meaning that who eats only, water, that who eats only air. (Here though the word only or eva is not used yet a single word ap or water has the sense of only water). So the word ‘siddha’ means only siddha, i.e. eternal and not sadhya or karya, i.e. product.33 Bhāṣya. athavā pūrvapada-lopo’tra draṣṭavyaḥ- atyanta-siddhaḥ siddha iti. tad yatha- devadatto dattaḥ, satya bhāmā bhāmeti. athavā “vyākhyānato viseṣa-pratipattir na hi sandehad alakṣṇam” iti nitya-paryaya-vācino grahaṇam iti vyākhyāsyāmaḥ. kim punar anena varnyena? kim na ma- hata kaṇṭhena nityasabda evopatto yasminn upādīyamāne’sandehaḥ syāt? Or it should be understood that there is an elision of the previous word. (The word ‘sidhha’) is supposed to be formed of the two words atyanta and of Brahman which is the underlying in them. The word ‘siddha’ means nitya or eternal. The words are regarded as eternal because they refer to eternal objects.
33. Patanjali is here discussing the meaning of the phrase “siddhe śabdârthasam- bandhe” of Kätyāyana, the Värttikakāra. His interpretation is that ‘siddha’ here means eternal, i.e. the words, their objects and their relations are all eternal. He at first draws this meaning by reference to the conclusive decision on the subject arrived at in the Sangraha (possibly by Vyādi). His next interpretation is based on the suggestion that siddha means siddha eva, i.e. sabdârthasambandhaḥ siddha eva. He refers to other words, abbhakṣa, vayu-bhakṣa which though single words are pregnant in the idea of eva, sarva eva apaḥ bhakṣayati iti ab-bhakṣaḥ. The sentence ‘apa eva bhakṣayati ‘is the explanation of the word ab-bhakṣa which contains within it the sense of eva in a pregnant manner. There is only one opposite to the word siddha meaning eternal and that is karya. Siddha eva means only siddha which negates karya, there being only two alternatives. 34 Mahābhāṣya of Patanjali siddha and the word atyanta is elided. The word siddha or accomplished is used in the sense extremely accomplished, such as the word Datta is used for Devadatta and Bhāmā for Satyabhāmā. Or the particular meaning is to be comprehended in accordance with interpretation and from doubt (it should not be thought) that there is no definite meaning. We shall interpret the siddha as being taken as an equivalent to ’nitya’ or eternal. But (what is the good) again of such long explanation? Why not loudly in a loud voice the word ’nitya’ had been uttered, which being done there would be no doubt? Bhāṣya. mangalārtham. mangalika ācāryo mahataḥ sastraughasya mangalartham siddha-sabdam aditaḥ prayunkte-mangaladini hi śāstrāṇi prathante vira-puruṣāņi ca bhavanty ayuṣmat-puruṣāņi câdhyetāraś ca siddhartha yathā syur iti. The auspicious teacher desiring auspiciousness for the extensive sastra at first uses the word siddha. (There is a maxim) the sastras that begin with auspicious words are traditionally continued and invest (their readers) with heroism and are associated with long life, and so that those who read (such sastras) attain their fruit.34 Bhāṣya. ayam khalu nitya-sabdo nâvaśyam kūṭastheṣvavicāliṣu bhāveṣu vartate. kim tarhi? abhikṣṇye’pi vartate. tad yatha-ni- tya-prahasito nitya-prajalpita iti. yāvatā’- bhikṣṇye’pi vartate tatrapy anenaivârthaḥ syāt “vyākhyānato višeṣa-pratipattir na hi sandehad alakṣṇam” iti. paśyati tvācāryo ‘mangalarthaścaiva siddha-sabda aditaḥ prayukto bhaviṣyati, śakṣyāmi cainam nitya- paryāya-vācinam varṇayitum’ iti. ataḥ siddha-sabda evopatto na nitya-sabdaḥ.
34. Though Katyāyana might have used the word ’nitya’ instead of the word ‘siddha’ he chose the latter word as an auspicious one in consonance with the maxim that only those sastras which begin with auspicious words are traditionally continued through time and those who study them cannot be defeated by others. First Ahnika 35 This word ’nitya’ again is not always to be used to denote those entities which are unmoved like an anvil. What then? It is also used in the sense of repetition. Thus ’nitya-prahasita’ means often laughing and ’nitya- prajalpita’ means always talking. It may be objected since the word can also be used in the sense of repetition why not take that sense. (The reply will be) that here also since there is a doubt between (two meanings) the proper meaning to be accepted (is that which is) according as it is explained (and as to the cause of this doubt one should not think that there is) no definite explanation. The teacher (Katyāyana) thinks that the word ‘siddha’ in the auspicious sense should be used at the commencement. It is possible for us to describe it as being equivalent to nitya or eternal. For this reason the word ‘siddha’ has been used and not nitya. Bhāṣya. atha kam punah padartham matvā eṣa vigrahaḥ kriyate- ‘siddhe śabde’rthe sam- bandhe ca’ iti? ‘ākṛtim’ ityāha. kuta etat? ākṛtir hi nitya, dravyam anityam. atha dravye padarthe katham vigrahaḥ kartavyah? ‘siddhe śabde artha-sambandhe ca’ iti. nityo hy arthavatām arthair abhi-sambandhaḥ. athava dravya eva padarthe eșa vigraho nyayyaḥ: siddhe śabde arthe sambandhe ce- ti. dravyam hi nityam ākṛtir anityā. To what does the signification of words refer? In the compound Sabdartha-sambandhe, where the compound means the words, their objects and the relations being eternal, the reply is-(the reference is to the class-notion). Why so? (Because) the class-notion is eternal and the particulars are non-eternal. Now if the signification refers to particulars, how should the com- pound be made? (The reply is) the words being eternal, and their relations to objects being eternal, the relation of significant words with objects is eternal. Or, rather, if signification of words refers to particulars it would be as follows, the words, their objects and the relations being eternal. The particulars are eternal and the class-notion is non-eternal.35
35. In the view in which particulars are regarded as non-eternal the relation of objects with words are regarded as eternal. But if objects are non-eternal, how can relations with them be eternal? The reply is that even though objects may be non- 36 Bhāṣya. Mahābhāṣya of Patanjali katham jñāyate? evam hi dṛśyate loke mrt kayacid akṛtyā yukta pindo bhavati, piṇḍākṛtim upamṛdya ghatikāḥ kriyante, ghațikākṛtim upmṛdya kundikāḥ kriyante. tatha suvarṇam kayācid ākṛtyā yuktam pindo bhavati, pinḍākṛtim upamṛdya rucakāḥ kriyante, rucakākṛtim upamṛdya kaṭakaḥ kriyante, kaṭakākṛtim upamṛdya svastikāḥ kriyante. punaravṛttaḥ suvarna-pindah punar aparaya"kṛtya yuktaḥ khadirāngāra-savarne kundale bhavataḥ. ākṛtir anyā cânya ca bhavati, dravyam punas tad eva. akṛty upamardena dravyam evâvasi- syate. How is it known? So it is seen among people that from certain points of view earth may be regarded as lump. By demolishing the lump-form it is made into jugs. By demolishing the jug-forms, pitchers are made. So also when gold is associated with certain forms, it becomes a lump. By demolishing these lump-forms, certain neck-rings can be made and these again may be changed into bracelets. These again can be changed into a svastikā. These again may be reformed into a lump of gold and may be made into other forms. The substance is the same but the form may be this or that. With the demolition of the form the substance remains the same. Bhāṣya. akṛtāvapi padartha eșa vigraho nyāyyaḥ- ‘siddhe sabde’ rthe sambandhe ca’ iti’. nanu coktamākṛtir anityeti? eternal yet the relation of signification may be eternal. There is a further relation between words and objects that certain words permanently denote certain objects. In the view in which particulars are regarded as eternal the eternity refers to the underlying Brahman. (This is of course the interpretation of Kaiyața). In such a view the class-notion is non-eternal for when there is a realisation of the Brahman everything becomes illusory. This constant reference to Brahman in discussions of grammar by Kaiyața seems to have no basis in the Bhāṣya itself. The Bhāṣya seems to refer to different perspectives and tries to explain the status of signification with the varying perspectives, wherein the aspect of universals or of particulars may alter- nately be emphasised. First Ahnika naitad asti. nitya"kṛtiḥ. katham? na kvacid uparateti kṛtvā sarva- troparata bhavati. dravyāntarasthā tūpalabhyate. athava nedam eva niiya- lakṣṇam-‘dhruvam kūṭastham 37 avicalyanapayopajana-vikary-anutpatty- avṛddhy-avyaya-yogi yat tan nityam’ iti. tad api nityam yasmimstattvam na vihanyate. kim punas tattvam? tasya bhāvas tattvam ākṛtāvapi tattvam na vihanyate. athava kim na etena- ‘idam nityam idam anityam’ iti? yannityam tam padartham matvaiṣa vigrahaḥ kriyate - ‘siddhe śabde’- rthe sambandhe ca’ iti. The compound may be so formed that the signification may refer to the form (the class-notion), the word, the object and the relation being eternal. But it has been said that the class-notion is non-eternal. But this is not so, (says Patanjali). The class-notion is eternal. It cannot be said that because it is not exhausted anywhere, it is exhausted everywhere. It can be perceived in other objects. Or, rather, this not the only definition of eternity. Eternal is that which is unchanging, stays like an anvil, immovable, devoid of decay, growth or change, is not produced, does not grow and has not loss. That also is eternal in which the reality is not destroyed. What is reality? The reality is its essence. In the case of class-concepts also the essence is not destroyed. Or, rather, what is the use of entering into this question-this is eternal and this is non-eternal? The compound has been so formed as to refer to the idea of eternity whatever it may mean-the word, the object and its relation being eternal. katham punar jñāyate-‘siddhaḥ śabdo’rthaḥ sambandhas ca’ iti? Bhāṣya. Värttikka. lokataḥ. Bhāṣya. yalloke’rtham artham upadaya sabdan prayuñjate, naiṣām nirvṛttau yatnam kurvanti. 38 Vārttikka. Bhāṣya. Mahābhāṣya of Patañjali ye punaḥ kārya bhāvā nirvṛttau tāvat teṣām yatnaḥ kriyate. tad yatha-ghaṭena kāryam kariṣyan kumbha-kara-kulam gatvā"ha- ‘kuru ghatam kāryam anena kariṣyāmi’ iti. na tadvac chabdan prayuyukṣamāņo vaiyāk- araṇakulam gatva"ha ‘kuru sabdan prayokṣye’ iti. tavaty evârtham upādāya sabdan prayuñjate. yadi tarhi loka eşu pramāņam kim śāstreņa kriyate? lokato’rthaprayukte sabdaprayoge śāstreņa dharma-niyamaḥ. lokato’rthaprayukte sabda-prayoge śāstreņa dharma-niyamahḥ kriyate. kim idam dharma- niyama iti? dharmaya niyamo dharma- niyamaḥ, dharmârtho va niyamo dharma- niyamaḥ, dharma-prayojano vā niyamo dharma-niyamaḥ. How is it known? The word, the object and the relation are eternal. Here the Varttika intervenes and says lokataḥ, i.e. from the behaviour of the people. In the world, the people take objects one after another and use words to signify them. But they make no effort in producing them. Those entities which are produced require effort to be made for their production. If one has to do with a jug, he goes to the potter’s house, and says: ‘Make a jug that I have something to do with it.’ But wishing to use words, he does not go to the grammarian and says: ‘Make words. I wish to use it.’ But he takes whatever objects he likes and uses words. If people’s behaviour is to be regarded as valid, what is effected by the śāstras? (Here the Värttika intervenes and says) words are used to denote objects and the śāstra makes restrictions for the accruing of merit. To this (the Bhāṣya says) when people drawn by objects use words, restriction is made by the śāstra (grammar) for the sake of merit. (A questioner asks) what is the meaning of ‘dharma-niyama’ (restriction for the sake of merit). The Bhāṣya remarks that it may be compounded in different ways- First Ahnika 39 restriction for the sake of merit or restriction for the need of merit or restriction which has for its object merit.36 Värttikka. Bhāṣya. yatha laukika-vaidikeṣu. priya-taddhitā dākṣinātyāḥ-yathā loke vede ceti prayoktavye ‘yatha laukika-vaidikeṣu’ iti prayuñjate. athava yukta evâtra tad- dhitarthaḥ-‘yatha laukikeṣu vaidikeṣu ca kṛtānteṣu’. loke tavat-‘abhakṣyo gramya-kukkuṭaḥ’, ‘abhakṣyo gramya-sukaraḥ’ ity ucyate. bhakṣyam ca nama kṣutpratighātārtham upadiyate. śakyam cânena śvamāmsādibhir api kṣut-pratihantum. tatra niyamaḥ kri- yate-‘idam bhakṣyam idam abhakṣyam’ iti. tathā-khedāt strişu pravṛttir bhavati. samānaś ca kheda-vigamo gamyāyām câ- gamyāyām ca. tatra niyamaḥ kriyate-iyam gamyeyam agamye ti. Such as it is found among the people and in the Vedas, the southerners. are fond of (applying) taddhita suffixes. Instead of saying ’loke’ and ‘vede’ they use ’laukika-vaidikeṣu’. Or, rather, the uses of taddhita is justifiable as in the passage ’laukikeṣu vaidikeṣu ca kṛtānteṣu.’ (Here laukikeṣu means drawn from popular usage and vaidikeṣu means drawn from the Vedas and from other decisions). Among the people it is said that the village fowl should not be eaten, the village pig should not be eaten. But all that is taken for removing hunger should be regarded as eatable and on this principle one might remove his hunger even with the meat of dogs.
36. The phrase ‘drawn by objects’ in the translation has been used in the sense impelled by the necessity of knowing objects. When people want that others should understand a particular meaning they use words; when these words are articulated, people know things. For the use of words, the science of grammar directs the nature of the stem, the suffixes and the like; when with such consciousness people use words they attain merit. Nageśa points out that Prabhakara thinks that by the use of the imperative or the subjunctive when certain actions are dictated to be performed, a person is impelled to do an action and this is called niyoga and this is dharma.40 Mahābhāṣya of Patañjali Therefore, a restriction is made-this ought to be eaten and this ought not to be eaten. So through sexual urge one turns to women. This urge can well be satisfied by going to women to whom one may legitmately go and to others to whom one may not legitmately go. There a restriction is made- this woman is legitimately worthy of cohabitation and this woman is legitimately not so.37 Bhāṣya. vede khalvapi-“payo-vrato brāhmaṇo yavāgū-vrato rājanya amikṣā-vrato vaisyah” ity ucyate. vratam ca namâbhyavahārārtham upādīyate. śakyam cânena śālimāmsādiny api vratayitum tatra niyamaḥ kriyate. tathā “bailvaḥ khādiro va yūpaḥ syād” ity ucyate. yupaś ca nama paśu-bandhârtham upādīyate. śakyam cânena yatkiñcid eva kāṣṭham ucchṛityânucchṛitya va pasur anubandhum. tatra niyamaḥ kriyate. tathāgnau kapālāny adhiśrityâbhiman- trayate “bhṛgūņām angirasām dharmasya tapasā tapyadhvam” iti. antareṇāpi mantram agnir dahana-karmā kapālāni santapayati. tatra ca niyamaḥ kriyate-’evam kriya- māṇam abhyudayakāri bhavati’ iti. evam
37. The sentence in the Bhasya “athava yukta evâtra taddhitârthah” etc. wherein the use of the taddhita suffix in the words laukika and vaidika has been justified, has another annotation by Kaiyața on the words laukika and vaidika. Kaiyata says that laukika here means as embodied in the smṛti-sastra and vaidika means as embodied in the Vedas. This is to be distinguished from the use of the word ’laukika-vaidikeṣu’ in the previous sentence where the words laukika and vaidika have been used in the sense of the ’the people as a whole’ and ’the Vedas as a whole’. The usage of taddhita in that passage is therefore only optional. Instead of saying ‘among the people’, the speaker says, among the people as a whole’. Herein the speaker only shows his vagary in his fondness for using the taddhita suffix whereas in the latter case the use of the taddhita suffix in the word laukikais justifiable on the ground that here the word ’loka’ means smṛti-sastra and the taddhita suffix means ‘codification’. These two examples are introduced for showing that there are some usages where certain suffixes are not exactly justifiable but not incorrect while in other cases they are quite justifiable. First Ahnika 41 ihāpi-samānāyām arthâvagatau śabdena câpaśabdena ca dharma-niyamaḥ kriyate- ‘sabdenaivârtho’ bhidheyo nâpaśabdena’ iti. ’evam kriyamāṇam abhyudaya-kāri bhavati’ iti. In the Vedas also it is enjoined that a Brahmin should restrict his diet to milk, a Kṣatriya to rice-gruel and a Vaiśya to curd mixed with water. A restriction is made among all things that a man may be desirous of eating. One may eat good rice and meat and restriction is made there. So also there is the injunction-a sacrificial post should be made of the bilva or the khadira wood. A sacrificial stake (yupa) is the post at which the animals are tied. One can take any kind of wood and make it smooth or rough and tie the animal at it. Here the restriction is made. So also putting on the fire the halves of a jug one utters the mantra- “Grow heated with the ascetic fervour of the heat of Bhrgus and Angirasas.” But one might as well put the jug-shells on the burning fire without uttering such mantras. A restriction is here made and if it is so done it becomes auspicious. So here also when meaning can be understood equally from correct and incorrect words a restriction for merit is made- object should be indicated by correct words and not by incorrect words. If it is so done it becomes meritorious. Varttika. Bhāṣya. astyaprayuktaḥ. santi vai sabda aprayuktāḥ. tad yatha- ‘ūşa’, ’tera’, ‘cakra’, ‘peca’ iti. kim ato yat santy aprayuktāḥ? prayogad hi bhavañ chabdānām sādhutvam adhyavasyati. ya idānim aprayuktā nāmi sadhavaḥ syuḥ. idam tavad vipratiṣiddham yad ucyate- ‘santi vai sabda aprayuktāḥ’ iti. yadi santi nâprayuktāḥ, athâprayukta na santi, ‘santi This discussion is introduced on the word dharma-niyama referred to in the previous section of the Bhāṣya. Words may diversely be used and in all these usages they may be significant. But it is only in certain restricted usages as enjoined by grammar that they may contribute to merit. Any food may appease hunger but only some food is prescribed as contributing to merit and others are prohibited. So some forms of speech as enjoined by grammar contribute to merit and others do not. 42 Mahābhāṣya of Patanjali câprayuktas ca’iti vipratiṣiddham. prayuñjāna eva khalu bhavan aha-‘santi śabda aprayuktāḥ’ iti. kaś cedānim anyo bhavaj-jātiyakaḥ puruṣaḥ śabdānām prayoge sādhuḥ syāt? — naitad vipratiṣiddham. ‘santi’ti tavad brūmo yad etañ śāstra-vidaḥ śāstreṇa anuvidadhate. ‘aprayuktāḥ’ iti brūmaḥ. yat loke ‘prayuktāḥ’ iti. yad apyucyate - ‘kaścedānim anyo bhavaj-jātīyakaḥ puruṣaḥ śabdānām prayoge sadhuḥ syad’ iti. na brūmo’smābhir aprayuktā iti. kim tarhi? loke ‘prayukta’ iti. nanu ca bhavan apy ābhyan- taro loke? ābhyantaro’ham loke, na tvaham lokaḥ. A Varttika sūtra (raising an objection) intervenes that there are words which are not in usage such as ūṣa, tera, cakra, peca. What does it matter if there are words which are not in usage? (This is said) that because it is only through usage that you can determine the correctness of words. The words that are not now used cannot be regarded as correct. (To this) a further objection is raised-this is all contradictory to say that there are words which are not in usage (for if the words exist it cannot be that they are not in usage, and if they are not in usage they cannot be said to exist. It is contradictory to say that they exist and are not in usage. It is by using the words that we say that these words are not in use). Who is the other better man than like you by whose use the words would be correct? (This is a joke-we are using the words and yet saying that the word is not in use). Your very use of the word is an instance of using it and that contradicts you. It is when men use a word that the word is said to be in usage. What special type of a man we should imagine by whose use of the word, should we speak of the word as being in usage and how can we exempt you from belonging to that category? The supposed reply is-this is not contradictory. (When we say that) the words that the words exist, we mean those words that the grammarians enjoined by rules. (By referring to words) that are not used we mean that they are not used by people. To your question-who is there the other kind First Ahnika 43 of person etc. we do not say that we have not used it. What then? (We mean that they are not used by people. Further reply to this position is) are you not included in ’the people’? (The reply is) yes-I am included within the people but I am not the people. Värttika. Bhāṣya. Värttika. Bhāṣya. asty aprayukta iti cennârthe śabda- prayogāt. asty aprayukta iti cet? tanna. kim kāraṇam? ‘arthe sabda-prayogat’- arthe śabdāḥ prayujyante. santi caiṣām śabdānām arthā iam yeṣvartheşu prayujyante. aprayogaḥ prayogānyatvāt. aprayogaḥ khalvapy eṣām sabdānām nyāyyaḥ, kutaḥ? ‘prayoganyatvāt’. yad eṣām śabdānām arthe’nyañchabdan prayuñjate. tad yathā-‘ūṣa’ ityasya sabdasyârthe- ‘kva yuyam uṣitāḥ’, ’tera’ ity asyârthe ‘kva yuyam tirṇāḥ’, ‘cakra’ ity asyârthe-‘kva yuyam kṛtavantaḥ’, ‘peca’ ity asyârthe- ‘kva yuyam pakvavantaḥ’ iti. Here again a Varttika intervenes-it is wrong to suggest that there are words which are not in use, for the words are used to denote objects. It is suggested that there are words which are not used. (The reply is) it is not so. What reason is there? Words are used to denote objects. These words refer to objects to signify which they are used. — Here another Värttika intervenes criticising the previous Varttika- the word ‘aprayoga’, i.e. not in use (means) that they are other than the words which properly denote the objects. (The Bhāṣya commenting on the above says) it is proper that these words should not be used; for they are other than the words which really denote the objects (i.e. words that are used other than those that would denote the objects). Thus in the place of the word ‘uṣa’ there is the phrase where did you stay (kva yuyam uṣitāḥ)? In the sense of the word ’tere’-where did you cross (kva yuyam tirṇaḥ)? In the sense of the word ‘cakra’-where did you do? In the sense of ‘peca’-where did you cook? 44 Vārttika. Bhāṣya. Värttika. Bhāṣya. Mahābhāṣya of Patanjali aprayukte dirgha-satravat. yady apy aprayuktaḥ, avaśyam dirgha- satraval lakṣaṇenânuvidheyāḥ. tad yathā dirgha-satrāņi vārṣaśatikāni vārṣasaha- srikāņi ca. na câdyatve kaścid apy aharati. kevalam ṛṣi-sampradayo dharma iti kṛtvā yājñikāḥ śāstreṇânuvidadhate. sarve deśāntare. sarve khalvapy ete sabda deśantareṣu prayujyante. na caiva upalabhyante. (In the matter of the consideration of) words not in usage the maxim of long sacrifice applies. (To this the Bhāṣya comments)-there are indeed words which are not in usage (but in this matter) the maxim of long sacrifice must be followed. Thus there are indeed long sacrifices extend- ing over hundred or thousand years but no one in these days takes to them. The sacrificial priests give instructions about them only because there is merit in the study of the Vedas. (Another Varttika again intervenes)-they all exist in other countries. (The Bhāṣya comments)-all these words exist in other countries. (But how is it) that they are not to be found? Bhāṣya. upalabdhau yatnaḥ kriyatām. mahān sabdasya prayogaviṣyaḥ. sapta-dvīpā vasumati trayo lokās catvāro vedāḥ sāngāḥ sarahasya bahudha bhinna ekasatam adhvaryuśākhāḥ sahasra-vartmā sāma-vedaḥ ekavimśatidhā bāhvṛcyam, navadha”- tharvano vedaḥ vakovākyam itihasaḥ purāṇam vaidyakam ity etāvān sabdasya prayoga-viṣayaḥ. etavantam sabdasya prayoga-viṣayam ananuniśamya ‘santy aprayuktāḥ’ iti vacanam kevalam sahasa- matram eva. etasmimś câtimahati sabdasya First Ahnika 45 prayoga-viṣaye te te sabdās tatra tatra niyata-viṣaya dṛśyante. tad yatha-savatir gati-karma kambojeṣveva bhāṣito bhavati, vikara enam ārya bhāṣante ‘sava’ iti. ham- matiḥ surastresu. ramhatiḥ pracya-ma- dhyeṣu. gamim eva tvaryāḥ prayuñjate. datir lavanarthe pracyeṣu, dātram udicyesu. ye câpy ete bhavato ‘prayuktā abhimatāḥ śabdā eteṣām api prayogo drśyate. kva? vede. tad yatha-“saptāsye revati revadūṣa” (R. V. 4.51.4), “yadvo revati reva- tyām tamūṣa”, “yanme naraḥ śrutyam brahma cakra” (R. V. 1.165.11), “yatrā naścakra jarasam tanūnām” (R. V. 1.89.9) iti. Make efforts in discovering them. The field for the use of words is very large. There are the seven islands of the earth, three worlds, the four Vedas with their accessories and esoteric doctrines diversely varied. There are the hundred branches of the Adhvaryuveda and thousand ways of Samaveda, twenty-one-fold Bähvṛcya, ninefold Atharvaṇaveda, the dia- logue literature, legends and purāņas, medical science, so much is the field of the use of words. Without examining such a large field of the application of words, to say that there are words which are not used, is merely boldness. In this large field of the use of words, specific words are used in special senses. Thus the root ‘Sava’ is used in the sense of movement in the Kamboja country. But the Aryans use it in the sense of decomposition; and the word ‘hammati’ is used in Surastra, ‘ramhati’ in the Middle East, but the Aryans use the word ‘gama’. The root ‘dati’ is used in the sense of tearing in the East, and in the same sense the word ‘datra’ is used in the North. Even those words, that you regard as being out of usage, are found to be used. Where? In the Vedas. Thus the word ‘uṣa’is found in the Vedic passage ‘saptasye revati revadūṣa’,‘yadvo revati revatyām tamūṣa’ (The) word ‘cakra’ is used in the passage ‘yan me narah śrutyam brahma cakra’, and in the passage, ‘yatra naś cakrā jarasam tanūnām’ Bhāṣya. kim punaḥ sabdasya jñāne dharmaḥ, āhosvit prayoge? kaś câtra viseṣaḥ? 46 Mahābhāṣya of Patanjali Vārttika. Bhāṣya. jñāne dharma iti cet tathâdharmaḥ. ‘jñāne dharmaḥ’ iti cet tatha adharmo’pi prāpnoti. yo hi sabdañ jānāty apaśabdān apy asau jānāti. yathaiva sabda-jñāne dharma evam apaśabda-jñāne’py adharmaḥ. athavā bhūyan adharmaḥ prapnoti. bhūyāmso hy apaśabdāḥ, alpiyamsaḥ śabdāḥ. ekaikasya sabdasya bahavo’pabhramśaḥ. tad yatha- ‘gaur’ ity asya gavi- goņi-gotā-gopotalikety evamādayo’pabhramśāḥ. (Now comes the question of the production of merit by the knowledge of words). Is merit produced by the knowledge of words, or by their use? What is the difference? (Here the Värttika intervenes)- if there is merit in knowledge, there is also demerit. (The Bhāṣya comments)- if there is merit in knowledge, then there is also demerit. He who knows correct words, also knows incorrect words. As there is merit in the knowledge of correct words, there is also demerit in the knowledge of incorrect words. Or rather there will be more demerit. There are more vulgar words and fewer correct words. For each correct word, there are many corrupt forms. Thus the word ‘Gauḥ’ has many corrupt forms such as, gāvi, goni, gotā, gopotalikā, etc. Värttika. Bhāṣya. Värttika. Bhāṣya. ācāre niyamaḥ. ācāre punar rṣir niyamam vedayate, “te’surā helayo helaya iti kurvantaḥ parābabhūvuḥ” iti. astu tarhi prayoge. prayoge sarvalokasya. yadi prayoge dharmaḥ, sarvo loko’bhyuda- yena yujyeta. kaś cedānim bhavato matsaro, yadi sarvo loko’bhyudayena yujyeta? na khalu kaścin matsarah. prayatnanartha- kyam tu bhavati. phalavata ca nāma prayat- nena bhavitavyam. na ca prayatnaḥ phalad Värttika. Bhāṣya. First Ahnika vyātirecyaḥ. nanu ca ye kṛta-prayatnās te sadhiyaḥ śabdan prayokṣyante, ta eva sādhiyo’bhyudayena yokṣyante. 47 vyatireko’pi vai lakṣyate. dṛśyante hi kṛta- prayatnaś câpraviṇaḥ, akṛta-prayatnaś ca praviṇāḥ. tatra phala-vyatireko’pi syāt. evam tarhi nâpi jñāna eva dharmo nâpi prayoga eva. kim tarhi? śāstra-pūrvake prayoge’bhyudayas tat tulyam veda-sabdena. śāstra-pūrvakam yaḥ śabdan prayunkte so’bhyudayena yujyate. tat tulyam veda- śabdena. veda-sabda apy evam abhivadanti- “yo’gni-ṣṭomena yajate ya u cainam evam veda,”,“yo’gnim nāciketam cinute ya u cainam evam veda”. apara aha-’tat tulyam veda-sabdeneti. yathā veda-sabda niyama-pūrvam adhītāḥ phalavanto bhavanty evam yaḥ śāstra- pūrvakam sabdan prayunkte so’bhyudayena yujyate’ iti. The restriction is about usage. (The Bhāṣya comments)- in the matter of usage again the Vedas instruct us the restriction,-the asuras were defeated, as they pronounced ‘helayah’, ‘helayah’. Let it be then in usage. (Here the Värttika intervenes)- in that case it would be so for all people. (The Bhāṣya comments)- if there is merit, in the use of words, then all people should attain so. Well, why should you be jealous, if all people attain merit? No one is becoming jealous. Only there would be no meaning in efforts. Efforts ought to be fruitful. Efforts cannot be dissociated from fruits. Well, then, those who have made proper efforts, will use correct words and they will excel and be associated with merit, but the reverse is often seen (i.e. sometimes there are some who make efforts, but do not attain fruit). We often find that there are some who make efforts, and yet 48 Mahābhāṣya of Patañjali are not well-versed, and there are some, who do not make efforts, but yet they are well-versed. There must be difference in fruit as well. Well, then, there is not merit merely in knowledge or in use. What, then? (Here the Värttika intervenes)- it is only in the use of words with proper scientific knowledge (that merit is attained) as it is said in the Vedas. (Here the Bhāṣya comments)- he alone is associated with merit who uses words in accordance with grammatical science. This is equal to the Vedic passage. The Vedic passage also says, “he, who sacrifices by the agnistoma, and he, who knows it to be such”, “he, who collects the naciketa fire and knows it to be such.” The above Varttika passage ’tat tulyam’ etc., is also differently interpreted thus: The other interpreter, (says the Bhāṣya), says in explain- ing the Varttika phrase, ’tat tulyam vedaśabdena’, says, just as the Vedic words produce fruit when they are studied according to proper rules, so he, who uses words according to the grammatical science, is also associ- ated with merit. Bhasya. Värttika. Bhasya. athavā punar astu, ‘jñāna eva dharmaḥ’ iti. nanu coktam, ‘jñāne dharma iti cet tatha’- dharmah’ iti. naisa dosah; sabda-pramāṇakā vayam, yac chabda aha tad asmākam pramāņam. sabdaśca śabda-jñāne dharmam āha, nâpaśabda-jñāne’dharmam. yac ca punar asiṣṭâpratiṣiddham, naiva tad doṣāya bhavati, nābhyudayāya. tad yathā hikkita- hasita-kaṇḍūyitāni naiva doṣāya bhavanti, nâbhyudayāya. abhyupāyo vā’paśabda-jñānam sabda- jñāne. atha va’bhyupaya evâpaśabda-jñānam sabdajñāne. yo hy apaśabdañ jānāti sabdān apy asau jānāti. tad evam ‘jñāne dharmaḥ’ iti bruvato’rthād apannam bhavati, ‘apaśabda- jñāna-pūrvake sabda-jñāne dharmah’ iti. Värttika. Bhāṣya. First Ahnika kūpa-khānakavad vā. 49 atha vā kūpa-khanakavad etad bhavisyati. tad yathā kūpa-khānakaḥ kūpam khanan yady api mṛdā pāmsubhis câvakirņo bhavati. so’psu samjātāsu tata eva tam guņam āsādayati. yena ca sa doṣo nirhanyate; bhūyasă câbhyudayena yogo bhavati. evam ihāpi yady apaśabda-jñāne adharmas tathāpi yas tvasau śabda-jñāne dharmas tena ca sa doso nirghāniṣyate bhūyasă câbhyudayena yogo bhavisyati. Or, rather merit may accrue from knowledge. Well, it has been said that if there is merit in knowledge, there is also demerit. (To this an objection is raised which will be settled by the next Värttika). This is no fault. We follow the authority of the Vedas. Whatever the Vedas say is valid for us. The Vedas say that there is merit in the knowledge of correct words and there is demerit in the knowledge of vulgar words. Those words which are not current among the cultured, nor prohibited, would produce neither merit, nor demerit, just as hiccuping, laughing, or itching is neither of merit nor of demerit. (Here the Varttika intervenes)-or, rather, the knowl- edge of vulgar words would be the means for knowing the correct words. (Here the Bhāṣya comments)- or, rather, the knowledge of vulgar words is the means of the knowledge of correct words. He who knows the vulgar words also knows the correct words. So when it is said that merit accrues from knowledge, it may imply that when one knows the correct words after knowing the vulgar words, merit accrues to him. (Here another Värttika intervenes) - or, it may be like the digger of a well. (The Bhāṣya comments)- or, rather, it should be like the digger of a well, i.e. just as the digger of a well in digging a well becomes besmeared with mud and dust. But when the water is produced, it derives from that such quality by which that impurity is removed and he gets much good out of it. So here also though there may be demerit in knowing the vulgar words yet the merit, that accrues by the knowledge of correct words, would remove that demerit and he would attain much good. Värttika. ācāre niyama iti ced yājñe karmaņi sa niyamo’nyatrâniyamaḥ.50 Bhāṣya. Mahābhāṣya of Patanjali yadapy ucyate-‘ācāre niyamaḥ’ iti. yājñe karmaņi sa niyamo’nyatrâniyamaḥ. evam hi śrūyate-‘yarvāņas tarvāņo nāma ṛṣayo babhūvuḥ pratyakṣa-dharmāṇaḥ parā- parajñā vidita-veditavya adhigatayāthāta- thyāḥ.’ te tatrabhavanto ‘yad vā nas tad vā nah’ iti prayoktavye ‘yarvāņas tarvāṇaḥ’ iti prayuñjate, yājñe punaḥ karmaṇi nâpa- bhāṣante. taiḥ punar asurair yājñe karmaṇy apabhāṣitam tatas te parābhūtaḥ. (Here another Värttika intevenes) if it is suggested that the rule (regarding the correct use of words) applies only in Vedic duties, then the rule applies only in sacrificial work, and not elsewhere. (Here the Bhāṣya comments) with reference to what is said that the restriction applies only to Vedic duties, that the rule applies only to sacrificial work and not elsewhere. It is also said in the Vedas that there were sages, Yarvāṇa and Tarvāṇa who directly realised merit and knew all the mundane and extramundane things, and all that was worth knowing, and had realized the truth. They, the masters, instead of pronouncing the words, ‘yad vā nas tad vā naḥ’ used to pronounce it as ‘yarvāņas tarvāṇaḥ’ (and hence probably were nicknamed as yarvāṇas-tarvāṇas), but in the sacrificial work, however, never used a vulgar word. The Asuras, however, used vulgar words in the sacrificial work and hence were defeated. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. atha ‘vyākaraṇam’ ity asya sabdasya kaḥ padarthaḥ? sūtram. sūtre vyākaraṇe şaşthy artho’nupa- pannaḥ. sūtre vyākaraṇe şaşthyartho nopapadyate- ‘vyākaraṇasya sūtram’ iti. kim hi tad anyat sūtrād vyākaraṇam yasyâdaḥ sūtram syāt? sabdâpratipattiḥ. sabdānām câpratipattiḥ prapnoti, ‘vyākara- ṇāc chabdan pratipadyamahe’ iti. nahi Värttika. Bhāṣya. Värttika. Bhasya. Värttika. Bhāṣya. First Ahnika 51 sūtrata eva sabdan pratipadyante. kim tarhi? vyākhyānatas ca. nanu ca tad eva sūtram vigṛhitam vyākhyānam bhavati. na kevalāni carca-padāni vyākhyānam- ‘vṛddhiḥ, āt, aij’ iti. kim tarhi? udāharaṇam, pratyudāharaṇam, väkyâdhyāhāra ityetat samuditam vyā- khyānam bhavati. śabde lyuḍ-arthaḥ. yadi sabdo vyakaraṇam, lyuḍ-artho nopa- padyate ‘vyākriyante sabda anena’ iti vyākaraṇam. nahi sabdena kimcid vyā- kriyate. kena tarhi? sūtreṇa. bhave ca taddhitaḥ bhave ca taddhito nopapadyate, ‘vyākaraṇe bhavo yogo vaiyakaraṇaḥ’ iti. nahi sabde bhavo yogah. kva tarhi? sūtre. proktādayaś ca taddhitāḥ. proktādayaś ca taddhitā nopapadyante, Pāṇinina proktam pāṇiniyam, apisalam, kāśakṛtsnam iti. nahi Pāṇininā śabdāḥ proktāḥ. kim tarhi? sūtram. kimartham idam ubhayam ucyate- “bhave ca taddhitaḥ”, “proktādayaś ca taddhitaḥ” iti. na ‘proktādayaś ca taddhitaḥ’ ity eva bhave’pi taddhitas coditaḥ syāt? purastād idam ācāryeņa dṛṣṭam-“bhave ca taddhitaḥ” iti, tat pathitam. tata uttara- 52 Mahābhāṣya of Patanjali kālam idam dṛṣṭam, “proktādayaś ca taddhitaḥ” iti, tad api paṭhitam. na cedānim mācāryāḥ sūtrāņi kṛtvā nivartayanti. Now what is the meaning of the word ‘vyākaraṇa”? (The answer is) the sutras. (Here the Värttika intervenes)-in the ‘sutra’ and in the ‘vyakaraṇa’ the application of the meaning of the sixth-case is not justifiable. (The Bhāṣya comments) in the expression-the sutra of vyakaraṇa (grammar), the meaning of the sixth-case is not applicable. What is that in grammar which are not the sutras, that it can be said that of that is the sūtra (a grammar being wholly made up of sutras, there cannot be any separate existence of grammar to which the sutras belong. The sutras themselves are the grammar, and, therefore, there cannot be any sixth- case). (Here the Värttika intervenes)-(Well in that case) there cannot be any comprehension of words. (Herein the Bhāṣya comments)-in such a case one cannot understand the correct words (and they say)- we know words from grammar, but words cannot be known merely from the sutras. What then?- (Well) from interpretations as well. But then the sutras themselves when properly analysed, give the interpretation. (To this the Bhāṣya objects and says) merely by analysing the words, such vṛddhi, at, aic (in the sutra, vṛddhir- ad-aic), can we get the interpretation of the sutra? What then? Examples, counter-examples supplying of elisions-all these make interpretation. (Another objection is raised for introducing the next Varttika.) Well then, the words (may be called) grammar. (Here the Varttika intervenes)-in words (there cannot be) any signifi- cance of the suffix ’lyut’. (The Bhāṣya comments on it and says)-if it is supposed that vyakarana is nothing but words, then the meaning or the suffix ’lyut’ (in the word vyakarana) becomes unjustifiable. The word ‘vyakarana’ is derived thus, the words are analysed by it. The words do not analyse anything. By what are they analysed then? By the sūtras. (Another Värttika introduces another objection)-A taddhita suffix in the sense of ‘produced out of it’ (becomes inapplicable). (The Bhāṣya comments) the taddhita suffix in the sense of ‘being produced’ be- comes inapplicable. (As one says), the vaiyakarana is he whose wisdom is produced from grammar. No wisdom is produced from the words. What is then? In the sūtras? First Ahnika 53 (Another objection is introduced by a Varttika)- the taddhita suffix takes place in such senses as ‘have been spoken’. (The Bhāṣya comments)- the taddhita suffixes in the sense of ‘have been spoken etc.’ become inapplicable. (In the sense) ‘spoken by Pāṇini’, we have ‘paniniyam’, so we have ‘apisalam’, ‘kāśakṛtsnam’. Pāņini did not speak words. What then? The sutras. What is the good of referring to two sūtras: “bhave ca taddhitah”, “proktadayaś ca taddhitaḥ”? Would not a reference to ‘proktadayaś ca taddhitaḥ’ involve within it a reference to ‘bhave ca taddhitaḥ”? (The reply to such an objection is then suggested in the Bhāṣya). At first the sutra, “bhave ca taddhitaḥ” struck the teacher and he read it. Later on the rule “proktadayaś ca taddhitaḥ” also struck him, and he read that also. The teachers do not stop after making the sutras (i.e. just as Panini uses the method of definition and elaboration, so by way of the latter, Katyāyana also gave two instances of the taddhita suffix. Bhāṣya. ayam tavad adoṣo, yad ucyate- ‘sabde lyuḍarthaḥ’ iti. nâvaśyam karaṇādhikara- nayor eva lyuḍ vidhiyate. kim tarhi? anyeṣvapi kārakeṣu, “kṛtya- lyuto bahulam” (Pa. 3.3.113) iti. tad yathā- praskandanam, prapatanam iti. atha va sabdair api sabda vyākriyante. tad yatha-gaur’ ityukte sarve samdehā nivar- tante ’nāśvo na gardabhaḥ’ iti. ayam tarhi doṣaḥ, ‘bhave (ca taddhitaḥ’), ‘proktādayaś ca taddhitaḥ’ iti. This is no criterion that (the word vyakaraṇa has two meanings of the lyut suffix and there can be no such) meaning of the lyut suffix in words (or rather if the word sabda be regarded as standing for the science of grammar). There is no such obligation that the lyut suffix should be used only in the instrumental (karana) and in the locative (adhikarana). The lyut suffix is used in the sense of other kärakas as well, such as ‘praskandana’ or ‘prapatana’ (where the suffix is used in the sense of the ablative). Or, rather, words may also explain words. When one pronounces the word ‘cow’, all doubts are dispelled-that this is neither a horse nor an ass. (The idea is that as when a word is pronounced, it signifies only one object, and no one would have other objects indicated by it, so when the example is given, all words of that type are indicated). 54 Mahābhāṣya of Patanjali This is a defect-(the reference to two types of taddhita suffix)- taddhita in the sense of bhava and in the sense of prokta, as referred to. Värttika. Bhāṣya. evam tarhi lakṣya-lakṣaṇe vyākaraṇam. lakṣyam ca lakṣaṇam caitatsamuditam vyākaraṇam bhavatiti. kim punar lakṣyam, kim vā lakṣaṇam? sabdo lakṣyaḥ, sūtram lakṣanam. evam apy ayam doṣaḥ-samudaye vyākaraṇa-sabdaḥ pravṛtto’vayave nopa- padyate. sūtrāṇi câpyadhiyāna iṣyate- ‘vaiyākaraṇa’ iti. naișa doṣaḥ. samudayeṣu hi śabdāḥ pravṛtta avayaveṣvapi vartante. tad yatha- ‘pūrve pañcālāḥ’, ‘uttare pañcālāḥ’, ’tailam bhuktam’, ‘ghṛtam bhuktam’, ‘suklaḥ’, ’nilaḥ’, ‘kapilaḥ’, ‘kṛṣṇaḥ’ iti. evam ayam samudaye vyākaraṇa-sabdaḥ pravṛtto’- vayaveṣvapi vartate. atha vā punarastu sūtram. nanu coktam- “sūtre vyakaraṇe sasthyartho ’nupapannah” iti. naiṣa doṣaḥ. vyapadesivadbhāvena bha- visyati. Grammar (involves within it) the object of definition and the definition itself (on two above Värttika, the Bhāṣya comments as follows). The object of definition (i.e. words) and as well as the definitions (i.e. the sūtras) together belong to grammar. What is the object to be defined and what is definition? The words are objects of definition and the sutras are the definitions. This also is a fault (says the supposed objector) that the word ‘vyākaraṇa’ being used for the whole is also used for its parts. Even he, who reads the sutras, is called a ‘vaiyakarana’ or a grammarian. (In refutation of the supposed objection it is said). This is not a fault. The 1 e d e e n n e First Ahnika 55 words, that are used to denote the whole, may also denote the parts. Thus (they say) the Eastern Pañcālas, the Northern Pañcālas, eaten oil, eaten clarified butter, white, blue, brown, black. So also the word, that indicates the whole, the word vyakaraṇa (grammar), may also be used to denote the parts. The eastern or the northern parts of the Pañcala country are only parts of the Pañcāla country and to signify these parts the word ‘Pañcāla’ which denotes the whole of the Pañcala, is used. So the medicated oil and clarified butter which contains other elements than the oil or the clarified butter, go by the name ’taila’ and ‘ghrta’ signifying oil and clarified butter which are only partial elements of those medicated mixtures. Even if an object has parts which are not white or not blue, brown or black may be called white, blue, brown or black only if they are largely so. So the word ‘vyākaraṇa’ (grammar), though it is applicable to the whole connection of sūtras, may be used to denote its constitutive parts-the sutras. Or, rather, the sutras may also be called ‘vyakarana’ (grammar). But it has been suggested that the meaning of the sixth-case is not applicable to the word ‘vyākarana’ (i.e. it cannot be said that the sutras belong to grammar on the contention that sutras are the grammar). (To this supposed objection the Bhāṣya replies)-this not a fault. A word may be used both in the sense of a name and in the sense disclosing the meaning involved in that name.38 Bhāṣya. yadapyucyate “sabdapratipattir” iti. nahi sūtrata eva sabdan pratipadyante. kim tarhi? “vyākhyānataś ca” iti. parihṛtam etat-tad eva sūtram vigṛhitam vyākhyānam bhavati’ iti
38. Thus we use the words ‘rahoḥ śirah’ (the head of Rahu)—the Rāhu is nothing else but a head. So in the same object an imaginary reference may be introduced as indicating a name and its meaning. The word ‘vyakaraṇa’ may be used technically in the sanse of grammar and also in its etymological sense-’that by which anything is explained’; and the word ‘sutra’ may be used in the sense of the whole grammar (as a collection of sūtras), as also to denote its parts, the individual sutras. In one aspect then, viz. that of explaining the words, the sutras may be identical with grammar, and in another sense it may mean the individual sutras forming a part of the whole collection of sutras called grammar or vyākaraṇa. 56 Mahābhāṣya of Patañjali nanu coktam “na kevalāni carca-padāni vyākhyānam-vṛddhiḥ at aij’ “iti; kim tarhi? “udāharaṇam pratyudaharaṇam vākyādhyāhāra ity etat samuditam vyākhyā- nam bhavati” iti. avijanata etad evam bhavati. sūtrata eva hi sabdaḥ pratipadyante. ataś ca sutrata eva. yo hyutsūtram kathayennādo gṛhyeta. It has been objected that (in such an interpretation of the word ‘vyakarana’ as meaning the sutras) one may not understand the nature of words; for one does not understand the nature of words only from the sutras, what then? From interpretations as well. This (objection) has already been refuted. Those sutras themselves when analysed become interpretations. Now it has been stated that by merely analysing the sutras we do not get any interpretation, but only with examples, counter-examples, supplying of elision, etc. that the explanation as a whole is possible. (In replying to this objection the Bhāṣya says such an objection) can be raised only by the ignorant. They understand words only from the sutras. For this reason from the sutras alone (the words are explained). He who says anything besides the sutras, only makes a sound. Bhāṣya. Värttika. Bhāṣya. atha kimartho varṇānām upadeśaḥ? vṛtti-samaväyârtha upadeśaḥ. vṛtti-samaväyârtho varṇānām upadeśaḥ. kim idam. vṛtti-samaväyârtha iti? vṛttaye samavayo vṛtti-samavayaḥ. vṛttyartho vā samavāyo vṛtti-samavāyaḥ. vṛtti-prayojano vā samavāyo vṛtti-samavāyaḥ. kā punar vṛttiḥ? śāstra-pravṛttiḥ. atha kaḥ samavāyaḥ? varṇānām ānupūr- vyeņa sanniveśah. atha ka upadeśaḥ? uccāraṇam. kuta etat? diśir uccaraṇa-kriyaḥ. uccārya hi varṇān āha-upadiṣṭā ime varṇā iti. Värttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. First Ahnika anubandhakaranârthaś ca. anubandha-karaṇârthaś ca varṇānām upa- deśaḥ- ‘anubandhan asankhyāmi’ iti. nahy anupadiśyavarṇān anubandhaḥ śakyā āsanktum. sa eṣa varṇānām upadeso vṛtti- samaväyârthas cânubandha-karaṇârthas ca. vṛtti-samavāyaś cânubandhakaranam ca pratyāhārârtham. pratyāhāro vṛttyarthaḥ. işṭa-buddhyarthaś ca. 57 iṣṭabuddhyarthaś ca varṇānām upadeśaḥ- ‘iṣṭān varṇān bhotsyamahe’ iti na hy anupa- diśya varṇān iṣṭā varṇāḥ śakya vijñātum. iṣṭa-buddhyarthaś ceti ced udāttânu- datta-svaritânunāsika-dirgha-plutānām upadeśaḥ. işṭa-buddhyarthaś ceti ced udattânudatta- svaritânunāsika-dirgha-plutānām apy- upadeśaḥ kartavyah. evamguņā api hi varṇā isyante. akṛtyupadeśāt siddham. Värttika. Bhāṣya. avarṇākṛtir upadiṣṭā sarvam avarṇa-kulam grahiṣyati. tathevarṇākṛtiḥ. tathovarṇākṛtiḥ. Värttika. Bhāṣya. ākṛtyupadeśāt siddham iti cet samvṛtādinām pratiṣedhah. ‘ākṛtyupadeśāt siddham’ iti cet? sam- vṛtādinām pratiṣedho vaktavyaḥ. ke punaḥ samvṛtādayah? samvṛtaḥ kalo dhmāta enikṛto’mbukṛto’rdhako grasto 58 Värttika. Mahābhāṣya of Patanjali nirastaḥ pragita upagitaḥ kṣvinno romasa iti. 66 apara aha- “grastam nirastamavalambitam nirhatam ambūkṛtam dhmātamatho vikampitam/ sandaṣṭamenikṛtamardhakam drutam vikirṇametāḥ svara-doṣa-bhāvanāḥ //” iti ato’ anye vyañjana-doṣāḥ. naisa doṣaḥ. gargādi-bidādi-pāṭhāt samvṛtādinām nivṛttiḥ. Bhāṣya. gargādi-bidādi-pāthāt samvṛtādinām nivṛttir bhaviṣyati. Vārttika. Bhāṣya. Vārttika. Bhāṣya. Värttika. Bhāṣya. of samudaya-sādhutvârtham tu gargādi- bidādi-pāṭhaḥ. asty anyad gargādi-bidādi-pāṭhe prayojanam. kim? samudāyānām sādhutvam yathā syād iti. pratyapatti-vacanam. evam tarhy aṣṭādaśadha bhinnām nivṛtta- kalādikām avarṇasya pratyapattim vakṣyāmi. sā tarhi vaktavyā. lingârthā tu pratyāpattiḥ. lingârthā sā tarhi bhavati. tat tarhi vaktavyam. yady apy etad ucyate, athavaitarhi anekam anubandha-śatam noccāryam it-samjñā ca na vaktavyā, lopaś ca na vaktavyaḥ. yad-anu- bandhaiḥ kriyate tat kalādibhiḥ kariṣyate. siddhyatyevam. apāṇiniyam tu bhavati. First Ahnika 59 (In introducing another topic brought forward by the Värttikakāra the Bhāṣya says) -Well then, why there are instructions also about syl- lables? (The correctness or incorrectness of words cannot be explained by instructions about syllables, so what is the good in instructing them?) (Here the Varttika intervenes)- This instruction as the utterance of syllables is for the sake of the introduction to the grammatical science. (The Bhāṣya comments)- it is for introducing the śastras the syllables are instructed in order (vṛtti-samaväyârtham varṇānām upadeśaḥ). What is the meaning of the word ‘vṛtti-samaväyârtha”? Their utterance in order is for the sake of the introduction of the grammatical science. What is again the meaning of ‘vrtti’? Introduction of the śastra. What is the meaning of ‘samavāya’? The utterance of syllables in order. What is the meaning of ‘upadeśa’? Articulation. The root ‘disi’ means ’to utter’. It is by uttering the syllables that one says that these syllables have been instructed. (upadiṣṭā ime varṇāḥ). (Here the Varttika intervenes)-Also for making the attachments. (The Bhāṣya comments)-The syllables are instructed for the sake of interpreting the attachments (anubandha); so that we may connect the attachments. It is not possible to connect the attachments without instruct- ing the syllables. The instruction or utterance in order of the syllables is then both for the cohesion of the introduction of the śastras and for the sake of the knowledge of the attachments. The commencement of the sastras as involving the utterance of syllables in order is for the sake of the attachments and for forming the pratyāhāras. The pratyāhāras are for the sake of introducing the śāstras in a simple manner. Pratyāhāra means the comprehension of a series of letters or affixes into one syllable effected by combining the first member of the series, without its indicating letter or letters, with the indicatory final consonant of the last member. Thus the pratyāhāra ‘ac’ is the technical term for all the vowels and hal for all the consonants. (Here the Varttika intervenes)-(The syllables are instructed) also for the sake of indicating the desired articulation of the syllables. (The Bhāṣya says) -The syllables are instructed so that we may indicate the desired articulation of the syllables. It is not possible to indicate the nature of the desired articulation, of the syllables without instructing properly about the syllables. (Here the Varttika intervenes with a supposed objection)-If the above view is accepted, then the instruction ought to indicate the udatta, the anudatta and the svarita accents as well as the nasal, the long and vocative prolongation in each case.60 Mahābhāṣya of Patanjali (The Bhāṣya repeats the Varttika sentence and says) that the syllables of all these qualities since each syllable has varied forms according to their accent, length and the like; the instruction should then specify and indicate all the characters. (Here the Värttika intervenes with a conclusive answer)-By the in- struction (of the syllables) in their universal aspect, this is effected. When the universal class of ‘a’ is instructed, all ‘a’-s may be comprehended. So the universals of i and u syllables. (Here another Varttika intervenes)-If the purpose is served by the instruction of the syllables as universals then there should be no instruc- tion about the closed or open utterances. (The Bhāṣya repeats the Varttika sentence and raises the question)- What are these closed and open utterances? (The reply is)These are the closed distorted (when a syllable is uttered from an improper place of articulation, there is a distortion of it resembling the sounds of birds), the aspirate (by too much aspiration even a short sound appears as long), indistinct when a syllable is not properly pronounced it cannot be distinguished from another syllable. Inarticulate (this means a sort of articulation in which though a syllable is pronounced, it appears as if it was in the throat of the speaker), half-pronounced (a long sound appears as a short one), the struck (syllables are so pronounced that they are arrested at the root of the tongue), harsh, singing tone associated with the singing articulation of the neighbouring syllables, shaky, too voluminous, too swift, hoarse, extended (melting to other syllables). These are the defects of articulation.39 Apart from these, there are other defects of proper expression. In introducing the next Värttika, the Bhasya says that the faults of too closed utterances and the like (the defects just pointed out above) are removed by the utterance of the list of words coming under the gargādi or the bidādi list. To this another objection is raised by the Varttika- the utterance of the gargādi and the bidadi is for demonstrating the correctness or articu- lation of the separate words falling under that list (i.e. the utterance of the words of those lists serve to indicate the proper articulation of number of words, but not as syllables of a class).
39. Kaiyața remarks that the defects of articulation are limitless as may be due to the inability of proper articulation and error. This is again repeated in the Bhāṣya as a verse which is not translated for fear of reduplication. First Ahnika 61 (The Bhāṣya says)-Another purpose is served by the utterance of the gargādi and bidādi list. That is done. (Those lists have been uttered for demonstrating the correctness of the whole words). (Another Värttika intervenes)-It is for this reason that the proper nature of the articulation of syllables has been indicated. In this way we shall describe the eighteen different kinds of the pronunciation of ‘a’ as free from the defects of indistinct pronunciation and the like as describing the proper nature. (The Bhāṣya raises another objection)-This then requires a separate enumeration. If after once describing the nature of the syllables we have again to describe it, that is reduplication. A provisional answer to such an objection is introduced by the Varttika. This description of the nature of syllables may be for indicating the various characteristics as there may be defects of articulation in the anubandha syllables. So there may be similar defects in the articulation of the syllables in the notes etc. The instruction of the nature of the syllables is not merely for indicating the defects of the syllables in their universal aspect, but also in the aspects in which they appear in various combinations and therefore such a redescription of their nature is not a duplication. (The Bhāṣya repeats the Värttika line and says that), then that ought to be described. (The Bhāṣya here suggests a supposed objection). It has. been suggested that one should rather not introduce hundreds of attachments (anubandha) or direct it or evasion and other elisions. What may be done by the anubandhas may as well be done by the different forms of articulations, indistinct, nasal, etc.40
40. It is suggested here, just as there are adhikara-sutras in Panini in which certain roots are marked out as belonging to a class for the sake of different operations, so all operations in grammar might have been based upon a system of different kinds of defects in articulation which might mark out particular class of roots or words for particular operations. Thus, instead of forming such a rule as “anudatta-ñita”, i.e. roots which are articulated in the anudatta accent, or has an attachment of ñ, should take ātmanepada, one might have formed a rule “kalādātmanepadam”, i.e. roots pronounced in a particular indistinct manner, should take ātmanepada. The suggestion here made is that instead of introducing the system of anubandhas and it-s it was possible to make the same sort of classification by means of different kinds of tones. Probably grammars were made under such a system and Panini introduced a reform in introducing the anubandha system where articulate syllables replaced the tones and thus attained celebrity. But it cannot be denied that the remnants of the old system existed even in Panini’s system, as he also used accents and nasal tones for the classification of roots in certain operations. 62 Mahābhāṣya of Patanjali (The reply is)- It will, no doubt, serve the purpose, but it would not be in accordance with the system of Panini. Then let it be as it is. Bhāṣya. nanu coktam-“akṛtyupadeśat siddham iti cet samvṛtādinām pratiṣedhaḥ” iti. parihṛtam etat “gargādi-bidādi-pāṭhāt samvṛtādinām nivṛttir bhaviṣyati’ iti. nanu cânyad gargādi-bidādi-pāṭhe prayojanam uktam. kim? samudāyānām sādhutvam yathā syād iti. evam tarhy ubhayam anena kriyate- pāṭhas caiva viseṣyate, kalādayaś ca nivart- yante. katham punar ekena yatnenobhayam labhyam? labhyam ityaha. katham? dvigatā api hetavo bhavanti. tad yatha-āmrāś ca siktāḥ pitaraś ca priņītā iti. tathā vākyāny api dviṣṭhāni bhavanti-‘śveto dhavati’, ‘alambusānām yātā’ iti. atha va idam tāvad ayam praṣṭavyaḥ- kveme samvṛtādayaḥ śruyerann iti? agameṣu. agamaḥ śuddhaḥ pathyante. vikāreṣu tarhi? vikārā api śuddhāḥ pathyante. pratyayeṣu tarhi? pratyayaḥ api śuddhaḥ paṭhyante. dhātuşu tarhi? dhātavo’pi śuddhaḥ pathyante. prātipadikeṣu tarhi? prātipadi-kany api śuddhāni pathyante. yani tarhy agrahaṇāni prātipadikani? eteṣām api svara-varṇānupūrvi-jñānârtham upadeśaḥ kartavyah. ‘śaśaḥ’ ‘ṣaṣaḥ’ iti mā bhūt, ‘palāśaḥ’ ‘palāṣaḥ’ iti ma bhūt. ‘mañcako’ ‘mañjakaḥ’ iti mā bhūt. First Ahnika “āgamāś ca vikārāś ca pratyayaḥ saha uccāryante tatasteṣu neme praptāḥ 63 dhātubhiḥ/ kaladayaḥ”// iti śrimad-bhagavat-patanjali-viracite vyākaraṇa- mahābhāṣye prathamâdhyāysya prathame pāde pra- thamam āhnikam. It has been suggested that if the instruction of syllables is in their universal aspect, then the closed or open utterances need not be instructed. But this critcism has already been refuted by the utterance of the gargādi or bidādi list, the prohibition of the defects of closed or open utterances can be removed. But to this also it has been pointed out that there are other purposes to be fulfilled by the utterance of the gargādi list. What was that purpose so that whole words may be correctly pronounced?Now it is being suggested that both the purpose may be served by the utterance of such a list. Thus, by the utterance of such lists the nature of the articulation of the whole words is shown, and their mispronouncing tones such as indistinct and the like are also prohibited. (But the question arises) how can both the purposes be served by the same means? The answer is that there are causes which go two ways. The mango-trees have been sprinkled and the forefathers have been satisfied. So there are propositions which may be interpreted in two ways. Thus, the sentence ‘śveto dhavati’ may be interpreted both ways (The dog is running away from here-śvā ito dhavati; or the white one runs). Read the sentence-alambusānām yātā, it may be interpreted in both ways- (He has gone to the Alambusa country; or busānām yātā alam- this one which has attained the colour of straw, is now fit). —

41. “a, i un” is a technical rule and should be read along with “adirantyena saheta” (Pa. 1.1.71) which means that in the case of a pratyāhāra (an, in, ac, etc.) the first syllable together with the attachment (as an) would signify the initial and intervening syllables. The point raised is that why the syllables should not have suffixes and why should not the a i or u combine in euphonic combination. Bhattoji Dikṣita in commenting on the “ai un” sutra says that this is a samjñā sūtra as it is to be interpreted along with the rule “adirantyena saheta” which means that in the sutra"a i un” the ‘a’ with the attachment n will indicate the initial, the ending and the middle vowel. Some say that whether we take a here as the original syllable or its imitative sound, since it has no meaning it has no suffix or declension, or as others have said the suffix sup has been dropped, as the sutras are to be treated as Vedic forms. But be that as it may, why should not there be a sandhi (euphonic combination between a i and u). On this point Bhattoji repudiates the tentative suggestion that the rules of sandhi cannot be operative while the instruction on the syllables are being given. His own reply is that the euphonic combination is only possible when there is contiguity. The rule is that contiguity between vowels and consonants has to be assumed in the same pada, between the dhātu and the upasarga and in the case of the compounds. But in the case of a vakya or a sentence, the contiguity is optional. Since “a i un” forms with “adir- antyena saheta” one vakya, the contiguity is optional and hence there is no sandhi. The condition of contiguity has been admitted in grammar to provide for such cases. Second Ahnika 67 ‘atimāla’ ity atrāntaryato vivṛtasya vivṛtaḥ prāpnoti, samvṛtah syad ity evamarthā pratyapattiḥ. naitad asti. naiva loke na ca vede’kāro vivṛto’sti. kas tarhi? samvṛtaḥ, yo’sti sa bhaviṣyati, tad etat pratyapatti-va- canam jñāpakam eva bhaviṣyati ‘vivṛtasyo- padisyamānasya prayojanam anvākhyāyate’ iti. kaḥ punar atra viseṣaḥ ‘vivṛtasyopa- disyamānasya prayojanam anvākhyāyeta, samvṛtasyopadisyamānasya va vivṛto- padeśaś codyeteti? na khalu kaścid viseṣaḥ. ahopuru- ṣikāmātram tu bhavan āha-samvṛta- syopadiśyamānasya vivṛtopadeśaś codyate’ iti. vayam tu brūmo-‘vivṛtasyopadiśya- manasya prayojanam anvākhyāyate’ iti. (This instruction that an a, though naturally contracted in usage, should be regarded as open in articulation) is not an indication of any particular maxim (i.e. it does not imply the maxim that a vivṛta or open a-kāra in usage, should be regarded as contracted. It cannot be suggested that if it is not enjoined that a here should be taken as open, then it could not be taken as a savarna or homogeneous with a sound and in such a case the shortening of a in atikhaṭvā as atikhatva and atimālā as atimāla would be inadmissible, for if a is not made homologous with a there will be no vowel like a; but this criticism is invalid as this is effectuated by the indicatory suggestion contained in the rule “udicāmātaḥ…” (Pa. 7.3.46). There is some other purpose served in it. What is that? (The reason is that) on account of proximity the a sound as open can only be a substitute for a sound, but the last rule at the end of Panini’s work again instructs that what has been instructed as a should be regarded as relapsing back to a. — Kaiyata notes that the distinction between vivṛta and samvṛta (open and contracted) with the length of time required in pronouncing a vivṛta and samvṛta sound, is that the latter characteristic (the length of time required) is external and that the former, since its efforts begin in the mouth, is internal. 68 Mahābhāṣya of Patanjali But this is not true, for neither in popular usage nor in the Vedas a is pronounced widely, i.e. as open. How is it then pronounced? As closed or contracted? Then that form ought to be articulated which is in usage but the last rule of Panini, a relapse of a into a suggests that throughout the work the a has to be taken as open, as a (for otherwise there is no meaning in suggesting a relapse from a into a). This is a reason why a has been instructed here as having a wide and open articulation. What is the difference here between the two expression, ’the reason for the instruction of wide articulation being stated here’, and the expression that that which is instructed as contracted is being instructed has been wide? (The reply is) there is no difference. It is only vanity. You suggest that which is instructed as contracted, is here instructed as wide. We, however, say the reason for instruction of a as wide is being stated here. Värttika. Bhāṣya. tasya vivṛtopadeśad anyatrâpi vivṛto- padeśaḥ savarna-grahaṇârthaḥ. tasyaitasyākṣara-samamnayikasya vivṛto- padeśad anyatrâpi vivṛtopadeśaḥ kartavyaḥ. kvânyatra? dhātu-prātipadika-pratyaya- nipātasthasya. kim prayojanam? savarṇa-grahanârthaḥ. ākṣara-samāmnāyikenāsya grahaṇam yathā syat. eva. kim ca kāraṇam na syat? vivāra-bhedad ācārya-pravṛttir jñāpayati- bhavaty- ākṣara-samāmnāyikena dhātvādisthasya grahanam iti, yad ayam “akaḥ savarne dirghah” (Pa. 6.1.101) iti pratyāhāre’ko grahaṇam karoti. katham kṛtvā jñāpakam? nahi dvayor- ākṣara-samāmnāyikayor yugapat samavas- thānam asti. naitad asti jñāpakām. asti hy anyad etasya vacane prayojanam. kim? yasyākṣara- Second Ahnika 69 samāmnāyikena grahaṇam asti tadartham etat syat ‘khatvāḍhakam’, ‘mālāḍhakam’ iti. sati prayojane na jñāpakam bhavati. tasmād vivṛtopadeśaḥ kartavyaḥ. ka eşa yatnaś codyate-vivṛtopadeso nāma. vivṛto vopadiśyeta samvṛto vā. ko- nvatra viseṣaḥ ? sa eşa sarva evamartho yatnaḥ kriyate. yany etāni prātipadikany agrahaṇāni teṣām etenābhyupāyenopadeśaś codyate, tad guru bhavati tasmād vaktavyam- ‘dhātvādisthaś ca vivṛtah’ iti. The instruction of wide articulation of the syllables should be extended to other sphere for including the homogenous syllables (Varttika). (In the interpretation of this, the Bhāṣya says) the instruction of wide articulation of a as advised regarding the a occurring in the syllabic order should be extended elsewhere. Where is the elsewhere? (The answer is) in the akāra of dhātus, pratipadikas, pratyayas and nipātas. What is the reason? For including the homologous syllables so that it may be affiliated with those articulated in the syllabic order. Why should it not be so?42 Because the open articulation is different from the closed one. The usage of Panini indicates that the syllables articulated in the syllabic order should include the vowel in the beginning of a root. Thus in the rule “akaḥ savarne dirghah”, i.e., the syllables included under ak, i.e. aiuṛl would be lengthened when followed by homologous vowels. In this pratyāhāra ak has been taken.43 (A question is raised by Patañjali)- how can this indicate that? The a of the syllabic order cannot at the same time take two
42. The open a-kāra is advised in the pratyāhāra involving a different effort from the closed a-kāra, this closed a-kara would not be a savarna of the open variety and consequently the rules for their conjunction and other functions would not be operative. So for effecting the homologousness of all a-karas they are all to be regarded as open.
43. The idea is that when by the combination ak Panini includes the a of the pratyāhāra, a i un; this a is not only a vivṛta or open sound as a, but also the closed sound as a, otherwise the sutra would have been “ikaḥ savarne dirghah”. Moreover, if a as closed sound is not at all included then there will be no provision for making the euphonic combination as danda + agram = dandâgram. Thus by advising the70 Mahābhāṣya of Patanjali forms, the a and a. Therefore the objection may be raised that this has no indicatory value. There was some other necessity for it. What is that? The a that is read in the syllabic order should be read as a in order to provide for the euphonic combination khaṭṭvā +āḍhakam, mālā + āḍḥakam. If there is a necessity for instructing the open sound, it cannot be said to have any indicatory value. Therefore the a of the pratyāhāra should be in- structed as having an open sound. But what is the special use in making much ado about this? (Since a is advised in the pratyāhāra, this a may be articulated as an open sound or a closed sound. There is no difference in that). What does it matter if the open or the closed a be instructed. What is the difference? (In reply to this one-sided objector, Patañajali says)- this instructive effort is made for this purpose. (If this effort would not have been made, it would have been more cumbrous for otherwise) — it should have to be advised that all the a-kāras and the pratipadikas have to be instructed as having an open sound in the method prescribed. This would be more cumbrous. For this reason it has to be separately an- nounced that the initial a of the dhatus should also be regarded as having a vivṛta or open sound.44 a-kāra as open, it is indicated (jñāpaka) by Pāṇini that the closed one is also to be included in it.
44. On the above discussion, Bhaṭṭoji remarks as follows: a is closed, but the other vowels are open. E and ai are still more open, and ai and au are further more open than e or ai, and a is the openest of all. The words vivṛta and samvṛta, open and closed, refer to the similar effort of the mouth, and also to the vowels which are so articulated, i.e, on the one hand, open would mean the sort of opening of the lips by which a particular sound is articulated, and it may be also used adjectively to the vowel sounds articulated with such an open prayatna (effort). But though a has a closed sound yet in order to make it homologous with the long and the long-drawn (pluta) sound, it has been regarded as open for the formative operations of grammar. In order that in actual usage one may not use it as an open sound, Panini makes his final advice in the last sutra"a a”, i.e. the a advised as having an open sound, should be regarded as a closed one. It is this last sutra “a a” that suggests or indicates that throughout the whole of grammar a is to be treated as an open sound. But a as open sound cannot be regarded as homologous with a which requires a muh more open effort. For this reason the Varttika “akarasya vivṛtopadeśa akāragrahaṇārthaḥ.” The idea suggested is that the a sound should be taken not only as open but as wide open as in a-kāra. According to the rule “tulyasya-prayatnam savarnam”, i.e. those only are homologous vowels what are articulated from the same place of the mouth with the same effort; and syllables pronounced with different degrees of effort cannot be regarded as homologous and this without this instruction the closed a could not be regarded as homologous with the widely open a, and there would be no euphonic combination between them. Värttika. Bhāṣya. Second Ahnika dirgha-pluta-vacane ca samvṛta- nivṛttyarthaḥ. 71 dirgha-pluta-vacane ca samvṛta-nivṛttyartho vivṛtopadeśaḥ kartavyaḥ. dirgha-plutau samvṛtau mā bhūtām iti. ‘vṛkṣābhyām’, ‘devadatta’ 3 iti. naiva loke na ca vede dirgha-plutau samvṛtau staḥ. kau tarhi? vivṛtau. yau stas tau bhaviṣyataḥ. 66 ‘sthani prakalpayed etavanusvaro yathā yaṇam/ samvṛtaḥ sthāni, samvṛtau dirghaplutau prakalpayet //” anusvāro yatha yaṇam. tad yathā - sayyantā savvatsaraḥ yallokam, tallokam iti- anusvāraḥ sthani yanam anunāsikam sthānī prakalpayati. viṣama upanyāsaḥ. yuktam yat satas tatra praklptir bhavati, santi hi yaṇaḥ sānunāsikā niranunāsikāś ca. dirgha-plutau punar naiva loke naiva ca vede samvṛtau staḥ. kau tarhi? vivṛtau. yau staḥ tau bhaviṣyataḥ. evam api kuta etat tulya-sthānau prayatna-bhinnau bhaviṣyataḥ, na punas tulya-prayatnau sthānabhinnau syātām īkāra ūkāro veti?

45. The word ‘anuvṛtti’ in the Bhāṣya and the Varttika is used in the peculiar sense of anukula-vṛtti for an imitative articulation. The vowels mentioned in the syllabic order are that had been instructed in the sastra. But the a i ur! involved in the epigrammatic phrase ak are similar to the vowel sounds mentioned in the pratyāhāra ai un. The objection raised is that in such a supposition, since the vowels mentioned in the epigrammatic phrases such as ak etc. not being the same (but only being similar) they would not include the homologous vowels. The pratyahāra sastra was applied to a, but in the śāstra it is not the same a, but only similar a. Therefore the reference in the pratyāhāra would not apply to the vowels referred to in the directive sutras like akaḥ savarne dirghaḥ. 76 Mahābhāṣya of Patanjali there? The words ‘kirina’ and ‘girina’ (formed out of the words ‘kiri’ and ‘giri’) would have a terminal high accent, they being monosyllabic. Here again is the case of the word ‘ghata’ in the formation ‘ghatena tarati ghatikah’ (ghatika is one who swims with a pitcher) the than suffix which is applied to disyllabic words, would not be applicable.46 (Another objection is raised by the Varttika)-the usage also ought to be like things. (The Bhāṣya says)-the usage ought to be like things just as among things many persons cannot do their work with one jug at the same time, so not many persons can pronounce the same a at the same time. (Now follows the refutation of the first objection made by the Varttika). (The difference of the same vowel sounds) is accompanied in different contexts owing to their being marked differently (by their attachments). (The Bhāṣya explains it as follows) that in the different contexts the same a appears with different marks as in the rules “karmanyan”, “ato’nupasarge kah” it is known thereby that there is no confusion of attachments (if the same a is meant then it would be useless to mark it differently with different attachments) or the a may be the same which is pronounced with all kinds of attachments (leading thereby to different grammatical operations).47 (A further objection is raised in the Bhāṣya)-this (attachment) has no such indicative value. It is so (that is the application of the attachment is) for the sake of making the technical name it. It cannot be supposed that the a is covered all over with the attachment like a porcupine, and the formation of the technical term it would be adversely affected. The technical term itrefers to the two. Which two? The first and the last.48 (A reply to this objection is now introduced in the Bhāṣya)-The indicative nature can be known through the application of different marks
46. The words ‘kiri’ and ‘giri’ are obviously disyllabic; but if the supposition is that there is only one i sound, then the two is of ‘kiri’ and ‘giri’ become one and therefore the words become monosyllabic and not disyllabic, and so also ghata become mono-syllabic, because the as in gha and ta are one.
47. The word ‘guna’ in the Bhāṣya literally meaning quality, refers to the anubandhas or attachments; that is the same a is pronounced with different attach- ments for different operations.
48. The purport of the objection is this: it cannot be said that the attachments are intended for distinctive operations of the same vowel, for in the epigrammatic word an the n attachment does not direct any special operation for the vowels included in it or of the a. The attachment is applied only in justification of the formation of the technical term it. Second Ahnika 77 or attachments in different contexts. It is seen that in different contexts the a is differently marked, as the rule “pragdivyato’n” (the an suffix should be added to all roots occurring up to the rule (divyato’ etc.)), or “Sivadibhyo’n” (an suffix should be applied to roots falling under the sivadi list). This is indicative of the fact that there should not be any confusion of operation due to the ignoring of the value of the attachments (anubandha); for had it been so, there will be no meaning in marking the vowels again with the anubandhas.4 49 Or rather the explanation would be that the difficulty is solved by marking the vowels differently in different contexts. (A further objection is raised)-It has been suggested that the attach- ments are added for the formation of the epigrammatic name of it. (To this a further reply is made)-This is no criticism. It follows from popular usage. So among the people, one would speak to Devadatta as follows: here, shave away your hair, or here get your locks tangled, or here have only one tuft of hair at the middle of the head. So whatever mark is suggested the person appears there with that mark. So with whatever the a may be advised in any particular context in whatever mark it appears there.50 Bhāṣya. Värttika. Bhāṣya. yadapy ucyate-“ekajanekāj-grahaneṣu câ- nupapattiḥ” iti. ekājanekāj-grahaṇeṣu cavṛttisan- khyānāt. ekājanekāj-grahaneṣu cävṛtteḥ sankhyānād anekāctvam bhaviṣyati. tad yatha-“sapta- dasa samidhenyo bhavanti “iti, “triḥ
49. If there is any chance of confusion of operations owing to the attachments, then instead of prescribing “sivadibhyo’n”, the prescription might have been “Sivadibhyaḥ a”. It is suggested that the attachment of n signifies that the operations should be of n-attachment and not of n-attachment. It must be said in reply that it is a case of vidhi or prescription and not of niyama or selection.
50. This is an answer to the previous supposition that a should be uttered with all the attachments. It is here suggested that a should be pronounced with that particular characteristic that may be advised in any particular context. An a with a particular characteristic attachment may be advised in any particular context, the object be in that different kinds of grammatic operations are indicated by those attachment marks. 78 Värttika. Bhāṣya. Värttika. Bhāṣya. Mahābhāṣya of Patanjali prathamām anvāha triruttamām” ityāvṛttitaḥ sapta-daśatvam bhavati. evam ihāpyāvṛttito’- nekāctvam bhaviṣyati. bhaved āvṛttitaḥ karyam parihṛtam. iha tu khalu ‘kiriņā’, ‘giriņa’ity ekaj-lakṣaṇam antodattatvam prāpnoty eva. etad api siddham. katham? lokataḥ. tad yatha-rṣi-sahasram ekām kapilām ekai- kaśaḥ sahasrakṛtvo dattvā tayā sarve te sahasra-dakṣiṇaḥ sampannāḥ. evam ihāpi anekāctvam bhavisyati. yad apy ucyate-‘dravyavaccopacārāḥ prāpnuvanti’ iti. bhaved yad asambhavi kāryam, tan naneko yugapat kuryāt. yat tu khalu sambhavi kāryam aneko’pi tad yugapat karoti. tad yatha-ghatasya darśanam sparśanam vā. sambhavi cedam kāryam akārasyoccaraṇam nāma. aneko’pi tad yugapat kariṣyati. ānyabhāvyam tu kāla-sabda-vyavāyāt. anyabhavyam tvakarasya. kutaḥ? kāla- śabda-vyavāyāt. kāla-vyavāyāc chabda- vyavāyāc ca. kāla-vyavāyād ‘danda-agram’. sabda-vyavāyāt ‘daṇḍaḥ’. na caikasyātmano vyavayena bhavitavyam. bhavati ced bhavatyānyabhavyam akārasya. yugapac ca deśa-prthaktva-darśanāt. yugapac ca deśa-pṛthaktva-darśanān manyāmahe-‘anyabhavyam akārasya’ iti. yad ayam yugapad deśa-pṛthaktveṣūpa- labhyate-aśvaḥ, arkaḥ, artha iti. na hy eko Second Ahnika Moore on devadatto yugapat srugghne ca bhavati mathurāyām ca. 79 It has been suggested that there is a difficulty in settling the problem as to whether a word is monosyllabic or multi-syllabic. (The Varttika intervenes and repudiates the objection and says that) the problem of deciding as to monosyllabic or multi-syllabic character is to be settled by counting the number of times that the sound is repeated. (The Bhāṣya interprets it as follows)-the matter of deciding the question as to the monosyllabic or multi-syllabic character, the multi- syllabic character is to be decided by counting the number of times that the sound is pronounced. So here also a word is multi-syllabic even if the same syllabic is pronounced more than once. (A further objection is raised in the Bhasya as it says that) by the number of types that a vowel may be produced, the main purpose may indeed be served; but in the examples ‘kiriņa’ and ‘girina’, the high accentuation of the last syllable characteristic of monosyllabic words, would certainly be applicable.51 This can be decided. How? According to the popular usage. Just as one thousand rṣis by giving away the same cow thousand times (by purchasing each time the cow that is given away) succeeded in donating thousand cows, so here also the same vowel when taken many times by way of articulation could render the word multi- syllabic. (As regards the third objection the Bhāṣya says that) the treat- ment should be in accordance with the manner of ordinary articles (wherein a single article cannot be used by many persons at the same time), the answer is that it would be so if the action was impossible in which case many persons could not simultaneously do it, but there are actions which though one may be performed by many at the same time. Thus many people can see the jug or touch it at the same time. This action, namely, that of articulating the a sound, can be done by many persons at the same time and therefore many may do it simultaneously. (Here the Värttika intervenes with the viewpoint of the objector that the a is not one but many. The Värttika thus says)-it may be supposed that each a may be different from the other, because each is separate from the other by time and by syllables.52
51. The idea is that since the word involving the pronunciation of i which is all one is indeed monosyllabic, the mere repetition of the same vowel cannot render it di- syllabic in nature.
52. Thus, in the case of a i un there is a pause between a, i and un by virtue of which they retain their individual characteristic and do not enter into euphonic combination.80 Mahābhāṣya of Patañjali (The Bhāṣya in interpreting the Varttika says)- the a should be regarded as different. Why so? Because of the intervention of time and other syllables. The first case is exemplified in ‘danda’ ‘agram’ (here there is a pause or lapse of time between the a of ‘danda’ and the a of ‘agram’). The intervention of syllables as in ‘dandah’ (here the a of da is distinct from a of nda, because the syllables nda separate them. An entity cannot remain the same when there is an intervention in the midst. So if there is an intervention, one a becomes different from the other. (Here the Varttika intervenes and emphasises the view that a is not one but many. Thus the Varttika says) that a is many, because it is simulta- neously different in different places. (The Bhāṣya in explaining it says that) we consider one a as different from another, because it can be pronounced at the same time in different places (i.e. words). Thus one a can be articulated at the same time in different words or places such as aśvaḥ (horse), arkaḥ (sun), arthaḥ (the object). Now the same Devadatta cannot be at the same time at the cities of Srughna and Mathura. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. yadi punar ime varṇāḥ- sakunivat syuḥ. tad yatha-śakunaya aśu-gāmitvāt purastād utpatitāḥ paścãd dṛśyante. evam ayam akaro ‘da’ ity atra dṛṣṭo ’nda’ ity atra dṛśyate. naivam sakyam. anityatvam evam syāt. nityāś ca śabdāḥ. nityeṣu ca sabdeṣu kūṭasthair avicālibhir varṇair bhavitavyam anapayopajana-vikāribhih. yadi câyam ‘da’ ity atra drsto ’nda’ ity atra dṛśyeta, nâyam kūṭasthaḥ syāt. yadi punar ime varṇaḥ- adityavat syuḥ. tad yatha eka adityo’nekādhikaraṇastho yugapad deśa-pṛthaktveṣupalabhyate. Again in the word drti the ris made distant from i by the intervention of the sound ta between rand i. See the Bhāṣya on” aṇuditsavarṇasya” (Pa. 1.1.69) in this connection. Värttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. heir Second Ahnika 81 viṣama upanyāsaḥ-naiko draṣṭā ādityam anekādhikaraṇastham yugapad deśa-pṛtha- ktveṣu upalabhate. akāram punar upalabhate. akāram api nopalabhate. kim kāraṇam? śrotropalabdhir buddhinirgrahyaḥ prayogena abhijvalita ākāśa-deśaḥ śabdaḥ. ekam ca punar ākāśam. ākāśadeśā api bahavaḥ. yāvatā bahavaḥ, tasmād anyabhāvyam akārasya. ākṛtigrahaṇāt siddham. avarṇākṛtir upadiṣṭā sarvam avarṇakulam grahiṣyati. tathevarṇākṛtiḥ, tathovarṇākṛtiḥ. tadvacca tapara-karaṇam. evam ca kṛtva taparāḥ kriyante-ākṛti- grahaṇenâ tiprasaktam iti. nanu ca savarna-grahanenâtiprasaktam iti kṛtvā taparāḥ kriyeran. pratyakhyāyate tat “savarne’an-grahaṇam aparibhāṣyam akṛtigrahaṇād ananyatvāc ca❞ iti. hal-grahaneṣu ca. (hal-grahaneṣu ca). kim? “ākṛti-grahaṇāt siddham” ityeva. “jhalo jhali’’(Pā. 8.2.26)- avāttām, avattam, avatta. yatraitan nâsti- an savarṇān gṛhṇāti iti. rūpa-sāmānyād vā. rūpa-sāmānyād va siddham etat. tad yatha- ’tan eva śāṭakān acchādayāmo ye mathura- yam’, ’tān eva sālīn bhuñjmahe ye maga- dheṣu’, ’tad evedam bhavataḥ kārṣāpaṇam yan mathurāyām grhitam.’ anyasmimś 82 Mahābhāṣya of Patanjali cânyasmin rupa-sāmānyāt ’tadevedam’ iti bhavati. evam ihāpi rupa-sāmānyāt siddham (aiun). (The Bhāṣya anticipates the view of those who hold that the syllables are one and says by way of introducing a Varttika)-(If these syllables again) were like the birds, that is, the birds who for quickness of flight when they are flying, appear as going backwards, so the one a sound appearing in da also appears at the end of the conjunction n and d as nda. This cannot be explained in this way, for then the varṇas would be transitory and yet the words are known to be eternal. If the words are eternal, then the syllables must also be devoid of production and destruc- tion and immovable like an anvil. If the same a is seen in da and ṇḍa then it cannot be unchangeable. (Now the Bhāṣya introduces the view of the Mimāmsakas, and in that way introduces a Varttika with the words-if again these syllables) be like sun (Vārttika). Just as the sun being reflected from various places may appear at the same time at different places, (so the varnas may also be the same and yet may appear as different in different contexts and conjunctions). (A further objection is raised in the Bhāṣya)-the analogy does not fit in. It is not possible that one perceiver should perceive the sun as reflected from different places, and thus simultaneously perceive it in different places, but one cannot perceive a in that manner. (To this a further criticism is considered, in the Bhāṣya)-it is not possible to perceive a in that manner. Why so? The ear is the place where sound is grasped and then it is comprehended by intelligence; the sound is roused by making in the ākāśa and ākāśa is one. But the ākāśa has also many positions (and it is again objected) that in accordance with the diversity of the positions of the ākāśa. One as may be different from one another. (Here the Värttika intervenes by refuting the criticisms against those that regard a-kāra as particularistic). (Thus the Varttika says)- the problem is solved by taking a in the sense of akṛti or class-notion. (The Bhāṣya says)-instruction is regarding a as universal a which would include all a-s. So also there is the universal of the i-syllable and u-syllable. Second Ahnika 83 (Here the other Värttika comes and holds that)-the ta-attachment is introduced in the same manner. (The Bhasya says) having done in this way, the syllables are ap- pended with ta. (The appendage of ta-kara is indicative of the fact that a was taken as universal and was all-pervasive. This is hinted by the line of the Bhāṣya as follows)-the syllables are appended with ta, because by taking the syllables in their universal aspect that universal aspect was applicable everywhere (i.e. no means was left for indicating any particu- lar aspect of the syllable). (But it may be objected that) the application of a was limitless, because it included all the homologous sounds, and it is for that reason that particular aspects have to be denoted by the appendage of ta. (To this the Bhāṣya replies that) this view has been refuted in the rule “anudit- savarnasya” in connection of which (the Bhāṣya says that) in compre- hending the homologous sounds an should not be regarded as sufficient, for each vowel is referred to in its universal aspect and as one. (There is however another argument in favour of the universalistic aspect of the syllables. The Värttika says that the universalistic aspect of the syllables is further understood when) take the scope of comprehension of the epigrammatic term hal. (The Bhāṣya in explaining it says)-what is this? The problem is settled by taking the syllables in universalistic sense. When the syllables included within jhal follow similar syllables, then those syllables, that is, the latter are elided. Thus we have the forms-avattam, avattam, avatta. (The lun form of the root vas). Here there is no such rule that an includes the homologous syllables).53 (Here the Varttika intervenes and explains the situation by accepting the particularistic view and says) or it may be because of the similarity of form (Vārttika). (The Bhāṣya in interpreting it says) or rather the problem is solved on account of the similarity of forms. Thus people say, ‘We cover those clothes which are in Mathura, we eat those rice which are in Magadha, it is the same kārṣāpaṇa that was taken in Mathura’. Even if the things are
53. In the particularistic view the ta-kāra in kha pha cha tha tha ca ta tav, is only an imitative variety of one ta-kāra, and the two ta-karas (the sic between which is elided on the supposition that sic is jhal which comes after a jhal and is followed by a jhal) and the two ta-karas are not both included within the epigrammatic term jhal and if it is not so, sic would not be elided. (Thus if we hold the particularistic view the forms of avattam, avatta, etc. cannot be explained. 84 Mahābhāṣya of Patanjali different, they are regarded as one on account of their similarity of form. Here also it would be on account of the similarity of forms.54
54. Now after much discussion it was admitted that the a articulated in the syllabic order, is open, but the a of the syllabic order is not the same as the a in the dhātus, pratipadikas and the nipatas. The latter is of the closed type and as such there should not only be no homologousness between the two types, but latter, that is, the a of the dhātus, pratipadikas, etc. cannot be counted under ac, the a in ac being of an entirely different type; and this would lead to much confusion in the operation of Panini’s sutra. Some instances of such confusion have been exemplified by Bhattoji which I cannot discuss here for want of space. It is further pointed out there that since the a referred to in “asya cvau” is a samvṛta variety, it is not the a-kära which is included in an and thus since the a referred to in “asya cvau” is a samvṛta variety such as the ain aiun, the direction that a becomes i before cvi, can refer only to the samvṛta variety of a as suklibhavati, but not to the vivṛta variety as exemplified in the exemplification mālībhavati. It cannot also be suggested that by virtue of the fact that the a of the syllabic order has been declared as open, the a-s that are similar to it in the prätipadikas and the dhātus will have the a of the syllabic order as open, is well-satisfied in rendering an euphonic combination of words where a as initial follows a as the ending vowel, e.g., khaṭṭvāḍhakam. The force of the declension of the syllabic a as vivṛta having been exhausted here, it may well be held that it cannot have any further indicative value by which we can extend its scope to the a sounds similar to it and to other a-kāras found in dhātus and pradipadikas. It has been further declared that the a in aiun has been taken to mean the open a-sound in order to effectuate its homologousness with the long and the long-drawn sounds, by which alone (that is savarṇata) the euphonic combinations can be made as in khaṭṭvāḍhaka. But it is suggested that it is directed that not only the a in the pratyahāra should be articulated as open, but whenever a occurs in the grammar should be regarded as a (iha Sastre yāvān akāraḥ sa sarvo’pi prakriyāyām vivṛto boddhavyah). But even then, since with every articulation we have different syllable (that is each articulated ‘a’is different from every other articulated a pronounced a second later by the same person or by other persons). These ‘a’-s not being included within the pratyāhāra aṇ (aiun), these ‘a’-s would not be homologous with each other; in the rule, ‘akaḥ savarne dirghah” the ‘a’is an imitative variety of the pratyahāra a, and therefore the homolo- gousness should be limmited to the pratyahāras only and not elsewhere. Here it is suggested that there is only one a and the same a that we find in the pratyahara is found elsewhere also, and it is futile to think that the ‘a’ -s occurring in other places are different from the ‘a’ in the pratyahāra an. This does not involve any difficulty, for we can quite conceive that the same ‘a’may be pronounced with an udatta accent, and that this accentual character will be manifested by the articulated sound. The provision of attachments can also be explained after the analogy of popular usage. Thus the same man may be asked to keep a tangled hair at a certain place under certain circumstances or to shave it at other place and under qther circumstances. So the same ‘a’ may induce different grammatical operations under different circumstances and these are distinguished by the difference in attachments or anubandhas which do not Second Ahnika 85 make the ‘a’ different, but only indicate the nature of grammatical operations that would be induced by the same ‘a’ under different circumstances. But it is again objected that if ‘a’ is the same, then in the word ‘ghata’ we should not take into consideration the two articulated a-kāras associated with gha and ta, and regard them as identical and therefore consider the word as monosyllabic, but such a suspposition would prevent us from appyling the ’than’ which can only be done through disyllabic words. The reply is that since ‘a’ is articulated twice in the same word ‘ghata’, the word ‘ghata’ should be regarded as disyllabic. So from these considerations ‘a’ should be regarded as one but to this particularistic object. They say that the same ‘a’ cannot have different accentual characteristics. They further say that the varnas are the characters of sound, but if that is so then ka, ga, a, i, o may all be regarded as characteristic of sound and may, therefore, be regarded as one, and we hear also a pada is a whole or a vakya is a whole. Moreover, in such a view, the varnas are manifested through the sounds and may not be regarded to have any real existence. We may serve our purpose by sticking to the manifesting power of sounds. Thus Bhatṛhari has said the universal that is manifested by many particulars, is its sphota. Others, however, regard the sounds as manifesting the particulars. So if we can explain only the manifesting power of sound, we have the view of jäti-sphota, otherwise the various varņas are all to be regarded as different even in their different accentuations. So leaving aside the view of the Mimamsakas, we may accept that the varņas are infinite, and we may take them as standing for the universal ‘a’, a universal ‘i’ and so forth. Though this view has its justification, yet it does not appear that this view was sponsored by Panini who denied the universal character of a and regarded that it has 18 varieties. Thus there is difficulty both in regarding a as universal or as particular. As a matter of fact there are three views contending with each other. One view is that the syllables are one as manifested by sound (dhvani). The other view is that though they are many their significance is in their class-notion and the third view is that each syllable is as many as there are ways of articulation in each articulated a-kāra. Bhattoji holds that Panini’s rule can be explained on any of these suppositions. Thus the a-kāra though articulated with different accents or by different persons or in different contexts, may be recognised as one, and this will serve to explain why the a-kara in “Ato’nupasarge ka” and “karmany an” should induce different operations; and in order to extend the operation of the rule “asya cvau” to a and a for providing for the cases of suklibhavati and mālībhavati, the short a has been declared to be open. The objection is invalid that the a in “asya cvau” is not the same as a in an, for the a in an includes both the closed and the open variety. These two characters do not belong to a alternately as mundo bhava and jațilo bhava, but simultaneously as syamo dirghas caturvahuḥ. The difference in accentuation or in the nasal and non-nasal character, is not a real difference as has been explained in the sutra “vrddhir adaic” (Pa. 1.1.1). Again in the view where a is regarded as infinite and in the view where a is regarded as having 18 varieties, it is no doubt admitted that the universal character of a both as implying hrasva, dirgha, pluta jointly and severally as hrasva, dirgha and pluta. According to this view the open character has been ascribed to one universal a implying hrasva, dirgha and pluta in the aiun pratyahāra. Therefore both the short and the long variety are included within it and there will be no difficulty in the scope Mahābhāṣya of Patañjali 86 Pāṇini. ṛļk. 2 Bhāṣya. Värttika. Bhāṣya. atha l-kāropadeśaḥ kimarthaḥ? kim viseṣena l-kāropadeśaś codyate, na punar anyeṣām api varṇānām upadeśas co- dyate? yadi kimcid anyeṣām api varṇānām upadese prayojanam asti l-kāropadeśasyâpi tad bhavitum arhati. ko vā viseṣaḥ? ayam asti viseṣaḥ-asya hi l-kārasya alpiyāms caiva prayoga-visayaḥ. yaścâpi prayoga-visayaḥ so’pi klpisthasyaiva. kṛ- peśca latvam asiddham. tasyâsiddhatvād r-kārasyaiva ac-karyāņi bhaviṣyanti nârtha l- kāropadesena. ata uttaram pathati- 1-kāropadeśo yadṛccha-śakti- jānukaraṇa-plutyādyarthaḥ. 1-kāropadeśaḥ kriyate yadṛccha-sabdârtho’- śaktijānukaraṇârthaḥ, plutyadyarthaś ca. yadṛcchāśabdârthas tavat -yadṛcchaya kaścid Įtako nama. tasminn ackāryāṇi yathā syuḥ-dadh-y-ltakāya dehi, madh-v-ltakāya dehi, udańnltako’gamat, pratyannĮtako’ gamat. catuṣṭayi śabdānām pravṛttiḥ-jāti- śabdāḥ, guṇa-sabdāḥ, kriyā-śabdāḥ, yadṛccha-sabdaś caturthāḥ. aśaktijānukaraṇârthaḥ-aśaktyā kayācid brāhmaṇyā ‘ṛtaka’ iti prayoktavye ‘Įtaka’ iti prayuktam, tasyânukaraṇam ‘brāhmaṇy-ltaka of the application of the rule, “asya cvau”. The same explanation applies to the formation of the words-avattām, avattam, avatta involving the elision of the sof sic. Otherwise if the two ta-kāras were not regarded as being included in jhal form would have been avatstām. Värttika. Bhāṣya. biedig Värttika. Second Ahnika ityāha, ‘kumar-y-ltaka- ityāheti. plutādyarthaś ca lkāropadeśaḥ kartavyaḥ. ke punaḥ plutyadayaḥ? pluti-dvirvacana- svaritāḥ. klpta3-śikha, klpptaḥ, prakṬptaḥ. plutyādiṣu kāryesu kṛper latvam siddham. tasya siddhatvād ac-kāryāņi na siddhyanti. fotasmād l-kāropadeśaḥ kriyate. naitāni santi prayojanāni. 87 nyāyyabhāvāt kalpanam samjñādiṣu. nyayyasya ṛtaka-sabdasya bhāvāt kalpanam samjñādiṣu, sadhu manyante-‘ṛtaka evāsau, na Įtakaḥ’ iti. apara aha- ’nyāyya ṛtaka- sabdaḥ śāstrānvito’sti. sa kalpayitavyaḥ sādhuḥ samjñādiṣu-ṛtaka evāsau na Įtakaḥ. ayam tarhi yadṛccha-sabdo’parihāryaḥ— !phidaḥ lphidḍaśceti. eşo’pi rphidaḥ, rphidḍaś ca. katham? arti- pravṛttiś caiva loke lakṣyate, phiḍa-phiḍḍāv auṇādikau pratyayau. trayi ca sabdānām pravṛttiḥ-jāti-sabdaḥ, guṇa-sabdaḥ, kriyā- śabda iti. na santi yadṛcchā-sabdāḥ. anyathā kṛtvā prayojanam uktam, anyathā kṛtvā parihāraḥ. ‘santi yadṛcchā-śabdāḥ’ iti kṛtvā prayoja- nam uktam, ’na santi’ iti parihāraḥ. samāne cârthe śāstrānvito’sastrānvitasya nivartako bhavati. tad yathā devadatta-śabdo devadiņņa-sabdam nivartayati. na gāvyādīn. naiṣa doṣaḥ. pakṣāntarair api parihārā bhavanti. anukaraṇam śistâśistâpratiṣiddheṣu yathā laukika-vaidikeṣu. 88 Bhāṣya. Mahābhāṣya of Patanjali anukaranam hi siṣṭasya vā sadhu bhavati, aśiştâpratiṣiddhasya va, naiva tad doṣāya bhavati, nâbhyudayāya. yatha laukika- vaidikeṣu. yatha laukikeṣu vaidikeṣu ca kṛtānteṣu. loke tavad ya evam asau dadāti ya evam asau yajate, ya evam asau adhite iti. tasyânukurvan dadyac ca yajeta câdhiyīta ca so’py abhyudayena yajyate. vede’pi-“ya evam viśva-srjaḥ satrany adhyasate” iti, teṣām anukurvan tadvat satrany adhyāsīta. so’py abhyudayena yujyate. aśiştâprati- ṣiddham yatha-ya evam asau hikkati, ya evam asau hasati, ya evam asau kaṇḍūyatiti. tasyânukurvan hikkec ca hasec ca kaṇḍūyec ca, naiva tad doṣāya syan nâbhyudayāya. yas tu khalv evam asau brāhmaṇam hanti evam asau surām pibatīti tasyânukurvan brāhmaṇam hanyāt surām vā pibet so’pi manye patitaḥ syāt. viṣama upanyāsaḥ. yaś caivam hanti, yas cānuhanti, ubhau tau hataḥ. yaś capi pibati, yaś cânupibati, ubhau tau pibataḥ. yas tu khalv evam asau brāhmaṇam hanti, evam asau surām vā pibatīti tasyânukurvan snātānulipto malyaguna-kanthaḥ kadali- stambham chindyāt payo va pibet, na sa ma- nye patitaḥ syāt. evam ihâpi ya evam asau apaśabdam prayunkte iti tasyânukurvann apaśabdam prayuñjita, so’py apaśabdabhāk syāt. ayam tvanyo’paśabda-padarthakaḥ śabdo, yadartha upadeśaḥ kartavyaḥ. na câpaśabda- padarthakaḥ śabdo’paśabdo bhavati. Värttika. Bhāṣya. Second Ahnika 89 avasyam caitad evam vijñeyam-yo hi man- yate ‘apaśabda-padarthakaḥ śabdo’paśabdo bhavati’ iti, ‘apaśabda’ ity eva tasyâpasabdaḥ syat. na caiṣo’pasabdaḥ. ayam khalv api bhūyo’nukaraṇa-śabdo’- parihāryaḥ, yadartha upadeśaḥ kartavyaḥ- ‘sādhv ļkāram adhite’, ‘madhv Įkāram adhite’ iti. kvasthasya punar etad anukaraṇam? klpisthasya. yadi klpisthasya klpeś ca latvam asiddham tasyâsiddhatvad r-kāra evâckar- yāņi bhaviṣyanti. bhavet tadarthena nârthaḥ syāt. ayam tvanyaḥ klpistha-padârthakaḥ sabdaḥ, yadartha upadeśaḥ kartavyaḥ. na kartavyah. idam avaśyam kartavyam- “prakṛtivad anukaraṇam bhavati” iti. kim prayojanam? ‘dviḥ pacantv ityaha’, “tinnatina” (Pā. 8.1.28) iti nighāto yathā syat. ‘agni ityāha’. “id-ud-ed-dvivacanam pragṛhyam”(Pa. 1.1.11) iti pragṛhya-saṁjñā yathā syāt. yadi ‘prakṛtivad anukaraṇam bhavati’ ity ucyate, apaśabda evâsau bhavati- ‘kumaryļtaka ity aha’, ‘brāhmaṇyltaka ity aha’. apaśabdo hy asya prakṛtih. ’na câpaśabdaḥ prakṛtiḥ. nahy apaśabdā upadisyante, na cânupadiṣṭā prakṛtir asti. ekadeśa-vikṛtasyânanyatvāt plutyādayaḥ. ’ekadeśa-vikṛtam ananyavad bhavati’ iti plutyādayo’pi bhavisyanti.90 Värttika. Bhāṣya. Mahābhāṣya of Patañjali yadi ’ekadeśavikṛtam ananyavad bhavati’ ity ucyeta, “rājñaḥ ka ca” (Pa. 4.2.140) rājakiyam, “allopo’nah” (Pa. 6.4.134) iti lopaḥ prāpnoti. ekadeśa-vikṛtam ananyavat şaṣṭhi- nirdiṣṭasya. yadi ‘ṣasthi-nirdiṣṭasya’ ity ucyate, ‘kl3pta-sikha’ iti pluto na prapnoti. na hy atra ṛkāraḥ ṣaṣṭhinirdiṣṭaḥ. kas tarhi? rephah r-kāro’py atra şaṣṭhi-nirdiṣṭah. katham? avibhaktiko nirdeśah-kṛpa uḥ raḥ laḥ- “krpo ro laḥ” iti. athavā punar astv aviseṣeņa. nanu coktam-“rājñaḥ ka ca” rājakiyam “allopo’nah” iti prāpnotiti. naiṣa doṣaḥ. vakṣyaty etat- “śvādinām samprasaraṇe nakaranta-grahaṇam ana- kārānta-pratiṣedhârtham” iti. tat prakṛtam uttaratrânuvartiṣyate “allopo’naḥ” nakārāntasyeti. iha tarhi-‘kl3ptaśikha’ “anṛta” iti pratiṣedhaḥ prāpnoti. ravat pratiṣedhāc ca. ravat-pratiṣedhāc caitat siddhyati. “guror aravataḥ” iti vakṣyāmi. yadi ‘aravataḥ’ ity ucyate, hotr+rkāraḥ ‘hotṛ3kāraḥ’ atra na prāpnoti. “gurorararavato hrasvasya” iti vakṣyāmi. sa eșa sutra-bhedena Ikāropodeṣaḥ plutyadyarthaḥ san pratyakhyāyate. saiṣā mahato vamśa-stamballaṭṭvānukṛṣyate. (rik) (The Bhāṣya in introducing the above Vārttika says)-why has Įkāra been (separately instructed)? What is the special significance of the Second Ahnika 91 instruction of l-kara? But, well, there ought not to be any further instruc- tion of any of the other syllables; whatever necessity there may be in the instruction of the other syllables the same necessity applies with regard to the instruction of 1-kara. What is the difference? There is this difference. There is very little scope for the use of l-kāra. Whatever use there is, applies to the root klpi. The sound of kṛpi is invalid, and the vowel changes that are to be affected should refer to the -kara in the kṛpi and, therefore, there is no use in advising 1-kara. So the reply is given (in the following Värttika). The instruction of l-kara is for underived words and ineffective imita- tions and for the long-drawn sound of pluta (Varttika). (The Bhāṣya says) the instruction of underived words and for in- effective imitations and for the long-drawn pluta sound by the word yadrccha’ (translated here as underived words but which really means words or names formed at will). Thus according to one’s wish we may have a name Įtaka. The advice regarding ! is given in order that vowel- change may be affected with regard to such names as in the examples ‘dadhyltakāya dehi’, ‘madhvltakāya dehi’, ‘udańňltako’gamat’, ‘pratyannltako’gamat. Words are used in four ways: words signifying universals, signifying qualities, signifying activities and signifying just whatever one might wish to have.55 (The 1-kara has been advised also for ineffective imitation.) Thus some Brahmin woman or wife being unable to pronounce the word Įtaka where rtaka should have been used and this may further be imitated, and one may say-brāhmaṇyltaka ity aha (the Brāhmaṇi said Itaka) or kumaryltka ity ähaiti (the maiden pronounced it as Itaka). The instruction of l-kara is also for the long-drawn pluta and the other intonations. What are these pluta and other intonations? The pluta consists in the doubling tone, such as kl3pta-sikha, klptaḥ, prak!ptah. In the pluta and other intonations, the rof krp becomes 1, and unless is advised, no vowel-changes can be effected. For this reason 1-kāra has to be advised. (The Bhāṣya in raising the objection says) there are none of these purposes (to be served). (This introduces the Värttika)-since there are correct words, these only should be used as names. (The Bhāṣya says)-since there is the correct word ‘rtaka’, they are to be used in forming names as correct; and that is ‘rtaka’, and not ‘Itaka’.
55. When words are used according to the wish of the speaker without any reference to meaning, they are called yadṛccha śabdas, such as ‘dittha’, etc. 92 Mahābhāṣya of Patanjali (This Vārttika has been differently explained). This another interprets it thus the word ‘rtaka’ is justifiable and is found in the sastras and this only should be used as correct in forming names, and that word is ‘ṛtaka’ and not ‘Itaka’. But there is a self-willed word (yaḍrccha śabda) which cannot be avoided, such as Iphida, Iphidda, but this is also pronounced as ‘rphida and ‘rphidda’. How ‘phida’ and ‘phidda’ are auṇādika suffixes, and they may be applied to root r by the people. The words may be used in a three- fold manner: as universals, qualities and actions. There are no self-willed ones. (To this the Bhāṣşya replies)-the need of the 1-kāras is explained in one manner, and it is avoided in another manner. The need was shown on the supposition that there existed self-willed words. It is avoided on the supposition that there are no such self-willed words. When the meaning is the same, the word that is sanctioned by the śāstras, rules out the word that is not sanctioned by the śāstras. Thus the word ‘Devadatta’ rules out the word ‘devadinna’, but not such words as gavi, etc. (Those that favour the non-instruction of l-kara are supposed to say as follows)-this is no fault. One may avoid an objection by taking other alternatives. (The Varttika intervenes and says)-an imitation of the cultured is correct, but an imitation of the uncultured, and that which is not prohi- bited, is neither bad nor good as among the people and Vedic usage. (Now the Bhāṣya explains it)-the imitation of the cultured may be correct, but an imitation of the uncultured or that which is not prohibited would be neither good nor bad as among ordinary people and Vedic people. Among the people there is an usage-‘he who gives like that’, ‘he who sacrifices like that’, ‘he who studies like this’, he who imitates him, would make donations, perform sacrifices and study, and will attain good. In the Vedas also, it is said he who performs in this way the viśva sacrifices, when one imitates such persons, he also sacrifices; he thus attains good. In the case of the uncultured and unprohibited, one may say-‘he who hiccoughs like this’, ‘he who laughs like this’, ‘he who itches like this’, when one imitates that, he would hiccough, laugh and itch, and that would be neither good nor evil. He who kills a Brahmin like this, he who drinks liquor like this, here by such an imitation if one kills a Brahmin or drinks liquor, would certainly be fallen. (To this the Bhāṣya further says) the analogy is contrarily put. He who kills and he who follows him in killing, both are performing the operation of killing. When one drinks and the other follows him, both of Second Ahnika 93 them drink. So when he drinks wine, imitating such a person, one will merely perform a consequent action like ‘one first bathed and then besmeared his body with sandal paste and put garlands on his neck’? One may cut down the trunk of a banana tree or drink milk, he will not be fallen thereby 56 So here also when one uses the vulgar word and another imitates him and utters a vulgar word, he becomes guilty of uttering a vulgar word. But this is quite another point that there are words of the same meaning as the vulgar words, for which reason the instruction is made. The word does not become an apaśabda (vulgar), because it has the meaning of an apaśabda. It must be so understood. He who thinks that a word becomes apaśabda when signifying the meaning of an apaśabda (is wrong). A vulgar word or apaśabda becomes vulgar, because it is a vulgar word (that is, not because it denotes the meaning of a vulgar word). But this word is not vulgar. Again here also imitation is unavoidable for which imitation is necessary. ‘sadhvļkāram adhita’, ‘madhvlkāram adhita’ (one is studying 1-kara well, one is studying 1-kāra sweetly). (The the next question comes)-of where is it an imitation? In the word ‘klpi’. If it is in ‘klpi’, then the 1-kara in ‘klpi’ will be incorrect. The vowel- changes should be made in r-kāra. (The necessity of advising the 1-kara is adduced by the Bhāṣya). May be, no purpose is served in that way, but this is another word having the same meaning as that of the word ‘klpi’ for which instruction is necessary. Such an instruction is unnecessary. This must be said that an imitation is after the original pattern. What is the necessity of imitation? “He said twice the word pacantu””, so that the accent may be lowered by the rule “tinnatinah”. Again, ‘he said agni’, so that agni may become pragṛhya under the rule ‘i-kāra, ū-kāra, e-kara’ in the dual number, are called pragrhya.’ (If a syllable is pragṛhya, it does not enter into an euphonic combination).
56. It is not by the fault of that which is imitated that the imitation becomes faulty and in the imitation of wrong deeds there is demerit, and in the performance of good deed there is merit, when one imitated another in doing good or bad deeds, one does not attain merit and demerit for imitating, but because he performed the enjoined action or committed the prohibited one. So it is not by imitating ṛtaka as Įtaka that any fault accrues to anyone. That is, it may be an imitation or anukarana word for which instruction is necessary, though it may not be the actual root ‘klpi’. The Bhāṣya further urges. 94 Mahābhāṣya of Patanjali (Again a further objection is raised)-If an imitation follows the original, then we must say that when the maiden said ‘Itaka’ or the Brahmin’s wife said ‘Itaka’, that was a vulgar word. (It cannot be an imitation of the word ‘ṛtaka’, as the maxim is that an imitation should follow the original. So the form ‘Itaka’ must be an imitation of the vulgar word ‘Itaka’.) (The reply given is)-a vulgar word cannot be the original. The vulgar words are not instructed in the sastras and there are no originals which are not instructed in the śästras.57 (Here the Värttika intervenes)-slight partial changes seem to keep the thing unaffected and so it may provide for the pluta intonation. (The Bhāṣya interprets it as follows)-Since change in part leaves the thing as if it were the same, the accents pluta etc. may take place. 58 (To this an objection is raised in the Bhāṣya which says that)-the partial modification of anything leaves it the same. Then in the rule “rajñaḥ ka ca” (there will be ka-suffix to the word ‘rajan’) forming the word “rajakiya’, we should expect the application of the rule “allopo’nah” (the a of a word ending in an should be elided). (To this a further reply is suggested)-The maxim that partial modifi cation leaves the word undisturbed, applies to a word ending in the sixth- case 59 (The Bhāṣya further says)-again if the condition that it should be enjoined by the sixth-case, be insisted on them, there would be no pluta in the word ‘k13pta-sikha’, for in the rule “krpo ro lah” the r-kara has not been mentioned in the sixth-case and so it cannot be the same as 1. (To this the Bhāṣya further replies)-that the rhere is also in the sixth- case for the form of the sutra “krpo ro lah” is to be analysed as kṛpa, u, ra, la. Only the declension suffixes have not been shown in the sutra. (The Bhāṣya again says) or rather, let there be no restriction (with reference to the maxim that a partial change is no change; that it should be enjoined in the sixth-case). But it has been said that in the formation of
57. The purport of the argument introduced in the Bhāṣya is that l-kara has been introduced in order to provide for the inefficient imitation of the word ‘rtaka’.
58. When the r-kara in the pluta accent becomes changed to l-kara, this partial change may be supposed to leave the word undisturbed.
59. The idea is that in the case of a universal application of the maxim the slight change is no change; we may treat the word rajan and rajñaḥ as if they were the same and in the formation of the word rajakiya we may have the word “rājña-kiya’, which is impossible. Second Ahnika 95 the word ‘rājakiyam’ by the rule “rajñaḥ ka ca”, there is an application of the rule “allopo’nah” (i.e. the a of an an-bhāgânta word should be elided). (The Bhāṣya returns to reply to this objection and says)-this is no criticism. It will be said later on that in the expansion of the words like ‘évan’ etc., the injunction that it should end in n, is only to rule out all words that do not end in n. This context is drawn on to the later sūtra “allopo’nah”, i.e. of the words ending in na.60 (The Bhāṣya further raises the point of view of the objector and says)- What is to be done here in ‘k13pta-sikha’. Since there is a prohibition by the rule “anṛta”, i.e. r-kāra should not have the pluta as 1-kāra. (Here a further objection comes in by the mouth of the Värttika)-Also because the prohibition applies to those that have r. (The Bhāṣya in explaining it says that) this will be effected by the prohibition that applies to words ending in r.61 We shall say later on “guroraravatah”. If it is suggested that the word should be ‘aravataḥ’, i.e., those that do not contain ra, then there would be prohibition in the lengthening of the ṛ-kara in the euphonic combination hotr+r-kara = hotr2kāra. (The Bhāṣya comes to the final solution and says)-The prohibition of pluta or the long drawn applies to the short sound and not to the long. So the introduction of a special sūtra for the instruction of l-kara for the pluta intonation is unnecessary. It is like scaling over a big bamboo for drawing the insignificant fruit called laṭvā.62
60. The point in question here is that when the ka-suffix is enjoined, the word ‘rajan’ may also be regarded as slightly changed and as ending in n-kara in accordance with the maxim that slight change does not distrub the identity. The reply is that by the word ekadeśa or part referring to the partial change, it is held that it is only a whole word that is regarded as the same, when there is a slight change in the part which appears in the whole. But a part that is changed partly, is not regarded as identical in spite of the change. Thus a dog with its tail partly cut may be regarded as the identical dog, but the cut-tail is not identical with the uncut-tail. This is all according to popular usage. Thus the slight difference in the word ‘rajan’ may leave the word undisturbed, but to say that ending in an is the same as ending in n cannot be tolerated.
61. That which always has ṛ, is called ravat. Therefore the ṛ-kara is ravat.
62. The idea is that so much ado about the instruction of l-kara is quite unnecessary. There is no significant ground for giving this special instruction about l-kāra. 96 Mahābhāṣya of Patanjali Panini. e oǹ. 3. Bhāṣya. Värttika. ai auc. 4. idam vicāryate-imani sandhy-akṣarāṇi ta- parāni vopadiśyeran, ’et ot n’, ‘ait aut c’ iti, ataparāņi vā yathānyāsam, iti. kaś câtra viseṣaḥ? sandhyakṣareṣu ta-paropadeśaś cet ta- paroccāraṇam. sandhyakṣareṣu ta-paropadeśaś cet ta- Bhāṣya. paroccāraṇam kartavyam. plutyādiṣv aj-vidhiḥ. Värttika. Bhāṣya. Värttika. Bhāṣya. Varttika. Bhāṣya. plutyādiṣv aj-āśrayo vidhir na sidhyati. ‘go3trata’, ’nau4trata’ ity atra “anaci ca” (Pa. 8.4.47) ity aca uttarasya yaro dve bhavata iti dvirvacanam na prāpnoti, iha ca ‘pratyannai4tikāyana’, ‘udannau4pagava’ iti “aci” (Pa. 8.8.32) iti namuḍāgamo na prā- pnoti. pluta-samjñā ca. pluta-samjña ca na sidhyati. ‘aiзtikāyana’, auзpagava. “ūkāloj-hrasva-dirgha-plutah” (Pa. 8.3.32) iti pluta-samjña ca na prāpnoti. santu tarhy ataparāṇi. atapara eva igghrasvadeśe. yady ataparāṇi “eca igghrasvadese” (Pa.1.1.48) iti vaktavyam. kim prayojanam? eco hrasvadeśa-śāsaneṣv ardha ekāro’rdha okāro vā mā bhūd iti. Värttika. Bhāṣya. Second Ahnika nanu ca yasyâpi ta-parāni, tenâpy etad vaktavyam. imav aicau samāhāra-varṇau— mātrā’varṇasya, matrevarṇovarnayos tayor hrasvadeśa-śāsaneṣu kadācid avarṇaḥ syāt, kadācid ivarṇovarṇau, mā kadācid avarno bhūd iti. 97 pratyakhyāyata etat “aicos cottarabhūy- astväť’ iti. yadi pratyākhyāna-paksaḥ, idam api pratyakhyāyate-“siddhamenaḥ sasthana- tvāť’ iti. nanu cainaḥ sasthanataravardha ekāro’- rdha okāraḥ. na tau staḥ. yadi hi tau syātām tāveva ayam upadiset. nanu ca bhoś chandogānām satyamu- griraṇāyaṇiyā ardham ekāram ardham okāram câdhiyate-“sujate eśvasūnṛte”, “adhvaryo odribhiḥ sutam”, “sukram te en- yad”, “yajatam te enyat’ iti. pāriṣadakṛtir eṣā tatra-bhavatām. naiva hi loke nänyasmin vede ardha ekāro’rdha okāro vā’sti. ekādeśe dirgha-grahaṇam. ekādeśe dirgha-grahaṇam kartavyam- “adguno dirghaḥ”, “vṛddhir eci dirghaḥ” iti. kim prayojanam ? antaryatas trimatraca- turmātrāṇām sthaninām trimātra-caturmätrā ādeśā mā bhūvann iti. khaṭvā indraḥ khatvendraḥ, khaṭvā udakam khaṭvodakam, khaṭvā iṣā khatveṣā, khaṭvā ūḍhā khaṭvoḍhā, khaṭvā elaka, khaṭvailaka, khaṭvā odanaḥ 98 Mahābhāṣya of Patanjali khatvaudanaḥ, khaṭvā aitikāyanaḥ khaṭvaitikāyanaḥ, khaṭvā aupagavaḥ khatvaupagava iti. tat tarhi dirghagrahaṇam kartavyam? na kartavyam. upariṣṭād yoga-vibhagaḥ kariṣyate- “akaḥ savarne” eko bhavati. tato “dirghaḥ” dirghaś ca sa bhavati yaḥ sa “ekaḥ purvaparayoḥ” (Pa. 6.1.84) ity evam nirdista iti. ihâpi tarhi prāpnoti- paśum, viddham, pacantiti. naiṣa doṣaḥ. iha tavat paśum iti. “amyekaḥ” iti iyatā siddham. so’yam evam siddhe sati yat pūrva-grahaṇam karoti tasyaitat prayo- janam-yathājātīyakaḥ pūrvas tathājātīyaka ubhayor yatha syad iti. ‘viddham’ iti. ‘pūrva’ ity evânuvartate. athavā ācāryapra- vittir jñāpayati ’nânena samprasaraṇasya dirgho bhavati’ iti. yad ayam hala uttarasya samprasaraṇasya dirghatvam śāsti. ‘pacanti’ iti. “ato gune parah” (Pa. 6.1.97) iti iyatā siddham. so’- yam evam siddhe sati yadrūpagrahaṇam karoti tasyaitat prayoj- anam — ‘yathājātīya-kam parasya rūpam tathājātīyakam ubhayor yatha syad’ iti. iha tarhi ‘khaṭvārśyo, mālarśyaḥ’ iti dirgha-vacanãd akāro na, anāntaryād ekāraukārau na. tatra ko doṣaḥ? vigṛhitasya śravanam prasajyeta.

63. The word ‘gotrataḥ’ is the vocative case of the word ‘gotratri’. So also the other word is ’nautrāṛ”” is the vocative case. The pluta accent is due here by the rule100 Mahābhāṣya of Patanjali (Here the Varttika intervening says)-Even the name pluta could not be ascribable. (The Bhāṣya here says) -The name pluta becomes invalid, as in ‘aitikāyana’ or ‘aupagava’. The technical name pluta as per Panini’s rule “ūkālojjhrasva-dirgha-plutah” (1.2.27) would not be valid.64 (The Bhāṣya in introducing the next Värttika raises the other side of the issue)-Let the ’en’ and ‘on’ be without the ’ta’ (and then the Värttika intervenes and says) - if ’en’ and ‘on’ be without a ’ta’, then when it is “guror-anṛto’nantyasyäpyekaikasya pracam” (8.2.86) which means that in the case of calling from distance, a long vowel which is neither r-kära, nor is at the end, may take pluta accent optionally. It should be noted that while a short vowel is regarded as having one unit of time, the long is supposed to have two units and the longdrawn or the pluta is supposed to have three units of time. Sometimes the pluta have four units of time. In the Varttika rule above and in the Bhāṣya, we have the word ‘plutyadi’, i.e., pluti etc. Here the word ‘adi’or etc. includes the pluti accents with four units of time or matras. The number of matras in a pluta accent may therefore be indicated by a mathematical number. Thus according to this rule you may have pluta accent after ‘go’ and ’nau’ in ‘go-trataḥ’ or ’nau-trātaḥ’. Now, the objection raised is that if ’et’ and ‘ot’ be taken as ending in ’ta’, they cannot indicate their pluta variety. So in gotrataḥ and pluta ‘o’ in ‘go’ cannot be regarded as a vowel and, therefore, the rule that says that when a consonant follows, then after the preceding vowel the syllables included in the pratyahara ‘yar’, could be duplicated, would not apply. Now, the pratyāhāra ‘yar’ includes in it all consonants excepting ‘ha’, and in accordance with the rule “anaci ca” we expect a doubling of the consonant ’ta’ followed by consonant ‘ra’ and preceded by the vowel ‘o’ in ‘gotrataḥ’, of which the optional form should be ‘gottrataḥ’. Now if pluta accent in ‘go’ is not included in ’en’, when instead of ’en’, we have ’et’ (i.e. ending in ’ta’), then the pluta in ‘go’ cannot technically become a vowe! and that there can be no duplication of ’ta’ in ’tratah’. Again we have the rule “nimo hrasvad aci namun nityam” (8.3.32) which means that when a pada ends in a nasal preceded by a short vowel and followed by a vowel, then a corresponding nasal will arise before the latter. Thus, we have pratyan-ātmā = pratyannātmā. Now, according to the above rule when we have a pluta accent in the word ‘ai’ as forming a part of the word pratyan-naitikäyana, then this ‘ai-kara’ in ‘aitikäyana’ being a pluta accent could not technically be a vowel and there would not therefore be the addition of ’n’ according to the above rule (8.3.32) whch could be operative only if ‘ai’ was a vowel.
64. The idea is that if ’et’ was substituted for ’en’, then we should take the long variety by the rule “taparas tat-kalasya”, i.e. when a vowel is followed by ’ta’, it indicates that the vowel indicates only one variety such as the short or the long. ‘Et’ would thus have only the long variety and no pluta accent would be admissible to it. The rule “ukālo” etc. shows a sort of pronunciation by the gradual increase of the units of time that would form the short, the long and the pluta accent as involving one unit, two units or three units of time. Second Ahnika 101 enjoined that they would be short, it ought to be specified that they should be ‘ik’. (The Bhāṣya says) - If without the ’ta’ then a separate instruction has to be made that when ’ec’ is to become short, it would be ‘ik’. What is the necessity? (The reply is) that when ‘ac’ is to become short, it might not become half ’e’ or half ‘o’.65 Now the Bhāṣya comes in with the objection that may be raised against those that prefer to associate ’ec’ with ’ta’. (The Bhāṣya thus says) - But even those that associate with ’ta’ must enjoin as follows: there two that are included in the ’ec’ are compound vowels, involving the unit of time of ‘a’ and ‘i’ and ‘u’. Now, when it is enjoined that they should be short, sometimes they may be ‘a’ and sometimes they may be ‘i’ and ‘u’, but they should never become ‘a’.66 (The Bhāṣya however says- This objection will be refuted because in ‘aic’ there is a greater portion of the later part (i.e., ‘i’ or ‘u’).67 (The Bhāṣya now gives a solution of those that do not like the associate ’en’ with a ’t’). If the refutation side is to be followed then it may be refuted on the ground that ‘i’ and ‘u’ have the same place of origin as ’en’. (There are no half e-kara or half o-kara. The e-kara is palatal, the o-kāra is labial,
65. If ’e’ or ‘o’ be made to end with a takara it would only include the long variety involving two units of time. But it cannot become i-kāra or u-kāra whch would involve only one unit of time and as such cannot be the substitute of e-kara or o-kāra which being associated with ’ta’ can only be of the long variety with two units of time. If there was no ta-kāra, then there would be no limitation to units of time and even half e-kara and half-i-kāra could melt together and be regarded as a vowel. Therefore when it is associated with the ta-kära, a separate rule has to be made.
66. The idea is that ‘ai’ and ‘au’ being compound vowels, are composed partly of ‘a’, ‘i’ and ‘u’. If ‘ai’ and ‘au’ be associated with ’ta’ they only comprehend the long variety and may be regarded as having ‘a’, ‘i’ and ‘u’, as their elements; as such, if it is enjoined anywhere that ‘ai’ or ‘au’ should be hrasva or short, either ‘a’ or ‘i’ or ‘u’ might take place. But in the pronunciation of ‘ai’ or ‘au’ no ‘a’ is realised and the half of ai-kara or of au-kara would not be vowels, for through the association ’ta’, ‘ait’, and ‘aut’ can be vowels. Moreover, owing to the fact that ‘ai’ and ‘au’ are both palatals, no ‘a’ which is a guttural can be substituted in its place. So even if ‘ai’ and ‘au’ be associated with ’ta’, Panini has to make the rule ’eca ig-hrasvadese’ in order to provide for the coming of ‘ik’, i.e. ‘i’ or ‘u’ when ’ec’ is prescribed to be hrasva.
67. There will be only half unit of ‘a’ which will be overflooded by the half unit of ‘i’ or ‘u’ and consequently ‘i’ or ‘u’ will have greater claim for their appearance at the time of prescription of hrasva than ‘a’. 102 Mahābhāṣya of Patanjali and in their place the palatal i-kāra and the u-kāra which is labial come in. (The word sasthana in the Bhāṣya means samāna sthāna). (The Bhāṣya further raises an objection and says)-But the half e-kara and the half o-kara have a closer proximity to ’en’. (To this objection the reply is) - But they do not exist, if they had existence, they would have been enjoined. (The Bhāṣya again raises a further objection that half e-kara and half o-kāra also exist. Thus the Bhāṣya says)-But the Satyamugrirāṇāyaniya branch of the Samaveda read a half e-kara and a half o-kāra in the passages: ‘sujate eśvasūnṛte’, ‘adhvaryo odribhiḥ sutam’, ‘Sukram te enyat’, ‘yajatam te enyat’. Here in pronouncing e aśva as eśva or o adribhiḥ as odribhiḥ and e anyat as enyat, the e-kara and o-kāra following e-kära and o-kāra are pronounced with half sound. (To this the reply is) - This pronunciation of those holy sages is due to their singing in tune. Neither among the people nor among the other Vedas, there is any sound as half e-kara and half o-kāra.68 (Now the Värttika intervenes and says that) - When one sound is enjoined as the substitute for two sounds in conjunction, that is, an ekādeśa, a long sound is to be taken.69 (In the case of ekādeśa the Bhāṣya in interpreting it says) - In the case where one syllable is substituted in the place of two, i.e. an ekādeśa, it should be specially instructed that the syllable, so substituted, is long. Thus in the case of “ad guṇaḥ” one ought to say “ad guno dirghah” or “vṛddhir aci dirghah.” What is the reason? Because according to the rule of proximity a vowel having three or four units of time, there may not be a substitution of vowels of three or four units of time. Thus khaṭvā indraḥ khaṭvendraḥ, khaṭvā udakam-khatvodakam, khaṭvā iṣā-khatveṣā, khaṭvā ūḍhā = khatvoḍhā, khaṭvā elakā = khaṭvailakā, khaṭvā odanaḥ = khatvaudanaḥ, khaṭvā aitikāyanaḥkhatvaitikāyanaḥ, khaṭvā aupagavaḥ -khatvaupagavaḥ etc.7 70
68. The word ‘parişadakṛti’ in the Bhasya is explained by Kaiyața as giti-vasät tathoccaraṇam. We, however, think that the word ‘parişadakṛti’ actually means- ‘done by the members of an assembly’. Sama songs were sung by a number of priests who may be regarded as forming a sacrificial assembly. The hymns were sung in a chorus. Pāriṣadakṛti thus means the ’e’ or ‘o’ is pronounced with half sound, because it was effected by the chorus of the assembly of the Sama songsters.
69. The idea is that in the case of technical terms of guna and vṛddhi there is association with ’t’ and therefore we have to take the long variety as in “ad guṇaḥ” “vrddhir ad aic”.
70. Bhattoji Dikṣita here remarks as follows: Though among the Sama singers of particular sakhas the half e-kara or half o-kara Second Ahnika 103 Then the injunction that there should be the long vowel (when a single substitute in place of the preceding and the succeeding is advised in ekādeśa). (An one-sided objector’s view is anticipated in the Bhasya as fol- lows) No, it should not be done. Later on, they would divide the rule into two parts, such as “akaḥ savarne” (when a homologous vowel follows ‘a’, T’, ‘u’, ‘r’, ‘’), they become one. Then comes the word ‘dirghah’ (long); are actually pronounced, yet this sound is not universal. If as in the case of a-kara, we may have 18 varieties, so in the case of ’e’, ‘ai’, ‘o’, ‘au’ or ’ec’, also we could have a short variety involving one unit of time or one matra, then ’ec’ should have been advised as being of one matra. It may be contended that the rule, “eca igghrasvadese” was formulated for preventing the coming in of the half e-kara and half o-kara. But the reply is in the negative. The enunciation of the rule has been necessary for preventing the coming in of ‘a-kara’ which is a constituent of e-kara. The first part of ’e-kara’ is like ‘a’ and the second part is like ‘i’. So as ’e-kara’ has a similarity with them both, there is a chance of a-karas coming in; so in order to provide for the coming in of i-kāra and u-kära and ruling out a-kara the rule was enunciated. But this rule is actually regarded as unnecessary by the Värttika which holds that e-kära and o-kära being palatal and labial, may well be replaced by palatal i-kära and the labial u-kāra. Again, in the view that e-kara contains the half-matra of a-kara and one and a half mātrās of i-kāra, i-kāra has more justification of coming in than a-kära. There is, of course, a view that e-kara contains half a-kāra and half i-kāra. Another point is that if ’e’ and ‘o’ were associated with ’ta’, then the e-kāra and the o-kara could not have a short variety and it might be regarded as impertinant to prescribe a short variety. But the reply is that in such a case the pluta is also to be described. The various rules that effect the euphonic changes in such words as ‘gotrataḥ’ and ‘pratyannaitikāyana’ would not take place. There are other objections also which have been detailed above in the translation of the Bhāṣya. Again if the ’en’ is not regarded as being associated with ’ta’ then in such euphonic combinations as ‘gangodaka’ or ‘ramesa’ wherein ganga+udaka we have ‘a’ and u’ making three mātrās or three units of time, and in ‘rameśa’, we have ‘ramā + iśa, i.e. ‘a and ‘i’ making four matras or four units of time and therefore the euphonic substitutes in their places the ‘o’ and ’e’ should have three and four matras. The reply that in the rule “akaḥ savarne dirghah”, the injunction is sufficient to restrict the ’e- kara’ and ‘o-kara’ (which is enjoined as dirgha) to two matras, for then in the case of pašu + am = paśum, the u-kara, which is a substitute for ‘u’ and ‘a’ (being ekādeśa, like ’e’ in ramesa) ought to be dirgha and of two matras. The justification of the letter, however, is to be found in the maxim that as is the preceding vowel so should be the later one. This maxim is drawn from the fact that in Panini (6.1.84) we have the rule “ekaḥ purva-parayoḥ” (there should be one substitute for that which precedes and follows). This is an adhikara sutra and is extended up to Panini 6.1.102 “prathamayoḥ pūrva- savarṇaḥ”. (In a vowel, in the first and the second case the rule “akaḥ savarne dirghaḥ” 104 Mahābhāṣya of Patanjali that one substitute which comes in place of the preceding and the succeeding, becomes long-this is how it has been instructed (as an interpretation of the rule “akaḥ savarne dirghah”).71 (The Bhāṣya runs on as follows raising an objection and refuting it). Then here also there is the likelihood of a long substitute-paśu + am= paśum viddham, pacanti. This is no criticism. The word ‘paśum’ (i.e. paśu + am) is solved by the rule ‘amyekah’ - (The actual rule is “pūrvaḥ”- Panini 6.1.107, which means that when ‘am’ follows the ‘a’ of ‘am’ together with the preceding vowel is replaced by the preceding vowel). When by the preceding rule “ekaḥ pūrva-parayoh” (6.1.84) the injunction that there should be one ‘adeśa’ in place of the preceding and the succeeding, the fact that the word ‘preceding’ (pūrva) which was running contextually from the former to the latter sutra, was yet again uttered in the present sutra as ‘pūrva’ in the ‘ami pūrvaḥ’, is explained as leading to the suggestion that the vowel that is enjoined to come in place of the preceding and the succeeding, must be of the type of the preceding. Then again in the case of the word ‘viddham’ we have the contextual reference to ‘purva’ in the rule “samprasaranac ca” (6.1.108) on pūrva which is separately taken again, or rather Panini seems to instruct us that though ‘purva’ is coming only contextually here and not taken again separately, yet there will be no lengthening of the vowel obtained by samprasarana (when in place of ya, va, ra, la, ‘i’, ‘u’, ‘T’, T is enjoined, such injunction is called samprasarana). The case of the word ‘pacanti’ is sovled by the rule “ato gune” (6.1.97).72 would apply). The repetition of the word ‘purva’ suggests that the substitute would be of the preceding vowel. We have also the rule Panini, 6.1.107 “ami pūrvaḥ” (when ‘am’ follows, then there should be one substitute of the nature of the preceding in the place of the preceding and the following). Here also the word ‘purva’ is taken. The next rule is “samprasaraṇāc ca” (when a vowel follows after a samprasaraṇa, there would be one substitute for the two which would be of the nature of the preceding. These and other rules will solve the difficulty and obviate the necessity of the substitutes being dirgha in such cases as ‘paśum’, ‘viddham’, ‘pacanti’. It is therefore permissible to have the rule “akaḥ savarne dirghah” as interpreted by ‘yoga-vibhāga’, as will be found in the translation of the following Bhāṣya.
71. The word ‘yoga-vibhāga’ in the Bhasya means that the sutra is split it into two parts. Thus the rule “akaḥ savarne dirghah” is interpreted firstly by taking “akaḥ savarne”, and then the meaning obtained is associated with the word ‘dirgha’.
72. The rule means that where a guna follows an a-kira which is not at the end of a pada, then there will be one vowel in place of the preceding and the succeeding and this one vowel will be of the type of the succeeding one, because though the injunction Second Ahnika 105 (In order to support the Varttika the Bhāṣya says as follows): We do not say that where there is any fault in doing the operation (i.e. of ’ekādeśa’ as in the rule “vṛddhir eci” we should do it (enjoin the dirgha). But where there is no fault in such an operation, we should do it. Where is there no such fault? (The reply is)- In the case of forming technical names, as in “vṛddhir ad aij dirghaḥ”, “aden guno dirghah”. Then there the term ‘dirgha’ should be pronounced. No, it should not be pronounced. Why, on account of the maxim of proximity in place of vowels of three or four matras, ādeśa should not be of three or four mātrās also ? (The reply is) - The gunas and the vṛddhis are followed by ’ta’. What is the meaning of followed by ’ta’ (taparah)? Does it mean that ‘which is followed by ta’? No, it also means ‘after the ta’, i.e. preceded by ’ta’. (The objection comes) - If it means that also which is after ’ta’ then it should apply only to such rules as “ṛdorap”, but not in the case of ’lavaḥ’, ‘pavaḥ’. (But again the conclusive answer comes) - But it is not ’ta-kara’, but ‘da-kara’. But what is the utility of ’ta-kara’ ? If ’ta-kara’ is for certainty, ‘da-kara’ may as well be that. If ’ta-kara’ is for facility of pronunciation, ‘da-kara’ may as well be so.73 leading to ’ekādeśa’ was secured, yet the word ‘rupa’ was added to ‘para’ as ‘pararupam’ in the rule ’eni pararupam’ (6.1.94) (from which the word ‘para-rupam’ is extended to the rule “ato gune” by contextual continuity) and this was done to indicate that the adeśa that comes in the place of the preceding and succeeding, should be of the nature of the succeeding. In the case of Khaṭvā+rsya and mala+rsya, we have ‘khatvarsya’ and ‘malarśya’. There cannot be any a-kara here, because dirgha has been enjoined. From want of proximity there cannot be any e-kara or o-kāra. But then, what is the harm? Then there ought to be no conjunction of vowels and instead of ‘khatvarsya’, we should have ‘khatvarsya’, i.e. we should have to utter the word as unconjoined in euphonic combination.
73. Bhaṭṭoji in his Sadba-kaustubha explains the general idea of the Värttika and the Bhasya as follows: The question was whether the conjunct syllables should be pronounced with a supplementary ’ta’ or not. Now, if there is no supplementary ’ta’ in the case of such vowel combinations as is found in ‘gangodaka’, we have a combination between ‘a’and ‘u’and in ‘rameśa’, ‘a’ and ‘I’, i.e. in the former we have three mātrās and in the latter four matras. The adeśa that is to take place in their place by the rule “ad gunah” (6.1.87). (When a vowel follows ‘a’, then the two together would be replaced by one guna adeśa, i.e. by ’e’ or ‘o’ or ‘ar’ or ‘al’) will be as said above, ’e’ or ‘o’, or ‘ar’ ‘al’ and there being no further definitive direction, we may expect that in place of a trimātra, the adeśa will be trimatra and in the case of 106 Mahābhāṣya of Patanjali caturmatra the adeśa will be of four matras. The objection that according to the rule “akaḥ savarne dirghah” interpreted according to the yoga-vibhāga principle would enjoin only a long vowel and therefore the adeśa of a guna in the rule “ad guṇaḥ” will be dirgha and therefore the former supposition that a trimatra should be followed by a trimätra and caturmatra by a caturmatra, will not hold, for in the case of pasu + am in place of two matras we have only one matra which tallies with the preceding vowel. We have seen that the supplementary word ‘purva’ in ‘ami pūrvaḥ’ is supposed to indicate (by virtue of the fact that though ‘purva’ was received in the rule ‘ami’ from its contextual continuity with ‘purva’ in “ekah pūrva-parayoḥ”, the additional word ‘pūrva’ in ‘ami purvaḥ’ suggests that the adeśa should be of the type of the preceding vowel) that the adeśa will be of the type of the preceding vowel. In the rule “samprasaraṇāc ca” we have also the contextual continuity to the word ‘purva’ and consequently there is no dirgha here also and we have the form ‘viddha’ from the root vyadh+kta. This, however, can also be obtained from the indicatory suggestion contained in the rule “halah” (6.4.2), which means that a ‘samprasarana” of conclud- ing vowel becomes long after a consonant, as in ‘hutah’. From this the ‘samprasarana’ of ‘vyadh’ being an initial one, is clearly out of the scope of the rule and, therefore, short and thus explaining to us the form ‘viddha’. In connection with rule “ato gune” (6.1.97) also we have the contextual continuity with the ‘para-rupam’ from the previous rule “eni para-rupam” (6.1.94). In this latter rule simple ‘para’ might have served the purpose; the additional redundant ‘rupam’ is supposed to suggest that the vowel that replaces the preceding and the succeeding vowels, should be in the nature of the succeeding one, and this explains why in paca + anti in place of two matras, we have only an ‘a’ of one matrā giving us the form ‘pacanti’. So there is no fault in enjoining dirgha by yoga-vibhāga in the rule “akhaḥ savarne dirghah”. But there is a further objection-how on such a supposition we may have kṛṣṇa + rddhi = kṛṣṇarddhi? For, here the ’ekādeśa ‘being dirgha, we cannot have ‘a’, and for absence of proximity we cannot have ‘a’ or ‘o’. So it may be said that to interpret the rule “akah savarne dirghah” as enjoining a dirgha ekādeśa will be faulty. If this objection holds, then we fall back upon the old contingency that the trimätra should be succeeded by the trimatra, and the caturmatra by the caturmatra. But to this a further objection can be made that a caturmatra does not exist. To this the reply is that it exists traditionally among the Sama songesters and the Bhasya itself admits it in the rule “plutā vai ca idutau” (8.2.106). With reference to this whole discussion Bhattoji says that the appendage of ’ta’ is admitted only with reference to guna samjñā and vṛddhi samjñā. But if it is admitted that the appending of ’ta’ applies only to the samjña of guna and vṛddhi, how can ’en’ and ‘aic’ refer to their own units of time? It may be pointed out that in the rule “sahi- vahor od avarnasya” (6.3.112), when ‘dha’ and ‘ra’ are elided there, the ‘a’ of ‘saha’ and ‘vaha’ become ‘o’. The maxim, that ’tapara’ means that which is followed by ’ta’ and preceded by ’ta’, makes it possible to enjoin that a-varna here, means only the short variety of ‘a’, and not the long or the pluta variety. Here, therefore, both ‘o’ and ‘a’ refer to their own units of time or matra, so ‘aiu’ and ‘aic’ that may come after ’ta’ may be supposed to indicate the long variety. The rest, however, should naturally include other varieties as well, as they cannot be regarded as coming after ’ta’ as these Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Second Ahnika 107 idam vicāryate-ya ete varneṣu varṇaikadeśā varṇāntara-samānākṛtaya eteṣām avayava-grahanena grahaṇam syād vā na veti. kutaḥ punar iyam vicāraṇā? iha hi samudāyā apy upadisyante, avayavā api. abhyantaraś ca samudaye’vayavaḥ. tad yathā vṛkṣaḥ pracalan sahâvayavaiḥ pracalati. tatra samudayasthasyâvayavasyâvayava-graha- nena grahaṇam syad vā na veti jāyate vicāraṇā. kaś câtra viseṣaḥ? varṇaikadeśā varṇa-grahaṇena cet sandhyakṣare samānākṣara-vidhi- pratiṣedhaḥ. varṇaikadeśā varṇa-grahanena cet sandhy- akṣare samānākṣarāśrayo vidhiḥ prāpnoti. sa pratiṣedhyah. agne, indra, vayo,.udakam. “akaḥ savarne dirghah” iti dirghatvam prāpnoti. dirghe hrasva-vidhi-pratiṣedhaḥ. dirghe hrasyāśrayo vidhiḥ prāpnoti, sa are not so pronounced. We have thus to admit that only in the case of guna samjñā and viddhi samjñā ’en’ and ‘aic’ do not include its homologous vowels or savarnas. The appendage of ’ta’ there is only for clarity. The view is a correct one, for, with reference to ‘a’ in the rule “vrddhir ad aic”, the a form not being included in the Sivasūtra “a, i, un”, it cannot be valid with its other homologous sounds, and, therefore, since ‘a’ cannot be extended to any of its savarnas the appendage of ’ta’ is unnecessary. In that case the association of ’ta’ will only be for the sake of ‘aic’. But here also the ’en’ and ‘aic’ according to our discussion cannot be confused with any of its savarnas and, therefore, the association of ’ta’ is unnecessary here also. To the point that in the rule ‘rdor ap’ (3.3.57) the appendage of ’ta’ would restrict the scope of the rule only to verbs ending in short ‘a’, and the verbs ending in long ‘u’ would not be within its purview, elicits the answer that the terminal be here is not ’ta’, but ‘da’. 108 Mahābhāṣya of Patanjali pratiṣedhyaḥ. ‘ālūya’, ‘pralūya’. “hrasvasya piti kṛti tuk” (Pa. 6.1.71) iti tuk präpnoti. naiṣa doṣaḥ, ācāryapravṛttir jñāpayati - ’na dirghe hrasväśrayo vidhir bhavati’ iti. yad ayam dirghac che tukam śāsti. naitad asti jñāpakam, asti hy anyad etasya vacane prayojanam. ‘kim”? “padāntad va” (Pa. 6.1.76) iti vibhāṣām vakṣyāmīti. yat tarhi yoga-vibhāgam karoti. itaratha hi “dirghāt padāntad vā” ity eva brūyāt. iha tarhi— ‘khaṭvābhiḥ’, ‘mālābhiḥ”, “ato bhisa ais” (Pa. 7.1.9) ity ais-bhavaḥ prāpnoti. ta-para-karaṇa-samarthyān na bhaviṣyati. iha tarhi - ‘yātā’, ‘vātā,” “ato lopaḥ” ārdha-dhatuke ity akara-lopaḥ präpnoti. nanu ca atrâpi tapara-karaṇa-samarthyād eva na bhaviṣyati. asti hy anyat tapara-karane prayojanam. kim? sarvasya lopo ma bhūd iti. atha kriyamāṇe’pi ta-pare parasya lope kṛte pūrvasya kasman na bhavati? paralopasya sthānivad-bhāvād asiddha- tvāc ca. evam tarhy ācārya-pravittir jñāpayati - ’nākārasthasyâkārasya lopo bhavati’ iti. yad ayam “ato’nupasarge kah” (Pa.3.2.3) iti kakāram anubandham karoti. katham kṛtvā jñāpakam? kit-karane etat prayojanam - kitity ākāralopo yatha syad iti. yady ākāra- sthasyâkāra-lopaḥ syāt, kit-karaṇam an- arthakam syāt; parasyâkārasya lope kṛte dvayor akārayoḥ para-rupe hi siddham rūpam syāt-‘godaḥ, kambaladaḥ’ iti. paśyati tvācāryo- nākārasthasyâkārasya Värttika. Bhāṣya. Värttika. Bhasya. Second Ahnika lopaḥ syad iti, ataḥ kakāram anubandham karoti. naitad asti jñāpakam. uttarârtham etat syat “tunda-śokayoḥ parimṛjâpanudoḥ” (Pā.3.2.5) iti. 66 yat tarhi “gaposṣṭak” (Pa.3..2.8) ity ananyârtham kakāram anubandham karoti. ekavarṇavac ca. 109 ekavarṇavac ca dirgho bhavati iti vaktavyam. kim prayojanam ? vācā taratiti dvyaj- lakṣṇaṣ-than ma bhūd iti. iha ca vāco nimit- tam “tasya nimittam samyogotpātau” (Pa. 5.1.38) ity anuvartamāne “godvyaca” (Pa. 5.1.39) iti dvyaj-lakṣano yan ma bhūd iti. atrâpi go-nau-grahaṇam jñāpakam- ‘dirghād dvyaj-lakṣano vidhir na bhavati’ iti. ayam tu sarveṣām eva parihāraḥ. nâvyapavṛktasya avayave tad vidhir yatha dravyeṣu. nâvyapavṛktaşyâvayavasyâvayavāśrayo vidhir bhavati yatha dravyeṣu. tad yatha dravyeṣu “saptadaśa samidhenyo bhavanti” iti na saptadaśâratnimatram kāṣṭham agnau abhyadhiyate. viṣama upanyāsaḥ, pratyṛcam caiva hi tat- karma codyate. asambhavaś câgnau vedyām ca. yathā tarhi “saptadaśa prādeśamātrir āśvatthiḥ samidho ‘bhyadadhita” iti na saptadaśa-prādeśa-matram kāṣṭham agnau abhyadhiyate.110 Värttika. Bhasya. Värttika. Bhāṣya. Mahābhāṣya of Patanjali atrâpi pratipraṇavam caitat karma codyate. tulyaś câsambhavo’gnau vedyam ca. yatha tarhi “tailam na vikretavyam”, “māmsam na vikretavyam” iti vyapavṛktam ca na vikriyate, avyapavṛktam - gāvaḥ sarṣapāś ca vikriyante, tatha “loma-nakham spṛṣṭvā saucam kartavyam” iti vyapavṛktam spṛṣṭvā niyogatahḥ kartavyam, avyapavṛkte kāmacāraḥ. yatra tarhi vyapavargo’sti. kva ca vya- pavargo’ sti? sandhyakṣareşu. sandhyakṣareşu vivṛtatvāt. (sandhyakṣareşu) yad atrâvarṇam vivṛtataram tadanyasmād avarṇāt. ye api ivarṇovarne vivṛtatare te anyābhyām ivarṇovarṇābhyām. athavā punar na gṛhyante. agrahaṇam cen nuḍvidhi-ladeśa- vināmeṣu ṛkāragrahaṇam. agrahanam cen nuḍvidhi-lādeśa-vināmeşu rkara-grahaṇam kartavyam. “tasmānnuḍ dvihalah” (Pā.7.4.71) rkāre ca iti vaktavyam. ihâpi yathā syāt-‘ānṛdhatuḥ’, ‘anṛdhuḥ’ iti. yasya punar grhyante “dvihalaḥ’ ity eva tasya siddham. yasyâpi na grhyante tasyâpy eșa na doṣaḥ. dvihal-grahanam na kariṣyate tasman nuḍ bhavati ity eva. yadi na kriyate, ‘aṭatuḥ’, ‘atuḥ’ ity atrâpi prāpnoti. aśnoti-grahanam niyamārtham bha- visyati-‘aśnoter evâvarṇopadhasya nânyasyâvarṇopadhasya’ iti. [nuḍvidhiḥ]. Second Ahnika lādese ca ṛkara-grahanam kartavyam. “krpo ro laḥ” (Pa.8.2.18) ṛkārasya ça iti vaktavyam. ihâpi yathā syāt - ‘klptaḥ’, ‘klptavān’ iti, yasya punar grhyante ‘ra’ ity eva tasya siddham. 111 yasyâpi na grhyante tasyâpy eșa na doṣaḥ. ṛkāro’py atra nirdiśyate. katham? avibhak- tiko nirdeśaḥ- krpa, uḥ, raḥ, laḥ = “krpo ro lah” iti. athavā ubhayataḥ sphota-mātram nirdiśyate-ra-śruter la-śrutir bhavati iti. [lādeśaḥ] vināme r-kara-grahaṇam kartavyam. “raṣābhyām no ṇaḥ samānapade” (Pa.8.4.1), “ṛkārāc ca” iti vaktavyam. ihāpi yathā syāt- ‘mātṛṇām’, ‘pitṛṇām’ iti. yasya punar grhy- ante “raṣābhyām” ity eva tasya siddham. na sidhyati. yat tad rephāt param bhaktes tena vyavahitatvān na prāpnoti. mā bhūd evam. aḍvyavaya ity eva siddham. na sidhyati. varṇaikadeśaḥ ke varṇa-gra- hanena grhyante, ye vyapavṛktā api varṇā bhavanti. yac câpi rephat param bhakteḥ, na tat kvacid api vyapavṛktam dṛśyate. evam tarhi yoga-vibhāgaḥ kariṣyate, “raṣābhyām no ṇaḥ samānapade”. tataḥ “vyavaye”, vyavaye ca raṣābhyām no no bhavati iti. tataḥ-“at-ku-pvannumbhiḥ’ iti. idam idānim kirmartham ? niyamārtham- ’etair evākṣarasamāmnāyikair vyavaye, nânyair’ iti. yasyâpi na grhyante tasyâpy eșa na doṣaḥ. ācāryapravṛttir jñāpayati -‘bhavaty ṛkārān no natvam’ iti. yad ayam kṣubhnādiṣu nṛna- mana-sabdam paṭhati. 112 Vārttika. Bhāṣya. Vārttika. Bhāṣya. Mahābhāṣya of Patanjali naitad asti jñāpakam. vṛddhy artham etat syat ’nārnamanih’. yat tarhi trpnoti sabdam pathati. yaccâpi nṛnamana-sabdam pathati. nanu coktam-‘vṛddhyartham etat syāt.’ bahirangā vṛddhir antarangam ṇatvam. asiddham bahirangam antarange. athava upariṣṭād yoga-vibhāgaḥ kariṣyate- ṛtaḥ no no bhavati. tataḥ “chandasy avagrahať”. ‘ṛta’ ity eva. plutāvaica idutau. [plutāvaica idutau] etac ca vaktavyam. yasya punar grhyante, ‘guroster’ ity eva plutyā tasya siddham. yasyâpi na grhyante tasyâpy eșa na doṣaḥ. kriyata etan nyasa eva. tulyarūpe samyoge dvivyañjanavidhiḥ. tulyarūpe samyoge dvivyañjanaśrayo vidhir na sidhyati - ‘ku2kkuṭaḥ’, ‘piзppali’,

74. In a-kāra there are parts like ‘a’; in r-kara and l-kāra there are parts like ‘ra’ and ’la’; in e-kara and o-kara and in ai-kāra and au-kāra there are parts like ‘a’, ’e’, or ‘o’. Now the question is whether vowel operations may be extended to the vowels which are constitutive parts of the compound words. The collective or samudaya consists of r, 1, e, ai, o, au and the parts a, i, u, ra and la. Now, when any rule refers to the collective vowels, though the constitutive members exist there, they are not within the operation of those rules, though they exist there. When again rules refer to the parts of the simple vowels, the collective ones have no status there. Thus consonants like ha, ya and va are associated with a-kara in normal utterence on account of the difficulty that is experienced in pronouncing them without the terminal vowel. But in ‘dadhyatra’ they enjoin ‘ya’ in place of ‘i’ as to be without the terminal a-kāra. Now an adeśa ‘ay’ etc. is enjoined in place of collective vowels. In such a case the simple vowels forming their part are invariably ruled out of operation. But when the case is such that there is the opportunity for the effectuation of vowel-change in the simple part of vowels as well as in the synthetic vowels, it is the former that finds preference. Thus, in the phrase ‘agna indra’ existing before the combination as ‘agne indra’, we have the euphonic combination ‘agnayindra’ and further the other form ‘agna indra’. No further euphonic combination takes place. The ‘a’ of ‘agna’ and the ‘i’ of ‘indra’ do not further enter into any euphonic combination. Here there is an opportunity for the operation both of the simple vowel and of the collective. Thus, in ‘agne indra’, the ’e’ is compound of ‘a’ and ‘i’; and since ‘i’ of ‘indra’ follows there is a case for the lengthening of the vowel ‘i’, as also the adeśa of ‘ay’ in place of ’e’. (This is ‘samudaya karya’ and the former was ‘avayava karya’). Now it is suggested that preference should be given to the ‘avayava karya’, i.e., operation of change in the constitutive vowels and again since we can know the whole only through the parts (i.e., of the change of the ‘i’ in ’e’ together with the following ‘i’ of indra) is more proximate than the vowel change of ’e’ in ‘ay’ which is a samudaya karya and of an external operation (bahiranga). But, as a matter of fact the change of ’e’ into ‘ay’ takes place and we have ‘agna ayahi’. The reply to such an objection is that the constitutive vowels have disappeared in the conjunct vowel and have no independence in making 114 Mahābhāṣya of Patanjali (The Bhāṣya in introducing the Varttika says as follows:) What is the difference? (Here the Varttika intervenes) - If the constitutive parts of a syllable be included within the syllable, then in a conjunct syllable a rule should be made for the operation of changes of similar syllables (a, a, i, i, u, u, r, I, 1, 1). (The Bhasya interprets it as follows:) If in enjoining an operation about syllables the consitutive vowels of a conjunct syllable be taken into account, then the operation should apply to all the similar vowels (a, a, i, i, u, u, r, r, 1, 1). This should be prohibited. ‘Agni’, ‘indra’, ‘vayo’, ‘udakam’ - (in such places the ‘i’ in ‘agni’ might combine with ‘i’ in ‘indra’, or the ‘u’ of ‘vayo’ might combine with ‘u’ in ‘udakam’ and become lengthened by the rule “akaḥ savarne dirghah”-the vowels, a, i, u, r, 1 when followed by homologous vowels, will be transformed into one long vowel; thus, there is a case here for the lengthening operation. (Another Värttika intervenes) - In the case of long vowels there should be prohibitory rules against shortening them. (The Bhāṣya runs as follows) - In the case of long operations of making short become applicable and this should be prohibited - ‘ālūya’, ‘proluya’. There is also the case here for the coming in of the augment ’tuk’ by the rule “hrasvasya piti krti tuk” when a kṛt suffix with a redundant ‘pa’ follows, the augment ’tuk’ follows after a short-vowel.75 (The Bhāṣya in refuting this objection says) - This is no fault. Panini’s rule “dirghat” (6.1.75, i.e. when a ‘cha’ follows a long vowel, the augment ’tuk’ comes in between the long vowel and the ‘cha’) provides for an operation of ’tuk’ after a long vowel and this implies that there is no further operation to be prescribed for the short constitutive vowel. (Now an objection is raised by the Bhasya against the above answer)— There is no scope for infering such an indicatory suggestion. Another purpose is served by this rule, for in the next rule “padantad vā” (6.1.76) it is enjoined that if the long vowel is at the end of a pada, the appearance of the augment ’tuk’ will only be optional. (To this the Bhāṣya further says)- But still there may be another purpose. It may be for the purpose of splitting up one whole into two a free combination (tirohitatvad avayavānām samudayakaryapäratantryāt). The e- kāra though constituted of a-kāra anbd i-kāra, has become a new vowel as it were. This is what has been referred to in the Bhasya by the phrase, ‘abhyantaraś ca samudayaḥ avayavaḥ’.
75. ‘ālūya’ = a + lū+ yap. Now if we consider the consitution of ‘u’ in the root ’lu’, we find that it is composed of two short ‘u’-s. Now, if the vowel operation is allowed to extend to the constitutive vowels, it would apply to the short vowel ‘u’ and as such it would come under the operation of the rule. “hrasvasya piti krti tuk”, when a ‘pit kṛt’ like ‘yap’ follows. But to prevent this there ought to be a prohibitory rule. Second Ahnika 115 by yoga-vibhāga, otherwise the sutra might have been ‘dirghāt padāntād va’.76 According to that principle then, in the instances of ‘khaṭvābhiḥ’ and ‘mālābhiḥ’ the ‘bhis’ ought to become ‘ais’ after ‘a-kara’ by the rule “ato bhisa ais” (7.1.9). If ‘mala’ be reducible into two vowels ‘a’ and ‘a’, then word ‘mala’ may be regarded as ending in ‘a’, and thus instead of ‘bhis’ there might be ‘ais’ as in ’naraiḥ’. (The reply is)-‘bhis’ does not become ‘ais’, because the rule “ato bhisa ais” provides only for words ending only in ‘a’, because the ‘a’ in the sutra “ato” etc. has a terminal ’ta’. (Another objection is raised in the Bhāṣya as follows): How then ‘yātā’, ‘vāta’? There is the rule “ato lopaḥ” (6.4.48, the a- kara is elided when an ardhadhätuka suffix follows) and according to this rule the ‘a’ of ‘ya’ might be elided. To this also it may be said that as in the rule ‘a’ has a terminal ’ta’ it only refers to the short ‘a’ and therefore the rule has no shortening application to the root ‘ya’. But there is also another purpose served in the terminal ’ta’ and this is to prevent the addition of the whole vowel ‘a’. (But it may be asked) that even though there is a terminal ’ta’ we may by virtue of it elide the second ‘a’, in ‘a’ in the root ‘ya’, and then why should not there be a further elision of the first ‘a’ (leading thus to the complete elision of ‘a’)? (The reply is)—that the elision of the second ‘a’ has to be regarded as implying that what has been eliding was as if it had not been elided, and moreover, this rule “ato lopaḥ” should be treated as if it had no application according to the rule “asiddhavad atrabhat” (6.4.22)-(which means that all the rules from 6.4.22 to the end of that chapter should be regarded as having no scope with reference to the operations that have been directed with reference that are technically called ‘bha’). Panini’s intention thus shows that the ‘a’ of ‘a’ is not elided and this intention is manifested from the rule “ato’nupasarge kah” (3.2.3)-(the ‘ka’ suffix is to be added to roots ending in a and not preceded by an
76. By the rule “dirghat” we have ’tuk’ after a long vowel followed by ‘cha’ and by the next rule “padantad va” we have the provision of an optional operation in case of a long vowel at the end of a pada and the suggestion here given is that the rule’ dirghat’ has been given not for prohibiting the vowel operation of the constitutive simple vowels but for enjoining the optional application of ’tuk’ in the case of a ‘padanta dirgha’ and the formation of the two rules-“dirghat” and “padāntad vā"has been by way of splitting up of one idea into sutras, and not for prohibiting the vowel operation of constitutive simple vowels. So far, then the result is that vowel changes ought to be applicable to the constitutive simple vowels. Other objections are being enunciated in the Bhāṣya that follows. 116 Mahābhāṣya of Patanjali upasarga), where a ‘ka’ has been added to ‘a’ as a supplementary attachment. How can this express the intention of Panini? Because this purpose is revealed in the attachment of the redundant ‘ka’ that on this account the final a should be elided by the rule “ato lopa iti ca” (6.4.64)- (when an ardhadhātuka vibhakti follows beginning with an initial ‘it’ or a suffix having a redundant attachment of ‘ka’ or ’n’ follows the a of a root ending in a is elided) the a of the roots that take ‘ka’ are elided. If it was the intention of Panini to drop the a-kara constitutive of a, then it would be unnecessary to have the redundant attachment ‘ka’ for by eliding the second ‘a’, and by advising that the first ‘a’ together with the ‘a’ of ‘ka’ will take the form of the latter ‘a’ of ‘ka’ by the principle of ‘para-rupa’, we can have the derived form as in ‘godaḥ’, ‘kambaladaḥ’. Thus Pāņini desires that the ‘a’ in ‘a’ should not be elided and for this reason he had advised the redundant ‘ka’ attachment to ‘a’ in the rule “ato’nupasarge kah”. It is not for this suggestion (the objector says)-that the redundant ‘ka’-attachment was made. This ‘ka’-attachment has been advised here for a later sūtra “tunda-śokayoh parimrjâpanudoḥ” (3.2.5, the roots ‘parimrj’ and ‘apanud’ take ‘ka’ when followed by ’tunda’ and ‘soka’). But in the rule ‘gapoṣthak’ (3.2.8)-(to the roots ‘ga’ and ‘pa’ the suffix ’thak’ is added and there is no upasarga). The redundant ‘ka’ has been added to ‘a’ only for the purpose of dropping the a of the roots ‘ga’ and ‘pa’ and not for any other purpose. (Here the Värttika intervenes)-the long conjunct vowel should be treated as a single one. (The Bhāṣya says)-a long vowel should be advised as being like a single vowel. What is the necessity? (We have) ‘vācā tarati’ (but not vācika) so that there may not be ’than’ which is suffixed to words having two syllables. Here again, in the rule “tasya nimittam samyogotpātau” (5.1.38) (a word takes the suffix ‘yat’ and ’than’ when a contact or an omen is signified as being for the sake of the object denoted by that word) ‘śatasya nimittam dhanapatina samyogaḥ śatyaḥ śatikaḥ, śatasya nimittam utpāto dakṣinākṣi-spandanam’. According to this rule, we may expect to have a ‘yat’-pratyaya to the word ‘vac’ according to the meaning for the sake of words, by the rule “godvyaco’samkhyā-parimāṇāśvāder yat” the ‘yat’ is suffixed to the word ‘go’, words containing two vowels or denoting number and measure and the word ‘aśva’ etc. in the context of the rule, 5.1.38. A dirgha vowel is to be regarded as a single vowel in order that we may not regard the word ‘vac’ as consisting of two vowels (a+a=a), so that they may not have the ‘yat’ suffix according to the rule Second Ahnika 117 5.1.39, just referred to. Again, here also the inclusion of the word go’ in the rule 5.1.39 along with ‘dvyac’ (words containing two vowels) shows that go is not regarded as ‘dvyac’ and it is regarded merely as a simple long vowel to which the rule does not apply. So also the inclusion of the word ’nau’ in the sutra, “nau-dvyacasthan” (4.4.7). (The Bhasya introduces a Värttika and says) That this is a refutation of all objections. (Thus the Varttika says)-the injunction may be with regard to the whole, it would not be applicable to any of its parts which have been fused together with it as one whole and this will be after the analogy of things. (The Bhāṣya interprets it as follows)-No injunction applying to parts should be applicable to such parts as are one with the whole as in the case of things. (This is a line in the Värttika which the Bhasya quotes for explanation) “there are seventeen logs”. This phrase does not mean that a log seventeen cubits long should be put into fire.77 (Hence the Bhāṣya says)-the analogy is contrariwise (for two rea- sons), for one with each Rk-mantra a log has to be put into the fire, two, it is impossible to put it in the fire on the sacrificial altar.7 78
77. ‘Samidheni’ is a word denoting the Ṛk-mantras. The word ‘samidheni’ denotes the Rk-mantra by the Värttika ‘samidhenyo manträḥ’ contained in Panini (4.3.120). It originates from the word ‘samidhena’ from ‘samidh’. From association it means the logs that are put in the sacrificial fire. “Saptadasa samidhenyo bhavanti”-” would thus mean that logs seventeen cubits long, should be put in the fire. The question is whether these should be one log 17 cubit long, or these should be 17 logs each one cubit long. The idea is that though each log a cubit long and taken 17 times, would make a log of 17 cubits; yet, since the injunction speaks of a log 17 cubits long, we cannot break it up into 17 bits. These parts, the cubit bits, being one with the whole and the injunction being with reference to the whole, it cannot apply to the part. So also when an injunction is made with reference to a whole vowel, that injunction cannot apply to the constitutive parts of which it is made. But as a matter of fact not one log 17 cubits long is to be applied to the fire in the sacrificial altar. The analogy thus points the other way, i.e. the injunction with reference to the whole applies to the parts.
78. A sacrificial altar is a three-and-a half cubits square and it is impossible to put a log that is 17 cubits long, for it is prohibited that a log should not extend out of the limit of the altar. Therefore it is impossible to put a 17 cubits long log into the sacrificial altar and for the reasons stated above one cannot put the same log into the fire with 17 Ṛk-mantras; for these two reasons one cannot take the whole log in 17 cubits, but in parts. The point that was wanted to be established was that an injunction referring to the whole, cannot apply to the part with which it forms a whole. Now, therefore, the analogy is contrary to what was intended to be established, for here an injunction referring to the whole, applies to the part. The reply in the Bhāṣya is that 118 Mahābhāṣya of Patanjali

79. The idea is: that there should be 17 bits of aśvattha wood each as big as the span of the hand and not one bit 17 spans long. Here also each of the bits is to be put in the fire with the pranava mantra and it is impossible that one bit 17 spans long would be put into the fire, for in that case one cannot satisfy the injunction that with each bit of log the mantra is to be uttered and uttered 17 times.
80. The idea is that an injunction may apply to parts, but may not apply to the whole. Thus, the cow contains flesh and the mustard contains the oil; but the injunction against selling meat and oil does not apply to the respective wholes to which they belong, i.e. the cow and the mustard. Therefore the injunction does not apply against the selling of cows and mustard. Again, there is the injunction against touching hair and nails, but only to such hair and nails that are separated from human body. There is no injunction against touching hair and nails unseparated from the whole human body. Second Ahnika

81. The idea is that the a-kāra, i-kāra and the u-kāra in ‘ai’ and ‘au’ being wider and opener in other ‘a’, ‘i’, or ‘u’, there cannot be any homologousness between the two and therefore there cannot be any combination between constitutive ‘a’, ‘i’ or ‘u’ in the conjunct syllables ‘ai’ and ‘au’ and other varieties of ‘a’, ‘i’ and ‘u’.
82. It is only a mistake to suppose that e, ai, o and au are made up of other simple vowels, for they can be regarded as homologous in themselves.120 Mahābhāṣya of Patanjali a remodelling of the sutra we can do without including the constitutive vowels within the conjunct. Again in enjoining a la, ris to be specified- thus as a supplement to the rule “krpo ro lah” (8.2.18- the sound of the root kṛp should be changed into la) we should have to say ‘r-karasya ca’ (r-kara also should be taken in) so that we may get the forms ‘klptaḥ’, ‘kiptavān’. For those who would include (simple vowels) the ‘r’ in the rule will be sufficient. Those who would not include (the simple vowels) may also pass the rule as being without a fault, for here r is also included; how? The words in the sutra will then have to be regarded as being without the declension suffixes such as krpa uh rah laḥ= krpo ro laḥ, or in both cases only the joint whole sound is to be taken; the ra sound becomes the la sound.83 Again in the change of cerebral na the r-kara has to be mentioned separately (as a constitutive factor of ‘ra’). (Thus the Bhāṣya says) in the change of cerebral na the r has to be specially mentioned. The rule for enjoining the change of dental into cerebral na is “raṣābhyām no ṇaḥ samanapade’ (8.4.1-if the dental na is changed into cerebral na after ra and șa) and with it one has to add ‘r-kārāc ca’ also after r, so that we have cerebral na in ‘mātṛṇām’ and pitṛṇām’. Those who include r within ra, may allow the rule ‘raṣabhyam’ without any supplementary addition. (But an opponent urges) - that if by ‘ra’ of ‘raṣabhyam’, the mere ṛ- sound is meant, then there is an intervention of ‘a’ between this ṛ-sound and the dental na that follows and consequently one may expect that the dental na will not be changed. (To this the Bhāṣya replies) - it cannot be so, becasue it has been enjoined by another rule, “at-kupvan-num-vyavaye’pi” (8.4.2- the change into cerebral na will be affected even if there is an intervention of at, ku or the ka-group, pu or the pa-group, an and num) that the change into cerebral will take place even if there is an intervention by ‘at’. (Now, the Bhasya raises the view of those who would not include the simple vowels as effective constituents of the conjunct vowels and says) - it would not do what syllables being parts of vowels may be regarded as constitutives of vowels, only those that when separated will also remain as whole vowels. To say that in ‘ra’ there is a division between the r-sound (repha) and a is meaningless, for there is no vowel of the nature of a repha or r which is used separately.
83. The idea is explained in the Uddyota in detail. There is one class of sound as implying both ra and la. Second Ahnika 121 (Again the Bhāṣya raises the points of those who are in favour of including the constitutive vowels and say) — In that case it is possible to make a split. (The first part standing as “raṣabhyam no ṇaḥ samānapade”

84. There is the rule “kṣubhnādiṣu ca” (8.4.39-the na of the words belonging to the ‘kṣubhnadi’ list shall not be changed into cerebral) and in this list there is a word ’nṛnamana’ (the name of a man); this suggests that it is expected that the second na in nṛnamana’ should have been changed into cerebral since it comes after r, but it would not be so changed because it is included in the ‘kṣubhnadi’ list. But this is no indicatory suggestion (jñāpaka) as has been suggested, for it may as well be said to be a provision for the reviving cerebral na in the case of vṛddhi, we may have a dental na in ’narnamani’ from ’nṛnamana’; but the word ’trpnoti’ is also read in the ‘kṣubhnadi’ list (and this proves that it was the intention of Panini that na after ṛshould be cerebral and this ris contained in the ra of ‘raṣābhyam no nah’). In this connection we may think that the reading of the word ’nṛnamana’ also indicates the same interpretation. But it has been suggested that the word ’nṛnamana’ has been read in the ‘kṣubhnadi’ list for providing a cerebral na in the case of vṛddhi as in the word ’narnamana’.
85. There is a maxim “asiddham bahirangam antarange”; the operating cases directly constitutive to the desired operation are called ‘antaranga’. The cases or conditions not having a direct bearing to the constitutive operation is called ‘bahiranga. Thus there is a rule “kharavasanayor visarjaniyah” (8.3.15-i.e. the r-terminal to a 122 Mahābhāṣya of Patanjali (Here the Varttika intervenes)-(in the view that simple vowels are not constitutive of conjunct ones, the Varttika says as follows) -When one is regarded as pluta or long-drawn intonation the i and the u within it are to be regarded as pluta. (The Bhāṣya interprets it as follows)- The above rule has to be separately enjoined according to those who would not put simple vowels as operative parts in the conjunct ones. For those who hold the reverse view, it is enough to have the sutra formed as ‘gurosteḥ’-in the case of long vowel the pluta intonation should be ascribed to the ’ti’ (i.e the part of a word beginning with the last vowel is called ’ti’-aco’ntyādi tt 1.1.64). 86 pada is changed into aspirate (visarga) when followed by tenuis consonants and tenuis aspirates and sibilants and stoppage. Thus the rafter vṛkṣa becomes changed into an aspirate vṛkṣaḥ, when the cha of chadayati follows and we have vṛksas chadayati. But we have nrkutyah bhavaḥ= närkutyah, here the r of nrkuta has been changed into nar by the application of taddhita suffix to the whole word nṛkuta. It is on account of this taddhita suffix that nṛhas become när. The ra-terminal of nar from nṛhas been caused by virtue of the taddhita suffix which is not constitutive to the word ’nrkuta’; it is true that here we have a terminal rain när followed by a tenuis ka and might have expected here that the r of the nar changed into an aspirate as in ‘vrkṣas chadayati’; but this change of nṛ into nãr being caused by conditions and causes belonging to the suffix is external to the constitution of the word ’nrkuta’. In the formation, therefore, of the changes it would have been due to the internal constitution of the word nrkuta-the ka’s following the r, the operation caused by the taddhita suffix, viz., that of changing nṛ into när, should be regarded as not having happened as it is an external influence not constitutive to the word. Now, it is suggested in the Bhasya that vṛddhi in ’narnamani’ is external to the cerebralisation of na in namani, and for this reason the är of narnamani should be regarded as not having happened, and consequently the inclusion of nṛnamana in the kṣubhnādi list cannot be for the prohibition of the cerebralisation of na in narnamani as that prohibition is already effected in accordance with the maxim, “asiddham bahirangam antarange”. The other alternative suggestion is that the rule “chandasy rd avagrahat” may be split up into two parts as ‘chandasy avagrahat’ and ‘rtah’ so that it may be definitely specified that the na after r should be cerebralised, in the case ris not included within ra of ‘raṣābhyam no nah’ and the provision for the cerebralisation of na is made by a specific enjoining that the na after I should be cerebralised.
86. The idea is that there is a rule-“plutävaica idutau” (8.2.106). The rule means that when in calling or singing, the pluta accent is enjoined with regard to ‘ai’ or ‘au’. The accent is to be given to their constitutive i or u. This rule would be unnecessary for those who would admit the operative function of the constitutive vowels; for then it would be enough to say that in the case of long syllable about which the pluta accent has to be made, the accent is to be ascribed to the portion technically called ’ti’, i.e. on the portion commencing with the last vowel and the rule should be ‘gurosteh’. Second Ahnika 123 (But the Bhāṣya here says)-even in the latter case there is no harm in having this sūtra and the sutra is therefore put. (Kaiyata here says that even in the view in which the operative function of the constitutive syllables is admitted the sutra “plutāvaica idutau” has to be made, for the constitutive parts being wider and opener than the normal variety they would not be homologous with the normal ones and therefore they could not be styled as normal i or u). (Here again the Värttika intervenes) - when the connection is of the same type, i.e. when consonants combine in the same manner as the vowels, then according to the negative view the injunction with regard to double consonants would not apply. (The Bhāṣya explains it as follows:)-when the vowels are of the same form as parts they may be regarded as forming according to one view a simple homologous vowel, and according to the other view, a compound vowel which on account of quickness of articulation appears as one. (The Bhāṣya says) - the compounding being of the same form, the rules of pluta accents regarding double consonants should not apply, as it is formed to do in the pluta accents in words involving such double consonants, such as ‘ku2 kkuṭaḥ’, ‘piзppali, ‘pi зttam’. According to the positive view there are two ‘pa’s and two ’ta’s.87 (The above objection raised in the Varttika is repudiated in the Bhāṣya as follows:) - Those who do not hold the positive view, have also to admit the double consonants as double and two ‘kas’, two ‘pa’s, two ’ta’s in the above example. How so?- There is evidently some difference of time; but there is a further objection that consonants have no difference in time. When such differences have not been instructed, how can we know that, and if they do not exist, how can we understand that? (A further objection has been raised in favour of the positive view in the Bhāṣya)-though it may be said here that in the case of double consonants such difference in timing cannot be known; yet, just as in the case of pratyāhāra ‘an’, each syllable includes within it its homologous ones, so here also it may be expected that the consonants also have some timing in articulation which determines their homologous character.
87. The principle contained in the problem of regarding conjunct syllables as made up of operative constitutives or not, would apply also to the case of articulation of double consonants, and in their case also unless the double consonants are regarded as made up of two operative consonants, they (the double consonants) cannot be regarded as double and therefore the rule regarding their pluta accent would not apply. 124 Mahabhāṣya of Patanjali Otherwise how can we have such examples as ‘sayyanta’, ‘savvatsaran’ ‘yallokam’, ’tallokam’. (In the above cases ya, ra, la, va are supposed to have nasal varieties which argues for the supposition that the consonants also have homolo- gous varieties. In the rule “halo’nantaraḥ samyogah” (1.1.7) it is said that consonants unintervened by vowels are to be regarded as compound consonants and this shows that it is admitted that the two consonants may join together to form compound consonants). Where such conditions prevail, the consonants would intrude in them their homologous variety. Here also then there is timing in articulation of consonants, as in the case of ‘sayyanta’ etc. But it has been said that the consonants have no timing in articulation. Unless so advised, how can it be known, and if it does not exist, how can we understand it? Panini. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. ha ya va rat 5. sarve varṇāḥ sakṛd upadiṣṭaḥ, ayam hakāro dvir upadiśyate-pūrvaś caiva paraś ca. yadi punaḥ pūrva evopadiśyeta, para eva vā. kaś câtra viseṣaḥ? hakārasya paropadeśe’ ad-grahaṇeṣu hagrahanam. hakārasya paropadese ad-grahaneṣu ha- grahaṇam kartavyam-“ato’ți nityam” (Pa. 8.3.3), “śaśchoti’ (Pa. 8.4.63), “dirghād ați samana-pāde” (Pa. 8.3.9), “hakāre ca” iti vaktavyam. ihâpi yatha syat-‘maha hi saḥ’. uttve ca. uttve ca hakara-grahaṇam kartavyam. “ato ror aplutad aplute” (Pa. 6.1.113), “haśi ca” (Pa. 6.1.114), “hakāre ca” iti vaktavyam. ihâpi yathā syāt-‘puruso hasati’, ‘brāhmaṇo hasati’ iti. astu tarhi pūrvopadeśaḥ. Värttika. Bhāṣya. Värttika. Bhāṣya. Second Ahnika 125 pūrvopadese kittvakseḍ-vidhayo jhal- grahaṇāni ca. yadi pūrvopadeśaḥ kittvam vidheyam-‘sni- hitvā’, ‘snehitvā’, ‘sisnihiṣati’, ‘sisnehișati’. “ralo’ vvyupadhad halādeḥ” (Pa. 1.2.26) iti kittvam na prāpnoti. ksa-vidhiḥ-ksaś ca vidheyaḥ- ‘adhukṣat’, ‘alikṣat”.“sala igupadhād aniṭaḥ ksaḥ” (Pa. 3.1.45) iti kso na prāpnoti. iḍ-vidhiḥ. it ca vidheyaḥ- ‘rudihi’. valādi-lakṣaṇa in na prapnoti. jhal-grahaṇāni ca. kim? a-hakārāni syuḥ. tatra ko doṣaḥ? “jhalo jhali’ (Pa.8.2.26) itiha na syāt― ‘adāgdham’, ‘adāgdham’. tasmāt pūrvas caivopadeṣṭavyaḥ paras ca. yadi ca kiñcid anyatrâpy upadese prayojanam asti, tatrâpy upadeśaḥ kartavyaḥ. idam vicāryate-ayam repho ya-kara-va- kārābhyām pūrva evopadiśyeta-“haya- varat” iti, para eva va yathānyāsam iti. kaś câtra viseṣaḥ? rephasya paropadeśe anunāsika-dvirva- cana-parasavarṇa-pratiṣedhaḥ. rephasya paropadeśe anunāsika-dvirvacana- parasavarṇānām pratiṣedho vaktavyaḥ. anunāsikasya ‘svarnayati’, ‘prātar- nayati’. “yaro’nunasike anunāsiko va” (Pa. 8.4.45) ity anunāsikaḥ prāpnoti. dvirvacanasya-‘madra-hradaḥ’, ‘bhadra- hradaḥ’. “yarah” iti dvirvacanam prāpnoti. parasavarṇasya-kundam rathena’, ‘vanam rathena’. “anusvarasya yayi’ (Pā. 8.4.58) iti para-savarṇaḥ prāpnoti. 126 Mahābhāṣya of Patanjali All the varnas have been pronounced in the Śiva-sutras only once; but this ha has been pronounced twice, once first and the other in the end. What would be the difference if it would have been pronounced only at the commencement or only in the end? (Here the Värttika intervenes)- if ha was pronounced only in the end, then for including ha within the pratyāhāra at, a special direction would be necessary. (The Bhāṣya interprets it as follows:)-If ha was pronounced only in the end, then its special direction is necessary that those directions that are applicable in the case of ‘at’ are also applicable in the case of ‘ha’. Thus there is the rule “ato’ti nityam” (2.8.3-i.e. when ‘at’ follows, then the a- kāra preceding ‘ru’ should be made nasal, e.g. maha asi; the ru is effected by the rule “dirghad ați samana-pāde” - the terminal na-kāra of a pada preceded by a long vowel should have ‘ru’ after it when ‘at’ follows, if the two occur in the same line of a verse). Then there is the rule"sas choti” (8.4.63-when at follows then the sa preceded by ‘jhay’ would have ‘cha’ after it). In the case of all these rules one should have to direct that the operations referred to by the above sūtras should take place also when ha follows, so that the derived nasal change may occur in such an example as ‘maha hi sah’. (Here the Värttika intervenes)-For the formation of ‘u’ also. (The Bhāṣya explains it as follows:)-If the first ‘ha’ is not pro- nounced, then for the formation of ‘u’ a separate direction should be given for ‘ha’. There is a rule “ato roraplutad aplute” (6.1.113i.e. after ‘a’ which is not of a long-drawn nature, the ‘ru’ preceding an ‘akara’ which is not a long-drawn nature, would be changed into ‘u’-‘vrkso’ tra’). There is another rule “haśi ca” (6.1.114 when ‘has’ follows, then the ‘ru’ after ‘a’ becomes ‘u’-‘puruso hasati’). If the first ‘ha’ was not in the Siva-sūtra list, then it would not be possible to form the rule, “haśi ca” and a separate rule has to be made - ‘hakāre ca’, that when ha-kāra follows, then also the ‘ru’ is to be changed into ‘u’ so that we may have such forms as ‘puruso hasati’, ‘brāhmaṇo hasati’. Then let the first instruction of ‘ha’ in the Siva-sutra be regarded as justified. (Here the Varttika intervenes pointing out the difficulty for not recognising the instruction of the later ‘ha’ in the Siva-sūtra as justifiable): (Thus the Varttika) - if only the instruction of the first ‘ha’ is admitted, then in the enjoining of ‘kit’, ‘ksa’ and ‘it’ and the pratyāhāra jhal’ would be in trouble. Second Ahnika 127 (The Bhāṣya runs as follows:)-If the ‘ha’ is only instructed in the beginning, then a separate rule has to be made in enjoining ‘kittva’ in such cases as the following: ‘snihitvā’ and ‘snehitvā’, ‘sisnihiṣati’ and ‘sisnehisati’, for there will be no ‘kittva’ by the rule- “ralo- vvyupadhāddhalādeḥ” (1.2.26-roots having ‘u’ or ‘i’ as their penulti- mate and beginning with a ‘hal’ and ending with a ‘ral’ would be optionally regarded as ‘set’ and optionally ‘kit’ when ‘san’ and ‘ktvāc’ follow-dyutitva, and ‘dyotitva’, ‘didyutiṣate’ and ‘didyotiṣate’-there is a rule “kniti ca” (1.1.5) which means that when a suffix has redundant (it), ‘k’, ’n’, then on their account there will be no guna (change of ‘i’, ‘u’ and ‘r’ into ’e’, ‘o’ and ‘ar’) and vṛddhi (change of ‘a’, ‘i’, ‘u’, and ‘r’ into ‘a’, ‘ai’, ‘au’ and ‘ar’). Now ‘it’ is an additional suffix often added as a sup- plementary suffix before other suffixes. ‘set’ means to be associated with ‘it’ and ‘kittva’ means the state of being associated with a ‘kit’ suffix, i.e. to be debarred from the normal change of guna or vṛddhi suffixes which are not be nature ‘kit’, but may be declared by a rule to be ‘kit’ in order to enjoin the prescription that there should be no guna or vṛddhi when such suffixes follow. ‘Hal’ means all the consonants including ‘ha’and ‘ral’ means all consonants with the exception of ‘ya’ and ‘va’. Now, in the case of the root ‘dyut’, we have a root beginning with ‘da’, a ‘hal’, and we have ‘u’ in the upadha and it ends in ’t’ which is a ‘ral’. Therefore, it comes under the operation of the rule just mentioned and there is guna of the penultimate ‘u’ only optionally, becasue according to the rule such roots when ‘san’ and ’tva’ follow, should be regarded as optionally ‘kiť. But if ‘ha’ is not included within ‘ral’, i.e. if the ‘ha’ is not instructed in the ‘Siva-sūtra for the second time, then the roots ‘snih’ when ‘san’ and ‘ktvac’ follow, we cannot have the optional forms of ‘snihitva’ and ‘snehitva i.e. having loved, ‘sisnihiṣati’ and ‘sisnehiṣati’. ‘Ksa’ has also to be separately enjoined for the providing for such uses as ‘adhukṣat’, ‘alikṣat’, for if ‘ha’ is not pronounced a second time in the Śiva-sutra there will be no ‘ksa’ by the rule, “Sala igupadhadaniṭaḥ ksaḥ” (3.1.45-the rule means that roots that do not have ‘it’ as a supplementary suffix and which have ‘i’ ‘u’r ‘I’ (ik) as their penultimate and ending in ś, s, ș and ‘ha’ (sal) get their suffix ‘cli’ in the ’lun’ changed into ‘ksa’ as we had adhukṣat’ from the root ‘duh’ and ‘alikṣat’ from the root ’lih’). (If the second ‘ha’ is not taken into consideration then the pratyāhāra ‘sal’ would not include within it ‘ha’ and therefore the root ‘duh’ or ’lih’ would not be a root ending in ‘sal’ and would not therefore come under the operation of the above rule 3.1.45 and would not have the forms ‘adhukṣat’ and ‘alikṣat’). 128 Mahābhāṣya of Patanjali It also has to be separately enjoined for the formation of such a word as ‘rudihi’, for there would be no ‘it’ here which is enjoined to roots beginning with the consonants included within the pratyahāra ‘ral’. (There is a rule: “ardhadhātukasyeḍvaladeḥ”, 7.2.35- according to which it is enjoined to suffixes beginning with all the consonants with the exception of ‘ya’ of an ardhadhātuka nature. Now, if ‘ha’ is not taken a second time then ‘hi’ would not be a valādi suffix and we cannot have the supplementary suffix ‘it’ for the formation of ‘rudihi’). ‘Ha’ has to be separately enjoined. Why? Because the ‘jhal’ would not include ‘ha’ within it. What is the harm? There would be no scope for the operation of the rule: “Jhalo jhali” for the formation of such words as ‘adāgdhām’ and ‘adagdham’. Therefore both the first ‘ha’ and the second ‘ha’ are justifiable. (There is a rule: “jhalo jhali”’- 8.2.26. The rule means that a s-kāra coming after the consonants included in ‘hal’, i.e. all the consonants excluding ‘ya’, va’, ‘ra’, ’la’, when followed by consonants included in ‘jhal’, shall be elided. If the second ‘ha’ is not taken then the root ‘dah’ would not be regarded as a root ending in ‘jhal’ and as such the ‘sic’ suffix in lun would not be elided, when another ‘jhal’ suffix followed, as in the case of adagdhām and adagdham, and should have to make another rule, “hacca”. (It should be noted in this connection that the enunciation of vowels and consonants in the Siva-sutras is not for giving a list of alphabets but for providing for such future use of them as may be regarded as necessary or economical in the interest of the sutras of Panini that follow). (The Bhāṣya continuing says:) If it was necessary to articulate any syllable at any other place, the same may be done (because the articulation of syllables in the Siva-sutras is for the advantage of making the sutras as has been just stated). (Then another problem arises about placing of the ‘ra’ and the Bhāṣya on the subject introducing a Varttika follows:)- This is being discussed. Should this ‘ra’ in the ‘ha-ya-va-rat’ be placed before ‘ya’, ‘va’ or after them as it stands. What is the difference? (Here the Värttika intervenes)-if ‘ra’ should stand after ‘ya’ and ‘va’ then there will be no nasal nor duplication, nor the amalgamation of a vowel with a following homologous one. (Here the Bhāṣya follows:)-if ‘ra’ is allowed to stand after ‘ya’ and ‘va’, as in the Śiva-sutra “ha-ya-va-raf’, then there would be no nasalising and no transformation into a homologous sound. Thus, in ‘svarnayati’ and Second Ahnika 129 ‘prātarnayati’, there would be a nasal by the rule “yaro’nunāsike’ nunāsiko va” (8.4.45 when consonants included within ‘yar’, come as the last member of a pada and an ‘anunäsika’ follows, then those consonants will be nasalised optionally. Now, if ‘ra’ is regarded as coming after ‘ya’, ‘va’ as it stands in the Siva-sutra “ha-ya-va-raf’ then it would include also ‘ra’ for the consonants that come under ‘yar’ are really all the consonants excepting ‘ha’ and therefore ‘ra’ being included within it, there ought to be a nasalising of it into cerebral na, for the cerebral na is the appropriate nasal of the cerebral ‘ra’). Then also again by the rule, “aco rahābhyam dve” (8.4.46-the con- sonants that come under ‘yar’ preceded by ‘ra’ and ‘ha’ which are preceded by a vowel is doubled) which refers to the consonants that come under ‘yar’, we shall expect a doubling of ‘ra’ after ‘ha’ in such words as “madra-hrada’ and ‘bhadra-hrada’. Again by the rule, “anusvarasya yayi parasavarṇaḥ” (8.4.58- i.e. when the consonants included in the ‘yay’ i.e. all the consonants excepting the sibilants and ‘ha’ follow, then the anusvära becomes changed into homologous consonants; e.g. ‘śańkita’ we should expect the transformation of the anusvara in ‘kundam’ into the homologous syllable n when followed by ‘rathena’ or in the case of ‘vanam rathena’. Bhāṣya. Värttika. Bhāṣya. astu tarhi pūrvopadeśaḥ. pūrvopadese kittva-pratiṣedho vyalopa- vacanam ca. yadi pūrvopadeśaḥ kittvam pratiṣedhyam. ‘devitvā’, ‘didevișati’-“ralo’vyupadhād” iti kittvam präpnoti. naisa doṣaḥ, naivam vijñāyate “ralaḥ-vyupadhād” iti. kim tarhi? ‘ralaḥ-av-vyupadhād’ iti. kim idam- ‘av vyupadhād’ iti? avakārāntad vyupadhād av-vyupadhād iti. ‘vyalopa-vacanam ca’. vyoś ca lopo vaktavyaḥ. ‘gaudheraḥ’-‘paceran’, ‘yajeran’. ‘jive radānuḥ’-‘jiradānuḥ’, valiti lopo na prāpnoti.130 Mahābhāṣya of Patanjali naisa doṣaḥ, repho’py atra nirdiśyate- “lopo vyor vali”’ (Pa. 6.1.66) iti. rephe ca vali ceti. atha vā punar astu paropadeśaḥ. nanu coktam-“rephasya paropadese anunāsika- dvirvacana-parasavarna-pratiṣedhaḥ” iti. anunāsika-parasavarṇayos tavat prati- şedho na vaktavyah, rephoṣmaṇām savarṇā na santi. dvirvacane’pi. nemau rahau karyiņau dvirvacanasya. kim tarhi? nimittam imau ra- hau dvirvacanasya. tad yatha-‘brāhmaṇā bhojyantām, Mathara-Kaunḍinyau pari- veviṣātām’ iti. nedanim tau bhuñjāte. Then let ‘ra’ be pronounced immediately before ‘ya’. (Here the Bhāṣya intervening objects to it and says:)-If it is supposed to be pronounced before ‘ya’ then there will be no ‘kittva’, nor the elision of ‘i’ or ‘u’. — 99 (Here the Bhāṣya runs as follows:)-If it is articulated before ‘ya’, then ‘kittva’ has to be prohibited, as in the examples ‘devitvā’, and ‘dideviṣati’; ‘kittva’ is enjoined by the rule: “ralo vyupadhad dhaladeḥ samśca (1.2.26 meaning already explained). This is no criticism, for the meaning of the sutra is not “ralaḥ vyupadhat’. What is it then? - “ralaḥ avvyupadhat’. What is the meaning of the phrase, “avvyupadhat”? The meaning is: that the rule applies to those roots which do not end in ‘va’ and which have ‘i’ and ‘u’ in the penultimate. (The objection was that if ‘ra’ was placed before ‘ya’ and ‘va’ then the root ‘div’ would be a root ending in ‘ral” and would therefore be optionally ‘kit’ by the rule “ralo” etc. (1.2.26) and we should have a form ‘divitva’ and not merely ‘devitva’. But if the rule is read as ‘ralah avyupadhat’ then the roots like ‘div’ which end in ‘va’ should not come under its operation and should not be optionally ‘kit’ and we should therefore have only one form - ‘devitva which shows that the root ‘div’ cannot be a ‘kit’ within the sphere of the application of the rule 1.2.26). Then again there is a rule, “lopo vyor vali” (6.1.66-a dhātu or a substantive ending in ‘va’ or ‘ya’ should have these consonants elided when the consonants included in ‘val’ follows.). Now, if ‘ra’ is placed Second Ahnika 131 before ‘ya’, ‘va’, then ‘val’ should not include ‘ra’ and as such the elision of ‘va’ and ‘ya’ that are expected to take place when ‘ra’ follows, as in ‘gaudherah’, ‘paceran’, ‘yajeran’ would not take place; and for securing the elisions of ‘va’ and ‘ya’ that take place when ‘ra’ follows, a separate rule has to be made that ‘va’ and ‘ya’ should be elided when ‘ra’ follows and the elision of ‘va’of the word ‘jiva’ when the suffix ‘radānuk’ follows, cannot take place by the rule, ’lopo vyorvali’, for ‘ra’ being placed before ‘ya’, ‘va’, ‘val’ would not include ‘ra’ and for this a separate rule has to be made. (It is urged in the Bhasya that)-this criticism can be avoided by splitting up the sutra as “lopo vyorvali” and thus only provision can be made for the elision of ‘va’ and ‘ya’ when ‘val’ follows: and the sutra would then mean:- when ‘ra’ and ‘val’ follows, the preceding consonant should be elided. But in such a case the word ‘rvali’ would have to be interpreted as - when ‘ra’ and the consonants included in ‘val’ follow, ‘va’ and ‘ya’ are elided. Then, let us suppose that ‘ra’ is articulated after ‘ya’, ‘va’. (To this however we are reminded of the former objection). Thus, it has been already objected that if ‘ra’ is articulated after ‘ya’, ‘va’, then the nasalisation, the doubling and enjoining of the homologous in consonance with what follows, would be inapplicable. (The reply is that:) There is no necessity of enjoining nasalisation and transformation of the homologous in consonance with what follows, for there are no homologous sounds to anunāsika (nasals), and the uşma sounds have no homologous consonants. In the case of doubling also the ‘ha’ and ‘ra’ are not productive of the doubling (in the rule, “aco rahābhyam dve” 8.4.46 see above for mean- ing). They are but reasons for the doubling. (We have a popular illustra- tion:) Feed the Brahmins but let Mathara and Kaundinya distribute.87
88. The point of discussion here is as follows: The Siva-sutra gives the pratyāhāra ‘ha-ya-va-rat’. If the pratyāhāra stands in this form, then the pratyāhāra ‘yar’ would include ‘ra’ within it and therefore three difficulties arise-according to the rule “yaro’nunāsike’nunāsiko va’ (8.4.45) and “aco rahābhyam dve” (yarah) -8.4.46, and “anusvarasya yayi parasavarṇaḥ” (8.4.58). According to the first rule the ‘ra’ in ‘svarnayati’ would become optionally nasal. The nasal appropriate to ‘ra’is ’na’, and we should expect ‘svarnayati’ to be optionally ‘svannayati’ which is not a possible form. According to the second rule, in the form ‘madra-hrada’, the ‘ra’ after ‘ha’ should be duplicated and this duplication being in the tripādi, a later rule (8.4.46) is to be regarded as asiddha with reference to the rule-“rori” (8.3.14) and therefore one 132 Mahābhāṣya of Patanjali of the ‘ra’s cannot be elided by the previous rule (8.3.14) and should expect the form-‘madra-hrradaḥ’ which is an impossible form. In the case of the third rule, since ‘yay’ would include within it ‘ra’, in the form-kundam rathena, we should expect the doubling of the ‘r’, i.e. kundam rrathena, and “rori” (8.3.14) being a previous rule in the uipadi, would not apply to “anusvarasya yayi parasavarṇaḥ (8.4.58), a later rule in tripādi. 99 Again, if with the pratyahara we put ‘ra’ before ‘ya’ then we have the form ‘ha- ra-ya-vat’. This saves us from the previous criticisms just mentioned; but other objections creep in. If ‘ra’ is placed before ‘ya’, then the pratyahāra ‘ral’ would include ‘ya’, and difficulties arise with reference to the rule “ralo vyupadhāt” (1.2.26), which means that a ‘vyupadha’ root ending in ‘ral’ and beginning in a consonant will be optionally ‘kit’ and always a set when ‘ktva’ and ‘san’ follow. Thus the root ‘div’ would fulfil the condtions of this rule and we are barred from making a guna of the verb ‘div’, i.e. transforming it into ‘dev’ in the forms ‘devitva’ and ‘didevişati’ (for a ‘kit’cannot have a guna). This objection can be overruled by supposing the above rule (1.2.26) to read as “ralo’ vvyupadhat”, i.e. ‘ralah avvyupadhat’ in which case an exception would be made in the case of not ending in ‘va’, and thus the root ‘div’ would be naturally excluded and difficulties obviated. But another difficulty may creep in, in the case of the rule, “lopo vyorvali” (6.1.66) which means that a terminal ‘va’ and ‘ya’ will be elided when ‘val’ follows. Here if ‘ra’is placed before ‘ya’ and ‘va’ then the pratyāhāra ‘val’ would not include ‘ra’ and we cannot have the word ‘jiradanu’ by suffixing the suffix ‘radanu’ to the word ‘jira’ for, the formation of the word ‘jiradānu’ would require us to drop ‘va’ of ‘jiva’, when the ‘ra’ of ‘radanuk’ follows. Such an elision was only possible to the above rule, “lopo vyorvali” if ‘val’ would include ‘ra’ which it does not according to the above supposition that ‘ra’ should be placed before ‘ya’, ‘va’. But this objection can also be obviated if the rule is read as ’lopo vyorvali’ in which case ‘ra’ will be separately added as a condition along with ‘val’. Weighing the two alternatives, the author of the Bhasya decided in favour of placing ‘ra’after ‘va’, i.e. keeping it as ‘ha-ya-va-rat’ as it exists in the pratyāhāra. The objections urged in the case of possibility of doubling or ‘ra’ by the rule “aco rahābhyam dve” is ignored by the supposition that ‘ra’ ought not to perform two functions at the same time. In the rule “aco rahābhyam” it is suggested that the consonants included in ‘yar’ should be doubled when it follows ‘ra’, and it so happens that ‘ra’ is a consonant included in ‘yar’. It is suggested, therefore, that as in the popular maxim in which all Brahmins are asked to dine and the two Brahmins called Mathara and Kaunḍinya are asked to distribute [serve], we cannot expect that the two Brahmins can both dine and distribute [serve]. So here also, when the doubling is enjoined after ‘ra’, ‘ra’ should perform in this context its special function only and ra which is included only in a general way within the pratyahāra ‘yar’ should not come within the scope of rule and doubled. The supposition that ‘ra’ should undergo an operation of doubling, is obviated by the fact that the ‘ra’ plays the part of an operator (rephasya karyatvam nimittatvena vadhyate). But, as Bhattoji points out, the same difficulty may arise in the rule “iko yanaci” (6.1.77), which means that when any vowel follows, ‘ï’, ‘u’, ‘r’, ‘, will be changed into ‘ya’ ‘va’ ‘ra’ ‘1’. Here ac includes Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Second Ahnika 133 idam vicāryate-ime’yogavāhā na kvacid upadisyante, śruyante ca. teṣām kāryārtha upadeśaḥ kartavyah. ke punar ayoga-vāhāḥ? visarjaniya-jihvāmūliyopadhmanya-anu- svara-nāsikya-yamāḥ. katham punar ayoga- vāhāḥ? yad ayukta vahanti, anupadiṣṭāś ca śruyante, kva punar eṣām upadeśaḥ kartavyaḥ? ayoga-vāhānām atsu, natvam. ayoga-vāhānām atsu upadeśaḥ kartavyaḥ. kim prayojanam? natvam-‘urahkeṇa’, ‘uraxkena’, ‘uraḥpena’, ‘uraxpena’. “aḍvyavaye” (Pa. 8.4.2), iti natvam siddham bhavati. śarşu jaśbhāvaṣatve. śarşu upadeśaḥ kartavyah. kim prayojanam? jaśbhāva-şatve. ayam ubjirupadhmāniyopa- dhaḥ pathyate, tasya jaśtve kṛte ubjitā ub- jitum ity etad rupam yathā syāt. This is being discussed- these ‘ayogavāhas’ that pass without being enjoined had nowhere been instructed, nor do we hear of them. So they ought to be instructed for the operations they are to undergo. Who are they which have not been enjoined? These are: ‘visarjaniya’ ‘jihvāmūlīya’, ‘upadhmaniya’, ‘anusvara’ and the nasal ‘ya’ and ‘ma’. These are current though not enjoined, and are heard though not instructed. But where should they be instructed? (Here the Värttika intervenes)-these that are current without being enjoined, should be instructed within the pratyāhāra ‘at’ for cerebralisa- tion of ’na’. i, u, r, 1, and since ‘i’ is both an operator and being operated upon, we cannot expect ‘i’ to turn into ‘ya’ according to the former analogy. To this Bhattoji replies that since ‘yan’ has been performed and i, u, r, Į, individually mentioned in ‘k’, it should not be arrested from undergoing a change after the former analogy. But the difficulty regarding ‘madra-hrada’, where the above analogy cannot apply, remains as before. 134 Mahābhāṣya of Patanjali (The Bhāṣya explains it as follows:) These ‘ayogavāhas’, i.e. those that are current without being enjoined, should be instructed within the pratyāhāra ‘at. What is the necessity? For cerebralisation of ’na’, as in urahkeṇa’, ‘uraxkena’, ‘uraḥpena’; ‘uraxpena’, so that we may have the crebralisation of ’na’ through the intervention of ‘at (according to the rule, “at kupvan-num-vyavaye’pi” 8.4.2- i.e. the dental na becomes cerebralised when it comes after ‘ra’ and ‘sa’ even when intervened by ‘at, ‘ku’, ‘an’, and ’num’). We know that when there is an intervention of ‘visarga’, ‘jihvāmüliya’ and ‘upadhmaniya’, the cerebralisation take place. It has, therefore, been suggested that ‘visarjaniya’, ‘upadhmānīya’ and ‘jihvāmūlīya’ would take their place within the pratyahāra ‘at’. (Here the Varttika intervenes and suggests:)- That they should have their place within ‘sar’ for effecting ‘jastva’ and ‘satva’. (The Bhāṣya explains it as follows:)- The above ayogavaha-s should also be included within the pratyāhāra ‘sar’. What is the use? The root ‘ubja’ is regarded as having ‘upadhmaniya’ in its upadha. This ‘upadhmaniya’ becomes changed to ‘va’ by the rule “jhalam jaśi”. (Thus there was upadhmaniya after ‘u’ and this was followed by ‘jita’ and ‘jitum’ and we have thus the forms ‘ubjita’ and ‘ubjitum’). If the ‘upadhmānīya’ is included within ‘jhal’, it will be placed within ‘sar’ and thus it is changed to ‘jas’. (But with this view of the Varttikakāra, the Bhāṣya does not agree, and he says as follows:) The ayogavāhas should be included within ‘Sar’. What is the necessity? In order that ‘jas’ may be enjoined and the cerebrals may be prescribed. The word ‘ubji’ is to be regarded as having for its penultimate an upadhmaniya (the form being ‘uji’ and if this upadhmaniya is regarded as having its place within the ‘sar’ pratyāhāra then it will be a ‘jhal’ and therefore be liable to be changed into ‘jas’ (the relevant jaś being here ‘va’) by the rule “jhalam jaśo’nte” (8.2.39). (The ‘jhal’ conso- nants occurring at the end of a pada, is transformed into a relevant ‘jas’ consonant. Thus we may have the forms ‘ubjita’, and ‘ubjitum’. [The Second Ahnika stops here in the manuscript.-Editor] Pāṇini. THIRD AHNIKA vṛddhir ādaic. 1.1.1 At and ‘aic’ are called vṛddhi. The Bhāṣyakara does not attempt any formal explanation of the sutra, but he starts at once with the different enquiries. The Uddyota first says that the sutra is composed of two padas, ‘vṛddhiḥ’ and ‘adaic’, for from an indication of the Bhasya (kṛtam anayoḥ sadhutvam) it is unnecessary to support the correctness of the two) we find that by the word ‘anayoh’ in the dual number it has been suggested that the sutra consists of two members. If it is supposed that the sutra does not consist of two members, then the word ‘vṛddhi’ should have to be repeated once in the case of ‘at’ and again in the case of ‘aic’ i.e. we have to interpret as ‘at’ is called ‘viddhi’ and ‘aic’ is called ‘vṛddhi’. If it is supposed to be of two members, we may say, ‘at’ and ‘aic’ are called ‘vrddhi’. In that case ‘adaic’ will be a ‘dvandva’ compound, but yet we need not suppose that there should have been ’tac’ as the samāsānta suffix by rule 5.4.106, for owing to the indication contained in the rule 5.4.115, the samāsānta rule is not obliga- tory. On this point Bhaṭṭoji in his Sabdakaustubha introduces further subtle considerations in support of the view that the concluding suffix at the end of a samāsa is not obligatory. The Uddyota, however, refers to the rule 6.2.193 when it is supposed that even when ‘rajan’ becomes the last member of a tatpuruşa, it may not have so-called obligatory ’tac’, and in such a case when the word is preceded by ‘prati’, the last member is associated with the udatta accent ‘pratiraja’. This suggests that the samāsānta rule is not obligatory. The alternative interpretation that the sūtra consists of three members, has also the support of the Bhāṣya, for in Paspaśā he definitely splits up the sutra as ‘vṛddhiḥ’at’ and ‘aic’. In such a case the word ‘vrddhi’ has to be taken twice as shown above, and the reference by the word ‘anayoh’ has to be taken as referring to the two parts of the sutra - one part is ‘vrddhi’ and the other part is ‘adaic’ consisting of the parts ‘at’ and ‘aic’ and the case is not one of a compound. Bhāṣya. kutvam kasman na bhavati “coḥ kuḥ”, (Pā.8.2.30) “padasya” (Pa.8.1.16) iti? bhatvāt. katham bha-samjñā? “ayasmayādīni 136 Värttika. Bhāṣya Vārttika. Bhāṣya. Mahābhāṣya of Patanjali cchandasi” (Pa. 1.4.20) iti. ‘chandasi’ ity ucyate, na cedam chandah. chandovat sūtrāņi. chandovat sūtrāņi bhavanti. yadi bha-samjñā, “vṛddhir ad aij”, “aden gunah” (Pa.1.1.2) iti jaśtvam api na prāpnoti. ubhaya-samjñāny api cchandāmsi. ubhaya-samjñāny api cchadāmsi dṛśyante. tad yatha- “sa susthubha sa rkvatā gaṇena”. padatvāt kutvam, bhatvāj jaśtvam na bhavati. evam ihâpi padatvāj jaśtvam, bhatvāt kutvam na bhavisyati. kim punar idam tadbhāvita-grahaṇam- vṛddhir ity evam ye akāraikāraukārā bhavyante teṣām grahaṇam, ahosvid ad- aijmātrasya? kim cataḥ ? yadi tadbhāvita- grahaṇam, ‘śāliyo’, ‘maliya’ iti vṛddha- lakṣaṇaś cho na prāpnoti. ‘āmramayam’, ‘śālamayam’, vṛddhilakṣano mayan na prāpnoti. ‘āmraguptayaniḥ’, ‘śālagupta- yaniḥ’, vṛddhalakṣaṇaḥ phiñ na prāpnoti. Why should not there be ka-group, that is why the ‘ca’ of ‘aic’ should not be changed into ‘ka’ when it is at the end of a pada. by the rule that a ka-group will replace the ca-group.88 (Now the ca in ‘aic’ being at the end of a pada ought to be transformed into ka by this rule-this is the question. The answer is that) Because it is ‘bha’. Why the bha-nomenclature? The words like ‘ayasmaya’ etc. (1.4.20) are to be regarded as correct as having the bha-nomenclature in the Vedas. It is said that it is to be so in the Vedas. This is not Vedas. — 88.coḥ kuḥ (8.2.30)-a cavarga or ca-group becomes transformed into a kavarga or ka-group when consonants of the ‘jhal’ follow or at the end of a pada, e.g. paktvā, vāk, etc. Third Ahnika (Here the Varttika runs)-That the sutras are like the Vedas. 137 (This Varttika is explained by the Bhāṣya as follows:)-The sutras are to be regarded as being like the Vedas. If, however, there is bha- nomenclature, then in the rules, “vrddhir ad aij; aden gunah’ (1.1.1 and 1.1.2) the ‘c’ of ‘aic’ should not be replaced by a ‘jas’ as a ‘ja’.” 89
89. These comments and discussions have no direct grammatical bearing of any importance. They are intended to show that the words used in the sutras have been correctly used. These discussions have elicited further discussions from commenta- tors like Kaiyața, Nageśa and Bhṭṭoji Dikṣita which are based on the later developed principles of the logical language. It is worthwhile showing the essential points of their arguments. Thus Kaiyața says that a word has many functions and when it is used as a nomenclature, its function is limited to a particular object. But in spite of such limitation, the potency of a word to be used as a nomenclature remains always valid. The whole point of the question here is as to whether the words used here should be regarded as one in the ordinary language or not, and in the former case whether it is to be regarded as being like an imitative word (anukarana śabda) or a class-name, or merely as a proper name as technical words signifying certain objects. Commenting on this point Nageśa says that first of all the relations between words and their objects are to be regarded as constant and eternal, and in the usage of words to denote particular objects by convention should also be regarded as always valid. Even though in such cases the connotation of the word is limited by convention, yet once accepted the potency of the word to denote its own special object remains un- restrained and at the time of denoting such an object the other functions of the word become arrested and its other meanings cannot operate. Nägeśa further says that words may have various meanings. Thus the word vṛddhi has many meanings, but in the context of the grammatical sutras of Panini the word vṛddhi has specific meaning as denoting ‘ad aic’, and at the time of such denotations these other meanings are arrested. Some say that words are of different functions in accordance with the objects with which they are associated; others say, that the divergence of function or potency depends upon its divergent powers of denotation. It is admitted that the relation of words with their objects is beginningless and as such the human behaviour ought not to produce any change in the denotatíve power of words and there is a dictum that all words may mean unrestrictedly all objects. Thus though words may have many meanings, the real power, potency or function of a word rests on an identity with the object it denotes. The point that is insisted on in the present situation is that words used in special senses in the grammatical literature have the force of popular usage. On one alterna- tive it may be regarded as limitative word. Thus ‘aic’ is a pratyahāra denoting ‘ai’ and ‘au’ and the word ‘aic’ in limitation of it stands as a separate word and may yet through association denote the same two alphabets ‘ai’ and ‘au’. According to others, however, the technical terms used in grammar having specific denotation and connotation, have the force of popular class-names. Again the word vṛddhi means growth, expansion and as such it might have meant any long or pluta sound. This samjñā rule is made in order to restrict the meaning of 138 Mahābhāṣya of Patañjali (Thus the idea introduced by the Varttika that in Vedic usage a word may have two nomenclature as a ‘bha’ and ‘pada’ at the same time and the Bhāṣya interprets it as follows:) Vedic words may have two nomencltures, i.e. they may enjoy the privilege of transformation of two nomenclatures at the same time. (Thus the Vedic passage runs as follows:) “sa susthubha sa ṛkvatā gaṇena”. Here with reference to the word ‘rkvata’ it is said, because it is a pada, therefore c becomes k and because it is ‘bha’ therefore it does not come under the group ‘jas’ (and therefore it does not become g). So here also on account of being pada, it can become ‘jas’ (there is change of ca into ja with a the word vṛddhi only to a, ai, and au. Such an interpretation, however, is objected to on the ground that such an analysis would not apply to its sister nomenclature guna means quality and at + en cannot be by the largest stretch of imagination brought under the ordinary meaning of guna as a quality or a rope. Bhattoji does not contribute anything to the discussion. The association of ’ta’ with a limits the long vowels a, ai, au to two mātrās only. These vowels when pronounced with the third and fourth matras are hereby excluded from the nomenclature, i.e. vṛddhi is ā, ai, and au of the second matra only. On the utility of associating ’ta’ for restricting the time of the sound to a particular model, see “taparas tatkalasya” (1.1.70). The whole point here is the problem that the ‘c’ of ‘aic’ does not become ‘k’ but become ‘j’ by phonetically combining with a of the next rule “aden gunah”. The ‘c’ is to become ‘k’, or for the matter of that a ca-varga is to be replaced by a ka-varga at the end of a pada by the rule 8.2.30. (‘coḥ kuh’). In the case of such a transformation we should have the rule “vṛddhir ādaik” (there c becoming k). The danger is avoided by supposing that ‘aic’ is not to be regarded as a pada, but as ‘bha’ after the analogy of the rule, “ayasmayadinic chandasi” (1.4.20) ‘bha’ is a technical term which is applied to words with certain suffixes and which contrives thereby to suspend the operation of all rules that would be set loose by the word becoming a pada. The bha- nomenclature is affected by the rule “yaci bham” (1.4.18) and this bha-nomenclature is extended towards ‘ayasmaya. So if aic has to be treated as ‘bha’, it has to be regarded as a Vedic word. On this point the Värttika comes to our help and says that the words of the rules of Panini are to be regarded as Vedic words. But then if it is a Vedic word, then c would not become kindeed, but in that case the ccannot be jwhich is effected only if cis regarded as a padanta letter by the rule “jhalam jaso ’nte”(8.2.39). But there is a general rule that in Vedic usage all rules are flexible and thus we have Vedic words which in the same use once enjoy the privilege of being a ‘bha’ and at another time of being a padanta. Thus in the passage “sa susthubha sa ṛkvatā gaṇena” in the word ‘rkvata’ the c becomes k as it is allowed to have the privilege of a padanta, and the cis transformed into k by the rule “coh kuh” (8.2.30). But when it were further to enjoy its rights of being a padanta by change of k into g by the rule “jhalam jaśo’nte” (8.2.39), it suddenly begins to behave as a ‘bha’ and the transformation of k into g is not effected (padatvāt kutvam bhatvāt jaśtvam na bhavati). Third Ahnika 139 following a by the rule “adengunah” and becaue it is ‘bha’ therefore it is not transformed into a ka-group, i.e. c changed into k). (Another point referred to by Nageśa on the treatment of sutra words as being of Vedic usage. It is said that grammar is a principal accessory (anga) to the Vedas. The words of the sutra do not follow the Vedic usage, because the sutras were written by a Rṣi, but because according to the instruction of the Bhasya, we sometimes find words in the sutras the same elasticity as we notice in the Vedic usage). (A new point of discussion arises. The point is, whether the ā, ai, and au are original forms or transformatory forms. Thus the Bhāṣya runs:) Are these forms to be regarded as having been effected by grammatical change when ā, ai and au are produced under the instruction that there should be a vṛddhi? Now, should the a and ai and au be formed through modification under the application of vṛddhi or are they to be regarded as the simple vowels a, ai and au? What is the idea? If it means those which are produced by modification, then such forms as ‘saliya’ and ‘maliya’ formed by the suffix ‘cha’, that is enjoined after long words, would not be justifiable, and the words like ‘amramaya’ and ‘salamaya’, where ‘mayat’ is applied to long words, would not also be justified. So also the ‘fin’ suffix which is enjoined after long words, should not be justifiable to form the words ‘amraguptayani’ and ‘salaguptayani’.90
90. Panini has made two samjñās or nomenclatures, ‘vṛddhi’ and ‘vṛddha’. We are discussing the nomenclature ‘vṛddhi’ here. The term ‘vṛddha’ has been defined in Panini’s rules 1.1.73, 1.1.74 and 1.1.75. The rule 1.1.73 “vrddhir yasyâcām ādis tad vrddham” means that a word is called vṛddha when out of the vowels that form its constitution, the first one falls on the nomenclature of ‘vṛddhi’. The rule 1.1.74 is “tadadini ca”. The words ’tad’ ‘yad’ ’etad’, etc. are also called ‘vṛddha’. The rule 1.1.75 is “en prācāmdeśe” meaning where of the composite voewls forming a word, the vowels of the group ’en’ come first, the word is called ‘vṛddha’ when it is used to denote a country, e.g. ‘bhoja-kotiya’, ‘gonardiya’. The necessity of forming the nomenclature ‘vṛddha’ is the enjoining with the suffix ‘cha’ or ‘cya’ by the rule “vṛddhac chah”. (4.2.114). Other occasions for the use of the nomenclature ‘vrddha’ is to be found in the rule “nityam vṛddha-śabdädibhyah” (4.3.144), and the rule, “udīcām vṛddhād agotrāt” (4.1.157). By the rule 4.3.144 the ‘mayat’ suffix is enjoined as in ‘amramaya’, ‘śālamaya’ referred to above and by the rule 4.1.157, the suffix ‘phiñ’, is to be enjoined. Bhattoji notes in his Sabdakaustubha that there is a non-distincion ‘abheda’ between ā-kāra, ai-kāra and au-kāra and vṛddhi. What is ā-kāra, ai-kāra and au-kāra and vṛddhi? This view finds justification in the fact tht Panini has himself designated ā-kāra, ai-kāra and au-kāra as vṛddhi in the rules 1.1.73 and 1.1.75. But a further140 Bhāṣya. Mahābhāṣya of Patanjali atha ad-aij-mātrasya grahaṇam, sarvo bhāsaḥ sarva-bhasa iti, “uttara-pada-vṛddhau sarvam ca” (Pa. 6.2.105) ity eṣa vidhiḥ prāpnoti. iha ca tavati bhārya yasya tāvadbhāryaḥ, yāvad- bhāryaḥ. “vṛddhi-nimittasya” (Pa. 6.3.39) iti pumvadbhāva-pratiṣedhaḥ präpnoti. astu tarhi ad-aijmātrasya grahaṇam. nanu coktam sarvo bhāsah sarvabhāsa iti, “uttarapada-vṛddhau sarvam ca” ity eșa vidhiḥ präpnoti iti. naiṣa doṣaḥ. naivam vijñāyate-uttara- padasya vṛddhir uttarapada-vrddhir uttara- pada-vṛddhau iti. katham tarhi? “uttara- padasya” ity evam prakṛtya ya vṛddhis tad- vaty uttarapade ity evam etad vijñāyate. avaśyam caited evam vijñeyam. tadbhāvita- grahane saty apiha prasajyeta-sarvaḥ kārakaḥ sarvakāraka iti. difficulty arises from another point of view. Vṛddhi is indeed equivalent to ā-kāra, ai- kāra and au-kāra, but are these vowels to be regarded as the long vowels that are read in the alphabetic list, or are they to be regarded as similar forms of vowels arising out of grammatical transformation? In this connection it is worthwhile noting that I have taken some licence in translating the word vṛddha as long. But the word vṛddha is a technical word and a technical nomenclature which though applicable to words having long a, ai and au as the initial vowel, there are cases in which words like ‘yad’, ‘idam’, etc. which have no such characteristic, are also called ‘vṛddha’. Now the point raised in the Bhāṣya is that since the word ‘vṛddha’ has been used merely to designate the simple a-kāra, ai-kāra and au-kara, these cannot be taken as being equivalent to the equivalent forms of the vowels as produced through gram- matical transformations as a result of the enjoining of vṛddhi in certain conditions. Thus the special suffixes ‘cha’, ‘mayat’, ‘phiñ’, etc. enjoined to words which may be styled vṛddha, should not be applicable to their respective spheres. If the word vṛddhi as used here, should mean only the a-kāra, etc. formed by grammatical modifications as a result of an operation of vṛddhi, then the pure long vowels that have been called vṛddhi in the rules for forming the technical nomenclature of vṛddha, would not be applicable to those words, for according to hypothesis vṛddhi can only have the designation to mean the vowels transformed under the injunctions of vṛddhi operation. Third Ahnika 141 yad apy ucyate “iha tāvati bhārya yasya tāvad-bhāryaḥ, yavad-bhāryaḥ” iti. ca “vrddhi-nimittasya” iti pumvadbhava- pratiṣedhaḥ prāpnoti iti. naiṣa doṣaḥ. naivam vijñāyate-vṛddher nimittam vṛddhinimit- tam vṛddhinimittasya iti. katham tarhi? vrddher nimittam yasmin so’- yam vṛddhinimittaḥ, vṛddhinimittasya iti. kim ca vrddher nimittam? yo’sau kakaro ñakāro ṇakāro vā. athava yaḥ kṛtsnāyā vrddher nimittam. kaś ca kṛtsanāyā vṛddher nimittam? yas trayāṇām ākāraikāraukāraṇām. (The Bhāṣya now takes up the other alternative, viz. the view of treating at and aic as mere simple long vowels and demonstrates the apparent objections. Thus the Bhāṣya says:)- If only at and aic (simple long vowels) are taken (as the proper meaning) then in the case of ‘sarvo bhāsaḥ sarvabhāsaḥ’ the rule “uttarapada-vṛddhau sarvam ca” (6.2.105) where a vṛddhi is advised to uttarapada, there is the latter member by virtue of adhikara, the words ‘sarva’ and ‘dis’ induce the udatta accent as the concluding accent of the word containing ‘sarva’ and words signifying ‘dik’, i.e. a quarter of the sky, such as east, west, etc.91 Here again the phrase, ‘he who has so much (tāvati) wife’ yields the compound ’tāvadbharyah’, analogously ‘yavadbharyah’ (one possessing
91. It should be noted here that there is a rule “susarvärdhaj janapadasya” (7.3.12) and a rule “diso’ madrāṇām” (7.3.13) which carry with them as ‘adhikara’ the word ‘uttarapadasya’ from the rule “uttarapadasya” (7.3.10). The rule “susarvärdhāj jana- padasya” thus means that when a word denoting a place is preceded by the word ‘su’, ‘sarva’ and ‘ardha’ then that word denoting a city which is also the last member of a compound, should have its first vowel lengthened by ‘vṛddhi’. There is another rule “uttarapada-vṛddhau sarvam ca” (6.2.105) meaning that in the case of words whose last members are affected by vṛddhi according to the rules 7.3.12 and 7.3.13, the words that contain the word ‘sarva’ and words denoting ‘dis’ or direction should be associated with the terminal udatta accent. Now, the point here is that the words ‘sarvabhasa’ and ‘sarvakaraka’ consist of each of two members ‘sarva’ and ‘bhasa’, ‘sarva’ and ‘karaka’. ‘Bhasa’ and ‘karaka’ have both a long a, and a vṛddhi means the long’ ‘a’, then ‘sarvabhasa’ and ‘sarvakaraka’ are words which have vṛddhi in the last member, and as such the words ‘sarvabhasa’ and ‘sarvakaraka’ should be associated with the terminal udatta. This generates the suspicion that vṛddhi does not mean ‘a’ and ‘aic’ as suggested in the alternative interpretation under consideration. 142 Mahābhāṣya of Patanjali as many wives). By the rule “vṛddhi-nimittasya ca taddhitasyâraktavikāre” (6.3.39- if a word having such a taddhita suffix that may cause vṛddhi provided also that such a taddhita suffix has not been enjoined in the sense of redness or modification, then a feminine suffix with which it may be associated, is not masculinised), there should be no masculinisation (in ‘yāvadbhāryaḥ’ and ’tāvadbhāryaḥ) as might have been expected.92 Let us then take merely the simple vowels, at and aic. But it has been objected that in the case of ‘sarvo bhāsaḥ sarvabhāsaḥ’ the rule “uttara- pada-vṛddhau sarvam ca” becomes operative. This is no criticism. We do not understand that vṛddhi of the ‘uttarapda’ has been equivalent to “uttarapda-vṛddhi” and then (in the locative) ‘uttarapada-vṛddhau’ (i.e. in the case of the vṛddhi of the last member). How then? It means a vṛddhi which has been sanctioned by the rule “uttarapadasya” (with reference to a’uttarapada’ and the word ‘uttarapada-vṛddhi’ means a ‘uttarapada’ that contains such a ‘vṛddhi’. This is the way in which we understand it. Certainly this is the way in which we understand it, for even if we take vṛddhi in the modificatory sense (as against the simple sense) the same difficulty occurs with reference to ‘sarvaḥ’ (all)kārakaḥ (agent)— ‘sarvakārakaḥ’.93
92. There is a rule ‘a sarvanāmnaḥ’ (6.3.91). It means when the suffixes ‘drk’, ‘drs’ and ‘vatu’ follow, then a ‘sarvanaman word will have the augment a in place of all vowels and consonants excluding the first consonant. Thus in the case of ‘yavat’ and ’tāvat’ we have the suffix ‘vatu’ to the words ‘yad’ and ’tad’. This ‘vatu’ is a nimitta or in other words it causes the augment a by virtue of which ‘yad’ and ’tad become ‘ya’ and ’ta’ and then in association with ‘vatu’ they become ‘yavat’ and ’tāvať’ and in association with the feminine suffix they become ‘yavati’ and ’tavati’. Now if this simple ‘a’ may be styled a vṛddhi then it has to be admitted that the suffix ‘vatu’ is the ’nimitta’ of this ‘a’, i.e. it has caused the ‘vṛddhi’ (i.e. a). Consequently the phrase - ‘yāvati bhārya yasya saḥ’) comes under the purview of the rule “vṛddhi- nimittasya” etc. (6.3.39) which would be a bar to masculinisation. But yet there is no masculinisation.
93. Now we have the rule “uttarapada-vṛddhau sarvam ca” (6.2.105). If instead of this, the rule would have been ‘vrddhau sarvam ca’ and we had dropped the word ‘uttarapada’, then the meaning would have been a ‘uttarapada’ that contain the vṛddhi vowel. The fact that the word ‘uttarapada’ had been inserted in the sutra, would therefore have significance. This significance is interpreted to be that where vṛddhi is enjoined as to take place in the last member of the compound, the subtle point thus is that the word ‘uttarapda-vṛddhau’ means that it has reference only to those words where a vṛddhi has been enjoined to a ‘uttarapada’ by virtue of its being an ‘uttarapada’ and according to the rule ‘uttarapadasya’ or rather to the rules which carry the word ‘uttarapadasya ca’ as adhikara along with them. Now the word ‘bhasa’ Third Ahnika 143 Again it is said here he who has ’tavati’ (such like) ‘bharya’ is ’tāvadbhāryaḥ’ and analogously ‘yāvadbhāryaḥ according to the rule “vṛddhi-nimittasya” (6.3.39) which has no masculinisation here. This is no criticism. The meaning of the word ‘vrddhi-nimitta’ is not that which is the first of ‘viddhi’, but it means that which induces ‘vṛddhi’ or which may induce ‘vrddhi’. How then wherever there is a cause of vṛddhi that is called ‘vṛddhi-nimitta’ and then the genitive ‘vrddhi-nimittasya’? What is the cause of vrddhi? The suffix in which the k-sound has been elided, or from which the ’n’ or ‘ñ’ has been elided, or that which is the cause of all vṛddhis. Again what is the cause of all vṛddhis, namely that of the three-ākāra, aikara and aukāra.94. Värttika. Bhāṣya. Värttika. samjñādhikāraḥ samjñā-sampratya- yārthaḥ. “atha samjñā” ity evam prakṛtya vṛddhyā- dayaḥ śabdāḥ pathitavyāḥ. kim prayojanam? samjñāsampratyayārthaḥ. vṛddhyādinām śabdānām ‘samjñā’ ity eṣa sampratyayo yathā syāt. itarathā hy asampratyayo yathā loke. in ‘sarvabhāsaḥ’, no doubt, contains a vṛddhi vowel a, but this vṛddhi vowel has not been indicated in the ‘uttarapada’ by any special rule which speaks of ‘uttarapada- vṛddhi’ and takes its ‘uttarapada’ as an adhikara from the rule, “uttarapadasya” (7.3.10). In the case of ‘sarvaḥ karakaḥ sarvakarakaḥ’ the long vowel in the word ‘karaka’ is a modificatory vowel and as such it ought to have come under the rule (6.2.105), but it does not do so, for with reference to the rule 6.3.105 it is innocuous whether vṛddhi is taken as a modificatory vowel produced by a rule enjoining vṛddhi, or simple long- vowel. The point that matters is whether there have been any ‘uttarapada-vṛddhi or a vṛddhi of the last member of a compound by the adhikara taken from the rule ’ uttarapadasya’.
94. The question of ‘pumvadbhava’ or masculinisation of ’tāvadbhāryaḥ’ and ‘yāvadbhāryaḥ’ as against the rule “vrddhi-nimittasya taddhitasya araktavikare” (6.3.39) the whole difficulty is about the meaning of the word ‘vṛddhinimittasya’. The Bhāṣya says that the word ‘vṛddhi-nimitta’ does not mean the taddhita-suffix which is the ’nimitta’ or cause of vṛddhi. But the word “vrddhi-nimitta” means that in which the “nimitta” of vṛddhi, viz. such conditions that by their existence in a taddhita-suffix induce vṛddhi, such as suffixes that elide ‘k’, ‘ñ’ and ’n’. Thus in the word, 144 Bhāṣya. Vārtikka. Mahābhāṣya of Patañjali akriyamāṇe hi samjñādhikāre vṛddhyādinām ‘samjñā’, ity eṣa sampratyayo na syāt. idam idānīm bahusūtram anarthakam syāt. ‘anarthakam’ ity aha. katham? ‘yathā loke’ loke hy arthavanti cânarthakāni ca vākyāni dṛśyante. arthavanti tavat-‘de- vadatta gāmabhyāja suklām daṇḍena’, devadatta gamabhāyaja kṛṣṇām’ iti. anarthakani-daśa dāḍimāni, şaḍapūpāḥ, kuṇḍam ajajinam, palala-pindah, adha- rorukam etat kumaryāḥ, sphaiyakṛtasya pitā pratiśīnaḥ’ iti. samjñā-samjñyasamdehaś ca. ‘śraughnibharyah’ we have first the word ‘śraughni’-‘śrughnasya ayam iti śraughnah’ formed by the adding of the suffix an to ‘śrughna’ and then we had the feminine suffix ’nip’ to ‘śraughna’ forming the word ‘śraughni’. Now ‘śraughni is a word which is formed by a taddhita-suffix that possesses conditions for indicating a vṛddhi. The word ‘śraughna’ is formed by the addition of ‘an’ to ‘śrughna’, and ‘an’ being a suffix that elides ’na’, generated vṛddhi. But ’tavati-bhāryaḥ’ or ‘yavati- bharyaḥ’ does not contain any word where a vṛddhi has been initiated by the prescription of a suffix that has dropped its ‘k’, ’n’ or ‘ñ’. Therefore it is quite natural that the sutra “vṛddhi-nimittasya” etc. is quite unconcerned as to whether a vṛddhi means the simple ā, ai, etc. or the modificatory a, ai, etc. It only lays down provisions for the arrest of masculinisation in the case of words formed by taddhita suffixes that drop ‘k’, ’n’ or ‘ñ’. Now the Bhāṣya raises the question- does the ’nimitta’ means that which is the cause of all types of vṛddhi, i.e. vṛddhi as a-kāra, ai-kāra and au-kāra. On this point Bhaṭṭoji says that it cannot be supposed that this would create any trouble in the case of ‘vaiyakaraṇa-bharyah’ and ‘sauvaśya-bharyah’ for the ai-kara in ‘vaiyakaraṇī’ has not been generated entirely by the influence of the ‘an’ suffix (in the word ‘vaiyakaraṇī the transformation takes place by the rule 7.3.3), for it involves the application of the influence of samprasarana. But in the case of ‘pariṣadyabharyaḥ’, we have the negation of masculinisation. The ‘pariṣadya’ is formed by the rule ‘pariṣadah nyah’. Here, since the whole word ‘pariṣadya’ is formed by ’nyah’, ‘pariṣadya-bharya’ is akin to ‘śraughni-bharya’, because it contains the taddhita suffix which is responsible for the vṛddhi. The conclusion then is that vṛddhi is to be regarded as equivalent to the simple ā- kāra, ai-kāra and au-kara. This conclusion is endorsed both by Nageśa and Bhaṭṭoji. Bhāṣya. Värttika. Bhāṣya. Vārttika. Bhāṣya. OVEY Third Ahnika kriyamāṇe’pi samjñādhikāre samjñā- samjñinor asamdeho vaktavyah. kuto hy etat-vṛddhi-sabdaḥ samjñā, ādaicaḥ samjñinaḥ iti. na punar ādaicaḥ samjñā, vṛddhi-sabdaḥ samjñi iti? yat tavad ucyate-‘samjñādhikāraḥ kar- tavyaḥ samjñā-sampratyayārthah’ iti. na kartavyaḥ. ācāryācārāt samjñāsiddhiḥ. ācāryācārāt samjñā-siddhir bhaviṣyati. kim idam ‘ācāryācārād’ iti? ācāryāṇām upacārāt. yatha laukika-vaidikeṣu. 145 tad yatha laukikeṣu vaidikeṣu ca kṛtānteṣu. loke tavan mātā-pitarau putrasya jātasya samvṛte’vakāśe nāma kurvāte-‘devadattaḥ’ ‘yajñādattaḥ’ iti. tayor ūpacārād anyepi jānanti-‘iyam asya samjñā’ iti. vede yājñikāḥ samjñām kurvanti -‘sphyaḥ’, ‘yupaḥ’, ‘caṣālaḥ’ iti. tatrabhavatām upacārād anye’pi jānanti-‘iyam asya samjñā’ iti. evam ihâpi. ihaiva tavat kecid vyācakṣāṇā ahuḥ-vṛddhi-sabdaḥ samjñā, ādaicaḥ samjñinaḥ’ iti. apare punah-“sici vṛddhiḥ” (Pa.7.2.1) ity uktvā ākāraikāraukārān udāha- ranti. tena manyamahe-‘yayā pratyāyyante sā samjñā, ye pratiyante te samjñinaḥ’ iti, yad apy ucyate-‘kriyamāņe’pi samjñā- dhikāre samjñā-samjñinor asamdeho vaktavyah’ iti. (na vaktavyah). 146 Vārttika. Bhāṣya. Värttika. Bhāṣya. Vārttika. Bhāṣya. Mahābhāṣya of Patanjali samjñā-samjñyasamdehaś ca. samjñā-samjñinos câsandehaḥ siddhaḥ. kutaḥ? ācāryācārād eva. ukta ācāryācārāḥ. anākṛtiḥ. athavā anākṛtiḥ samjñā, ākṛtimantaḥ sam- jñinaḥ. loke’pi hy ākṛtimato māmsapin- ḍasya ‘devadatta’ iti samjñā kriyate. lingena vā. athava kimcil lingam asajya vakṣyāmi- ‘itthamlingā samjñā’ iti. vṛddhi-sabde ca tallingam karisyate, na ādaic- chabde. idam tavad ayuktam, yad apy ucyate- ‘ācāryācārād’ iti. kim atrâyuktam? tam eva upālabhya ‘agamakam te sūtram’ iti tasyaiva punaḥ pramāṇīkaraṇam ity etad ayuktam. aparituşyan khalvapi bhavan anena parihā- reņa “anākṛtir lingena va’’ ity aha. tac câpi vaktavyam. yady apy etad ucyate, athavā etarhi it-samjñā na vaktavyā. lopos ca na vaktavyaḥ. samjñālingam anubandheṣu kariṣyate. na ca samjñāyā nivṛttir ucyate. svabhāvataḥ samjñā samjñinam pratyāyya svayam nivartate. tena anubandhānām api nivṛttir bhaviṣyati. siddhyaty evam. apāņiniyam tu bhavati. yathānyāsam eva astu. nanu coktam “samjñādhikāraḥ samjñā- sampratyayatrtha itaratha hy asaṁpratyayo yatha loke” iti. na ca yathā loke tathā vyākaraṇe. pra- māṇa-bhūta ācāryo darbha-pavitra-pāṇiḥ Third Ahnika śucavavakāśe pranmukha upaviśya mahatā prayatnena sūtrāņi praṇayati sma. tatrâ- sakyam varnenâpy anarthakena bhavitum, kim punar iyatā sūtreņa. kim ato yad aśakyam? ataḥ samjñā-samjñināveva. 147 kuto nu khalvetat-samjñā-samjñināveva iti. na punaḥ sadhvanuśāsane asmiñ śāstre sādhutvam anena kriyate? kṛtam anayoḥ sadhutvam. katham? vrdhir asmaya-viśeṣenopadiṣṭaḥ prakṛtipāṭhe, tasmat ktin-pratyayaḥ. ādaico’py akṣara- samāmnāye upadiṣṭāḥ. prayoga-niyamartham tarhi idam syāt- ‘vṛddhi-sabdāt pare ādaicaḥ prayoktavyāḥ’, iti. neha prayoganiyama ārabhyate. kim tarhi? samskṛtya samskṛtya padany utsṛjyante teṣām yatheṣṭam abhisambandho bhavati. tad yatha-āhara pātram’, ‘pātram āhara’ iti. ādeśās tarhi ime syuḥ. vṛddhi-sabdasya ādaica ādeśāḥ. ṣaṣṭhinirdiṣṭasyādeśā ucyante. na câtra ṣaṣṭhim pasyāmaḥ. āgamās tarhi ime syuḥ. vṛddhi-sabdasya ādaica agamāḥ. agama api şaṣṭhinirdiṣṭasya eva ucyante, lingena vā. na câtra ṣasthim, na khalv apy agamalingam pasyāmaḥ. arbidam khalvapi bhūyaḥ sāmānādhika- ranyam eka-vibhaktitvam ca. dvayoś caitad bhavati. kayoḥ? viseṣaṇa-viseşyayor vā samjñā-samjñinor vā. tatraitat syad viseṣaṇa- visesye iti. 148 Värttika. Bhāṣya. Varttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Mahābhāṣya of Patanjali tac ca na. dvayor hi pratita-padarthakayor loke viseṣaṇa-viseṣyabhavo bhavati. na ca adaic chabdaḥ pratita-padarthakaḥ. tasmāt samjñā-samjñināveva. tatra tu etāvān sandehaḥ-‘kaḥ samjñi kā samjñā’ iti. sa câpi kva sandehaḥ? yatrobhe samānākṣare? yatra tu anyataral laghu sā samjñā, yad guru sa samjñi. kuta etat? anākṛtiḥ samjñā, ākṛtimantaḥ samjñinaḥ. loke’ pi hy ākṛtimato māmsa-pindasya de- vadatta iti samjñā kriyate. avartinyaḥ samjñāḥ. athavā āvarttinyaḥ samjñā bhavanti. vṛddhiśabdaś ca avartate, na ādaic chabdaḥ, tad yathā―itaratrâpi devadatta-sabda āvar- tate, na māmsa-pindah. pūrvoccāritaḥ samjñi paroccărită samjñā. athavā pūrvoccāritaḥ samjñi, paroccāritā samjñā. kuta etat? sato hi karyiṇaḥ kāryeṇa bhavitavyam. tad yatha- itaratrâpi sato māmsapinḍasya ‘devadatta’ iti samjñā kri- yate. katham ‘vrddhir ādaij’ iti? vṛddhirādaij iti mangalārtham. etad ekam ācāryasya mangalartham mṛṣyatām. mangalika ācāryo mahataḥ śāstraughasya mangalartham vṛddhi-sabdam aditaḥ prayunkte. mangalādīni hi śāstrāṇi prathante, vira-puruṣakāņi bhavanty ayuṣ- Värttika. Bhāṣya. Third Āhnika mat-puruṣakāņi ca adhyetāraś ca vṛddhi- 149 yukta yatha syur iti. sarvatraiva hi vyākaraṇe pūrvoccāritaḥ samjñi, paroccāritā samjñā- “aden gunah” iti yathā. doṣavān khalvapi samjñādhikāraḥ. aṣṭame’pi hi samjñā kriyate-“tasya param amreḍitam” (Pa. 8.1.2) iti. tatrâpi idam anuvartyam syat. athava’sthāne’yam yatnaḥ kriyate. na hîdam lokād bhidyate. yadi idam lokād bhidyeta tato yatnârham syat. tad yatha- agojñāya kaścid gām sakthani karne vā grhitvā upadiśati- ‘ayam gaur’ iti. na ca asmayacaṣṭe ‘iyam asya samjñā’ iti. bhavati câsya sampratyayaḥ. tatraitat syat-‘kṛtas tatra pūrvair abhi- sambandhaḥ’ iti. ihâpi kṛtaḥ purvair abhisambandhaḥ. kaiḥ? ācāryaiḥ. tatraitat syāt-‘yasmai tarhi sampraty upadiśati tasyâkṛtah’ iti. loke’pi hi yasmai sampraty upadiśati tasyakṛtaḥ. atha tatra kṛtaḥ, ihâpi kṛto draṣṭavyah. sato vṛddhyādiṣu samjñā-bhāvāt tadāśraya itaretarāśrayatvād apra- siddhiḥ. sataḥ samjñinaḥ samjñābhāvāt. tadāśraye samjñāśraye samjñini vṛddhyādiṣu itareta- rāśrayatvād aprasiddhiḥ. kā itaretarāśrayatā? satām ādaicām samjñaya bhavitavym, samjñayā ādaico bhavyante, tad etad itaretaraśrayam bhavati, itaretarāśrayāṇi ca kāryāņi na prakalpante.150 Värttika. Bhāṣya. Mahābhāṣya of Patanjali tad yatha-naur nāvi baddha netaratrāṇāya bhavati. nanu ca bho itaretarāśrayāṇy api kāryāṇi drsyante. tad yatha-nauḥ śakatam vahati’, ‘sakatam ca nāvam vahati’. anyad api tatra kimcid bhavati. jalam vā sthalam vā. sthale sakaṭam nāvam vahati. jale nauḥ śakatam vahati. yatha tarhi triviṣṭabdhakam, tatrâpy an- tataḥ sūtrakam bhavati. idam punah itaretarāśrayam eva. siddham tu nityaśabdatvāt. siddham etat. katham? nitya-sabdatvāt. nityāḥ śabdāḥ, nityeşu śabdeşu satām ādaicām samjñā kriyate. na ca samjñayā āda- ico bhavyante. (The Varttika starts with the remark that) the topic regarding nomencla- tures is for the purpose of proper understanding of the nomenclature or in other words for the proper understanding of the nomenclature a separate topic on nomenclature should be made. (This is interpreted by the Bhāṣya as follows:) The words vṛddhi etc. ought to be read after putting the characteristic form such as now, the topic of names. What is the good? For explaining the names; so that people may easily understand that the words vṛddhi etc. are but grammatical technical names. (The intervening Varttika says:) Otherwise as in the popular usage, it would not be so apprehended. (This is interpreted by Bhasya as follows:) If a topic for samjñā or names be not made people will not understand that the words like vṛddhi etc. has been introduced merely as names. In such a case a sastra would have often to contain two many sutras which would not strictly be useful.95
95. In the translation of the line ‘idam idānim bahusutram anarthakam syat’, I have rendered it as what appeared to me as the obvious meaning. The traditional meaning as has been imperfectly suggested in the Pradipa and the Uddyota seems to me to be rather lame. But at least I have not been able to appreciate its force so as to carry conviction. Third Ahnika 151 You say meaningless, why? as in the world we find sentences that have meaning, and sentences that have no meaning. Significant sentences are as ‘Devadatta drives the white cow with a stick.’ ‘Devadatta drives the black cow with a stick.’ Meaningless sentences are as follows: ’ten pomegranates,” the six cakes, the feet goat skin, lump of flesh, petticoat of this girl, the father of the abundantly made this limb.“97 (The Varttika intervenes and says)- some words have to be used to remove the doubt as to which is the name and which is the named. (The Bhāṣya says)-Even when a special topic for names is made there ought to be words to resolve the possible doubt as to what is the name and what is the named. How is that? The word vṛddhi is the name and ‘at’ or ‘aic’ is the bearer of the name and not that at or aic is the name and vṛddhi is the named. It has been suggested that for the proper understanding of the names there should be special topic of names- it should not be made. (The Värttika says)-The symbolic or technical names will be intelli- gible from the manner of the master’s treatment. (The Bhāṣya says) The treatment of names will be accomplished by the manner of the master. What is this? This manner of the master? By the usage of the master. (Here the Varttika says)-As in the popular and the Vedic words. (Here the Bhāṣya says) Just as in the case of the popular and Vedic doctrines. In the world the mother and the father give the name, ‘De- vadatta’, ‘Yajñadatta’, to the son in a closed or secreted place. From their usage others are to know. In the Vedas the sacrificers invent the names,
96. As there is no final vowel attached to it, it is difficult to understand anything about the ten pomegranates.
97. As a matter of fact these are not examples of vakya or significant sentences. Each sentence must be a completed relation of significant terms. But here only words have been taken to be called vakya, and want of significance is due to the introduction of more words without forming them into vakyas. The whole point of the agreement is that the technical words are merely used and if it is not indicated that these have been introduced only as technical names, people who are unacquainted with the unfailing authority of another, may feel confused and regard these to be meaningless. It is also to be noted that it is only for the particular kind of relation between significant terms so as to make these facts occur that there is a meaning to be attached to the propositions which expressed them. Neither names which are neither here nor there. They have no status in the world of meaningless matter. If technical terms are introduced without the specification that they are symbols, they would carry no sense. 152

98. The opponent suggests that the rule may have been formed for regulating the method of syntax in which one is allowed to say ‘vṛddhir ādaic’ and not ‘ādaij vrddhiḥ’ 154 Mahābhāṣya of Patanjali (The Bhāṣya replies) the augments also can only be in place of words marked with genitive suffix, for through other indicatory signs we do not find here any genitive, nor do we find any indication of an augment.99 (Again the Bhāṣya says) the rule is that the two should have the same case-inflexion and the same relation to each other as that of a substance or quality. This happens in the case of the two. Of which two? Either in the case of the symbol and the symbolised. May it not be there that here the relation is that of substantive and adjective. No. The relation of the substantive and the adjective can hold only with reference to the objects that are empirically perceived. At and aic are not empirically perceived objects. Therefore they are related as the symbol and the symbolised.100 Now there is this doubt which is the sybolised and what is the symbol? Where is the doubt? The two have the same number of letters. Where one is simple that is the symbol. That which is complex is symbolised. How do you know this? (Here the Värttika intervenes)-symbolisation is for the purpose of making things simpler. (To this the Bhāṣya says)-is not always that symbolisation directed to making simple what is complex? What then? To denote the unstructural character also.
99. Whenever an augment is prescribed, it is prescribed to take place with reference to a word in the genitive case. But there are cases as in “na yvābhyām padantābhyām” etc. (7.3.3) where an aic agama is prescribed to take place before the y and v of a word and there is no indication of the prescription of this augment by any genitive case-ending. But in such cases it has been specially enjoined. Without such special injunction, as a general rule the augment takes place only in place of the genitive.
100. The substance and a quality are supposed to reside in the same locus and therefore they are called ‘samanâdhikarana’. They have therefore the same case- ending. Others, however, think that in adjectival phrases like ’nilam utpalam’ (the blue lotus) the word ’nila’ is used to denote a quality, blue, and ‘utpala’ the class notion of lotus. The two are objectively in the same locus and this constitutes the ‘samanâdhikaranya’. The two views are on the question as to whether the relation of ‘samanâdhikaranya’ is a prospositional quality or quality of the objects. When the Bhasya used the word ‘samjñā and ‘samjñin’, the symbol and the symbolised, the word ‘samjñin’ or the symbolised means the prticular word or words of which a ‘samjña’ or name has been framed. Thus Nägeśa says ‘samjñi-padena samjñi-samarpakaḥ śabdo ‘bhipretaḥ’. Bhattoji also says that ‘vṛddhi’ etc. are words denoting nothing else but themselves and they are related with words ādaic etc. as the symbol with the symbolised-‘vrddhyādayo yatha Sabdaḥ svarupopanibandhanaḥ adaic-pratyayitaiḥ śabdaiḥ sambandham yanti samjñibhiḥ’. Third Ahnika 155 (Here the Värttika says) the symbol has no structure. The symbolised ones are structural. (The Bhāṣya says) in the world also a lump of flesh having structure is given the name ‘Devadatta’. (Here the Varttika intervenes) or rather the symbols are frequently repeated. (The Bhāṣya says)-the names are frequently repeated. Thus the word ‘vṛddhi’ is frequently repeated, but not the words ‘at’ and ‘aic’. So elsewhere also the name ‘Devadatta’ is repeated, but not the lump of flesh. (Here the Varttika intervenes)-what is pronounced first is the sym- bolised and what is pronounced later, is the name. (The Bhāṣya says)-or, rather what is first pronounced is symbolised and what is pronounced next is the symbol. How so? It is only with reference to existing things that any action can be prescribed. Thus elsewhere also it is because the lump of flesh exists that the name ‘Devadatta’ can be given to it. Why then the form “vrddhir ādaic"101 (Here the Varttika intervenes)-the form ‘vṛddhir ādaic’ has been chosen (the word ‘vrddhi’ has been pronounced first) to make it an auspicious sign. (The Bhāṣya says)-one may take this as the desire of the master for beginning his work with an auspicious word. Desiring the good of the readers the master used the word ‘vṛddhi’ at the commencement as an auspicious rite to the great collection of grammatical dictates. The scriptures beginning with auspicious words find a wide currency. They generate heroism and produce long life so that readers may become prosperous. Everywhere else in grammar the symbolised is spoken first and the symbol is spoken later as in “adenguṇaḥ”. It is wrong to suggest that there should be a topic of symbols (which must first enumeate all the symbols) for Panini makes symbols even in the Eighth Book as in the rule “tasya paramamreḍitam” (8.1.2)- (when a word or a phrase is pronounced twice for chastisement or other purposes, the second of these words or phrases is technically called ‘amreḍita’). (According to the suggestion that there should be a separate topic denoting only the symbols the difficulty of such a procedure would be that) in that case these sutras relating to ‘samjña’ should have to be carried forth as ‘adhikara’ even to this 8th Book.
101. It has been said that the symbolised is that which is pronounced first and the symbol is pronounced later; but in the present case the word ‘vṛddhi’ is regarded as a symbol. Why then it is pronounced first? 156 Mahabhāṣya of Patanjali This is much ado over nothing in a wrong place. This does not differ from the practice of the people in the world. Thus a person does not know a cow and in giving him a cow one might hold the cow by the ears or the bones and say ‘do not worry, here is the cow’ and the man recognised the cow. (And there is a recognition). 102 (Thus the Bhāṣya says) it may be that previous teachers had already defined this relation between symbols and symbolised, between vṛddhi and at and aic. Here also the previous people had already established such a relation. By which previous people? By the teachers (i.e. the teachers that flourished before Panini). Since the relation has, already been established by the teachers, it becomes new to those to whom it is taught now. In the world also, when even an already established name is instructed for the first time to some people, it is done to them for the first time then. If others are instructed these symbolic names before, well and good, it will be instructed now again. (Here the Värttika intervenes and says)-since the words vṛddhi etc., which are already existent, are being associated and being now known as a samjñā or symbol, then in this regulation of the symbol and the symbolised we have the form of mutual dependence (itaretaraśraya). (This is explained by the Bhāṣya)-it is the symbolised that is associ- ated with the symbol or the name.103
102. The idea is that even without the use of a word which is a symbol, people have a knowledge of the symbol through the popular usage. Thus no one has defined a particular structure and said that it should be called a cow; but when one sees that a person does not know a cow and goes to the cow and shows it as ‘here is the cow’ So symbols and names are known to people through popular usage and for that any definition of symbols is unnecessary. This would naturally presuppose that the words ‘at’ and ’ec’ have been known to people by popular usage called the vṛddhi, and therefore, they might know them from that without any such definition. This idea is expressed by the Bhasya as follows.
103. Thus where it is enjoined that in the ‘parasmaipada’ when there is a ‘sic’ there will be ‘vṛddhi’ (sici vṛddhiḥ parasmaipadeṣu-7.2.1.) one may expect that when in the ‘parasmaipada’ there is ‘sic’, the word ‘vrddhi’ would be introduced there. The point is that when the ā-kāra, ai-kāra and au-kāra are known as vṛddhi, the name and the vṛddhi are the same and are known by ā-kāra, ai-kāra and au-kāra. One is not certain whether the words a-kāra, ai-kāra and au-kāra should be introduced or the word vṛddhi should be introduced. Third Ahnika 157 When the symbolised is known only as the object of the symbol in the determination of the word ‘vrddhi’ etc., we have the fallacy of mutual dependence, and may not, therefore, be able to say, which of these is definitely meant. What is the mutual dependence? ‘At’ and ‘aic’ should be known through its name, and it is by the name again that we know the ‘at’ and ‘aic’. This is mutual dependence. Any direction that involves mutual dependence, cannot be put into action. A boat that is tied with another, cannot save anybody. But it may be asked that many actions are seen performed even though they involve mutual dependence. Thus, the boat carries the chariot and the chariot carries the boat, but in such cases there is some other condition such as water or land. Thus, one may say that in the land the chariot carries a boat, in the water the boat carries the chariot. But what about a tripod?104 But there are at least grooves or nails to connect them. But here it is only a case of simple mutual dependence. (Here the Varttika intervenes and says)-such a relation can hold because of the eternity of sound. (The Bhāṣya runs as follows)-the difficulty is solved. How? Because the words are eternal. When the words are eternal, at, aic, etc. are given a particular name. The name does not create or constitute ‘adaic’.105 yadi tarhi nityaḥ śabdāḥ, kimartham śāstram? Bhāṣya. Värttika. kimartham śāstram iti cen nivartakatvāt siddham.
104. A tripod is three-legged depending on the other, but inspite of its mutual dependence, they can produce any action viz. that of carrying and holding things- so it is not always right to say that things mutually depending on each other cannot be fruitful in action.
105. The objection of mutual dependence consisted in this that the opponent supposed that it is only through the name that ‘at’ and ‘aic’ are known and it is only through ‘at’ and ‘aic’ that the name is known. Each is known by the other and none has any separate root of existence. This opens the door for the charge of mutual dependence. But if words are existing eternally by themselves then the charge of mutual dependence falls to the ground, for then each of them has a separate root of existence. Thus when we have the form ‘marsti’ from ‘mrj’, then we have ‘ār’ from ?’ and one may naturally be in doubt as to whether this is correct. In such circum- stances the sutra enjoining vṛddhi makes it clear to all that the form is correct. 158 Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Mahābhāṣya of Patanjali nivartakam śāstram. katham? mrjir as- māyaviseṣeṇa upadiṣṭaḥ. tasya sarvatra mrji- buddhiḥ prasaktā. tatrânena nivṛttiḥ kriyate, mrjer aknitsu pratyayeșu mrji-pra-sange marjiḥ sadhur bhavati iti. pratyekam vṛddhi-guna-samjñā-va- canam. (pratyekam vṛddhi-guna-samjña-vacanam kartavyam). pratyekam vṛddhi-guna-samjñē bhavata iti vaktavyam, kim prayojanam? samudaye ma bhūtām iti. anyatra sahavacanāt samudaye samjñā- prasangaḥ. anyatra saha-vacanāt samudaye vṛddhi- guṇa-samjñayor aprasangaḥ. yatra icchati saha-bhūtānām kāryam karoti tatra ‘saha’- grahaṇam. tad yatha-“saha supā”, (Pa. 2.1.4), “ubhe abhyastam saha” (Pā. 6.1.5) iti. pratyavayavam ca vakya-parisamapteḥ. pratyavayavam ca vakya-parisamaptir dṛśyate. tad yatha-devadatta-yajñadatta- viṣṇumitra bhojyantām iti. na ca ucyate- ‘pratyekam’ iti. pratyekam ca bhujiḥ parisamapyate. nanu câyam apy asti dṛṣtantaḥ-‘samu- dāye vākyaparisamāptir’ iti. tad yatha- ‘gargāḥ śatam dandyantām’ iti. arthinaś ca rājāno hiraṇyena bhavanti, na ca pratyekam daṇḍayanti. Third Ahnika saty etasmin dṛṣṭante yadi tatra saha- grahaṇam kriyate, ihâpi pratyekam iti 159 vaktavyam. atha tatrântarena saha-grahanam sahabhūtānām kāryam bhavati, ihâpi nârthaḥ pratyekam iti vacanena. If words were eternal, what is the good of writing this śāstra? (Here the Varttika intervenes)-if it is questioned why then the sastra? (The answer is that) the function of the śästra is satisfied by the restrictions that are imposed. (The Bhāṣya explains it thus)-the śastra is prohibitive. How? The root ‘mrji’ is offered without any adjective. So everywhere it appears to be the same ‘mrji’. The sastra makes the prohibitory rule that when suffixes that have no redundant ‘k’ and ’n’, follow in place of ‘mrji’ ‘marji’ will be correct (by the rule ‘mrjer-vṛddhiḥ’ 7.2.114). (Here the Varttika intervenes)-the ascription of the name ‘vṛddhi’ and ‘guna’ should be made severally. (Here the Bhāṣya runs)-the names ‘vṛddhi’ and ‘guna’ should apply severally so that it may not apply to the whole.106 (Here the Värttika says) in other places since the ascription is col- lective, here the name should not be used collectively. (The Bhāṣya says)-In other cases (in places other than the ascriptions of ‘guna-vṛddhi’) since the ascription is made exclusively collective, here the names ‘vṛddhi’ and’ guna’ should not be used collectively. Wherever the master desires that they should act collectively, he used the word ‘saha’ or ’together with’ Thus there are the rules “sahasupa” (2.1.4, i.e. ‘subanta’ word should be compounded with a ‘subanta’ word-here both the subantas are taken collectively and therefore the word ‘saha’ or together has been used), “ubhe abhyastam” saha (6.1.5; i.e. the two together should have the name ‘abhyasta’).107
106. In the case of the ascription of vṛddhi as in the sutra “mrjer vṛddhiḥ” (7.2.114), the vṛddhi should apply to particular vowels and not to the whole ‘mrji’. Thus, when it is said ’the day on which the father and mother died’ does not mean that the day should be such that on which both of them should die- but it is enough if any one of them dies.
107. The idea is that the master whenever he intends to mean that the samjñā should apply to two together, he uses the words ‘saha’ and ‘ubhe’ as in 2.1.4 and 6.1.5. So where that device is not employed, no samjñā should apply to two terms. The word ‘viddhi’ should not, therefore, in absence of such a device, mean both ‘at’ and ‘aic’, but only severally, ā-kāra, ai-kāra and au-kāra.160 Mahābhāṣya of Patanjali (Here the Värttika intervenes)-becasue of the fact that a proposition or vakya is complete in each of its parts. (The Bhāṣya runs)-a vākya or proposition is seen to be completed in each of its parts. Thus in the phrase ’let Devadatta, Yajñadatta and Viṣṇumitra be fed’ - it is not said that they should be severally fed for the eating is completed in each one of them. But there are instances where a proposition is completed collectively. Thus it is said- let the Gargas be fined hundred rupees (here the fine is not singly, but collectively); the kings are in need of gold, but they do not fine each person. In this example that would be so, if the word ‘saha’ or together were introduced (as in “saha supa”). Here also it is necessary to say that the ascription is made severally. But there also, without the ‘saha’, collective action is seen to take place. (In the rule “saha supa” there is a yoga-vibhāga or splitting up of the rule for yielding two meanings at the same time. The word ‘saha’ has been put in there in order to facilitate the yoga-vibhāga, otherwise it was not indispensable. This shows that even without the use of the word ‘saha’ collective action may take place). So here also it is not necessary to use the word ‘pratyekam’ or ‘severally’.108
108. Kaiyața says here ‘vṛddhi-samjñā tu lakṣye pratyekam ādaicām darśanať, i.e. the name vṛddhi should apply to ‘at’ and ‘aic’ severally and not jointly. The whole argument has been to prove that ‘vṛddhi’ is not the name jointly of ‘at’ and ‘aic’. The reasons are two-fold: One, the technical reason that whenever any name applies jointly to two terms, the collective appellation is definitely declared by the use of the word ‘saha’ or ’together’. This implies (i.e this is a jñāpaka) that whenever such indicatory ‘saha’ is not used, the name applies, severally to ‘at’ and ‘aic’. But this method is subject to criticism that indicatory ‘saha’ is not indispensable for appella- tions that are used jointly to indicate two terms. For this Patanjali introduces his second reason illustrative from popular usage, every action can apply only severally to two agents when he says, ‘väkyârtha’ he really means ‘padârtha’. The propositional meaning is not indeed completed severally, but in the actual world of facts, actions are generally completed severally by several agents. Thus when we say let the people eat, then there is actually no collective eating, each person eats and his eating completes his action. But there are, of course, actions like that of the enacting of a drama which is the result of a collective action (as noted by Kaiyața). Kaiyata also points out another jñāpaka to justify the contention that the word ‘vṛddhi’ applies severally to ‘at’ etc. His argument is as follows: There is the Panini’s rule, “prasthe’ vṛddham akarkyādinām” (6.2.87). This rule means when ‘prastha’ is the last member of a compound and the first word is a word which not being of the class of ‘karkyādi’ does not also contain a vṛddha-vowel gets an ‘udatta’ accent in the first pada. Now the next rule is “mālādinām ca’ (6.2.88) which means that ‘prastha’ being the last Bhāṣya. Vārttika. Third Ahnika 161 atha kimartham ākāras taparaḥ kriyate? ākārasya tapara-karaṇam savarṇârtham. Bhāṣya. ākārasya tapara-karaṇam kriyate. kim prayojanam? savarṇârtham. “taparas Värttika. Bhāṣya. Värttika. Bhāṣya. tatkalasya” (Pā.1.1.70) iti tatkālānām savarṇānām grahaṇam yatha syāt. keṣām? udattânudatta-svaritānām. kim ca kārṇam na syāt? bhedakatvāt svarasya. bhedaka udattādayaḥ. katham punar jñāyate ‘bhedaka udattādayaḥ’ iti? evam hi dṛśyate loke ya udatte kartavye anudattam karoti khaṇḍikopadhyāyas tasmai capeṭām dadāti ‘anyat tvam karoşi’ iti. asti prayojanam etat? kim tarhi iti. bhedakatvad guṇasya. bhedakatvad guṇasya iti vaktavyam. kim prayojanam? anunāsikyam nāma guņas tad-bhinnasyâpi grahaṇam yatha syāt. kim ca kāraṇam na syāt? “bhedakatvād guṇasya.” bhedakā guṇāḥ. katham punar jñāyate bedakā guņā iti? evam hi dṛśyate loke, eko’ yam ātmā udakam nāma, tasya guṇa-bhedad anyatvam bhavati- anyad idam śītam, anyad idam uṣṇam iti. member of a compound and the words ‘mālā, ‘śāla’, etc. being the first member, the first vowel should be with the ‘udatta’ accent. Now, a vṛddha-vowel is equivalent to vṛddhi, and if vṛddhi meant jointly ‘at’ and ‘aic’ there would be no meaning in making an exception in the case of ‘maladi’ which has only a-kara as the initial vowel. 162 Mahābhāṣya of Patanjali nanu ca bho abhedaka api guņā dṛśyante. tad yatha-devadatto mundyapi jatyapi sikhyapi svām ākhyām na jahāti. tathā bālo yuvā vṛddho vatso damyo balivarda iti. ubhayam idam gunesu uktam-bhedakāḥ, abhedaka iti. kim punar atra nyāyyam? abhedakā guṇā ity eva nyāyyam. kuta etat? yad ayam-“asthi-dadhi-sakthy- akṣṇām anan udattaḥ” (Pa. 7.1.75) ity udatta-grahaṇam karoti, taj-jñāpayaty ācāryo’ bhedakā guņā iti. yadi hi bhedakā guṇaḥ syur udattam evoccārayet. yadi tarhy abhedakā guṇāḥ, anudattāder antodattac ca yad ucyate tat svaritādeḥ svaritāntāc ca prāpnoti. naiṣa doṣaḥ. aśrīyamāno guņo bhedako bhavati. tad yatha- ‘śuklam ālabheta’, ‘kṛṣṇam ālabheta’. tatra yaḥ śukla ālab- dhavye kṛṣṇam alabhate, na hi tena yathoktam krtam bhavati. asandeharthas tarhi takāraḥ. ‘aic’ ity ucyamane sandehaḥ syat-kim imau ‘aicau’eva, ahosvid akāro’py atra nirdiśyata iti? sandeha-matram etad bhavati. sarva- sandeheṣu cedam upatiṣṭhate-“vākhyā- nato višesapratipattir na hi sandehad alakṣaṇam” iti. trayāṇām grahaṇam iti vyākhyāsyāmaḥ. anyatrâpi hy ayam evam- jātīyakeṣu sandeheşu na kamcid yatnam karoti. tad yatha-“automśasor” (Pa. 6.1.93) iti. Third Ahnika 163 idam tarhi prayojanam- antaryatas trimātra-caturmātrāṇām sthāninām trimātra- caturmātrā ādeśā mā bhuvanniti. khaṭvā indraḥ = khatvendraḥ, khaṭvā udakam = khaṭvodakam, khaṭvā iṣā = khatveṣā, khaṭvā ūḍhā = khatvoḍhā, khaṭvā elakā = tib khaṭvailakā, khaṭvā odanaḥ = khatvaudanaḥ, khaṭvā aitikāyanaḥ = khatvaitikāyanaḥ. khaṭvā aupagavaḥ = khatvaupagava iti. atha kriyamāṇe’pi takāre kasmād eva trimātra-caturmātrāṇām sthāninām trimātra- caturmātrā ādeśa na bhavanti ? “taparas tatkalasya” iti niyamāt. nanu taḥ paro yasmat so’yam taparaḥ. ’na’ ity aha. tad api paras taparaḥ. yadi tad api paras taparaḥ “rdor ap” (Pa.3.3.57) iti ihaiva syat- yavaḥ, stavaḥ, lavaḥ, pava ity atra na syat. naiṣa takāraḥ, kas tarhi? dakāraḥ. kim dakāre prayojanam? atha kim takāre? yady asandehârthas takāraḥ, dakāro’pi. atha mukha-sukhârthas takāraḥ dakāro’pi. (vrddhir ādaic 1). Why is a (in vṛddhir adaic) associated with ta in the end? (Here the Varttika runs)—the ākāra is associated with ta as following it in order that vowels of the same measure (savarna) may be included within it. (The Bhāṣya runs) the ākāra is made with ta following it. What for? For including the savarnas. By the rule “taparas tatkalasya” (1.1.70); a vowel is followed by ta or which follows ta will include vowels of the same measure and same time. Thus in the rule, “ato bhisa ais” (7.1.9); i.e in the case of a word ending in ‘at’, ‘bhis’ will be changed into ‘ais’; thus nara+bhis is equal to ’narais’. This ‘at’ is one in which ‘a’ is followed by ’ta’ and for that reason it only includes within it the simple ‘a’ - nasal or 164 Mahābhāṣya of Patanjali unnasal, but not the long variety as a or the pluta variety which takes a longer time in pronunciation. Thus, when ‘bhis’ follows ‘mala’, the ‘bhis’ is not changed into ‘ais’ (and we have mālābhiḥ), all the vowels of the same measure and time are accepted. Which are they? Udātta, anudatta and svarita. (Though the association of ta excludes the long vowels, yet it includes the vowels of the same time of pronunciation though they may differ in accent). Why should it not be so? (Here the Värttika intervenes)- because of the variation of the accent. (The Bhāṣya runs) - the accents like udatta etc. distinguish the same vowel. How do you know that accents produce such a variation? So it is seen in the world, that he, who pronounces a word with udatta accent with anudātta, receives a slap on his face from his teacher saying, “You are mispronouncing it”. Is there any necessity? What is that? They produce qualitative difference. In the case of the nasal also there is a qualititative difference and this should also be included within the term ‘savarna’ referred to by the tapara vowel. Becasue it is the gunas that differentiate one from another. The gunas or qualities are separators. How do you know that the gunas are separators? So it is seen in the world that there is one substance water which on account of its difference in quality appears as different. This is cold, is something different, and this is not and is something different. But there are qualities which do not differentiate. Thus the same Devadatta with a shaven head, or with a tuft of hair, or with long-plaited hair does not lose his own name. So also the calf, the heifer and the bull are the same though they differ in age as too young, young and old. Both have been affirmed of the qualities. Some say they are separators, some say they are not. What is correct? It is correct to say they are not separators. Why so? That this “asthi-dadhi-sakthy-akṣṇām anan udattaḥ” (7.1.75, the augment anan with an udatta accent is prescribed to come after the words ‘asthi’ (bone), ‘dadhi’ (curd), sakthi (strength) and “akṣi’ (eye) in case-endings beginning with a vowel from the third case-ending onwards to all the other case-endings). Here the word ‘udātta’ has been specifically mentioned in this sūtra; by this the master indicates that the gunas i.e. the qualities (of accent etc.) cannot distinguish or differentiate. If these gunas could differentiate (instead of mentioning the word ‘udātta’, the master would have pronounced the word as ‘udatta. 109
109. The point raised is as follows: whenever a vowel is pronounced, it has to be pronounced with an accent. The question is whether such an accentuation has produced any such difference in the vowel that it can be distinguished from other Third Ahnika 165 If then the qualities cannot distinguish, then whatever has been prescribed about words having the anudatta accent or with an antodatta accent, should apply also to those that have an initial svarita or concluding svarita accent. This is no criticism. For when definitely declared, the qualities can distinguish. Thus, when it is advised that one should vowels or from the same vowel of other varieties. Thus when one says ā, he has distinguished it from its other variety a; so when one has pronounced i, he has distinguished it from its short variety i; but one pronounces an udatta accent and that accentuation does not so distinguish the vowel as to make it a recognisable variety. For this reason instead of pronouncing the anan in the rule 7.1.75 with an udatta accent, the master uses the word ‘udatta’ for definitely distinguishing this particular ‘anan’. From this indication it appears that accentual differentiation does not differ- entiate the individuality of a vowel. Another point is raised here. There is a rule “hano vadha lini’ (2.4.42), i.e. the root ‘han’ becomes ‘vadha’ when ‘ardhadhātuka’ suffixes of ’lin’ follow, the augment ‘vadha’ ends with the vowel a but this a is elided. But though it is elided, it is regarded as existent for purposes of ‘guna-vṛddhi’. Thus there is no optional vṛddhi here by the rule “ato haläder laghoh” (7.2.7) which means that when a ‘sic’ with an ‘it agama’ follows in parasmaipada, then a root beginning with a consonant and ending in akāra, should have vṛddhi optionally-as ‘akanit’, ‘akāņit’. For the purpose of this rule by the ‘sthänivadbhavaḥ’- maxim the agama ‘vadha’ is regarded as ending in a, though its a is elided. The maxim of ‘sthanivadbhava’ is derived from the rule 1.1.56 which declares that ‘adeśa’ should be regarded as having the same functions as that in place of which it has occurred. Now an accent of the ‘sthanivadbhava’ rule, ‘vadha’ means associated with an ‘anudatta’ accent and the ‘agama’ of it being prohibited thereby, there occurs a difficulty that no ‘it’ augment can be advised, and this would render the derivation of ‘akāņit’ impossible. For this purpose it has been said that here the ‘vadha’ should be regarded as udatta. We know that by the rule “ekaca upadese anudattāt” (7.2.10, i.e. a root having one vowel only in the original list should be regarded as anudatta and there should be no it-augment after that), there ought not to be any ‘it-agama’ after ‘han’ which is a one-vowel verb and therefore anudatta. “Vadha’ being an ‘adeśa’ in place of ‘han’ is anudatta’ by the ‘sthanivadbhava’ maxim, therefore it could not get ‘if’ after it by the present rule 7.2.19; but still without it we cannot have the form ‘akanit’. Therefore ‘vadha’ has to be regarded as ‘udatta’ by special declaration. Here also we find that when a vowel is pronounced as anudatta, it cannot differentiate itself as such, and when it is to be rendered udatta, it has to be specifically mentioned. The same rule applies to the nasal, i.e. nasality also cannot distinguish a vowel by itself, but has to be specifically mentioned. Thus by the rule “uñaḥ” (1.1.17) and “"(1.1.18) it is declared that the vowels included in ‘un’ should be regarded as being nasal. Thus, nasality cannot differentiate a vowel merely by pronounciation. It has to be declared separately. So also it is asked in the rule 7.2.10- which mono-vowelic root is anudatta? The reply is - those that have been so pronounced by the master. How to know which have been so pronounced by the master? These have been separately shown or collected in the verse list (karika) of roots to which ‘anit’ is prohibited. 166 Mahābhāṣya of Patañjali immolate the white cow or that one should immolate the black cow, one behaves otherwise, that is when the white is to be killed, he kills the black, he does not behave in the proper manner. The ta-kara is thus inserted for removing the doubt. In the case of ‘aic’ there is the doubt whether ‘aic’ means only ai and au or whether it includes the a-sound also. This is a mere doubt. In all cases of doubt the principle is that the special meaning can be obtained only by the interpretation and a sutra should not be left as meaningless because there is any doubt. In other places also in similar cases of doubt no further effort is made, as in the case of the rule “automśasoḥ” (6.1.93, i.e. when ‘am’ and ‘sas’ follow a word ending in ‘au’, should have the vowel a in place of au and the vowel following it - here also there may be a doubt as to whether ‘au’ is composed of a and au, but no attempt has been made in the sutras to remove the doubt. For this reason it should be noted that for such cases of doubt one should look to the traditional interpretation only.110

110. It is thus seen that the insertion of ta cannot be for the removal of any doubt. We have seen that the first alternative interpretation offered was that ta has been associated with a in order that a may include vowels of the same measure, i.e. vowels having the same time, but differing in accentuation and nasalisation. The discussion proves that accentuation and nasalisation are not distinctive marks of vowels and that in case where such distinctions are intended, it has to be specifically mentioned. This alternative explanation therefore for the association of ta fails. The other alternative interpretation is that ta stands as barrier between aand aic and it shows that the aic does not contain within it a. But the ta could not have been associated for the removal of such doubts, for in similar cases as in 6.1.33 there might have been a similar doubt, but the master is quite indifferent to this. Such doubts are to be removed through traditional interpretations and now a third alternative is proposed as the Bhāṣya, that follows, would show. Third Ahnika 167 etikāyana), au in Aupagava (the son of upagava). The point raised here in this- an ‘aic’ would mean ai and au and these may be of three measures or four measures, but the putting of aic after the ’ta’ or ‘a’ shows that in this context the aic after should be of two mātrās, and not of three or four, and incidentally this is to be regarded as an exception that though generally an augment vowel has the same measure as that in place of which it occurs, yet in these cases an augment in place of a vowel with three or four measures should not be of three or four measures). Well then even when takara is associated, why should not an agama occurring in the place of a vowel having three or four mātrās should not have three or four matras by the rule “taparas tat-kalasya” (1.1.70). Now what does ’tapara’ mean? That which has ta after it. No, that is not the only meaning. It also means that which is preceded by ta or that which follows ta. If it is to mean that which follows ta, then by the rule “Fdor ap” (3.3.57, a root ending in ror u should have the suffix ‘ap’) we should have only ‘yavah’ (barley), stavah (hymn), but not ’lavah’ and ‘pavaḥ’.111 Here the vowel short u is after la and not ta. What is the necessity of da? Well then, what is the necessity of ta? If ta is for the removal of doubts, so may also the da. If ta is for the facility of pronounciation, so may also be da. Pāṇini. iko guṇavṛddhi 1.1.3. The ‘guna’ and the ‘vrddhi’ will take place in place of ‘ik’. Here ‘ikah’ in the sixth case and the meaning of the sixth case in a sutra is given in the rule ‘şaşthi sthaneyoga’ (1.1.9). That sutra as well as the present one are ‘paribhāṣā sūtras or general maxims, ‘guna’ and ‘vṛddhi’ can take place only with reference to ‘ik’, (i, u, r, !) and no other word. The rule 1.1.49 determining the meaning of the sixth case means that in place when something is directed to take place with reference to words in the sixth case, then the meaning will be that whatever is directed to take place will occur in place of that which has the sixth case. Therefore when ‘guna’ and ‘vṛddhi’ are directed they are to occur in place of i, u, r, 1.
111. The words ’lavah’ and ‘pavaḥ’ have to be formed from the roots ’lu’ and ‘pu’ respectively. Now in “Fdorap” the u is preceded by ’ta’ and therefore it ought to include only the vowels of the same measure, i.e. vowels of the short variety and therefore we can have by this rule the ‘ap’ suffix to verbs ending in short u, but not verbs ending long u. How can we then have words like ’lava’ and ‘pava’ which are formed by ‘ap’ being added to roots ending in long vowels? 168 Bhāṣya. Värttika. Bhāṣya. Mahābhāṣya of Patanjali ig-grahanam kimartham? ig-grahaṇam at-sandhyakṣara-vyañ- jana-nivṛttyartham. ig-grahanam kriyate. kim prayojanam? ākāra-nivṛttyartham sandhyakṣara-ni- vṛttyartham vyañjana-nivṛttyartham ca. ākāra-nivṛttyartham tāvat-‘yātā, ‘vātā. ākārasya guṇaḥ prapnoti. ig-grahaṇān na bhavati. sandhyakṣara-nivṛttyartham-glāyati’, ‘mlāyati’. sandhyakṣarasya gunaḥ prāpnoti. ig-grahaṇān na bhavati. vyañjana-nivṛttyartham-‘umbhita’, ‘umbhitum’, ‘umbhitavyam’. vyañjanasya guṇaḥ prāpnoti. ig-grahaṇān na bhavati. ākāra-nivṛttyarthena tavan nârthaḥ. ācārya-pravṛttir jñāpayati-’nākārasya guņo bhavati’ iti yad ayam-“ato’nupasarge kah” iti kakāram anubandham karoti. katham kṛtvā jñāpakam? kit-karana etat prayojanam ‘kiti’ ityākāralopo yatha syat.yadi ca ākārasya guṇaḥ syāt kitkaraṇam anarthakam syat. gune kṛte dvayor akarayoḥ pararūpeṇa siddham rūpam syat - ‘godaḥ’, ‘kambala- dah’ iti. paśyati tv ācāryo-’nākārasya guno bhavati’ iti, tataḥ kakāram anubandham karoti. — sandhyakṣara-nivṛittyarthenapi nārthaḥ. upadeśa-samarthyāt sandhyakṣarasya guno na bhavati. vyañjana-nivṛttyarthenâpi nārthaḥ, ācārya- pravṛttir jñāpayati- ’na vyañjanasya guno Third Ahnika bhavati’ iti, yad ayam janer dam śāsti. katham kṛtvā jñāpakam? ḍit-karaṇa etat prayojanam-‘diti’ iti ți-lopo yathā syāt. yadi vyañjanasya gunah syad dit-karaṇam anarthakam syat. gune kṛte trayāṇām akārāṇām para-rupeņa siddham rūpam syat ‘upasarajaḥ’, ‘mandurajaḥ’ iti. paśyati tv ācāryo-’na vyañjanasya guno bhavati’ iti. tato janer ḍam śāsti. 169 naitāni santi jñāpakāni. yat tavad ucyate- ‘kit-karaṇam jñāpakam nākārasya guno bhavati’ iti. uttarartham etat syāt-“tunda- sokayoḥ parimṛjapanudoḥ” iti. yat tarhi “gapoṣṭak” ity ananyārtham kakāram anu- bandham karoti. yad api ucyate- ‘upadeśa-samarthyāt sandhyakṣarasya guno na bhavati’ iti. yadi yad yat sandhyakṣarasya prāpnoti tat tad upadeśa-samarthyād badhyate, āyādayo’ pi tarhi na prāpnuvanti. naiṣa doṣaḥ. yam vidhim praty upadeśo’- narthakaḥ sa vidhir badhyate. yasya tu vidher nimittam eva, nâsau badhyate. guṇam ca praty upadeso’narthakaḥ. āyādinām punar nimittam eva. yad apy ucyate-janer da-vacanam jñāpakam, na vyañjanasya guņo bhavati’ iti. siddhe vidhir arabhyamāņo jñāpakārtho bhavati. na ca janer gunena siddhyati. kuto hy etat janer guna ucyamāno’kāro bha- vati, na punar ekāro vā syād okāro vā’ iti? antaryato’rdhamātrikasya vyañjanasya mātriko’kāro bhaviṣyati.170 Mahābhāṣya of Patanjali evam apy anunāsikaḥ prāpnoti. pararūpeṇa śuddho bhaviṣyati. evam tarhi gamer apy ayam do vaktavyaḥ. games ca guṇa ucyamāna āntaryata okāraḥ prāpnoti. tasmad ig-grahanam kartavyam. Why does the sutra speak of ‘ik’? (Here the Varttika says)- ‘ik’ has been taken to secure that no gunavṛddhi can occur in place of ‘at’, a euphonic vowel or a consonant. (The Bhāṣya runs)-if the ‘ik’ is taken what for? For prohibiting a, an euphonic vowel or consonant. For prohibiting ā, such as ‘yātā, ‘vātā’, the guna should have taken place in place of a, but since it is said that guna and vṛddhi can occur only in place of ‘ik’ no guna occurs in place of ‘a’.112 For prohibiting the euphonic vowels-‘glāyati’, ‘mlayati’ the euphonic vowels of ‘glai’ and ‘mlai’ could have guna in the above cases; but, since there can be no guna to any vowel i, u, r, !, that does not occur. For prohibiting the consonants as in ‘umbhita’, ‘umbhitum’, and ‘umbhitavyam’ guna could have occurred in the place of the consonant, but it does not, because of speaking of ‘ik’ only in the sutra. In opposing the above preliminary thesis, as to the explanation as to why ‘ik’ has been spoken of in the sutra, the Bhāṣya runs as follows:- It is not for the sake of prohibiting ākāra. The intention of the master shows that guna cannot occur in place of a, because in the rule “ato’nupasarge kah” (3.2.3; i.e. roots ending in a having no upasarga and having a karma to it will take the suffix ‘ka’ in place of suffix ‘an’). How does it indicate anything? Because the suffix has been made to contain a ‘ka”? How does it then serve the purpose? So that by the rule “ato lopaḥ iti ca” (6.4.64; when suffixes with redundant ‘ka’ and ’n’ follow, the a of the root should be elided) it should not have been arranged that for securing the elision of the a, a suffix with a ka-redundant is to follow. If any guna could occur in place of a, then the elision through a suffix with redundant ‘ka’ would be unnecessary, for once the ā, is changed into a by guna, the a of the root and of the suffix could very well get changed into ‘akara’ (like words of the ‘karkandhu’
112. If ardhadhātuka suffix follows, there is general direction for vṛddhi. There is a special rule “ato lopaḥ” (6.4.48) which directs the elision of a, the short sound. The long sound a of the root ‘ya’ or ‘va’ could then have undergone guna or vṛddhi as the case may have been, had it not been for the fact that guna cannot occur in place of any other vowel than i, u, Į, Į. Third Ahnika 171 list) and we could have the forms ‘godah’ and ‘kambaladaḥ’. By this the master indicates that there should be no guna in place of a; for this reason he prescribes a suffix with a redundant ‘ka’ for the elision of a of the roots ending in ā. Ik has not also been taken for prohibiting guna to euphonic vowels. By force of the original introduction the euphonic vowels would not suffer any guna. (As Nāgeśa says the ‘in’ and ‘aic’ have been definitely declared to retain the specific character and therefore they cannot have guna in their case). It is not also for prohibiting the consonants, for the manner of the master shows that in place of the consonants there cannot be any guna. This is done when he advises da-suffix to ‘jana’. How does this indicate it? By virtue of the redundant ‘da’ by the rule “teh” (6.4.143). It has been prescribed that the ’ti’ of a ‘bha’ should be elided when a suffix with a da- redundant should follow.113 If guna could occur in place of a consonant, the provision of the suffix with the da-redundant will be unnecessary. For, if guna was advised then the three ‘a’s could have combined to form an a and that would have served the purpose. In forming such words as ‘upasarajah’ (a calf born at the first pregnancy of the cow), ‘mandurajaḥ’ (insect born in a mat), the master wishes to indicate that no guna should occur in place of consonants. For this purpose he advises the rule “janer dah” (3.2.97; when a word in the 7th case precedes, then the root ‘jana’ should have ‘da’ as its suffix). These are not indicatory illustrations-that it is said that by advising a suffix with redundant-k it is to indicate that no guna will occur in place of a. It has a future utility in the rule “tunda- śokayoḥ parimrjâpanudoḥ” (3.2.5; when the words ’tunda’ and ‘soka’ in the accusative case are associated with the roots ‘parimrja’ and ‘apanuda’, then they will take the ka-suffix).114 There is however the rule “gapoștak” (3.2.8, i.e the root ‘ga’ and ‘pā’ with an accusative case in association with it and bearing no‘upasarga’ will take the suffix ’tak’, e.g. ‘samagaḥ’, ‘surapaḥ’). Where the redundant
113. ‘Ti’ means a part of a word commencing with the last vowel.
114. ‘Tunda’ means a protuberent belly and ’tunda-parimṛja’ is one who is in the habit of strolling his belly. The k-suffix was introduced in the above rule 3.2.5 and not so much for ordinary ākārānta roots. If it was introduced only for äkäränta roots, then the contention that it was indicatory to the fact that Panini had already shown that there will be no guna in place of ā. 172 Mahābhāṣya of Patanjali ka of the suffix may be regarded as demonstrating the same purpose (viz. that Panini indicates that there will no guna in place of a). It has been said that by force of the original instructions, the guna, does not occur in place of euphonic vowels. (To this the reply is) that if by force of the original instruction, such a rule applies to the euphonic vowels, then it ought to apply in the case of ‘ay’, av’, etc., (i.e. ai, au, etc. are always to remain as ai, au and ought not to be transformed into ay, av). This is no criticism. The rule with reference to which the original instructions would be regarded meaningless, would be negated. But where it becomes the occasion of the rules, it cannot be negated. With reference to guna the original instructions become mean- ingless, but with reference to ‘ay’, etc. it is the occasion.115 Why so? Where is it said ‘jana’ will have guna? We do not know whether there would be a-kāra, e-kāra or o-kāra. Well then (the reply is) on account of the rule of proximity in place of the consonant of half measure we may have the a of a single measure. Well, again this a may be nasal one. Even if it be a nasal one by its assumption of the form of the succeeding vowel (pararupa), the form will justify it.116 Well, then the ‘da’ has been advised also to the rule ‘gami’, that there should be guna in the place of ‘gami’ then ‘okara’ will be more pro- mixmate and therefore the desired form will not be possible. For this reason it is necessary that sutra should speak of ‘ik’.117
115. It is on the occasion of ‘ai’ that it becomes changed into ‘ay’ and therefore this rule cannot be negated. It is urged that the rule “janer daḥ"is indicatory that guna cannot occur in the place of a consonant. When the rule is already known there is a further enumeration of it. It would mean that such an enumeration must have some special indication. It is not enough here to say that the root ‘jana’ should have guna.
116. The rule of proximity is “sthane’ntaratamaḥ” (1.1.50) when there is a doubt as to which of the many possible ones any particular one is to take place on a particular occasion, the rule of proximity i.e., that which is nearest in the articulating effort (to that in whose place it is to occur) will take the role of precedence.
117. In actuality there is no such rule as “gamer daḥ”, but there is a rule “saptamyām janer daḥ” (3.2.97) and thereafter the other rule “anyeṣvapi drsyate” (3.2.101) and another Värttika rule ‘anyatrâpi dṛśyate’ in association with the sūtra “antātyantādhva-dūra-para-sarvananteṣu daḥ”(3.2.48) by which you may have ‘da’ to ‘gama’. The substance of the discussion then is that by prescribing a suffix with a redundant da’ Pāņini does not indicate any particualr’intention as suggested by the opponent. Bhāṣya. Vārttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya Värttika. Bhāṣya. Third Ahnika 173 yadi ig-grahanam kriyate, ‘dyauḥ’, ‘panthāḥ’, ‘sah’, ‘imam’ iti, ete api ikaḥ prāpnuvanti. samjñayā vidhāne niyamaḥ. samjñaya ye vidhiyante teşu niyamaḥ. kim vaktavyam etat? nahi. katham anucyamānam gamsyate? guna-vṛddhi-grahaṇa-samarthyāt. katham punar antarena guna-vṛddhi- grahanam iko guna-vṛddhi syātām? prakrtam guna-vṛddhi-grahanam anuvar- tate. kva prakṛtam? “vṛddhir ādaij”, “aden gunah” iti. yadi tad anuvartate ‘adenguno vṛddhis ca ity adenām vṛddhi-samjñā’pi präpnoti. sambandham anuvartisyate. sambandham anuvartisyate “vrddhir ādaic”, “aden gunah” iti vṛddhir ādaic. tataḥ-“iko guna-vṛddhi’’ iti. guna-vṛddhi-grahaṇam anuvartate, ādaij-aden-grahanam nivṛttam. 66 maṇḍukagatayo’dhikārāḥ. athavā maṇḍūkagatayo’dhikārāḥ. yathā maṇḍūkā utplutyotplutya gacchanti tādvad adhikārāḥ. ekayogo vā. athava ekayogaḥ kariṣyate-“vrddhir adaijaden guṇaḥ”. tataḥ “iko guṇavṛddhi’’ iti. na caikayoge’nuvṛttir bhavati. athava “anyavacanāc cakārâkaraṇāc ca prakṛtâpavādo vijñāyate, yathotsargeņa prasaktasyâpavādo badhako bhavati.” 174 Värttika. Bhāṣya. Mahābhāṣya of Patanjali anyasyāḥ samjñāyā vacanāc cakārasya cânukarṣaṇārthasya akaraṇāt prakṛtāyā vṛddhi-samjñāyā guṇasamjñā badhikā bha- viṣyati, yathā utsargeņa prasaktasyâpavādo badhako bhavati. athavā vakṣyaty etat- “anuvartante ca nāma vidhayaḥ. na cânuvartanād eva bhavanti. kim tarhi? yatnad bhavanti” iti. athava ubhayam nivṛttam. tad apekşi- ṣyamahe. kim punar ayam alo’ntyaśeṣaḥ, āhosvid alo’ntyāpavādaḥ? katham ca ayam tac-cheṣaḥ syāt, katham vā tad-apavādaḥ? yady ekam vākyam-tac cedam ca, ‘alo’- ntyasya vidhayo bhavanti’. “iko guṇa- vṛddhi, alo’ntyasya” iti. tato’yam tac-cheṣaḥ. atha nānāvākyam-taccedam ca, alo’ntyasya vidhayo bhavanti. “iko guna-vṛddhi, an- tyasya cânantasya ca” iti. tato’yam tada- pavādaḥ. kas câtra viseṣaḥ? vṛddhi-guṇāv alo’ntyasyeti cen midi- mrji-puganta-laghupadharcchi-dṛśi- kṣipra- kṣudreṣv ig-grahanam. vṛddhi-gunau alon’tyasyeti cen midi-mrji- puganta-laghûpadharcchi-dṛśi-kṣipra- kṣudreṣv ig-grahaṇam kartavyam. “mider guṇaḥ” (Pā.7.3.82), “ika iti vaktavyam” anantyatvad hi na prāpnoti. “mrjer vṛddhih” (Pa. 7.2.114) “ika iti Värttika. Bhāṣya. Third Ahnika 175 vaktavyam”, anantyatvād hi na prāpnoti. “puganta-laghūpadhasya guṇaḥ” (Pa. 7.3.86), “ika iti vaktavyam” anantyatvad hi na prāpnoti. nan rccher liți gunah, “ika iti vaktavym” anantyatvad hi na prāpnoti. “rdrso ’ni gunah” (Pa. 7.4.16), “ika iti vaktavyam”, anantyatvad hi na prāpnoti. kṣipra-kṣudrayor guṇaḥ, “ika iti vaktavyam”, anantyatvad hi na prāpnoti. sarvādeśaprasangaś cânigantasya. sarvādeśaś ca guno’ nigantasya prāpnoti. ‘yātā’, ‘vātā’. kim kārṇam? “alo’ntyasya” iti saṣṭhi caiva hy antyam ikam upasamkrāntā. angasya iti ca sthānaṣaṣṭhi. tad yad idānīm anigantam angam tasya guṇaḥ sarvādeśaḥ prāpnoti. naișa doṣaḥ, yathaiva hy alo’ntyasya iti şaşṭhi antyam ikam upasamkrāntā, evam angasya ity api sthana-ṣaṣṭhi. tad yad idānim anigantam angam tatra sasthy eva nâsti kuto guṇaḥ, kutaḥ sarvādeśaḥ? evam tarhi nâyam dosa-samuccayaḥ. kim tarhi? pūrvāpekṣo’yam doṣaḥ. hy arthe câyam ‘cah’ pathitaḥ-“midi-mrji-puganta- laghupadha-ṛcchi-drśi-kṣipra-kṣudreşv ig-grahṇam sarvadeśa-prasango hy aniganta- sya” iti. mider gunah, ‘ika’ iti vacanād an- tyasya na, “alo’ ntyasya” iti vacanād iko na. ucyate ca gunaḥ. sa sarvādeśaḥ prāpnoti, evam sarvatra. astu tarhi tadapavādaḥ. 176 Värttika. Bhāṣya. Mahābhāṣya of Patanjali inmātrasyeti cej jusisārvadhātukār- dhadhātuka-hrasvadyor guneṣvanantya- pratiṣedhaḥ. in-mātrasya iti cej jusi-sārvadhātukārdha- dhātuka-hrasvādyor guneṣu anantya- pratiṣedho vaktavyaḥ. jusi gunah-sa yatheha bhavati- ‘ajuhavuḥ’, ‘abibhayuḥ’, iti. evam- ‘anenijuḥ’, ‘paryaveviṣuḥ’, atrâpi prāpnoti. sārvadhātukārdhadhātukayor guṇaḥ-sa yatheha bhavati- ‘karta’, ‘harta’, ’nayati’, ’tarati’ iti. evam-‘ihita’, ‘ihitum’, ‘ihita- vyam’ ity atrâpi prāpnoti. hrasvasya gunah-sa yatheha bhavati- ‘he agne’, ‘he vayo’ iti. evam ‘he agnicit’, ‘he somasut’ ityatrâpi präpnoti. jasi gunah-sa yatheha bhavati- ‘agnayo vāyavaḥ’ iti. evam ‘agnicitaḥ somasutaḥ’ ity atrâpi prāpnoti. ṛtoni-sarvanāmasthanayor guṇaḥ-sa yatheha bhavati- ‘kartari, kartārau, kartāraḥ’ iti, evam ‘sukṛti, sukṛtau, sukṛtah’ ity atrâpi prāpnoti. gherniti guṇaḥ — sa yatheha bhavati- ‘agnaye’, ‘vāyave’ iti, evam ‘agnicite so- masute’ ity atrâpi prāpnoti. or gunah-sa yatheha bhavati- ‘bābhravyo māṇḍavyah’ iti. evam ‘suśrut- E o sauśrutaḥ’ ity atrâpi prāpnoti. Värttika. naisa doṣaḥ. puganta-laghûpadha-grahaṇam anan- tya-niyamārtham. Bhāṣya. Vārttika. Bhāṣya. Third Ahnika puganta-laghûpadha-grahaṇam anantya- niyamārtham bhaviṣyati-puganta-laghû- padhasyaiva anantasya, nânyasya anantya- sya iti. 177 prakṛtasyaiva niyamaḥ syāt. kim ca prakṛtam? “sārvadhātukārdhadhātukayoḥ” iti. tena bhaved iha niyaman na syat- ‘ihita’, ‘ihitum’, ‘ihitavyam’ iti. hrasvādyor gunas tvaniyataḥ, so’nantyasyâpi präpnoti. athâpy evam niyamaḥ syāt- puganta- laghûpadhasya sarvadhātukârdhadhātukayor eveti. evam api sārvadhātukârdhadhātu- kayor guno❜niyataḥ so’nantyasyâpi prāpnoti ‘ihita’, ihitum’, ‘ihitavyam’ iti.

118. We had first the rule “vrddhir ādaic”, then the rule ‘adengunah’; the words ‘vṛddhi’ and ‘guna’ would have been easily transmitted to the next sutra and in that case it was sufficient, if that sutra was read as merely ‘ikah’ in that case what is the meaning that the word ‘guna-vṛddhi’ should again be read with ‘ikah’. The interpre- tation is that those of which the names or symbols have been defined as guna and vṛddhi, will only take place in place of ik.
119. It is said here that since the rule runs as “vṛddhir ādaic” the continuity can only be of vṛddhi as associated with ādaic. Third Ahnika 187 (The Varttika says) or rather the three are connected in one. (The Bhāṣya runs) - or rather the three may be united together- “vṛddhir ādaic”, “adengunah” and “iko gunavṛddhi”. If the three are taken together as one, there cannot be continuity of ‘adhikara’.120 (The Bhāṣya further says)-in such a case there being no ca-kāra and to connect the two, the latter may be taken as the exception to the former just as what is prescribed by general rule is nullified by special rule acting as exception. There will be no ca-kāra from another samjñā. There cannot be a joining together of the two and it would as if the original instruction of vṛddhi is contradicted by the nomenclature of guna just as a general rule is nullified by a special rule. Or rather, as we shall say later on, it is the prescriptions that are continued. But merely by being continued nothing can be regarded as a prescription: for that one has to make a special effort. 121 Or rather both the words guna and vṛddhi may be regarded as having ceased (to operate in the next rule ‘iko gunavṛddhi’). So the word ‘ikaḥ’
120. The solution proposed was that the three sutras might be regarded as forming one whole, but there can be no internal unity of meaning between the sutras which can hold them together as forming parts of a complex proposition: for, “vṛddhir ādaic” cannot be united with ‘adengunah’ and neither of them can be united with “iko gunavṛddhi”. For this reason the proposed solution cannot work.
121. There is a rule “dhanyanam bhavane kṣetre khan” (5.2.1). The suffix ‘khan’ is added to special kinds of seeds to denote the field in which they are grown. After this in the same meaning ’thak’ and ‘yat’ are prescribed respectively to ‘vṛihi-śāli’ and ‘yava’, ‘yavaka’, ‘sasthika’. The fourth rule is “vibhāṣā tila-masa-umā-bhanga- anubhyah”-to the words ’tila’, ‘masa’, ‘uma’, ‘bhanga’ and ‘anu’ there will be optionally the ‘yat’- suffix to denote the field in which they are grown. Now, ’tila’ and ‘masa’ are seeds and they might have ‘khan’ by the rule “dhanyanām bhavane” etc. (5.2.1); but ‘umā’ and ‘bhanga’ not being grain seeds (‘uma’ means lincid and ‘bhanga’ means hemp) there could not be any ‘khañ’ to it. But here the ‘khan’ continues from the rule 5.2.1. But in the case of the previous two sūtras 5.2.2. and 5.2.3. where ‘dhak’ and ‘yat’ are prescribed for vrhi-sali’ and ‘yava-yavaka-şaşthika’ the ‘khan’ has no continuity, for there has been no effort on the part of the writer of sutra to carry on the continuity. But in “vibhāṣā tila-masa-uma-bhanga-anubhyah” (5.2.4) the application of the term vibhāṣā (i.e. optionally) is an explicit hint that ‘khan’ should continue in this sutra. So here also the word ‘vṛddhi’ would have continued to the rule ‘adengunah’ if there was any hint on the part of the master. But in the absence of that the word is not continued. 66 188 Mahābhāṣya of Patanjali requires the words guna-vṛddhi to complete its meaning (as there is no continuity of gunavṛddhi from the previous rules).122 (The Bhāṣya introduces now another question)-how are we to interpret the sixth case-ending in ‘ikah’. There is a rule “alo’ntyasya” (1.1.52)-when any augment is prescribed with reference to a word in the genitive case, that augment is to take the place of the last vowel or consonant of the word in the genitive case. Now, if this rule is applied there is the occasion that gunavṛddhi should take place in the place of the last consonant ‘ka’ of ‘ika’ which is in the genitive case in the rule ‘iko guna- vṛddhi’. Thus the Bhāṣya says-should this be regarded as taking place in place of the last consonant (of ‘ik’ by the rule “alo’ntyasya”) or should it be regarded as an exception to the rule “alo’ntyasya”. 123 (The Bhāṣya further continues)-how should it be that it refer to the last member and how should it be regarded as an exception? If the two form one sentence that and this, the prescription regarding genitive case should be in the place of the last alphabet of the word in the genitive case. ik should have gunavṛddhi with reference to the last alphabet of the word. In that case it becomes the best one.124 If the two are two different propositions- that and this, then the prescription takes place according to “alo’ntyasya”. There is the prescrip- tion that gunavṛddhi should take in place of ‘ik’ and this is supplemented
122. The word ‘apekṣiṣyamahe’ means that the word ‘ikah’ having been left alone is in await for some word to render itself significant.
123. The meaning of sasthi has been interpreted in two ways, as sthana-ṣasthi and avayava-yoga-saṣthi. The sthana-şaşthi is by the rule “şaşthi sthaneyoga” (1.1.43) and the special method of the application of this rule is determined by the rule “alo’ntyasya” (1.1.52). If sthana-șṣasthi cannot be applied here, then this must be in exception to the general rule which is always that of sthana-ṣasthi and then this would be avayava-şaşthi. This question is raised in the Bhāṣya, viz., should we regard it as a general case of sthana-ṣasthi or should it be an exception?
124. There is no difficulty in the above construction. Here the Bhāṣya attempts to explain in what particular way the prescription of gunavṛddhi should refer to the last alphabet of ‘ik’. He says this can be so if the two sutras “iko guṇavṛddhi”, and “alon’tyasya” be regarded as forming one compound or complex sentence for in that case alone the meaning of the phrase “iko gunavṛddhi” can be constructed in association with the phrase “alon ’tyasya” so that the prescription may be said to occur with reference to the last (sesa) alphabet of ‘ik’. Third Ahnika 189 by the further provision and also with reference to the last alphabet. In this case this is an exception of that.125 What is the difference? (There are a number of rules such as “mider gunah” (7.3.82), “mrjer vṛddhiḥ’ (7.2.114), “pugantalaghupadhasya ca” (7.3.82) in which guna and vṛddhi are advised. Now, these guna and vrddhi refer to ‘ik’ and this seems to be in opposition to the rule “alo’- ntyasya” according to which a prescription with reference to a word in genitive case should apply to the last alphabet. Thus referring to this discripancy the Varttika says)- (Värttika:) If the prescription of vṛddhi and guna is to be determined by the rule “alo’ntyasya” then the sutras that dictate guna vṛddhi with reference to mrdi, mrji, roots ending in ‘puk’ and having a short penulti- mate, the roots ‘rcchi’, ‘drśi’, ‘kṣipra’ and ‘kṣudra’ should read ‘ik’ along with it. (The Bhāṣya says)-If guna or vṛddhi is to be determined by the rule (“alo’ntyasya”) then in the case of the verbs ‘mrdi; ‘mrji’, roots ending in ‘puk’ and short penultimate, the roots ‘rcchi’, ‘drsi’, ‘kṣipra’ and ‘kṣudra’, the rules prescribing guna and vṛddhi should specifically mention ‘ik’ (i.e. that gunavṛddhi prescribed by the rule should apply to i, u, r, land not to the last alphabet of the root in the genitive case). The root ‘mrdi’ should have guna by the rule 7.3.82. Here ‘ik’ has to be mentioned for ‘ik’ not being the last member the guna would not apply to it.
125. The point is that if the two sūtras are supposed to have different fields of application, they cannot be joined together. The difference between sthānaṣasthi and avayavaşaṣṭhi is this: In the case of sthānaṣaṣthi to which the rule “alo’ntyasya” refers, the ṣasthi or the genitive case refers to the anga or the part, i.e. the last part or the last alphabet. A ṣasthi or the genitive case may have various senses, e.g., possession, proximity, collection, modification, parts, etc. The rule “sasthi sthaneyoga” (1.1.49) makes the niyama or restriction that though ṣasthi may be in various senses, yet it should be accepted in grammar only to have the sense of ‘sthaneyoga’, i.e. ‘in place of’. This ‘in place of’ is further modified by the rule “alo’ntyasya” which says that the augment should take place not in the place of the whole word, but in the place of the last alphabet of the word. But here since “alo’ntyasya” is not combined with “iko gunavṛddhi”; the rule “sasthi-sthaneyoga” as determining the meaning of the ṣasthi, also does not apply. It is therefore an exception in which we may accept the meaning of șasthi as avayava şaşthi and sthaneyoga ṣaṣthi. Avayava-şaşthi means that the genitive relation applies severally to the parts of alphabets composing the whole of ‘ik’. The parts of which ‘ik’ is composed are i,u,r,!.190 Mahābhāṣya of Patanjali ‘Mrji’ should have vṛddhi by 7.2.114. Here also ‘ik’ has to be men- tioned, for ‘ik’ not being the concluding vowel vṛddhi would not apply to it. There will be guna to roots ending in ‘puk’ and having a short penultimate. Here also ‘ik’ should be mentioned, for ‘ik’ not being the last member the guna would not apply to it. ‘Rcchi’ should have guna when ’lit’ suffixes follow. Here also ‘ik’ has to be mentioned, for ‘ik’ not being the concluding vowel guna would not apply to it. ‘Dṛśi’ would have guna if ’na’ follows. Here also ‘ik’ has to be mentioned for ‘ik’ not being the concluding vowel, guna would not apply to it..The roots ‘kṣipra’ and ‘kṣudra’ should have guna for ‘ik’ not being the concluding vowel, guna would not apply to them.126 (Thus the Varttika says)-in the case of those that do not end in ‘ik’ there is the occasion of the augment occurring in place of the whole word. (The Bhāṣya says)-in the case of the roots not ending in ‘ik’ the guna ought to take place in place of the whole word. ‘yātā’, ‘vāta- why so? The sixth case in the rule “alo’ntyasya” refers to the last ‘ik’ and we have the case of a sthāne ṣasthi here. So whatever roots do not end in ‘ik’ should have guna in the place of the whole of it. This is no criticism. Thus just as the genitive in the rule “alo’ntyasya” refers only to ‘ik’, so where it is said that the guna or vṛddhi occurs in place of the part by sthane ṣasthi, that part can necessarily be only the ‘ik’. Now that which does not end in ‘ik’, has no genitive case and so how can there be guna, or how can there be any substitution in place of the whole word? This is no criticism. Just as the genitive case would by the rule “alo’ntyasya” refer to the concluding ‘ik’, so would also the sthane ṣasthi refer to the part of the word. So if there is a part which though forming the last alphabet is not ‘ik’; then there is no genitive case referring to it and consequently no guna or no substitution for the complete word.127
126. It seems that the rule “alo’ntyasya” can have its scope only where the root ends in ‘ik’. In other cases, the prescription of an augment can only apply in place of the whole word bearing the genitive suffix.
127. The point is this: according to the rule “alo’ntyasya” an augment directed to occur in place of a word in the genitive case, will take place with reference to the last alphabet of the word. Now when gunavṛddhi is advised, gunavṛddhi being of the nature of the vowel- a, e, o, ā, ai, au can only occur in place of the vowels included with any ‘ik’. According to the principle of “sthane ’ntaratamaḥ” (1.1.50) when there is an occasion of the occurrence of many indiscriminately, the selection is made in Third Ahnika 191 This is then no new criticism. What then? This fault appears to relate to the first. Thus we get that with reference to the verbs not ending in ‘ik’ such as ‘mrdi’, ‘mrji’ roots ending in ‘puk’ or short penultimate, ‘ṛcchi’, ‘drsi, ‘kṣipra’ and ‘kṣudra’ ‘ik’ should be definitely read in the sutra and there is a chance of a substitution in place of the whole word. There will be guna in place of ‘mrdi’. Here ‘ik’ being taken, the guna will not be with reference to the last alphabet. By the rule “alo’ntyasya” it cannot take place in place of the ‘ik’ that with which the verb does not end. Yet guna is advised, so this guna ought to be in place of the whole word and so everywhere. Then let it be considered as an exception. 128 (Värttika:) If it is held that the ‘ik’ whether it be an ultimate vowel or not (should have the transformation by guna). Then in the case of the rules “jusi ca” (7.3.83), “sarvadhātukärdhadhātukayoḥ” (7.3.84), the guna that is advised with reference to short vowels should have a prohibition affecting the application of guna to ultimate vowels. (This is explained by the Bhāṣya as follows):- (Bhāṣya) If the guna is to apply to all ‘ik’ irrespective of its place, then in the ‘jus’ vibhakti and when suffixes of the ‘sarvadhātuka’ and ‘ardhadhātuka’ type follow, the guna, that takes place in place of the short vowel, should be declared to be of the type in which the prohibition against accordance with the principle that which is most proximate from the point of view of utterance, shall have preference in occurrence. Now, therefore according to what has been said in case of gunavṛddhi” alo’ntyasya” and therefore “sasthi sthaneyogā” will always require for its field of operation, a word ending in ‘ik’. In the case of words not ending in ‘ik’ neither the genitive case rule “şaşthi sthaneyoga”, nor the rule “alo’ntyasya” would have any application. Therefore in such a case there cannot be any occasion for guna or for substitution by the rule “sasthi sthaneyoga”
128. In the case of avayavaşaşthi there is no scope for the rule “alo’ntyasya”, for “alo’ntyasya” has reference only to sthāna ṣasthi but in that case the ‘ik’ which is not the concluding part of a verb, becomes also liable to guna transformation. But there are cases in which guna is advised with reference to the last ‘ik’, e.g. “hrasvasya gunah” (7.3.108). In the vocative case the last short ‘ik’ of a word becomes replaced by a guna. Here we have the scope of the rule “sasthi sthaneyoga” and “alo’ntyasya”. There is a rule “orgunah” (6.4.146)- a bha ending in u will have the u replaced by a guna when taddhita suffixes follows. In this rule also “şaşthi sthaneyoga” and “alo’ntyasya” have scope. From this point of view the replacement of an ‘ik’ which is not the ultimate vowel by guna should be regarded as an exception. The Värttika now intervenes to criticise their view that the transformation by guna of a ‘ik’ which is not ultimate, should be regarded as an exception. 192 Mahabhāṣya of Patanjali the transformation of the ultimate has been withdrawn. 129 When ‘jus’ follows, the ultimate ‘ik’ is replaced by guna and this leads to the forms ‘ajuhavuḥ’ (they offered oblations). ‘avibhayuh’ (they were afraid). So also ‘anenijuḥ’, ‘paryaveviṣuh’. Here also it applies. Then there is the rule that when ‘sarvadhātuka’ and ‘ardhadhātuka’ suffixes follow the ultimate ‘ik’ is replaced by guna 7.3.84 leading to the forms ‘karta’, ‘harta’, ’nayati’, ’tarati’. But then here also in ‘ihita, ‘ihitum’, ‘ihitavyam’- there ought to be guna. 130 There is a rule “hrasvasya guna” (7.3.108) when the vocative follows the ultimate short vowel, is replaced by a guna. Thus we have ‘he agne’, ‘he vayo’. But then we might have guna also in the case of the phrase ‘he agnivit’, ‘he somasut’. If there is guna when ‘jas’ follows, such as agnayoḥ, vāyavaḥ then there also ought to take place guna in ‘Agnivitaḥ’, ‘somasūtaḥ’. Then there is a rule “rto ni-sarvanāmasthanayoḥ” (7.3.110); where a sarvanāmasthāna-vibhakti follows and when ni follows, the ultimate ‘r’ of a word will be replaced by a guna, ni as in ‘mātari’, in sarvanāmasthāna as ‘kartarah’ by which we have the forms ‘kartari’, ‘kartārau’, ‘kartāraḥ’. But then there ought to be guna in ‘sukṛti’, ‘sukṛtau’, ‘sukṛtah’ (su+kr+kvip= sukṛt-the objection is that there ought to have been guna of the r-kara in sukṛti, sukṛtau, sukṛtah, for according to hypothesis ‘ik’ should have guna wherever it may occur). According to the rule “gher niti” (7.3.111), when a suffix having a redundant na follows words called ‘ghi”, i.e., words having ultimate i and u not being ‘sakhi’ have their ultimate ‘ik’ replaced by ‘guna’ (for ‘ghi-samjñā see “seṣo ghyasakhi” 1.4.7), guna occurs in place of the ultimate i and u and we have 66
129. If the description of guna is to be regarded as applying to all ‘ik’, then it has to be definitely stated that there is no ban to the guna occurring in place of ‘ik’ at the end of the verb as in the case of the rules 7.3.83 and 7.3.84 where changes of ‘ik’ are affected with reference to the last vowel.
130. The point of the augment is this-If the application of guna to non-ultimate ‘ik’ be regarded as being provided through exception then the injunction and exception together would make it such that guna will be applicable to all ‘ik’ ultimate or not. But in the examples are so given that where guna takes place in one case it does not take place in another. If in ‘karta’ we have guna of rinto ar, then why not the root ‘ihi’ should have guna and the second i be replaced by e? Third Ahnika 193 agnaye, vayave. But why then you have not guna in ‘agnicite’ ‘somasūte’. By the rule ‘or gunah’ (6.4.146: a bha ending in u should have its u replaced by guna when a taddhita suffix follows), we have vabhravyaḥ māṇḍavyaḥ. But why not in the case of ‘suśrut’ and sauśrutah’. This is no fault. (The Varttika here intervenes)- The reading of both ‘puganta’ and ’laghûpadha’ in the sutra “puganta-laghûpadhasya ca” (7.3.86) directs that the guna can also take place for ‘ik’ other than the ultimate ones. (The bhāṣya says) The sutra “puganta-laghûpadhasya ca” (7.3.86) is for the purpose of restriction so that we have only a non-ultimate ‘ik’ of the ‘puganta-laghûpadha’ type that will be replaced by a guna and by no other non-ultimate.131 The niyama or restriction should apply to the first instruction. What is the first instruction? ‘sarvadhātukärdhadhātukayoh’. Then it follows that on account of the niyama we should not have any guna as in ‘ihita’, ‘ihitum’, ‘ihitavyam’. In the case of roots beginning with a vowel the guna is unrestricted. (In that case) the guna can take place of the initial vowel
131. The concept of niyama should first be grasped. When there is a particular instruction to some particular effect and if there is again another instruction to the same purport, then it is to be understood that the first vidhi of instruction should have its application curbed or restricted to some extent. Guna was prescribed as taking place with reference to roots having ‘ik’ as its part when sarvadhātuka and ardha- dhatuka suffixes follow (sarvadhatukârdhadhātukayoh 7.3.84). The next rule 7.3.85 is “puganta-laghûpadhasya ca”, i.e. when sarvadhātuka and ardhadhātuka suffixes follow guna will occur in place of puganta and laghûpadha words. Thus the first rule was without any restriction, the second rule applies only to puganta-laghûpadha. If the first rule was unrestricted there was no necessity for the second rule for here also guna takes place when sarvadhātuka and ardhadhātuka suffixes follow. Consequently the conclusion is that the first rule is not unrestricted in its meaning. The restriction is to the effect that when guna takes place with reference to an ‘ik’ which is not ultimate that ‘ik’ must only be of verbs that end in ‘pa’ and have a short penultimate. So the rule “sarvadhātukärdhadhatukayoh”, is the vidhi and “pugantalaghûpadhasya ca” is niyama. We know from the Mimämsä vidhi and niyama. Thus ‘vrihin avahanyāt’, ‘vṛihi’ mean the husked paddy seed. The ‘vrihi’ could be unhusked in various ways, but ‘vrihin avahanyať’ is a niyama rule by virtue of which we are to understand that if ‘vrihi’ is to be unhusked for sacrificial purposes, it should not be unhusked in any manner that one pleases, but that it should be unhusked in an iron or wooden pit, and then striking it with a heavy-mouthed hammer. Vrihi unhusked for sacrificial purposes should be only of one type namely that obtained by hammering it in a pit. 194 Mahābhāṣya of Patanjali in an unrestricted manner (i.e. such guna can take place of roots other than puganta-laghûpadha).132 If such a restriction is made, guna in case of pugantalaghûpadha verbs should take place only when sarvadhātuka and ardhadhātuka suffixes follow. But even on such an interpretation, the guna prescribed with reference to sarvadhatuka and ardhadhātuka suffixes, would be unre- stricted and be applicable to non-ultimate ‘ik’- ‘īhita’, ‘ihitum’, ‘ihitavyam’. But if both the restrictions are operative, i.e. guna should take place when sarvadhātuka, ārdhadhātuka suffixes follow only of puganta- laghûpadha verbs and only with reference to sārvadhātuka and ārdhadhātuka suffixes guna is prescribed for puganta-laghûpadha verbs. Then also the guna when ‘jus’ follows, ought to be unrestricted and would also apply to non-ultimate ‘ik’. Thus we have ‘anenijuh’, ‘paryavevisuh’.133 (Thus after criticising the two views that sought to explain the rule on the basis of apavada or exception he takes up the criticism of those that want to explain it on a paribhāṣā basis. Thus the Bhāṣya says)-This is then not to be taken to mean only the ultimate nor also its exception. They are two different maxims which are independent of each other. The Krostriyas considering it a different paribhāṣā read a vārttika as follows. Owing to the restriction guna and vṛddhi takes place by the
132. The conclusion so far attained may be stated as follows. The ‘ik’ may be ultimate, penultimate or initial. About the ultimate and penultimate ‘ik’, guna takes place everywhere. There is no restriction. And it is only in the case of the penultimate ‘ik’ that the restriction applies and guna takes place only in the case. A distinction is to be made between the two niyamas ‘puganta-laghuadhasyaiva sārvadhā- tukärdhadhātukayoḥ’ and ‘sarvadhātukärdhadhātukayor eva puganta-laghupadhasya’. The first means- guna occurs only with reference to puganta-laghupadha verbs when sarvadhātuka and ardhadhātuka suffixes follow. The other one means that guna takes place only when sarvadhātuka and ardhadhātuka suffixes follow with reference to puganta-laghupadha verbs. The distinction is clear. The second one means that guna takes place of puganta-laghūpadha verbs that can only be with reference to sārvadhātuka and ardhadhātuka suffixes. The second niyama is now discussed in the Bhasya as follows.
133. The root in ‘anenijuḥ’ is a laghupadha and ‘jus’ is a sarvadhātuka suffix and therefore there is guna. In such words as ‘he buddhe– buddhayaḥ’ the upadha is not laghū and therefore the guna for the non-ultimate is not barred. For according to the second niyama puganta laghūpadha should have guna of the non-ultimate only in sarvadhātuka and ardhadhātuka, but nothing is said about what is not puganta- laghupadha. Therefore there is nothing to prevent the guna with reference to non- ultimate vowels. Third Āhnika 195 principle of supersession or vipratisṣedha. The principle of supersession is not proper. (For the rule about it is, in case of equal mutual opposition, the last one should prevail, “vipratiṣedhe param karyam” 1.4.2.). But here the prescription comes first and the niyama comes later. 134 The word ‘para’ in the rule 1.4.2. means that which is desirable. In the case of equal mutual operation that which is desirable, takes place. In such acase also the word ‘vipratiṣedha’ would be improperly used; vipratiṣedha means the occasion of two contrary operations. There is no occasion for ‘ik’ here to be liable to two contrary oppositions. But vipratiṣedha is not necessarily the occasion for two contrary oppositions. It also means ‘impossible’ and here it is impossible. What is the impossibility? Here is the impossibility-‘vrkṣebhyah’, ‘plakṣebhyah’.135 Here thus we have ‘vrkṣebhyah’, ‘plakṣebhyah’. There is once sthāni or term of operation and two augments. It is not possible that one term of operation should be replaced by two different augments. Here again ‘medyati’, ‘medyataḥ’, ‘medyanti’. There are two terms of operation and one augment. It is not possible that the augment should replace two terms of operation. This is the impossibility here. This being thus impossible it becomes vipratiṣedha or equal mutual operation of which the other meaning is impossibility.136
134. The utsarga and apavada, i.e. the general and the special rule and the principle of ‘antaranga’ and ‘bahiranga’, i.e. internal and external operation do not come under the principle of equal mutual operation.
135. “Ato dirgho yañi”, “supi” 7.3.101,102 when suffixes of the sarvadhātuka class beginning with ‘yan’ follow, then words ending in a are lengthened. There is another rule 7.3.103 “bahuvacane jhalyet” which means that when a suvanta suffix in the plural number beginning with jhal follows, then the ultimate a of the preceding word should be changed in e. Now in the formation of the word ‘vrkşebhyah’ there is occasion for the operation of the rule “ato dirgho yañi” etc. 7.3.101. In the case of the formation of the word ‘vrkşeşu’ there is occasion for the operation of the rule “bahuvacane” etc. 7.3.103. In the matter of the formation of the word ‘vrkṣebhyah’ there is occasion of the operation of both the rules (7.3.101 and 7.3.103). Conse- quently the second rule being the later one takes precedence and gets its operation.
136. The point raised is this: The root ‘midi’ has genitive case-ending. If this genitive case is an avayava-ṣasthi, then it will be replaced by guna. But if it is sthana- şaşthi then by the rule “alo’ntyasya”, guna should replace da-kara. Now one guna cannot at the same time perform these two functions. Therefore there is an impossi- bility of operation. And if the word ‘vipratiṣedha’ means impossibility of operation as suggested above, then the rule “vipratiṣedhe param karyam” ought to take its course here. 196 Mahābhāṣya of Patanjali But even in such a case it is improper to urge that here is a case of impossibility of vipratiṣedha. Vipratiṣedha takes place there where in the same field two different operations seem to be equally liable, but that is not the case here. But such an occasion can be easily met as in the instance, ‘cayanam’, ‘cayakaḥ’, ’lavanam’, ’lavakaḥ’. But here also it is the niyama of the rule “alo’ntyasya” that becomes operative. If the niyama could not be applied then the original prescription becomes applicable. Since without the niyama the original prescription would have been applicable, the niyama should be regarded as having the status of an exception. The principle of ‘vipratiṣedha’ cannot be applied to principles that are related as utsarga and apavāda, i.e. general and special rule. If it is supposed that somehow there is scope for the operation of the rule “iko gunavṛddhi” then just as through vipratiṣedha, guna is applicable in ‘medyati’, ‘medyataḥ’, etc. in place of ‘ik’. So here also in ‘anenijuh’ and ‘paryavevişuh’. This is behaving at pleasure. Once we qualify the ‘ik’ with the predicate advised and at another time we qualify the predicate advised by ‘ik’. The irregularity of behaviour is shown in this- in the case of the roots ‘midi’, ‘mrji’, the roots ending in pa and having a short penultimate, rcchi’, ‘drśi’, ‘kṣipra’ and ‘kṣudra’, we are qualifying the ‘ik’ with guna and say that the ‘ik’ of these verbs becomes guna. But here again when ‘jus’ follows and when sarvadhātuka and ardhadhatuka suffixes follow, we are qualifying the guna by the ‘ik’ and say that there is guna in the case of these verbs and they end in ‘ik’. Or rather in all these places it is the sthāne that is denoted. In the case of ‘mideḥ’, there is no suffix, it is ‘mida eḥ= mideḥ’, or rather it may be the sixth case. ‘mit i = midi’, then the genitive case mideḥ’. In the case of the rule “puganta laghûpadhasya ca” the interpre- tation is not that guns will take place with reference to roots having ‘puk’ in them and also a short penultimate. But then what it is? puki antaḥ pugantaḥ where the end is in puk and laghûpadha mean a short penulti- mate, then “pugantasya laghûpadha ca” where that which is puk in the end also a short penultimate, this must have to be interpreted in this manner. For if the existence of short penultimate were to qualify the verb as such, then guna ought to take place in ‘bhinatti’ and ‘chinatti’. In the case of ‘ṛccha’ also it should be read as rcchati, r, r, rtam, rcchatyṛtam. In the case of ‘drsi’ also, we have to split up. The rkäranta words have guna when an follows. (The sutra being “r-drśo’ni gunaḥ"7.4.16 have to be split up into ‘urani gunah’ and ‘drso’ni gunah’, uh is the genitive case of the root rin “r-drso’ni gunah”). So when an follows the r of ṛkäranta roots will be Third Ahnika 197 replaced by guna and we have the other rule by splitting up drśaḥ. When an follows dṛśa will be replaced by guna. This replacement is of the ror drsa 137 In the case of kṣipra, kṣudra also, the interpretation is that of which the yaṇādi comes later, undergoes guna. It is explained so far. As this was sufficient to explain the fact that the sutra again speaks of ‘purvasya ca gunah’, there will be guna of the previous- naturally suggests its purpose, viz. that the guna should be only of ‘ik’ and not of any alphabet other than ‘ik’ 138
137. There is a rule “sthula-dura-yuva-hrasva-kṣipra-kṣudrāṇām yaṇādiparam purvasya ca gunah” (6.4.156)-with reference to the above words when istha, iman, and iyas follow, the vowel of the first alphabet is replaced by guna and the ya, va, ra, la belonging to them is elided. So we have sthula + iṣtha = sthavistha, kṣipra + istha-kṣepistha. The phrase ‘yaṇādiparam’, means yan yaṇādir yasya tadrūpam param lupyate tataḥ purvasya ca gunah. This would mean that ya, va, ra, la will be elided and the vowel immediately preceding ya, va, ra, la will be replaced by guna. So in the case sthūla, la was elided and the vowel immediately preceding la, the long ū of sthū is replaced by guna. But what about kṣipra? The ra may be elided, but immediately preceding the ra there is pa and you cannot have a guna of pa. So the word ‘purvasya ca gunah’ has to be taken by yoga-vibhaga. In one case for the words sthula, etc. we have first the elision of ya, va, ra, la and then guna of the immediately preceding vowel. Then we have the words ‘purvasya ca gunah’, in the case of ‘kşipra’, ‘kṣudra’ etc. We have the first vowel that is to undergo the guna. In the case of the words ‘kşipra’, ‘ksudra’ the meaning of the phrase ‘yaṇādiparam’ is yaṇaḥ, ādiḥ (previous to ya, va, ra, la or yan)-in the case of ‘kṣipra’, ‘ksudra’ only pa and da are previous to the ‘yan’. The ‘yaṇadiparam’ means following yaṇādi. Since yaṇādi means pa and da in ‘kṣipra’, ‘kṣudra’, ‘yanadiparam’ means the following ra which is elided (yaṇādiparam lupyate). The phrase ‘purvasya ca gunah’ has reference to the purva or yaṇādi and what is purva of yaṇādi in kṣipra and kṣudra, the i and the u. These therefore undergo guna by the phrase ‘purvasya ca gunah’ which is a part of the rule. This is summarised in the Bhāṣya that follows.
138. The point suggested is this: The word ‘yaṇādiparam’ is interpreted by the Bhāṣya as having two different interpretations. In the one interpretation yaṇādi means yaṇaḥ adiḥ and yaṇādiparam means that which follows yaṇādi. The other interpreta- tion is that of which yaṇādi is the sequent. Now in kṣipra pa is yaṇādi and i of kṣi is yaṇādiparam. According to the meaning of yaṇādi in the first interpretation there is elision of ra in kşipra. According to the second interpretation yaṇādiparam means the i-kara of which yaṇādiparam is the sequent, and i and u suffer guna. So it was unnecessary for the rule to add the phrase ‘purvasya ca gunah’, for we get ‘purva’ by the interpretation of ‘yanadiparam’ as shown above. There is a maxim that if a direction is added where it is already patent it should be understood for a special purpose. Here the direction ‘purvasya ca gunah’ has been only repeated and therefore it indicates some special purpose. 198 Mahābhāṣya of Patanjali (The Bhāṣya says that)-That purpose is this that it directs if there is any guna, that guna should apply only to ‘ik’ and to no other vowel or consonant. (The opponent’s objection on the subject of vṛddhi is anticipated in the Bhāṣya. Thus the Bhāṣya says)-Why does the sutra speak of vṛddhi? But why do you ask particularly of vṛddhi and not of guna? If there is any utility in guna the same may be in the case of vṛddhi what is the difference? There is some difference (says the opponent). In the matter of the instruction of guna the sthänin (that which is replaced by guna) is not definitely mentioned. Therefore in every case for instructing the sthanin, the guna has to be mentioned. In instructing the vṛddhi, however, the sthanin has always to be mentioned, as in the following rules-“aco ñņiti” (7.2.115; words ending in vowels should have the vowel replaced by vṛddhi when suffixes having redundant ñ and cerebral n follow, there is vṛddhi in place of the penultimate a such as pacakaḥ), “taddhiteṣv acām ādeḥ” (7.2,117; when taddhita suffixes hav- ing redundant ñ and cerebral n follows, then vrddhi should be directed to the first vowel of the word). The answer is given by the Värittka. (The Varttika intervenes) - The word ‘vṛddhi’ has been taken for a later purpose. (The Bhāṣya says)-The vṛddhi has been taken. What for? For a later purpose. The master will speak of exceptions (of guna-vṛddhi) when suffixes that have the redundant ka and ǹ follow, so that it can apply also with reference to vṛddhi (so that the rule may apply to vṛddhi as well). What is the occasion of the vṛddhi when suffixes having the redundant ka and na follow? For vrddhi is directed to occur with reference to suffixes that have the redundant ñ and n. (Here the Varttika intervenes)-That is for the root ‘mrj’. (The Bhāṣya says)-The vṛddhi has been directed with reference to ‘mrj’ without any qualification and rule “kniti ca” (1.1.5) is there to prevent the occurrence of any vṛddhi when suffixes having redundant ka and n follow-mrstah, mrstavān (no vṛddhi here, because there is kta and ktavatu suffixes which have redundant ka). (The Varttika again says)-(vṛddhi has been taken) for this also. (The Bhāṣya says)- For this also, for the root ‘mrj’, vṛddhi has to be directed. There is no qualification for the vṛddhi of mrj. (A qualification is introduced) non-ik. vṛddhi must be of ‘ik’ and not of (Again the Värttika says)-If for the root ‘mrj’, then the meaning can be settled by the principle of ‘yoga-vibhāga’ or splitting up. Third Ahnika 199 (Bhāṣya:) If it is not of the root ‘mrj’, then let us split it up by ‘yoga- vibhaga’. Vṛddhi will occur in place of the vowel ‘mrj” “mrjer vṛddhih” (7.2.114; there is a vṛddhi of the vowels in ‘mr’ when suffixes having redundant ñ and n follow and this vṛddhi applies only to the vowels). A vṛddhi is to be spoken of vowels? Then how ’nyamart? The ‘at’ also should have vṛddhi. (Here Bhāṣya says)- What do you say? In the case of modifications not affecting the ultimate, there can be operations in regions near about the ultimate. 139 (Here the Varttika comes)- The unreasonableness of prohibiting vṛddhi comes from the fact of its having reference only to ‘ik’. Therefore vṛddhi is conditional to ‘ik’. (Now the Bhāṣya explains it)- The prohibition of vṛddhi is not reasonable. What is the reason? Because it belongs to the ‘ik’ context. The prohibition would be of ‘guna-vṛddhi’ of those that have ‘ik’. In such a case, there cannot be any vṛddhi or ‘ik’ belonging to ‘mrj’ should have vrddhi with reference to its ‘ik’. But here other grammarians have desired an optional vṛddhi with reference to ‘mrj’ when a suffix with an initial vowel followed parimrjanti, parimärjanti, parimṛjantu, parimärjantu, parimamrjatuḥ, parimamarjatuḥ- for solving these cases. So here also that has to be done and if that is to be done, one has to split up the sutra ‘mrjer vṛddhir acah’. The vowel of ‘mrj’ undergoes vṛddhi. Then we the rule ‘aci kniti’ i.e., the suffixes having the redundant ka and nia and beginning with a vowel, ‘mrj’ should have a vṛddhi-‘parimärjanti’, ‘parimärjantu’, ‘parimamarjatuḥ’. Why is this? For the purpose of restriction. In the case of ka-it and na-it suffixes, vṛddhi should take place only with reference to suffixes having an initial vowel. From where is the vṛddhi restricted? From mrstah, mṛṣṭavan. Then we have the option. It is only optionally that
139. This is a paribhāṣa which the Bhasya introduces in his interpretation of the rule 6.1.13 “antyasamanadeśa”- means the last alphabet and whatever is without any in conjunction with it. The point raised is in the case where there is an ‘at agama’ in vibhakti such as lan, lun, etc. vṛddhi should take place before the ‘at’ takes place and therefore when ‘at’ is made, there is no occasion for vṛddhi. But then again the paribhāṣā has been raised by which cases of modification that does not take place with reference to the ultimate, may as well take place with alphabets near about the ultimate. But Nāgeśa points out that if vṛddhi is to be effected with reference to ‘ik’ only, it cannot apply to ‘at’ naturally. The present paribhāṣā occurs as paribhāṣā No. 103 in the Paribhāṣendusekhara. We shall take up a more elaborate examination of this paribhāṣa in connection with the treatment of the interpretation of Bhasya 6.1.13.200 Mahābhāṣya of Patanjali when suffixes having the redundant ka and ǹa and beginning with an initial vowel follow that ‘mr’ suffers vṛddhi-‘parimrjanti’, ‘parimärjanti’. If the taking of vṛddhi is not necessary for ‘mrj’ then it ought to be for ‘sic’. When ‘sic’ follows vṛddhi has been enjoined in a general manner for ‘sic’ and that vṛddhi should refer to ‘ik’. In which case the non-ik has the occasion for vṛddhi? The akara has the occasion for vṛddhi as in ‘acikirṣiť, ‘ajihirṣit’. This is not so. The occasion for elision will arrest it. Well, then it may be for a-kāra as in ‘ayāsīt’, ‘avāsīt’. There is no difference here whether there is vṛddhi or not. Then it may be with reference to euphonic vowels. There are no roots ending in euphonic vowels. But it may be for that when dha is elided we have ‘udavoḍham’, ‘udavodham’, ‘udavoḍha’. Not for this. (Being a rule in the tripādi it should be regarded as not having taken place at all, so) the elision of dha is to be regarded as not having taken place. That being ineffective it cannot be regarded as being in the end. 140 Then the vṛddhi applies in the case of consonants such as abhaitṣit, acchaitṣit. But this would be arrested by the rule that prohibits vṛddhi with reference to words ending in consonants (by the rule “neti” 7.2.4-a verb ending in consonants would have no vṛddhi when ‘sic’ with the augment ‘if’ follows). But this prohibition is with reference to all that falls within the scope of the rule “neti” (7.2.4), akoṣit, amoșit. But this (’neti’) is a prohibition also of vṛddhi when ‘sic’ follows. How so? The prohibitory
140. The point is this: It was suggested that there was no root ending in a ‘sandhyakṣara’ (euphonic vowels), but taking the case of the word udvodham, ut + vaha + tām we have the at before vaḥa due to lun, we have ut + a + vaha +tām, then we have the rule “ho dhah” when a jhal follows and at the end of a pada the ha is changed into dha)- thus we have ut+a+ vaha +tam, then we have the rule ‘jhalam jaso’nte’ by which ta of tam is changed into dha, thus ut+a+ vaha + dham, then by “dho dhe lopah” the first dha is elided, then by “sahi-vahor odavarnasya” the a of vaha is changed into o and we have udavoḍham. Now, since the elision of dha as well as the transformation are both ineffective or asiddha being in the tripädi, i.e. the last three chapters of Panini (see purvatrâsiddham 8.2.1) we have the occasion of vṛddhi and the chance of ‘vaha’ being changed into ‘vaha’ by the rule “ato haläder laghoh” (7.2.7). Again when the a-kara of ‘vaha’ is changed into o, we should have the occasions of vṛddhi as applying to o. This was the contention of the previous Bhāṣya that the injunction of vṛddhi has been declared definitely with regard to ‘ik’ in order to prevent vrddhi of the euphonic vowels. It was first declared that no roots end in an euphonic vowel, but then even though they may not end so, there may be other occasions for effecting vṛddhi with reference to euphonic vowels as in the case of vṛddhi in place of o-kara in vodha. Third Ahnika 201 rule makes the general prohibition of vrddhi. But it does not stop there. It moves about and does not cease for a moment.141 Or rather by the rule “sici vṛddhiḥ parasmaipadesu” (7.2.1) we have the occasion of vṛddhi when ‘sic’ follows. This again is restricted with reference to verbs ending in consonant and this again is further restricted by the rule ’neti’, (with reference to those that have the augment ‘it’). But does it so occur in other places that when an exception is ruled the general rule is also inoperative? Yes, there is. ‘Sujate aśvasunṛte’. ‘Adhvaryo’ adribhiḥ sutam. ‘Sukram te anyat’. When the purvarupa is prohibited then the general rule prescribing ‘ay’, etc. also does not take place. 142 It may be that vṛddhi has to be mentioned with reference to ‘sic’ for a later purpose. Firstly, vṛddhi is advised to ‘sic’ without restriction. Then it is prohibited with reference to suffixes having the redundant ka and n- ’nyanuvit’, ’nyadhuvit’. There is no necessity for this. Here the fixing of the augment-uvan being more internal, the verb no longer ends in u and therefore there is no vṛddhi.143 But then it may be by the rule that prohibits vṛddhi with reference to the words ending in consonants. Here then ’nyastarit ’nyadarīt. When
141. The prohibitory rule “neti” may seem to prohibit only the vṛddhi. This is expressed by the word ‘dhvanati’ in ’lakṣaṇam hi nama dhvanati’. ‘Dhvanati’ means ‘speaks in an inarticulate manner’, i.e. it is not clear that it prohibits only the vṛddhi here. If it was only intended for that it could not make other prohibitions. But this prohibitory rule moves from one to the other. It not only thus prohibits vṛddhi, but prohibits it with reference to roots ending in consonants. This is the purport of the word ‘bhramati’. ‘Bhramati’ means ‘sarvatra vyapṛyate ubháyatrâpi vyāpārāt na ekatra viśramyati’.
142. In the Vedic passage the word ‘aśvasunṛte’ follows the word ‘sunṛte’ i.e. a follows e. But the rule “eco’yavāyāvaḥ”(6.1.78) there e was expected to be ‘ay’. This was overruled by the rule “erah padäntad ati” (6.1.109), i.e. when a follows the en takes the place of both the vowels en and a and we should have expected here sujāte Svasunrte (this being a later rule to eco’yavāyāvah). But this rule is again negated by the rule “pluta-pragrhya aci” (6.1.125), this being a still later rule-by this rule the pūrvarupa advised in 6.1.109 is negated. Thus here when one prohibition is cancelled by another prohibition but still the general rule in “eco’yavāyāvaḥ” does not become opertive and the e of ‘sujate is not changed into ‘ay’.
143. uvan- “aci śnudhatu-bhruvām yvor iyan-uvañau” (6.4.77). There are the roots ’nu’ and ‘dhu’ and in forming the ’lun’ we have the words ‘anuvit’ and ‘adhuvit.’ It is suggested that there in order to prohibit the vṛddhi of ‘sic’ when suffixes having the redundant ‘ka’ and ’na’ follow, the word vṛddhi has been taken in the rule “sici vrddhiḥ parasmaipadesu” but to this it is objected that the augment ‘uvan’ being prior to ‘sic’ for the purposes of ‘sic’ the verb is supposed to be ending in va and therefore the question of ‘vṛddhi’ does not occur. ‘uvan’ is antaranga with reference to ‘sic’ and we have the paribhāṣā-‘asiddham bahirangam antarange’. 202 Mahabhāṣya of Patanjali we make the guna and root ends in ar, the a of ar not being an ultimate, there is no vṛddhi. The vṛddhi of words ending in consonant, is prohibitetd by “neti”. May it be so, but it may be by the rule “ato Irantasya (7.2.2; the a takes vṛddhi when it is the ultimate and is associated directly with ra and la). Here then ‘alavit’, ‘ayavit’. Here when guna has once been made the root ’lu’ becoming ‘av’ can no longer be regarded as ending in u and there should be no vṛddhi. Vrddhi will be arrested as the word ends in consonant by the rule ’neti’. No, it would not be so. The rule “Irantasya” (7.2.2.) will apply. You speak of ‘Irantasya’, but this is no a Iränta. In the rule “Irāntasya”, ra is also included. But we do not find any va-kāra (va being elided by the rule “lopo vyor vali” 6.1.66). If so, then ‘mā bhavan aviť’ and ‘ma bhavan mavit’. Here also there is occasion. (In ‘sic’ the root ‘avi’ will get ‘at’ as an augment in the beginning and therefore it will be impossible to understand when there will be any vrddhi. or not and therefore in order to prevent the augment ‘at’, the Bhāṣya uses the word ‘man’. In the case of ‘mavi’, there was no such necessity of adding ‘man’, it was added only out of uniformity). It will be advised later on that there will be no vṛddhi. for the roots ‘avi’ and ‘mavi’ (by the rule “hmyantakṣaṇaśvasa-jagr-niśvyeditām” (7.2.5- roots ending in ha and ma and ya and kṣaṇa, śvasa, jāgr, ni, śvi, and those having a redundant e should not undergo any vṛddhi in the parasmaipada when a sic with an augment ‘it’ follows). This is a restriction to the rule “ato halader laghoh” (7.2.7) by which one could have vṛddhi optionally. It also restricts the vṛddhi by “ato Irantasya” (7.2.2). Then that ought to be said. No. Instead of ni, śvi ( in the rule “hmyantakṣaṇa-jāgr” etc.) one should read ‘avi’, ‘mavi’. Then one should have to declare separately the prohibition regarding ni, śvi. No. When guna is made and the e is changed into ‘ay’, then there will be prohibition by the rule “yāntānām na” (i.e. ‘hmyanta-kṣaṇaśvasa’ etc. 7.2.5) then the manner of the master shows that the ‘antaranga bahiranga’ rule does not apply in ‘sic’, and for this reason in the rule “ato halader laghoḥ”, a has been definitely mentioned. How does it indicate anything? For this purpose that there may not be any vṛddhi in ‘akoṣit’, ‘amoșit. If the antaranga rule had applied in ‘sic’, then the reading of a in “ato halader laghoh” would be unnecessary, for when guna is made the vowel would be long and therefore there will be vṛddhi. Therefore the master shows here (by taking a in “ato halader laghoḥ”) that there will be no antaranga rule in ‘sic’. Therefore he reads a in “ato halader laghoh”. This does not indicate anything. Let there be something else for which it is done. What is that? It is for those cases in which guna is Third Ahnika 203 prohibited. Thus it may be for ’nyakutit’, ’nyaputit’, that vṛddhi for ni, śvi has been prohibited, it is for the purpose of showing that with reference to ‘sic’, the antaranga principle does not apply. This is evident that the master speaks of a even when he has already mentioned that the operation should be of short vowels. (Here comes the conclusive Varttika)-For all these reasons vrddhi should be only with reference to ‘ik’. (The Bhāṣya explains)-therefore the vṛddhi should be with reference to ‘ik’. (The opponent’s view is again ushered in by the Värttika)-since ṣasthi has been marked as being sthāneyoga, there is prohibition of ‘ik’. If the genitive in ‘ikah’ is to be understood in the meaning of ‘sthāneyoga’, then everywhere the ‘ik’ ceases and here also it applies in ‘dadhi’, ‘madhu’. Why is again guna taken up (as in ‘mrder (gunah)’ ? (Now the Varttika says)-This may have been taken up for another purpose. (Bhāṣya:) It may be another purpose, so that in case of sarvadhātuka and ardhadhātuka vibhaktis there may be only guna. (The Varttika intervenes)-and in the case of expansion. (Bhāṣya:) In the case of expansion (samprasarana) all alphabets in- cluded in ‘yan’ should cease. Then it may be also in the case of ‘yātā’, ‘vātā’. (Another objection is anticipated). Then, why is it taken again? (The Varttika intervenes)- It may have been taken again for regulat- ing the nature of conditions. For regulating the nature of conditions as in “vacisvapiyajādīnām kiti” (6.1.15) only (-in the case of the roots vac, svap, yaj, etc. there is samprasaraṇa or expansion when a suffix with redundant ka - samprasarana means the replacement of ya, va, ra, la by i, u, r, 1, e.g. vac + kta = ukta). (Varttika:) For the rule “uran raparah” (1.1.51). (Bhāṣya:) In the rule “uran raparah” (1.1.51), all r should cease. Here also there is occasion, kartr, hartṛ. (Varttika:) The difficulty will be solved if şaṣthi is taken as adhikāra. (Bhāṣya:) The difficulty is solved. How? In the genitive as adhikara these further should be added in those places. These also should come in, whenever there is genitive case. Or rather when there is the sixth case there would be these two additional prescriptions. Or rather this has to be asked. If there is guna in the case of sarvadhātuka and ardhadhātuka suffixes, why not here in ‘yātā’ and ‘vātā “? Because it refers to “iko gunavṛddhi” 204 Mahābhāṣya of Patanjali As this awaits that, so here also we shall refer to the rule “sarvadhātukārdhadhātukayoḥ” (7.3.84).144
144. In “iko guṇa-vṛddhi’ it is first suggested that vṛddhi has been spoken of as referring to ‘ik’ and that is for advising vṛddhi in the case of ‘sic’ in lun. vṛddhi has been advised to ‘sic’ with reference to ‘ik’ and that should not take place in the case of k-it suffixes as in the lun of ’nu’ or ‘dhu’, e.g. ’nyanuvit’ and ’nyadhuvit’. But this view is rejected, for in both these cases, there is ‘uvan’ augment by the rule “aci-śnu- dhatu-bhruvām yvoriyan-uvarau”(6.4.77) and this being an antaranga, the prescription of vṛddhi in ‘sic’ being bahiranga, will not have any effect, i.e., the verb would be supposed to be ending in va of uvan and therefore the vṛddhi consequent upon ‘sic’ will not apply as it has application only to verbs ending in vowels. But then there is the rule “neti” (7.2.4) which prohibits vṛddhi to verb ending in consonants. Thus the provisional conclusion is that the vṛddhi in nynyvit and nyadhuvit is arrested by ’neti’ and it has no utility in the case of “sici vṛddhiḥ parasmaipadeşu” 7.2.1 as restricted by “kniti ca” (1.1.5). Provisionally it is also accepted that the antaranga-rule applies in the context of ‘sic’ (sic prakarana). Then it is suggested that vṛddhi has been taken for effecting the words ’nyastarit’, ’nyadarit’. By guna here we could have or and that would make the rule end in constant and hence make it liable to prohibition or vṛddhi by “net?” (7.2.4), but the opponent suggests that vṛddhi might be affected by the rule “ato Irantasya” (7.2.2). But how shall we then effect ‘alavit’ and ‘ayaviť? After effecting the guna the roots ends in ‘av’ and thus having an ultimate consonant has its prohibition of vṛddhi by “neti”. But in reply to it is suggested that vṛddhi will be effected by “Irantasya”, for the ra in ‘Iranta’ is supposed to include va. But then there ought to be vṛddhi also in the case of ‘aviť’ and ‘maviť. But here it is suggested that ‘avi’ and ‘mavi’ should be specially prohibited. But there is no such prohibitory rule. But the Bhāṣya suggests here the opponent’s opinion that in the rule “hmyanta-kṣaṇa-śvasa” etc. (7.2.5), ’ni’, ‘Svi’ should be replaced by ‘avi’ ‘mavi’. But then how to effect the prohibition in ’ni’ ‘śvi’. Thus in having the lun of ‘śvi’ as ‘aśvayit’, the ‘śvi’ has guna as ‘śve’ and e becomes changed in ay, then the root ending in ‘ya’ should be naturally ending in constant and has no vṛddhi. But this is in acordance with the provisional view that the ‘antaranga’ rule applies in ‘sic’, but we have the exactly opposite view that in the sic- prakarana antaranga maxim does not apply. To the question as to why ‘sic’ should be taken as a sphere (i.e. context) where the antaranga-bahiranga maxim does not apply. The answer is that the fact that in this rule “ato halader laghoh” a-kāra should be definitely mentioned in the sutra to prevent vṛddhi in ‘akoşit’ and ‘amoṣit’. Here guna is effected, we have ‘koşa’ and the o of ‘kosa’ not being a short vowel, there will be naturally no vṛddhi. But still that to prevent vṛddhi, the word ‘atah’ should be added, shows that the antaranga guna is regarded as imperative before the bahiranga sic. This, therefore, indicates that the antaranga rule does not apply to the sic-prakaraṇa. Now this view runs counter to the provisional acceptance that the antaranga rule applies in the sic-prakarana by which ‘aviť’, ‘maviť’, etc. was explained by which again ’ni’ and ‘śvi’ is suggested to be replaced by avi, mavi in the rule ‘hmyanta’ etc. Third Ahnika 205 The Bhāṣya here suggests that the word ‘atah’ in “ato halader laghoh” has been given to prevent vṛddhi in those cases where there is no guna as in ’nyakuțit’ and ’nyaputit’. But the fact that vṛddhi has been prohibited with reference to ’ni’, ‘śvi’, in the rule ‘hmyanta’ etc. shown that the antaranga rule does not apply in the sic- prakarana. The conclusion as declared by the Värttika is that vṛddhi has to be declared to happen to ‘ik’. We have seen that in ’nyanuvit’ the vṛddhi had a chance of occurring in place of ‘uvan’, yet the same vṛddhi was prohibited by ‘sici vṛddhiḥ’ where the direction was that vṛddhi must take place. The other point raised by the opponent is with reference to the genitive case- ending of the ‘ik’ in “iko guna-vrddhi”. The opponent urges that since ṣasthiis sthāne- yoga we have guna-vṛddhi in place of ‘ikah’. Though guna-vṛddhi could be got from the adhikara by “vrddhir adaic” and “adengunah”, it may have been for the purpose that no one should have guna-vṛddhi except that which is ‘ik’ and that whenever there is guna-vṛddhi enjoined, the word ‘ik’ should naturally go there. The same maxim may be followed with reference to the rule “ig yaṇaḥ samprasaranam” (1.1.45)- there should be ‘ik’ in place of ‘yan’ and this is called samprasarana. Again, we have the rule “vaci-svapi-yajādinām kiti” (6.1.15) which means samprasarana will take place with regard to the verbs-‘vaci’, ‘svapi’, etc. when a k-it suffix follows. This is limitation of the sphere. Having followed the same maxim one may apply the rule “uran raparaḥ” (1.1.51), which means that the ‘an’ occurring in place of ‘r’ should be followed by ra and may be used to mean that ṛ should cease and in place of it ‘an’ should come, and of the three vowels ‘an’ none being more in proximity to r than other, there is no rule which can determine which of the vowels of ‘an’ should take place in which context in all such cases the maxim ‘sthane antaratamaḥ’ applies, but there being no relative ‘antaratamya’ there is nothing for us to determine which will be ar, which ir and which ur. Therefore in all such cases where there is injuction by şaşthi, the special conditions have to be mentioned. There is a further point raised by the Bhāṣya that as the determining guna we have to mention the condition “sarvadhātukārdhadhātukayoḥ” so we have also to mention the fact that if guna- vrddhi is to take place it has to take place with reference to ik alone. SUMMARY The whole discussion of “iko guna-vṛddhi” passes through many stages. The first point raised, is-why the word ‘ik’ has been taken in the genitive case? The reply to this is that ‘ik’ has been taken in order to declare that guna-vṛddhi can take place only with reference to ‘ik’. Thus there is a “yoga-vibhāga” in the rule “iko guna-vṛddhi”-‘ikah’ and ‘guna-vṛddhi-guna-vṛddhi syātām-there will be guna-vṛddhi-this is an injuction. yadi guna-vṛddhi syātām- then it will occur in place of ‘ik’, tataḥ ika eva. This interpretation is derived from indications of other rules. The ‘ik’ therefore has been taken to prohibit any guna-vṛddhi with reference to ‘at’ and the euphonic vowel e, ai, o, au, and also the consonants. But now objections are raised by those who think that this sūtra is unnecessary. They say that it is certainly not for prohibiting guna with reference to a-kāra, for had it been the case, in the sutra “āto’- nupasarge kah” (3.2.3) the suffix should not have redundant ka, for if guna could be adduced to ā-kāra, a will be reduced to a and this a with the a of the suffix will combine into one a by para-rupa and we get such forms as ‘kambaladaḥ’. By associating the suffix with the redundant ‘ka’, the elision of a has been enjoined. This injuction would have been unneces- sary if guna could occur in place of a. The associating of the suffix with the redundant ka would be unnecessary. This latter fact thus indicates that guna should not occur in place of a-kara. The prohibition cannot apply with reference to the euphonic vowels, for that has been well done in the Sikṣā-prakaraṇa. It cannot be also for prohibiting the consonants. There is a rule “janer dah” (3.2.97)- the suffix here has a redundant da which indicates the elision of ’ti’, “aco’ntyādi ti” (1.1.64; the portion of a word beginning from the last vowel is called ’ti’, it is elided when a redundant da suffix follows, by the rule “teh” (6.4.143). If the na of ‘jana’ could undergo guna it would be a and the a of ja of jana, the a of na of jana and the a of the suffix would together form one a by pararupa and therefore the fact that the suffix has been associated with redundant da (as in the case of the redundant ka of the suffix in “ato’nupasarge kah” suggests that no guna can be ascribed to consonants. The Bhāṣya denies all this and justifies the existence of ‘ikah’ in “iko guna-vṛddhi”. The k-it suffix is not for indicating that guna would not apply to ā-kāra, but it is a pratyaya (i.e. a suffix) which is necessary in another rule (“tunda-śokayoḥ” etc. in 3.2.5). Being necessary in another Third Ahnika 207 rule the ka-suffix had to be coined. Being coined once it has been applied in “ato’nupasarge kah” (3.2.3). In the rule “gapoṣṭak” (3.2.8; the root ‘ga’ and ‘pa’ take the suffix ’tak’ when there is a karma before it) ’tak’ with a redundant ka has been prescribed to ‘ga’ and ‘pa’ preceded by an accu- sative case. The reference to this rule, however, rather strengthens the view of the opponents rather than the respondents. Again, if by force of the ‘sikṣa’, e, ai, o, au were to remain unchanged then the prescription of ‘ay’, ay, etc. in place of e, ai, etc. would be inapplicable. In the case of the consonants it cannot be said that the purpose of ‘janer dah’ would have been served by such a rule as ‘janer gunah’, for how do we know that the guna would be a-kara and not e-kara or o-kara for the rule of proximity does not apply in case of consonants as relating to vowels. So in the case of ‘gamer dah’ if we had said ‘gamer gunah’, then we should expect o-kara by the rule of proximity in place of ma and the form ‘ga’ would not be possible. For this reason the word ‘ik’ in “iko gunavṛddhi” may be regarded as barring gunavṛddhi to other vowels and consonants. The opponent again persists in his objection and says that in the rule “diva ut’ (7.1.84) ‘ut’ is prescribed in place of the va which is a consonant and au in a vṛddhi. To this the reply is that only those places where guna and vṛddhi are prescribed by the name guna-vrddhi that we hold that such guna and vṛddhi can take place only with reference to ‘ik’. Here the rule does not say ‘divah vṛddhiḥ’, but ‘diva ut. Vṛddhi has not been prescribed by the term ‘vṛddhi’ and therefore the condition of ‘ik’ does not apply here. Again the Bhāṣya says that if instead of the rule “iko guna-vṛddhi” we had only guna-vṛddhi, then of which should we think that guna-vṛddhi has been prescribed. So by force of the proposition the ‘ik’ has to be supplied of which the ‘guna-vṛddhi take place. Here another topic arises. The opponent says if the rules “vṛddhir ādaic”, “aden gunah” follow in the next rule, or even if guna and vṛddhi follow from the previous rule then the form would be ‘aden gunah vṛddhis ca’ which would mean that at-en are both guna and vṛddhi. The answer is that vṛddhi is taken with ādaic and therefore it cannot be separated from it and therefore it cannot relate itself to ‘aden’. Here leaving at and aic we have only the continuity of guna-vṛddhi, or we may suppose that vṛddhi and guna has jumped off the previous sūtra to the sutra” iko guṇa-vṛddhi”, or the three sutras may be taken together as occurring one after another, forming one whole “vṛddhir ādaic”, “aden gunah” and “iko guṇa-vṛddhi 208 Mahābhāṣya of Patanjali and therefore there is no contunuity of any word from the previous rules, as the three may be regarded to form one compact body. If vrddhi had come by adhikara, then vṛddhi and guna would have mutually contradicted each other and we could not get at any meaning. Moreover it is the instruction that continues, but there cannot be any instruction by mere continuity. (See note of the relevant passage). Or, we may think that both guṇa- vṛddhi have ceased to be continuous. Here another topic arises: ‘ikah’ is in the genitive case and of the various meanings of the genitive case the paribhāṣā rule “ṣaṣthi sthāneyoga” accepts only the meaning in place (sthāneyoga) of called sthane-ṣasthi. The meaning of ‘in place of’ has been further modified by the rule “alo’ntyasya”-i.e. where anything is prescribed in place of a word it means in place of last alphabet of the word. But in the case of the genitive case in “ikah” we cannot accept the sthāna- sasthi rule. We may have here the case of avayava-sasthi and the genitive case applies to i, u, r, I separately as forming the avayavas or parts of ‘ik’. Is this to be regarded as an exception to ‘sthane-ṣasthi’? In discussing this point the Bhāṣya says that if the rule “iko guna-vṛddhi” was within the purview of the rule “sasthi sthāneyoga” and “alo’ntyasya”- then in the case of the principle of guna-vṛddhi with reference to ‘mrdi’, ‘mrji’ and puganta-laghûpadha verbs ‘ik’ should have to be specially mentioned. Otherwise there will be no guna or vṛddhi or vṛddhi to the ikara of ‘mida’ or the ṛ-kara of ‘mrja’ for they are not ultimates and according to “alo’ntyasya” prescription can only take place with reference to ultimates. So also in the case of many other rules, such as “puganta laghûpadhasya”, “r-drso’ni’, etc. The difference between the two views of sesa and apavāda raised before, is this-that in the interpretation of sesa the two prescription “iko guna-vṛddhi” and “alo’ntyasya ca” are to be taken as one sentence in which the meaning turns to be the guna-vṛddhi prescribed in ‘ik’ which would take place only with regard to the ultimate. In the other interpretation the two are two different sentences. First we have the unrestricted guṇa-vṛddhi of ‘ik’ and then we have “alo’ntyasya ca” and also of the ultimate. In this case we have one general rule in “iko guna- vṛddhi” and a special rule in “alo’ntyasya ca”. One stands as the general rule and the other, the exception. The difference between the two views was affecting usage and the formation of sutras is pointed out and the Värttika and the Bhāṣya say that in the former view, i.e. when it means only the ultimate, in the rule mentioned above, i.e. ‘mida-mrja’ etc. ‘ik’ should be mentioned or otherwise the vowels affected thereby by guna or Third Ahnika 209 vrddhi being non-ultimate should not be liable to such modification. Moreover in the cases of verbs not ending in ‘ik’ we should expect guna to take place as in the verbs ending in a in ‘yata’, ‘vāta.There is another difficulty also: if, as in the seṣa-view, the rule “alo’ntyasya” be read as forming one sentence with “iko gunavṛddhi” then “alo’ntyasya” would have no application to cases not ending ‘ik’ and in those cases according to “ṣasthi sthāneyoga” the augment advised should have to occur in place of the whole word (sarvadeśa). The argument is refuted by the Bhāṣya on the ground that “alo’ntyasya” was read with ‘ik’ because ‘ik’ was in the genitive. So, ‘alo’ntyasya’ would join itself to every rule that have a word in the genitive case and thus there is no chance of sarvädeśa. But then there is another objection. By the rule “mider gunah” since ‘ik’ has to be taken here as specially mentioned, the guna cannot refer to the antya vowel, but by the rule “alo’ntyasya” guna cannot refer to ‘ik’ if ‘guna’ is prescribed. Then guna must therefore be in place of the whole word ‘midi’ (sarvādeśa) which is impossible. Thus criticising the sesa or the ultimate view the Bhāṣya is trying the apavāda view. Where in “iko guna-vṛddhi” ik is taken in a general manner applying to all ‘ik’ wherever it may occur and “alo’ntyasya” is taken as a apavāda restricting the guṇa- vrddhi only to the last alphabet. But in that case guna-vṛddhi might apply indiscriminately to all cases whenever the ‘ik’ might occur. If there is guna in kartā, hartā and sarvadhātuka suffixes follow then there ought to be guna also in ihita, ihitum. To this a partial reply is given at first. There is a rule “puganta-laghûpadhasya ca” (7.3.86), i.e. when särvadhatuka and ārdhadhātuka suffixes follow, all verbs ending in ‘puk’ and with a short penultimate will have guna. Since we could have guna in all such cases by the rule “sārvadhātukärdhadhātukayoh” (7.3.84), the further injunction in “puganta-laghûpadhasya ca” mean that no guna with reference to verbs having a long penultimate as in ‘ihita’. Now, therefore, it comes to this that a restriction is on the rule “sarvadhātukārdhadhātukayoḥ” i.e. the guna prescribed there, may apply to ultimates and non-ultimates, but not to chose having a long penultimate. This is a niyama or restriction on the guna. When sarvadhātutuka ardhadhātuka suffixes follow, guna is to occur only with reference to puganta-laghûpadha vowels. An alternative suggestion is made that guna will apply to puganta-laghûpadha vowels only when sarvadhātuka, ārdhadhātuka suffixes follow. In this interpretation guna will be unrestricted with reference to sarvadhātuka and ardhadhātuka suffixes and in this case there will be chance of guna even in ‘ihita’. The two interpretations in Sanskrit stand as follows: One,210 Mahābhāṣya of Patanjali ‘puganta-laghûpadhasya eva sarvadhātukārdhādhātukayoḥ’ and the sec- ond alternative is -‘puganta-laghûpadhasya sarvadhātukārdhatukayoḥ eva. The ’eva’ is gaken in the first case with ‘puganta-laghûpadha’ narrowing the circles of verbs only to puganta-laghûpadha’ and in the second case ’eva’ is taken with ‘sārvadhatukārdhadhatukayoḥ’ and as such no restriction is made to guna when sārvadhātuka and ardhādhatuka suffixes follow. Thus puganta-laghûpadhas may have guna as well as others. In such a case the objection is that there ought to be guna even in ‘ihita’ where there is a long penultimate. The objection against this two- fold niyama that the guna in the case of jas-vibhakti becomes unrestricted and applies also to the non-ultimate. Here the Kroștriyas are supposed to make their comments. They hold that there is not a case of ‘utsarga’ and ‘apavada’ or of ‘seṣa’, but a case of vipratiṣedha. In the later Bhāṣya after some discussion the principle of vipratiṣedha is refuted. (See discussion in the previous notes). The final suggestion of the Bhasya is that in each case it is definitely mentioned to which the guna or the vṛddhi would apply. Such as instead of “mider gunah” one should read ‘midah er gunah’ and so on. After this, another discussion arises regarding vṛddhi. The opponent starts the objection that there was a difference between vṛddhi and guna. Where guna is spoken of, the sthänin, i.e. in whose place guna is to be, is not mentioned always in the sutras. But in the case of vṛddhi the sthänin is always mentioned in the sutras as in “aco’ñniti” etc. Therefore the objection is that the word ‘vṛddhi’ should not have been mentioned in the rule “iko guṇa-vṛddhi”, for vṛddhi is always directed specifically in the sutras prescribing vṛddhi. They reply is that this is done for other sutras such as “kniti ca” (1.1.5) which enjoins the prohibition of ‘vṛddhi’. Then again in the case of “mrjer vṛddhiḥ” there is no specific mention of the sthanin. Then follows a discussion- if the rule “mrjer vṛddhiḥ” could be so explained that for securing the vṛddhi or ‘ik’ in it, one should not use the word vṛddhi in “iko guna-vṛddhi”. The conclusion is that there is no necessity of using the word vṛddhi in “iko guṇa-vṛddhi” for the sake of the word ‘mrj’. The next suggestion is that vṛddhi may be taken for ‘sic’ for in the rule “sici vṛddiḥ parasmaipadeṣu” the sthanin is not mentioned and there was a chance that if it was not mentioned that vṛddhi should take place of ‘ik’ in the cause of ‘sic’, one could have vṛddhi both for ‘ik’ and for ’non-ik’. This discussion continues and the conclusion is that even for ‘sic’ there is Third Ahnika 211 no necessity for using the word ‘vṛddhi’ in “iko guna-vṛddhi”. The use of the word vṛddhi in ‘sic’ is for other purposes, i.e. to rule out vṛddhi in cases of suffixes having ka or na redundant. Then arises a criticism of this necessity leading to the discussion as to whether the antaranga and bahiranga rule should apply in the sic-prakarana and the conclusion is that in the sic-prakarana the antaranga rule does not apply and therefore the word ‘vṛddhi’ in ‘sici vṛddhih’ is for the purpose of the rule “kniti”. In that case ultimately the mentioning of the word vṛddhi in ‘iko guna-vṛddhi’ is justified. FOURTH AHNIKA Panini. na dhatulopa ärdhadhätuke. 1.1.4. No guna-vṛddhi, when elision of dhātu and ardhadhātuka suffixes). The sutra plainly means that there will be no guna-vṛddhi when roots are elided and when ardhadhätuka suffixes follow. The rules that determined sarvadhātuka and ardhadhātuka are “tin-sit sarvadhatukam”-(the tinanta suffixes and suffixes having a redundant sa are called sārvadhātuka); “ardhadhātukam seṣaḥ” (all the other suffixes are called ardhadhātuka). Suffixes were therefore twofold; ’tin’ and ‘sit’ being called sārvadhātuka and the rest ardhadhātuka. Now the question is-is ardhadhātuka an adjective to dhātulopa or not, i.e. will the prohibition of guna-vṛddhi as prescribed by the sutra apply only to those cases where the dhatu has been elided on account of ardhadhätuka suffixes and when such suffixes follow? Or whether the prohibition of guna-vṛddhi should apply when- ever an ardhadhätuka suffix follows provided the dhatu has been elided? The elision of a dhatu generally means the elision of a part of a dhātu. To state it more clearly we know that the part of a dhatu may be elided either on account of an ardhadhātuka or for other reasons. The sutra is ambigu- ous. There is the phrase ‘dhatulope’ and the phrase ‘ardhadhātuka’ and we do not know if ardhadhātuka should be made adjectival to ‘dhatulope’, or the two should remain as if separate conditions. On one meaning the pro- hibition of gunavṛddhi takes place only when such ardhadhātuka suffixes follow that have been the cause of the elision of a part of the dhatu to which guna-vṛddhi is to apply. The other meaning is that there will be no guṇa- vṛddhi when a root has lost its part for whatever reasons, when ardhadhātuka suffixes follow. The whole of the Bhasya seeks to clarify the meaning by a reference to various usages. Bhāṣya. dhātugrahaṇam kimartham ? iha mã bhūt- lūñ-’lavitā’, ’lavitum’, puñ- ‘pavitā’, ‘pavitum’, ārdhadhātuka iti kimartham ? “tridhā baddho vrṣabho roraviti. (R. V. 4.58.3)” kim punar idam ardhadhātuka-grahaṇam 214 Vārttika. Bhāṣya. Vārttika. Mahābhāṣya of Patanjali lopa-viseṣanam-ardhadhātuka-nimitte lope sati ye guna-vṛddhi prapnutas te na bhava- taḥ’ iti, āhosvit-guna-vṛddhi- viseṣaṇam ardhadhātuka-grahaṇam ‘dhātulope saty ardhadhātuka-nimitte ye guna-vṛddhi prāpnutas te na bhavataḥ’ iti? kim câtaḥ? yadi lopa-viseṣaṇam-‘uped- dhaḥ’, ‘preddhaḥ’ atrâpi präpnoti. atha gunavṛddhi-viseṣaṇam-‘knopayati’ ity atrâpi prāpnoti. yathecchasi tatha’stu. astu lopa-viseṣaṇam. katham ‘upeddhaḥ’, ‘preddhaḥ’ iti? bahirango guṇaḥ, antarangaḥ pratiṣedhaḥ “asiddham bahirangam anta- range”. yady evam, nârtho dhātugrahaṇena. iha kasman na bhavati- lūñ ’lavitā’, ’lavitum’? ārdhadhātuka-nimitte lope pratiṣedhaḥ. na caivam ardhadhātuka-nimitte lopaḥ. athava-punar astu gunavṛddhi-vise- ṣaṇam. nanu coktam-knopayatity atrâpi prāpnoti iti. naiṣa doṣaḥ. nipatanāt siddham. kim nipātanam? “cele knopeḥ” (Pā. 3.4.33) iti. parigaṇanam kartavyam. yanyakkyava-lope pratiṣedhaḥ. yanyakkyavalope pratiṣedho vaktavyaḥ. yan -‘bebhiditā, marimrjaḥ’, yak-kuṣubhitā magadhakaḥ’. kya- ‘samidhitā dṛṣadakaḥ’ valope- jiradānuḥ. kim prayojanam ? numlopa-srivyanubandha lope’- pratiṣedhārtham. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Vārttika. Bhāṣya. Fourth Ahnika 215 numlope srivyanubandha-lope ca pratiṣedho mā bhūd iti. numlope - ‘abhaji’, ‘rāgaḥ’, ‘upabarhaṇam’. sriveḥ-asremāṇam’. anu- bandhalope-lūñ ’lavita’, ’lavitum’. yadi parigaṇanam kriyate, ‘syadaḥ’, ‘praśrathaḥ’, ‘himaśrathaḥ’ ity atrâpi pra- pnoti. vakṣyaty etat-“nipatanāt syadādiṣu” iti. tat tarhi parigaṇanam kartavyam? na kar- tavyam. numlope kasman na bhavati? ik-prakaraṇān numlope vṛddhiḥ. iglakṣaṇayor guna-vṛddhyoḥ pratiṣedhaḥ, na ca eṣā iglakṣaṇā vṛddhiḥ. yadi iglakṣaṇayor guna-vṛddhyoḥ prati- şedhaḥ, ‘syadaḥ’, ‘praśrathaḥ’, ‘himaśra- thah’ ity atra na prapnoti. iha ca prāpnoti- ‘avodaḥ’, ’edhaḥ’ ‘odmaḥ’ iti. nipātanāt syadādiṣu. nipatanāt syadādiṣu pratiṣedho bhavisyati, na ca bhaviṣyati. yadi iglakṣaṇayor guna-vṛddhyoḥ prati- ṣedhaḥ, srivyanubandha-lope katham ? pratyayasrayatvad anyatra siddham. ārdhadhātukanimitte lope pratisedhaḥ. na caiṣa ārdhadhātuka-nimitto lopaḥ. yady ardhadhātukanimitte lope pratiṣedhaḥ, ‘jiradānuḥ’ (RV. 5.83.1) atra na prāpnoti. raki jyaḥ samprasāraṇam. naitaj jive rūpam, raky etajjyaḥ sam-pra- 216 Mahābhāṣya of Patanjali sāraṇam bhavati. yāvata ca idānim raki, jiver api siddham bhavati. katham upabarhanam ? brhih prakṛtyan- taram. katham jñāyate-brhiḥ prakṛtyantaram iti? acīti hi.lopa ucyate, anajādāvapi dṛ- śyate-’nibṛhyate’. anițiti cocyate, iḍādāv api dṛśyate - ’nibrahita’, ’nibrahitum’ iti. ajdāvapi na dṛśyate-brmhayati’, ‘bṛmhakaḥ’. tasman nârthaḥ parigananena. yadi parigananam na kriyate - ‘bhe- dyate’, ‘chedyate’ atrâ’pi prāpnoti. naiṣa doṣaḥ. dhātulopa iti naivam vijñā- yate ‘dhātor lopo dhātulopo, dhātu-lope — iti. katham tarhi ? ‘dhātor lopo yasmims tad idam dhātu-lopam, dhātulope’ iti. tasmād iglakṣaṇā vṛddhiḥ. yadi tarhi iglakṣaṇayor guna-vṛddhyoḥ pratiṣedhaḥ ‘magadhakaḥ’, ‘drṣadakaḥ’-atra na prāpnoti. pāpacakaḥ’, ‘papaṭhakaḥ’, Värttika. allopasya sthanivattvāt. Bhāṣya. akāralope kṛte tasya sthānivattvād guna- vṛddhi na bhaviṣyataḥ. Värttika. anarambho vā. Bhāṣya. anarambho vā punar asya yogasya nyāyyaḥ. katham ‘bebhiditā’, ‘marimṛjakaḥ’, ‘kuşubhita’, ‘samidhita’, iti? atrâpy akara- lope kṛte tasya sthānivadbhāvād guṇa-vṛddhi na bhaviṣyataḥ. yatra tarhi sthānivadbhāvo nâsti tadartham ayam yogo vaktavyaḥ. kva ca sthānivadbhā- Vārttika. Fourth Ahnika 217 vo nâsti? yatra hal-acorādeśaḥ - ’loluvaḥ’, ‘popuvaḥ’, ‘marimrjaḥ’, ‘sarisṛpaḥ’ iti. atrâpy akāralope kṛte tasya sthanivad- bhāvād guṇavṛddhi na bhaviṣyataḥ. luki kṛte na prāpnoti. idam iha sampra- ’luk kriyatām allopaḥ’ iti. kim dharyam — atra kartavyam? paratvad allopaḥ. nityo luk. kṛte’py allope präpnoti, akṛte’pi präpnoti. lug apy anityaḥ. katham ? anyasya kṛte’l- lope prapnoti, anyasyâkṛte. “śabdantarasya ca prapnuvan vidhir anityo bhavati”. anavakāśas tarhi luk. savakāśo luk. ko’vakāśaḥ ? avasiṣṭaḥ. athâpi kathamcid anavakāśo luk syād evam api na doṣaḥ. allope yoga-vibhāgaḥ kariṣyate “ato lopah” (ato lopo bhavati). tato “yasya”, yasya ca lopo bhavati. ‘ata’ ity eva. kim artham idam ? lukam vakṣyati tadbadhanartham. tato “halah”, hala uttara- sya yasya ca lopo bhavati iti. ihâpi tarhi paratvad yoga-vibhāgād vā lopo lukam badheta-“kṛṣṇo nonava vrṣabho yadidam” (R. V. 1.79.2). nonuyater nonāva. samānāśrayo lug lopena badhyate. kaś ca samānāśrayaḥ ? yaḥ pratyayasrayaḥ, atra ca prāg eva pratyayotpatter lug bhavati. katham -‘syadaḥ’, ‘praśrathaḥ’, ‘himaśrathaḥ’, jiradānuḥ’, ’nikucita’ iti? uktam śeşe. 218 Bhāṣya. Mahābhāṣya of Patanjali (śeşe uktam) kim uktam ? “nipātanāt syadādiṣu”, “pratyayāśrayatvad anyatra siddham”, “raki jyaḥ samprasaraṇam” iti. nikucite’py uktam. kim? “sannipata-lakṣaṇo vidhir animittam tadvighātasya” iti. (na dhatulopa) Why is the word ‘dhatu’ (in the rule na dhatulopa etc.)? So that there may not be in such cases as lūñ, lavitā, lavitum, pūñ, pavitā, pavitum.145 Why has the word ‘ardhadhātuka’ been taken?“tridha baddho vrṣabho roraviti” (Bound in three directions the bull roars)146 Had the word ‘ardhadhātuka’ been taken to qualify lopa? (i.e. is it to be interpreted) as that there will be no guna-vṛddhi when there is a lopa in consequence of an ardhadhātuka suffix (or is to be interpreted as when there is a dhatulopa then the guna-vrddhi consequential to ardhadhātuka
145. There is a maxim that the grammarians consider it a source of great joy if they can do away with even half a syllable in a sutra. Brevity is the sole principle in the construction of the sutras. In the sutra here there is a word ‘dhatulope’. The question is: could not have been the word ‘dhatu’ left out? Even if the word ‘dhatu’ was not taken, the ardhadhātuka might have been taken as adjectival to lopa and the meaning would have been ‘when in cnsequence of an ardhadhätuka suffix there is an elision? This was sufficient for the purpose. The word ‘dhatu’ could not have been taken to exclude the directive redundant syllables (anubandha or it), for it has been decided later on that the elision of such anubandhas should not regarded as the elision of a part of a dhātu. Now in reply it may be said that the inclusion of the word ‘dhatu’ shows that the word ‘ardhadhātuka’ does not refer to ‘dhatulope’ at its adjective, but as the adjective to guna-vṛddhi, i.e. there will be no guna-vṛddhi when ardhadhātuka suffixes follow, provided there is dhatulopa. Or in other words ‘dhatulope sati na ardhadhātuke guṇa-vṛddhi’. Under such a meaning it is obvious that dhātu has been taken along with lopa. Otherwise there will be doubt as to what kind of lopa is it that is to be regarded as a condition of a guna-vṛddhi when ardhadhātuka suffixes follow. We know that an anubandha is not a part of a dhatu, for they have been separated from the dhatu by the fact that they have been designated as ‘it’ in the original instruction of dhatus. Therefore in the root ’lun’ the root is ’lu’ and there is a ’trc’ suffix. So also is the case with pūñ where ‘pu’ is the root. tre is an ardhadhatuka suffix; but still there is guna, for there has been no ‘dhatulopa’. It has been already said that the elision of the anubandha does not constitute a dhatulopa, for anubandha is not a part of the dhātu.
146. Now, ‘roraviti’ is from the root ‘ru’-‘rauti’ -‘roars’, and the word ‘roraviti’ is formed by a ’tin’ suffix which is sarvadhātuka and, therefore there has been guna. Had the ardhadhātuka been not mentioned then the prohibition of guna would have been extended to sarvadhātuka suffixes as well. Now again another criticism follows. Fourth Ahnika 219 suffixes will not take place. Here the word ardhadhātuka has been taken as an adjective to guna-vṛddhi.147 What then? If adjectival to lopa, the prohibition applies also in ‘upeddha’, ‘preddha’. If it is an adjective to guna-vṛddhi, then it ought to apply in such cases as ‘knopayati’.148 The guna is bahiranga and the prohibition is antaranga and there is the
147. The reader should note that the discussion alluded to in the preamble to our translation, commences from here. The views of two different parties have been clearly stated in the Bhāṣya.
148. Upa + idhi + kta = upeddha; pra + idhi + kta = preddha; ‘ñiindhi diptau’ = idhi means ’to shine’. The root ‘idhi’ gets a num augment by the rule “idito num dhatoḥ” (7.1.58) (the roots that have i as an anubandha gets the augment num of which na remains). The root ‘idhi’ now becomes ‘indhi’, upa + indhi or indh + kta; ‘kta’ is a k-it suffix and an ardhadhatuka and by the rule 6.4.24 the na is elided in association with an ardhadhātuka suffix that has a redundant k as in kta. The form is changed thus: upa+idh + kta = upeddha- by euphonic rules. Now the point is that the a of upa and i of iddha has both together been replaced here by a guna according to the rule “adgunah” (6.1.87; when a vowel follows a or a, than the vowel together with the a or a is replaced by one guna, tava + idam = tavedam). The commentators have raised points of discussion. The first point is that the prohibition of guna as directed here applies only to guna as occurring in place of ‘ik’, for there is a contininuity of the rule “iko guna-vṛddhi” in this rule. But it is restricted in its sphere by the present rule “na dhatulopa” etc. “iko gunavṛddhi iti sampūrṇam anuvartya yatredam tatra ’na dhatu’ ity upatisthate”-Uddyota). It therefore follows that there was no scope of guna here which could have been prohibited by this rule. It is true that in association with an ardhadhātuka suffix there has been a dhatulopa (as the na in ‘indha’ has been elided in association with ‘kta’ which is an ardhadhātuka). But the rule by which guna has taken place here, is, “ad gunah” by which a of upa and i of ‘indha’ have been replaced by a guna. This rule could not have prohibited guna by “ād gunah”, for that the ascription of guna there is not in place of an ‘ik’, and it is only the guna that occurs in place of an ‘ik’ that can be prohibited by the present rule. Hence this example is quite out of place. Guna could not have been arrested here under any circumstances. In defence of the illustration of the writer of the Bhāṣya it is said that this point need not be pressed, for the author of the Bhasya will give quite another explanation later on, and here is only a provisional criticism by the opponent. Dikṣita in his Sabda-kaustubha makes another relevant discussion. Assuming for argument that this rule could prohibit guna by “ad gunah” he raises the objection that the possible inadmissibility of ‘upeddha’ and ‘preddha’ here as raised in the Bhāṣya, may indeed be regarded as unjustifiable. Thus there may be two views: One, that the root ‘idh’ as first combines with the suffix forming the word ‘iddha’, the upasarga ‘pra’ then combines with and ‘preddha’ is formed; two, the upasarga ‘pra’ combines with the i of ‘idh’ and we have ‘predh’ and this combines with ‘kta’ to form ‘preddha’. In the first case the elision of na of ‘indha’ is due to ‘kta’ and is220 Mahābhāṣya of Patañjali maxim “asiddham bahirangam antarange” (Paribhāṣā 50 of the Paribhāṣendusekhara).” 149 If it is so, the word ‘dhatu’ should not have been taken (i.e. if the word ‘ārdhadhātuka’ is an adjective to dhatulope, then the word ‘dhātu’ should not have been taken). Then why not here? (i.e. why does not the prohibition apply here)- in forming ’lavita’ and ’lavitum’ in ’lūñ’. The prohibition is only according to the interpretation adopted with reference to ardhadhātukas therefore an antaranga vidhi. In the second case, guna becomes antaranga and in both such cases a guna cannot be arrested. In the second case, the elision of na is a bahiranga vidhi and therefore the prohibition depending upon the bahiranga is also bahiranga itself and it cannot therefore supersede the antaranga vidhi of a guna. In the first place the guna is bahiranga and therefore it is superseded by the antaranga and therefore it does not exist when ‘iddha’ has been formed. It is a later operation and therefore, since it does not collide with the antaranga, the consequential operation on the antaranga vidhi cannot affect it. So here also the guna takes place. It is true that there has been no guna of i at the time of the formation of the word ‘iddha’, but if ‘iddha’ later on combines with pra to form ‘preddha’, how can the application of the ardhadhātuka suffix arrest that guna which occurs in places of the a in pra + i in iddha. This combination is entirely irrelevant to the subject under discussion. Haradatta has suggested that this paribhāṣa of antaranga and bahiranga has no application here by the Paribhāṣā ’näjänantarye bahistva-praklptiḥ’, i.e. when there is an operation con- cerning the replacement of two alphabets of two padas by one alphabet, then that becomes a bahiranga-vidhi and for this reason the guna in preddha is a bahiranga- vidhi and there that bahiranga cannot be overridden by the antaranga vidhi. But Haradatta’s view is criticised by both Kaiyața and Dikṣita. Kaiyața says that this paribhāṣā, i.e. najanantarye’ should be so explained as to have application only in those places where a prescription is made regarding vowels and their prximity in a later operation and thus it has no scope here. It therefore has application in such sutras as “iko yan aci” (6.1.77). (See Pradipa and Sabda-kaustubha for greater elaboration of this discussion which for our purpose is unnecessary). We have now to turn to ‘knopayati’. There is a rule “arti-hri-vli-ri-knuyi-kṣmā- yyātām punnau” (7.3.36). These verbs get the augment puk when ’nic’ follows. The root ‘knuyi’ has i as it or anubandha. So the root stands as ‘knuy’, when ’nic’ follows as pa comes between the i of nic and after the ya of ‘knuya’. Then by the rule “lopo vyor vali"the yais elided. Therefore we have a case of dhatulopa and an ardhadhatuka suffix. Therefore according to the second interpretation (i.e. when dhätulopa and ārdhadhātuka are separate conditions or in other words, the ardhadhātuka is an adjective of gunavṛddhi) there is an occasion of the prohibition of guna-vṛddhi. But there is guna here. We shall not further discuss this point here, as we shall have to take it up immediately later.
149. This remark is with reference to ‘upeddhah’ and ‘preddhah’. The relevant discussion has already been made in the note of that word in the previous section. Fourth Ahnika 221 that produce an elision in the dhātu. But here in ’luñ’ the elision is not on account of a ardhadhātuka suffix. 150 Then let us take it (ardhadhātuka) as an adjective to ‘guna-vṛddhi’ (the second alternative meaning). But the objection had already been raised that in that case the prohibition should apply and we cannot have the form ‘knopayati’. This is not a fault. ‘Knopayati’ can be supported by the special nipatana rule. Where is the nipatana? In the rule “cele knopeḥ” (3.4.33; i.e. when the root ‘knuy’ in the sense of making noise or moisten and associated with the suffix ’nic’ has for its accusative a word denoting cloth. Then it takes the ’namul’ pratyaya, if by it the extent of rain is understood. Here the word ‘knopa’ has been specially allowed and arranged for by its being mentioned in the sutra. So the prohibition of guna here is arrested by force of this nipātana). It seems that it is desirable to mention the list of those suffixes which produce elision in the dhātu. (The Varttika intervenes and says) - that the prohibition applies with reference to those ardhadhātuka suffixes that produce the elision of ‘yan’, ‘yak’, ‘kya’ and ‘va’. (The Bhāṣya interprets it as follows) - The prohibition applies to the elision of ‘yan’, ‘yak’, ‘kya’ and ‘va’. In ‘yan’-’ vebhidita’,‘marimṛjah’, ‘yak’ -‘kuşubhita magadhakaḥ’, ‘kya’‘samidhita dṛṣadakaḥ’, ‘va’-jiradānuḥ’. 151 What is the use?
150. The ñ of ’lun’ is the antaranga and has been marked as ‘it’ in the Panini’s rule 1.3.3 “hal-antyam”. It is according to that rule that it is elided. It is thus really a case of anubandha-lopa and not dhatulopa and has already been stated to be so in the early part of the Bhasya of this sutra.
151. The va taken here for elision, is the a-käranta va and not merely the consonant. It is by this means that ‘srivi’ is excluded from the va-eliding group. The word ‘vebhidita’ is derived by applying ’trc’ to the word ‘vebhidya’, we have ‘it’ here and the ya of vebhidya is elided by the rule “yasya halah” (6.4.49). When ardhadhätuka suffixes follow then the ‘ya’ coming after a consonant is elided. In the word ‘marimrja’ we have first ‘yan’ by the rule “yaño’ci ca” (2.4.74)- when a suffix with a vowel initial follows then the ‘ya’ is elided. So from the yananta marimṛjya we have marimrjah. The words ‘kuşubha’ and ‘magadha’ fall under ‘kandvadi’ list, and we have ‘yak’ by the rule “kandvadibhyo yak” (3.1.27). We have ’trc’ suffix in the case of ‘kuşubhita’ and ’nvul’ in the case of ‘magadhakah’ and the ‘ya’ in both cases is elided by ‘yasya halah’ mentioned above. In the case of the words ‘samidhita and ‘drṣadakah’ the verbs ending in ‘kyac’ get the suffixes ’trc’ and ’nvul’ respectively and the ‘ya’ of ‘kyac’ is optionally elided by the rule “kyasya vibhāṣā” (6.4.50). 222 Mahābhāṣya of Patanjali (Varttika:) In order to restrict the prohibition in the cases of the elision. of ’num’, ‘srivi’, and the elision of anubandha. (The Bhāṣya interprets it as follows)-So that the prohibition may not apply to where the elision of ’num’ anubandha takes place. The case of ’num’-lopa is illustrated in ‘abhaji’, ‘ragaḥ’, ‘upabarhanam’. In the case of ‘srivi’-āsremāṇam. In the case of the elision of anubandha as in ’num’ -lavita, lavitum. 152 If this list is made then it applies here also. In syadaḥ (speed), praśrathaḥ (laxity), himaśrathaḥ (the moon).153 These will be provided for as the cases of ‘syadaḥ’, etc. are to be solved by nipatana. Then the list has to be made. No. Why there then the prohibition should not apply in the case of num-lopa. (The Varttika says)- Vṛddhi takes place in the case of num-lopa, because the prohibition only applies to vṛddhi in place of ‘ik’. (Bhāṣya:) The vṛddhi occurs only in place of ‘ik’, but this is not a vṛddhi in place of ‘ik’. If the prohibition is to take place with reference to the guna-vṛddhi of ‘ik’, then in ‘syadaḥ’, ‘praśrathaḥ’, ‘himaśrathaḥ’ the
152. The ’num’ is the name for na. The examples of ’num-lopa’ or na-lopa is to be made in the cases of ‘bhañja’ and ‘rañja’ by the rules “bhañjeś ca ciņi” (6.4.33), “rañjeś ca” (6.4.26) and “ghañi ca bhavakaraṇayoḥ” (6.4.27) i.e., the na of ‘bhañja’ and ‘rañja’ is elided when the suffixes ‘cin’ and ‘ghan’ follow. In the case of ‘upabarhanam’ the root is ‘bṛmhi’ - vṛddhau, it being an ‘id-it’ root, it gets ’num’ by the rule “idito numdhatoḥ” (7.1.58). By the Värttika rule “brmher acyaniti”-when an ‘anit’ vowel suffix follows, then the na of ‘brmhi’ is elided. The form ‘äsremāṇam’ is also read as ‘asremāṇam’. Here we have the ‘manin’-suffix to ‘srivi’ and the ‘ma’ is elided. In all these cases we have the elision of a part of the ‘dhātus, and if this list of cases where the prohibition does not apply, is mentioned, and also the cases where the prohibition applies, are stated, then there is no necessity of using the word dhātu in ‘dhatulope’. But in that case the non-prohibition in the case of anubandhalopa has to be specially mentioned. The va in ‘srivi’ being in upadha, should have been changed into ‘ut’ by the rule, “jara-tvara-srivy-avimavām upadhāyās ca” (6.4.20). But this being a Vedic usage, such a transformation does not take place here.
153. The word ‘syadaḥ’ has been formed from the root ‘syand’ with the suffix ‘ghan’ in the sense of speed by the rule “syadau jabe” (6.4.28) which also directs by nipatane, the elision of na and the prohibition of vṛddhi as an exception to the rule ’na dhatulope’ etc. The words ‘praśrathah’ and ‘himaśrathah’ are also formed by the rule “avodai-dhaudma-praśratha-himaśrathaḥ” (6.4.29) with the suffix ‘ghan’ and na is elided and the prohibition of guna as provided for “na dhatulope” etc. does not occur by virtue of this nipātana. Fourth Ahnika 223 prohibition of vṛddhi should take place and in the case of avodaḥ, edhaḥ, odmah no such prohibition should take place. (Varttika:) These cases of syadaḥ etc. are solved by nipatana. (Bhāṣya:) The prohibition should take place in the case of svadādi by nipäta, and it would not take place. 154 (Varttika:) Since the prohibition applies in the case ardhādhatuka suffixes, it does not apply elsewhere. (Bhāṣya:) The prohibition applies only to those cases where the elision of a part of a dhatu is consequent upon the ardhadhātuka suffix that follows. The elision of srivi and anubandha is not consequent upon an ärdhadhatuka suffix. If the prohibition applies only to the elision of verbal parts consequent upon an ardhadhātuka suffix, then in the case of ‘jiradānuḥ’ the prohibi- tion ought not to apply. (The Varttika says)- There is a samprasarana of the root ‘jya’. (Bhāṣya:) This cannot be derived from jiva. It is formed by the samprasarana (expansion) of the ‘jya’ when the suffix ‘rak’ suffix follows. If it can be formed from ‘jya’ with the ‘rak’ suffix, it can also be formed from jiva.155
154. Another objection is raised: if the guna-vṛddhi refers only with reference to vowels i, u, r, !, of the ‘ik’ order, then how do you manage the elision of ‘srivi’ and the anubandha?
155. An objection is raised that the prohibition of guna and vṛddhi is not really limited to the suffixes of the ardhadhātuka class that induces the elision of the part of a root; for in the formation of the word ‘jiradānu’, there is the prohibition of guna and vṛddhi and yet jiradānu is formed from ‘jiva’ with the suffix ‘radanu’. To this it is replied that jiradānu is formed from the root ‘jya’ with the suffix ‘rak’ by the samprasarana of jya (ig yaṇaḥ samprasaraṇam 1.1.45). But in the same way we can have jiradānu from the word ‘jiva’ by securing the elision of va by “lopa vyor vali’ (6.1.66) when the suffix ‘radānu’ followed. The objection raised is that the elision va, a part of the root ‘jiva’ is by the rule “lopo vyor vall”. This elision is not consequent upon an ardhadhātuka suffix, but still there is prohibition of guna. The reply was that the word ‘jiradānu’ is formed from the root ‘jya’ by samprasarana and hence the above objection does not hold good. Nägeśa comments here that we can form the word from ‘jiva’ also. The prohibition of guna-vṛddhi will then be due to the värttika, but in the case of the elision of ‘va’ (whatever may be the cause) there will be the prohibition of guna-vṛddhi. This further shows that the prohibition of guna-vrddhi should not be restricted only to those cases where the part of the root has been elided in consequence of an ardhadhātuka suffix. 224 Mahābhāṣya of Patanjali How to derive upabarhanam? This is another verb. How do you know ‘brhi’ is another verb ? The elision of na of ‘brmhi’ is advised before a suffix beginning with a vowel (brmher acy aniti), but we find the elision of na even before suffixes not beginning with a vowel as in nibrhyate, it is conditioned by the absence ‘it’, but the elision of na takes place here when the ‘it’ augment follows as in nibarhitā, nibarhitum. Again contrary to the direction the elision does not take place when it is followed by a suffix beginning with a vowel as in ‘brmhayati’, ‘brmhakah’. Therefore it is not mere counting any list (of operations where the part of the root is elided). But if such lists of operations were not counted, there ought to be prohibition of guna-vṛddhi in ‘bhedyate’, ‘chedyate’.156 This is no fault. Root-elision does not mean the elision of root as being equivalent to root elision and then in root-elision-dhatulope. How then? That in which there is elision of root: the vṛddhi is to refer to the vowels within the ‘ik’-group.157 If the prohibition of guna-vṛddhi must refer only to vowels of the ‘ik’- group, then the prohibition ought not to apply in such cases, as ‘papacaka’, ‘papathaka’, ‘magadhaka’, ‘drṣadaka’.158
156. The form ‘bhedyate’ is derived as bhid + nic + yak+ te. We have ‘yak’ (an ärdhadhātuka suffix by the rule “sarvadhatuke yak” 3.1.67) and on account of the i of ‘bhedi’, is elided. We have thus here a case elision of a part of the dhatu by an ärdhadhātuka suffix, but still there is no prohibition of guna as in bhedyate.
157. If it is taken in the bahuvrihi compound then the prohibition of guna-vṛddhi will take place under two conditions- firstly, that the elision of the root should be due to the ardhadhātuka suffix and secondly, the guna-vṛddhi can only be of the vowels i,u,r, or the ik-group.
158. In papacaka and papathaka the vṛdhiis prescribed by the rule “ata upadhāyāḥ” (7.2.116); the penultimate a of verbs suffers vṛddhi when suffixes having a redundant ñor n follow. Magadhaka is formed as magadha+yak (kaṇḍvādibhyah) + nvul, ya is elided by the rule “yasya halah” (6.4.49); drṣadaka is formed as drṣad+kyac+nvul, the ya of kyac is optionally elided by the rule “kasya vibhāṣā” (6.4.50). Let us take the problem in papacaka; the question is- there being no ‘ik’ here, vṛddhi cannot be arrested by “na dhatulopa” and therefore we should have vṛddhi by “ata upadhayaḥ”. To this the reply is that we have the yananta verb papacya+nvul; the mere consonant y is elided by the “yasya halah” and we have papaca - aka. The ultimate a of papaca is elided by the rule “ato lopah”. By the maxim of sthani the elision of a is to be regarded as not having taken place. The root should appear as if it were papaca and there is no upadha vowel under such circumstances and the rule “ata upadhāyāh” (7.2.116) has no field of application. If, however, in the root ‘papacya’ the ‘yan’ (and not merely the consonant) is elided, we have vṛddhi and the form with nvul’ is papacaka. Fourth Ahnika 225 (Varttika here says)-Because the elision of a is to be regarded as having the portion of a sthani. (Bhāṣya:) When a is elided, this elision having regarding as being a sthanin, there is no guna-vṛddhi. (sthāninā tulyam vartate iti sthānivat, i.e. that which has the same status as a sthānin. Sthānin is that in the place of which something is directed). (See also rule 1.1.56). (The Bhāṣya:) Or rather it is proper not to have made this sutra. How then bebhidita, marīmṛjaka, kuşubhita, magadhaka, here also when a is elided, that is in whose place it being a sthanin, there is no guna and vṛddhi. The sutra has then to be made for those places where there is no sthänivadbhava. Where the sthanivadbhāva does not apply? Where both the consonant and vowel are elided (as in the case of the elision of the whole ‘yan’) as in loluvaḥ, popuvaḥ, marimrjah, sarisrpaḥ.159 Here also since a has been elided, there is scope for sthanivadbhāva and consequently no guna-vrddhi.160 If ‘yan’ is elided, there is no scope for the elision of ya (by ato lopaḥ). This should be attended to- elision is directed, have the elision of a by “ato lopaḥ”, because it is the later rule.161 The luk must take place always, (i.e. it cannot be superseded by any other rule). Even if you have the luk of ‘yan’ (2.4.74) you have the elision of a by “ato lopah” (6.4.48). But even luk is not unobstructed, the luk is for one purpose and the elision of a is for another purpose. A prescription that applies to other words, cannot be obligatory. The luk therefore has no scope of operation. The luk has sphere of application. Where? Else- where.162
159. The words ’loluva’ etc. end in ‘yan’ and then where the suffix ‘ac’ by the rule “pacadyac” occurs, the ‘yan’ is elided by the rule “yoño’ci” (2.4.74). In such a case there is no scope of sthanivadbhāva.
160. There is a maxim “ekadeśavikṛtam ananyavat”- that which is changed in part, should be regarded as if it were unchanged and hence the ya of ‘yan’ may be looked upon as being as much a pratyaya or suffix as a, and hence the elision of ya need not be treated as an example of dhatulopa.
161. The rule for the elision of ‘yan’ is 2.4.74 and the rule for the elision of a is 6.4.48. The objector says that the last being a later rule, should have preference by the rule “vipratiṣedhe param karyam” (1.4.2).
162. The drift of all this argument is that the author of the Bhasya thinks that the sutra “na dhatulopa ardhadhātuke” is unnecessary, for its purpose can be served by the rule “ato lopah”. There is a maxim among the grammarians that of the three sages- Panini, Katyayana, and Patanjali that contributed to the completion of the 226 Mahābhāṣya of Patanjali If for some reason the luk has no sphere of application, then also there is no harm. (We have first) the elision of a, then of y by ‘yasya’ in “yasya halah”. By this the luk is contradicted (and its purpose served) and then ‘after the consonant’ (halah), i.e. a ya occurring after a consonant is elided. Then here also either on account of having a later rule or because of splitting the rule (yoga-vibhāga) the elision by the word ’lopa’ (in Pāṇinian school, the younger sages have greater authority than the older ones. Patanjali being the youngest, his authority is greater than both Katyayana and Pāņini and Katyāyana’s authority is less than that of Patanjali but greater than that of Pāṇini. (yathottaram muninām prāmānyam). The summary of the situation following the Sabda-kaustubha of Bhattoji Dikşita: The most important application of the rule “na dhatulopa ardhadhātuke” is in the formation of such forms as loluva from loluya + ac. The argument in favour of retaining the sutra is that the ‘yan’ of ’loluya’ is elided by the rule “yano’ci ca” (2.4.74). This ‘yan’ being a part of the dhatu and since it has taken place in consequence of the ardhadhātuka suffix ac, the condition for the prohibition of guna according to the rule “na dhatulopa” etc. has been satisfied and consequently the rule is justified, for without this rule we could not secure the prohibition of guna-vṛddhi here. The Bhāṣya in repeating the necessity of retaining this sutra says that the prohibition of guna here can be secured in another manner and therefore there is no necessity of having this rule. There are two rules-“ato lopaḥ” (6.4.48) and “yasya halah” (6.4.49). Now the suggestion of the Bhasya is that the rule “yasya halah” may be split up as yasya and halah; with yasya we can attach “ato lopaḥ” by continuity (anuvṛtti). The meaning then would be ‘yasya ato lopaḥ’, i.e. the a of ya is to be elided. There being later rules (6.4.48-49) to “yaño’ci ca” (2.4.74), it takes preference and the a of ‘yan’ is elided by this rule. By the rule “yano’ci ca” then only the consonant ya is elided. ‘yan’ was a part of the root ’loluya’, but not the mere consonant, then also that ya is elided not because it is immediately folowed by any ardhadhātuka suffix, but it has immediately after it only its own a which, though elided, should be regarded as existing by the maxim of sthänivadbhava. Thus guna will be prohibited as in vebhidita and the sutra “na dhatulopa” etc. will not be necessary. The uvan also has been directed in association with the a of yan which though elided, is regarded as being present in accordance with the ‘sthanivadbhāva’ maxim. The Bhasya had at first fought in favour of yan luk, but had eventually to give up the contention. His first argument in favour of yan luk was that a luk has a superior prerogative and it can even arrest an antaranga operation as “antarangan hi vidhin luk badhate”. But then “ato lopah” is a later rule and shows a rift to multiply the preceding rule ‘yaño’ci ca’. But then here is no case of multiplication ever, for “ato lopah” refers to the elision of a, whereby “yano ‘ci ca” the mere consonant ya is elided. The two are for two purposes and they do not contradict each other, but only restricts the scope of the other. The Bhasya points out that luk of ‘yan’ takes place even if there is the elision of a and if there is no elision of a. Such a prescription is not absolutely obligatory, for there is a paribhāṣā No. 43-“Sabdantarasya ca präpnuvan vidhir anityo bhavati”. This paribhāṣā means that a prescription which works in one connection as well as in other, is not obligatory. This paribhāṣā is particularly discussed in 1.3.60. Fourth Ahnika 227 “atolopaḥ’) will restrain the elision prescribed by luk (in “yaño’ci ca”) as in kṛṣṇo nonāva vṛṣabho yadīdam; nonuyate, nonāva. In such a case there will be no vṛddhi by “aco’ñņiti” (7.2.115)-when a suffix with a redundant ñ or ņ follows, the ultimate vowel of the verb undergoes vṛddhi, because the a, though elided, is to be regarded as existent by the ‘sthanivat’ maxim. In the same context one ’luk’ is restrained by ’lopa’. What is meant of ‘same context’ ?- As pertaining to some suffix. Here, however, the luk takes its place before the suffix comes. How then-syadaḥ, praśrathaḥ, himaśrathaḥ, jiradānuḥ, nikucitaḥ? (The Varttika says here)- it has been said about the rest. (Bhāṣya:) What has been said-The words ‘syadaḥ’, etc. are explained by nipatana, with the suffix ‘rak’, jya undergoes samprasarana. The maxim of nipātana will apply to all those places where the operation is in- dependent of the suffix. So also in the case of nikucita. A prescription, that is to cause a combination, cannot also be the cause of its description.163 Pāṇini. kniti ca 1.1.5. Also when suffixes with redundant k and ǹ follow (there is neither guna nor vṛddhi). The main purpose of the discussion in the Bhāṣya and the Värttika is to discover whether the prohibition of guna-vṛddhi is with reference to the immmediately preceding ‘ik’ or to ‘ik’ which is not immediately poxi- mate. Kaiyața raises the distinction between a prohibition as applying to the act of prohibition or to the objects of prohibition. In the first case it would have applied to the immediately proximate vowel, whereas in the latter case it would apply to wherever there is the occasion and as much would apply also to the ‘ik’ in the penultimate. The seventh case in ‘kniti’
163. It is suggested that in ’nonava’ the ‘yan’ is elided before the suffix is added- bahula-grahaṇānuvṛttyä anaimittiko “antarango luk prag eva pratyayotpatter bhavati”-Kaiyața. Nikucita-the na of ‘kuñca’ is elided by the rule “aniditām” etc. (6.4.24) when knit suffixes follow. Now when the ‘kta’suffix follows, the root ‘kuñc’ becomes ‘kuc’ and thus becomes ‘udupadha’ i.e., comes to have u in the penultimate and thus comes under the purview of the rule “ud-upadhad bhāvā- dikarmano’nyatarasyam” (1.2.21), i.e. a root with u in the penultimate, may be followed by a ’nistha’ suffix. But this ’nistha’ also should be regarded as becoming ‘kit’ only optionally. Here the root ‘vit’ which becomes ‘ud-upadha’ by virtue of being a ‘kit’ cannot itself be the case of its not being a ‘kit’ by the paribhāṣā “sannipata-lakṣaṇo vidhir animittam tad-vighatasya” a prescription which is the cause of a particular operation, cannot itself again be the cause of multiplying it. 228 Mahābhāṣya of Patanjali is ‘bhave saptami’ which means ‘consequent upon’. Prohibition thus refers to the object of prohibition, that is, the ‘ik’, the meaning of the seventh case is nimitta, i.e. causal agent. The meaning thus is when a suffix which is ‘knit’ follows as the causal agent, there will be guna and vrddhi of the relevant ‘ik’ whether immediately proximate or not. Värttika. kniti pratiṣedhe tannimitta-grahaṇam upadha-roravityartham. Kniti should be taken as a nimitta for extending the prohibition to the penultimates and for forming the word ‘roraviti’. Bhāṣya. Värttika. Bhāṣya. kniti pratiṣedhe tannimitta-grahaṇam kar- tavyam. knin-nimitte ye guṇavṛddhi prāpnutas te na bhavata iti vaktavyam. kim prayojanam? upadha-roravityartham. upadhārtham roravityartham ca. upadhār- tham tavat-‘bhinnaḥ’, ‘bhinnavān’ iti. kim punaḥ kāraṇam na siddhyati? ‘kniti’ ity ucyate. yatra knity anantaro guṇabhāvīgasti tatraiva syāt - ‘citam’, ‘stu- tam’. iha tu na syat- ‘bhinnaḥ’, ‘bhinnavān’. nanu ca― yasya guna ucyate, tam knitparatvena viseṣayiṣyāmaḥ. puganta- laghûpadhasya cângasya guna ucyate, ‘bhin- nam’, tac câtra knitparam. puganta-laghûpadhasya iti naivam vijñā- yate-pugantasya angasya laghûpadhasya ca iti. katham tarhi ? puki antaḥ, pugantaḥ. laghvi upadha laghûpadha. pugantas ca laghûpadha ca puganta-laghûpadham, puganta-laghûpadhasya iti. avaśyam caitad evam vijñeyam. angaviseṣaṇe hi sati ihâpi prasajyeta-bhinatti’, ‘chinatti’ iti. Fourth Ahnika 229 roravityartham ca. “tridha baddho vṛṣabho roraviti.” yadi tannimitta-grahanam kriyate, śacan- ante doṣaḥ-‘riyati’, ‘piyati’, ‘dhiyati’, ‘prādudruvat’, ‘prāsusruvat’, atra na prāpnoti. In the prohibition depending upon the following of ‘kniť it should not be declared that the ‘knit’ suffix here is the nimitta or a causal agent. The guna and vṛddhi dependent upon the ‘knit’ suffix, should not take place and this is for the purpose of extending the prohibition to the penultimate and for the formation of the word ‘roravi’. For the penulti- mate as in ‘bhinna’, ‘bhinnavān’. (Here the ‘ik’ of ‘bhid’ is a penultimate, but, thus though it is immediately proximate to the ‘knit’ suffix ‘kta’ yet there is the prohibition of guna). For what reason does not the prohibition apply. Because the sūtra ‘knit’, i.e. where there is occasion of guna in the ‘ik’ immediately preceding ‘kniť suffix, so the prohibition would apply only to ‘citam’, ‘stutam’. Now it may be asked-we shall consider the ‘knit’ to refer to all such cases where there is occasion for guna. Guna is prescribed for puganta- laghûpadha (roots ending in ‘puk’ and having a short penultimate- see 7.3.86). (Two ways of interpretation of the rule are attempted, and the first is repeated). The meaning of the rule “puganta laghûpadhasya ca” (7.3.86) is not to be considered as- ‘of puganta roots and of the roots ending in short penultimate’. What then? Puganta means ending in ‘puk’; laghûpadha means, short penultimate; puganta-laghûpadha means puganta and laghûpadha, belonging to puganta-laghûpadha vidhi. It must be so, for if it referred separately to parts, then there should be guna even in ‘bhinatti’ and ‘chinatti’. (The first point raised is that if it referred to the whole of puganta- laghûpadha, then it would not apply to where there is merely the ‘ik’. In ‘bhinatti’, ‘chinatti’, the ‘ik’ is not a penultimate). And also for ‘roraviti’ (the root is a yananta word, the guna is not due to ‘yan’, but to the ‘sarvadhatuka’ suffix. If it is taken in the sense of nimitta as suggested, then there will be difficulty in case of the words- ‘riyati’, ‘piyati’, ‘dhiyati’, ‘prādudruvat’, ‘prāsusruvat’. (The roots ‘rï’, ‘pi’, and ‘dhi’ are tudādi verbs and they get the vikarana ‘sa’. Here the i230 Mahābhāṣya of Patañjali of the root is separated from the tin suffix by the vikarana ‘sa’, but still the prohibition of guna applies. But the point as explained by the Bhāṣya in the immediately later sentence is that the augment of ‘iyan’ by the rule “aci śnu-dhātu” etc. (6.4.77) is antaranga and the ’tin’ is bahiranga. So ‘iyan’ first takes place and when it has once taken place, the i is no longer an upadha and therefore prohibition by this rule is quite unnecessary. (The objection was that if ‘knit’ be regarded as a nimitta, even then it cannot solve riyati etc.). (To this objection the Värttika says)- Värttika. Bhāṣya. śacañantasyântaranga-lakṣaṇatvāt [siddham]. antaranga-lakṣaṇatvad atra iyanuvanoḥ kṛtayor anupadhātvād guņo na bhaviṣyati. evam kriyate cedam tannimitta-grahaṇam, na ca kaścid doṣo bhavati. imāni ca bhūyas tannimitta-grahaṇasya prayojanāni— ‘hato’, ‘hathaḥ’, ‘upoyate’, ‘auyata’, ’lauyamāniḥ’, ‘pauyamāniḥ’, ’neniktaḥ’ iti. naitāni santi proyojanāni. iha tavat— ‘hato’, ‘hatha’ iti. prasaktasyânabhinirvṛt- tasya pratiṣedhena nivṛttiḥ śakya kartum. atra ca dhātûpadeśāvasthāyām evākāraḥ. iha ca ‘upoyate’, ‘auyata’, ’lauyamāniḥ’, ‘pauyamāniḥ’ iti. bahirange gunavṛddhi, antarangaḥ pratiṣedhaḥ. “asiddham bahir- angam antarange.” ’nenikta’ iti. pareņa rūpeṇa vyavahita- tvān na bhaviṣyati. upadharthena tavan nârthaḥ. dhātor iti var- tate. dhātum knit-paratvena viśeşayiṣyāmaḥ. yadi dhātur viseṣyate, vikaraṇasya na prāpnoti— ‘cinutaḥ’, ‘sunutaḥ’, ’lunitaḥ’, ‘punitaḥ’ iti. Fourth Ahnika naiṣa doṣaḥ. vihita-viseṣaṇam dhätugra- haṇam-‘dhātor yo vihitah’ iti. dhātor eva tarhi na prāpnoti. 231 naivam vijñāyate-‘dhātor vihitasya kniti’ iti. katham tarhi ? dhātor vihite kniti iti. athavā kāryakālam samjñā-paribhāṣam yatra kāryam tatra draṣṭavyam “puganta- laghûpadhasya guno bhavati’’ ity upasthitam idam bhavati “kniti na” iti. athava yad etasmin yoge knid-grahanam kriyate tadanavakāśam. tasyânavakāśatvād gunavṛddhi na bhavisyataḥ. athava"cāryapravṛttir jñāpayati- ‘bhavaty upadha-lakṣaṇasya gunasya pratiṣedhaḥ’ iti. yad ayam “trasi-grdhi-dhrși- kṣipeḥ knuḥ”(Pa. 3.2.140),“iko jhal, halantāc ca” (Pa. 1 .2.9-10) knu-sanau kitau karoti. katham kṛtvā jñāpakam ? kit-karana etat prayojanam-guṇaḥ katham na syād iti. yadi câtra guna-pratiṣedho na syāt, kit- karaṇam anarthakam syat. paśyati tvācāryo ‘bhavaty upadha- lakṣaṇasyâpi gunasya pratiṣedhaḥ’iti. tataḥ knu-sanau kitau karoti. roravity ity arthenâpi nârthaḥ. ‘kniti’ ity ucyate, na câtra knitam paśyāmaḥ. pratyaya-lakṣaṇena prāpnoti. “na lumatā tasmin” iti pratyaya-lakṣaṇa- pratiṣedhaḥ. athâpi “na lumatangasya” (Pa. 1.1.63) ity ucyate, evam api na doṣaḥ. katham? na lumatā lupte’ ngadhikāraḥ pratinirdiśyate. kim tarhi? yo’ sau lumatā lupyate tasmin yad angam tasya yat karyam tanna bhavati iti. athâpy angadhikāraḥ pratinirdiśyate. 232 Värttika. Bhāṣya. Värttika. Bhāṣya. Värttika. Bhāṣya. Mahābhāṣya of Patanjali evam api na doṣaḥ. katham ? “kāryakālam samjñā- paribhāṣam” yatra kāryam tatra draṣṭavyam. “sārvadhātukârdhadhātukayor guno bhavati” ity upasthitam idam bhavati “kniti na” iti. athavā chandasam etat. dṛṣṭānuvidhiś chandasi bhavati. athava bahirango gunaḥ, antarangaḥ pratiṣedhah. “asiddham bahirangam anta- range”. athavā pūrvasmin yoge yad ardhadhātu- kagrahanam, tad anavakāśam. tasyânavakāśatvad guņo bhaviṣyati. iha kasman na bhavati -’laigavāyanaḥ’, ‘kāmayate’? taddhita-kāmyor ik-prakaraṇāt. iglakṣaṇayor guna-vṛddhyoḥ pratiṣedhaḥ. na caite iglakṣaṇe. lakārasya nittvād adeseṣu sthanivad- bhāva-prasangaḥ. lakārasya nittvād adeseṣu sthānivadbhāvaḥ prāpnoti- ‘acinavam’, ‘asunavam’, ‘akara- vam’. lakārasya nittvād ādeśeṣu sthani- vadbhāva-prasanga iti ced yasuto nid- vacanāt siddham. yad ayam yasuto nid-vacanam śāsti taj jñāpayaty ācāryo- ’na nid-ādesā nito bhavanti’ iti. Fourth Ahnika 233 yady etaj jñāpyate, katham-“nityam nitaḥ” (Pa. 3.4.99), “itaś ca” (Pa. 3.4.100) iti? nito yat karyam tad bhavati, niti yat kāryam tan na bhavati iti. kim vaktavyam etat ? na hi katham anucyamānam gamsyate? yăsuța eva nid-vacanad. aparyāptas caiva hi yăsuț samudayasya nittve, nitam cainam karoti. tasyaitat prayojanam-

• nito yat

164. In the case of ‘pradudruvat ‘we have the root ‘dru’, but therefore guna can take place by the rule “sārvadhātukārdhadhātukayoh” (7.3.84), there is ‘can’ by “ni- śri-dru-śrubhyah” (3.1.48). By the rule “dvirvacane ‘ci’ (1.1.59), there is no ‘iyan’ and so it might be expected that before duplication ‘dru’ with ‘can’ should have guna as laghûpadha-guna, but being antaranga, the duplication takes precedence.
165. The word ‘auyata’ has been formed with the preceding ‘at’ to which the vrddhi is due. Vṛddhi is not due to ‘yan’ which is a knit-suffix and therefore cannot have vṛddhi by the present rule. The word ’lauyamani’ has been formed as ’luyamānasya apatyam’ and ‘pauyamani’ as ‘puyamānasya apatyam’. In both cases the vṛddhi is not due to ‘yan’ and hence any question of the inference of the scope of this rule does not arise. A discussion then arises about the word ‘hato’ and ‘hathah’. The root ‘han’ gets ‘sap’ by the rule “adi-prabhṛtibhyaḥ śapah” (2.4.72). The rule means that sap vikarana is elided after the roots ‘adi’ etc., and the elision is to be made by the word ’luk’. Now, 234 Mahābhāṣya of Patanjali be affected by prohibitory rule where something was expected to come, but was not prohibited. Here the a (in ‘han’ and ‘hatha’) exists in the state of the original instruction of the dhātu. In ‘upoyate’, ‘auyata’, ’lauyamāni”, ‘pauyamani’ the guna and the vṛddhi are external operations, and the prohibition is internal (antaranga), as there is a paribhāṣā — in case of internal operations, the external operations will be invalid- “asiddham bahirangam antarange” (paribhāṣā 50). In nenikta there will be no prohi- bition because of the intervention of its own latter part.166

166. The Bhāṣya enunciates the principle of prohibition as a general rule and as a special rule. The prohibition as a general rule leaves the scope open for the special cases and it is held that whatever was within the scope of the general rule, was already within the scope of the special rule and in such cases where in the special sphere the general prohibitory rule may not be applicable, the special rule comes in operation and effects the prohibition. But a vowel occurring as a constitutive part of an original dhātu or suffix (upadeśa) cannot come. In case of such words as ‘cita’ etc. (ci + kta) there could have been guna of the i of ci and that guna is prohibited by the “kniti ca”. But in the words ‘hata’ and ‘hatha’ the a is already there and it is not a result of transformation and there is no scope of the guna operation, only the alphabet a is technically called guna and hence this rule “kniti ca” can have no operation there. Nageśa in commenting on the above point says that the rule-“sarvadhātu- kārdhadhātukayoḥ” (7.3.34) by which guna is prescribed to ‘ik’ when sarvadhātuka and ardhadhātuka follow- is to be taken along with the prohibitory rule “kniti ca” as forming one complex proposition the meaning of which will be- except when Fourth Ahnika 235 It cannot be for providing for upadha or the penultimate, for we get of roots by continuity of adhikara and we shall make the following by ‘knit’ suffix as qualifying a dhatu or a root. The question of the prohibition of guna-vṛddhi in penultimates does not ococur. For we get the word ‘dhatoh’ by adhikara from the preceding rule. The rule thus interpreted will stand as ‘kniti ca dhator iko na guna- vrddhi. To be followed by a ‘knit’-suffix is a qualification of the ‘ik’ of a dhātua qualification that arrests the guna-vṛddhi of the ‘ik’ of a dhātu. Now, if the rule specifically refers to the ‘ik’ of a dhātu, there is no question of a penultimate, for since the prohibition refers to the ‘ik’ of a suffixes with a redundant k and ǹ follow, the sarvadhatuka and ardhadhātuka suffixes will reduce guna in the preceding ‘ik’. This prohibition is a paryudāsa prohibition as a general negative injunction and therefore the rule “kniti ca” may be regarded dis- tinctively leaving guna-vṛddhi in such occasion. But this cannot refer to the a of ‘han’ as enjoined in the original instruction of dhatu (upadeśa). Kaiyața in commenting on ‘hata’ says that here there was the ‘sap’ and the ‘sap’ is elided by the rule “adi- prabhṛtibhyaḥ sapah” (2.4.72). Now the a in ‘Sap’ may be expected to be elided by the/ rule “kniti ca” because a is technically called guna and therefore its elision cannot be the purport that prohibits guna (Sap bhavati ity asyam codanayam guṇa-rūpeṇa acodanāt. nâyam guna-nisedha-visayah). The word ‘upoyate’ has been formed from the root-upa + veñ with yak and the consequent samprasarana by which va becomes u. Thus we have upa+u+yak+ tan. Now the formation from ’ve’ is not an antaranga process, because according to the rule “grahi-jya-vayi-vyadhi-vasti-vicati-vṛścati- prcchati-bhrjjatīnām niti ca” (6.1.16), i.e. these roots will undergo samprasarana when ‘kit’ and ’nit’ suffixes follow. Thus the va becomes u when ‘yak’ a ‘kit’ suffix follows. The samprasarana having taken place as a change from va to u is an antaranga process and it is not allowed to nullify this ū-kāra by a rule prohibiting guna after ‘kit’ and nit’ suffixes, for that is external to it. In the case of ‘anyata’ we have an + veñ + yak+tan (lan). Here the vṛddhi is equal to at + u+ya+ ta and we have anyata. The vṛddhi is here affected by “atas ca” (6.1.90). When a vowel follows ‘at’ then the ‘at’ and the vowel together becomes a vṛddhi, e.g. aiyata. The ‘at’ has been directed by “āḍajādīnām” (6.4.72) which means that when suffixes of the lun, lan and Irn group follow, there will be the augment ‘at’ before verbs having an initial vowel. Here the augment at depends upon the coming of the suffix of the lar group and the vṛddhi takes place entirely as a bahiranga vidhi. In the case of ’nenikta’ we have nij + te= nij+te = ninij + te. Now the guna of the first ni in ninij cannot be restricted because the tan (te) that follows is an ’nit’ suffix, for between the ’nit’ suffix and the ’ni’ there is an intervention by the word part ’nij’. Therefore “kniti ca” cannot prohibit the guna of ni in ’nenikta’. All this shows that the formation of these words could not have been regarded by Panini as objectives when he composed the sutra “kniti ca”. Whether we take the meaning of the seventh case in ‘kniti’ as nimitta or otherwise, that would not in any way affect the fate of these words. 236 Mahābhāṣya of Patanjali dhātu, that ‘ik’ must be an ultimate ‘ik’ and not a penultimate (cf. the rules “sasthi sthāneyoga” (1.1.49) and “alo’ntyasya” (1.1.52) previously explained in detail). The word ‘dhatoḥ’ comes by adhikara from the sutra “na dhatulopa ardhadhātuke” (1.1.4). If ‘kniti’ is regarded as qualifying only to dhātu then it would not apply to the inserted suffixes like ‘śnu’ (i.e. the guna of these should not be prohibited when ‘knit’ suffixes follow). Thus we have cinutaḥ, sunutaḥ, lunitaḥ, punitaḥ, etc. But this is no criticism when (kniti) is regarded as qualificatory to the root, the root does not mean ‘of the root’, but it means ‘prescribed after the root’, i.e. that which has been prescribed after the dhātu. Then it should not apply to dhātu. The meaning is not this that when ‘knit follows (the prohibition applies) only to what has been prescribed to appear as an inserted suffix after the dhātu. What then is the meaning when ‘knit’ follows after a ‘dhatu”?167 Or rather a samjñā or symbol and a maxim becomes operative in its proper time. It comes in force when it is time for it to operate. There will be guna in case of verbs which are puganta and those which have been a short penultimate. In this field at once comes this rule “kniti na”- not in the case of ‘knit suffixes.168
167. The question raised here is this: we know that when ‘knit’ suffixes follow, guna is prohibited. But guna of what is prohibited? The reply is that the guna of the ‘ik’ part of the dhatu is prohibited. But such an answer is faulty, for the prohibition of guna has reference not only to the ‘ik’ part of the dhātu, but also to the ‘ik’ belonging to the vikaranas like ‘śnu’, ‘śna, etc., which are prescribed after dhātu. But if the ‘knit’ prohibition applies only to the dhātu, it cannot apply to the vikarana, and if it applies to the vikaraṇa, it cannot apply to the dhatu. How to include them both? If the interpretation was ‘dhātor vihitasya knitaḥ’ then ‘dhatoḥ’ being in the 6th case and vihitasya being in the 6th case, the prohibition applies only to the dhatus. The interpretation therefore proposed is ‘dhator vihite kniti’, i.e. when ‘knit’ is prescribed after a dhātu, then prohibition takes place. Here we do not mention either the dhātu or the vikarana and therefore we prescribe here the prohibition in a general manner. The purport there will be that when a ‘knit’ suffix is prescribed after a dhātu, then the guna of the ‘ik’ is prohibited with reference to that which is nearest to the ‘knit’ suffix, be it a dhatu or a vikarana. Therefore when there is an occasion for a vikarana to intervene, the prohibition of guna is directed to the ‘ik’ of vikarana; and when there is no vikarana it is directed to the ‘ik’ of the dhātu. By qualifying dhātu so as to mean ‘dhātor vihite kniti’, we can get the prohibition to apply to both the dhatu and the vikaraṇa.
168. There is a rule “puganta-laghûpadhasya ca” (7.3.86). It prescribes guna to verbs ending in p or puk and verbs having a short penultimate. There are two paribhāṣās “yathoddeśam samjñāparibhaṣām” (2) and “kāryakālam Fourth Ahnika 237 Or, rather in this specific place, there is no room for taking in ‘knit’. In absence of its having any room, there will be no guna-vṛddhi.169 samjñāparibhāṣam (3). The first means that the meaning of a samjñā and a paribhāṣā is to be understood from the context where it is first explained. There one only understands its proper meaning but does not know the scope of its application. There are some however who think that in understanding properly the meaning of a samjñā and a paribhāṣā, one should have in a flash before his mind all the possible occasions where they may have their fields of application and in the totality of that context as a unifying whole one could understand the meaning of samjñā and paribhāṣā when they are explained for the first time. The second maxim or the karyakala-paribhāṣā agrees with the first in every respect, but enjoins the further provision that they should have their application even in the tripādi, i.e. the last three chapters of the 8th book of Panini with reference to which the previous portion of Panini’s work, the first seven books and the first chapter of the 8th book are regarded as null (asiddha) by the rule “purvatrâsidham” (8.2.1). There is a further subtle point to be noted in this connec- tion. According to the first paribhāṣā where one grasps the meaning of a paribhāṣā as described in a particular place he has in the back of his mind the different cases of the application of the paribhāṣā throughout Panini’s work and each case with its distinctive peculiarity affects the understanding of the general meaning of the paribhāṣā as if it were a sweet and fragrant drink wherein with each sip of its taste one experiences the taste value of its constituent parts. Thus Kaiyața says the interpreta- tion of the Bhasya immediately following the present context-yathoddese pradha- nāni ātmasamskārāya sannidhiyamānāni guṇa-bhedam prayuñjate. In the second maxim however, the paribhāṣā as explained in the first context, appears immediately in the place where there is a field of its application and along with the context it expresses itself as unified with the meaning of that context- kāryakāle tu tad-tad- vidhi-pradese paribhāṣā-buddhyā ekavākyatā paribhāṣā - is there realized as be- longing to the place where it was first explained, but being there it simply travels to its new field of operation. Now as soon as we find the sutra, “puganta-laghûpadhasya ca” and find that guna is prescribed to the puganta verbs and verbs with a short penultimate, the rule “kniti ca” arrives there and shows its operating control and thereby where we have bhid + kta, the guna of the i-kära is prohibited.
169. In the preceding section we have regarded ‘kniti ca’ as a paribhāṣā and we have supposed that the paribhāṣā appears as a prohibitory qualification at any occasion when guna is prescribed as in “puganta-laghûpadhasya ca”. But here another explanation is suggested as follows: Let us suppose that “kniti ca” is not a paribhāṣā; still a prohibitory injunction has its scope of application to any and every place where there are objects of prohibition, and in the case of “puganta-laghûpadha- sya ca"the guna is object to prohibition, and ordinarily we should not expect the guna to be prohibited, for the prohibition is not to take place with reference to a penultimate ‘ik’ which is mediated by a consonant from ’nit’ suffix. But still there the prohibitory rule has its injunction and it prohibits and guna of the penultimate though it is mediated by a consonant. Or rather it may be assumed that the Bhasya here takes “kniti ca” in the sense of ‘yathoddeśam samjñā-paribhāṣam’ as explained above. 238

1. “Kniti ca” may be regarded as a paribhāṣā and enters into the context of the rule “puganta-laghûpadhasya ca”. In such a case the meaning of the rule “puganta- laghûpadhasya ca” would be- except ni such cases where ‘knit’ suffixes follow roots having an ultimate ‘puk’ and a short penultimate, shall be affected by guna.
2. Let us not take “kniti ca” as parabhāṣā, it is only a prohibitory rule and as such it has within its sphere of prohibition any and every case when guna is prescribed and as such it has no power to arrest the guna or the penultimate of a root as against “puganta-laghûpadhasya ca”.
3. ‘Kniti’ may be taken as a paribhāṣā in accordance to its first interpretation in the second maxim “karyakalam samjñāparibhāṣam”. In such a case the paribhāṣā ‘kniti ca’ has within its scope of meaning all the typical different cases when guna is prescribed and as such the guna of the penultimate is also taken within its meaning as an object of prohibition when the paribhāṣā was first explained–a paribhāṣā in the sense of “karyakālam samjñāparibhāṣam” and in the last interpretation it was taken in the sense of “yathoddeśam samjñāparibhāṣam”. Fourth Ahnika 239 should be no meaning in making ‘knu’ and ‘san’ ‘kit. It is not also for forming the word ‘roravi’ that the rule “kniti” has been made. (The condition of determining the prohibition of guna) is that ‘knit’ suffixes follow, as has been said. But here (roravi) we do not find any ‘knit’. When a suffix is elided the action demanded by the suffix would still hold good. But when the elision is made by any of the terms ’luk’, ‘śnu’ and ’lup’, the effect due to the immediate antecedent on account of the suffix shall not take place or rather there would be no effect on the part of a word due to a suffix when that suffix is elided of any of the terms ’luk’, ‘ślu’ and ’lup’. Thus there is no fault. Why? When anything is elided by any of the terms ’luk’, ‘ślu’, ’lup’, the effects due to the part of the word, are not referred back. What then? When that, which is elided by any of the words ’luk’, ‘ślu’, ’lup’, follows, there will be no action that was due to the part of the word on account of it. Even if it be suggested as the meaning that, that which was due to the part of the word, should be referred back, then also there is no fault. Why? A samjñā or a paribhāṣā is stirred into operation by the occasion. Wherever there is an occasion there, one notices its operation. There is a direction of guna when ‘sarvadhātuka’ and ‘ardhadhātuka’ suffixes follow and there does this provision “kniti ca” appear. 170
4. By way of conclusive suggestion as against the provisional suggestion mentioned in the previous section, the Bhasya says that the fact that the prohibition of guna refers to the penultimate ‘ik’ vowel also, can be deduced from the fact that the suffix ‘knu’ has been directed to verbs like ’trasi’, ‘grdhi’, ‘dhrsi’ and ‘kṣip’ (’trasi-grdhi-dhrsi-kripeḥ knuh’ 3.2.140, meaning already explained), leading to the formation of such words as ’trasnu’, ‘grdhnu’, ‘kṣipnu’. The object of deducting ‘knu’, i.e. nu with a knit has been deliberately intended to prevent the guna of the penultimate vowel in grdh and kṣip. This shows that it is the intention of the master that ‘knit’ suffixes can prohibit the guna of the penultimate vowel. A suggestion was made in the previous section that the seventh case in ‘kniti’ should be regarded as having a causal import and this will be useful also for the formation of the word ‘roravi’. The word ‘roravi’ has been formed from the root ‘ru’ by the suffix ‘yan’ which was elided by the rule “yano’ci ca” (2.4.74). which means that when a vowel follows, the ‘yan’ is elided by the term ’luk’. But here the Bhāṣya repudiates that provisional suggestion when it says that “kniti ca” is relevant towards the formation of ‘roravi’. There is practically no ‘knit’ suffix in ‘roravi’; ‘yan’ has indeed been elided, but ‘yan’ is a part of a dhatu. There is indeed a rule “pratyayalope pratyayalakṣaṇam” (1.1.62) which describes that when a suffix is elided, it still can affect such changes in the preceding word or note which was due to it, i.e. even when a suffix is elided its effects on the previous word remains unchanged as if no elision has taken place. But there is another rule “na lumatangasya” (1.1.63). This rule means240 Mahābhāṣya of Patanjali Or, rather ‘ik’ (roraviti-this word) is Vedic. A word like that is found to be used in the Vedas. Or, rather, the guna is bahiranga or external and the prohibition is internal. There is the maxim the external should be regarded as null (asiddha) and the internal is to be performed. Or, rather, in the previous sūtra there is no point in altering the word ‘ardhadhātuke’; that being out of point, there is no guna.171 that when an elision is effected by any of the terms ’luk’, ‘ślu’ or ’lup’, the effect of the suffix on any part of a previous word shall not take place, notwithstanding what has been said in the previous sutra “pratyayalope pratyayalakṣaṇam” (1.1.62). Com- menting upon the sutra” na lumatangasya” the Bhasya says that the meaning is that when an elision is made by ‘slu’, ’lup’ or ’luk’, then the original word immediately preceding is not affected by the action of the suffix, but the rule does not mean that all that was due to the original word, should all be suspended. Even if it is allowed to mean that in cases of such an elision, all that was due to be effected to the original word, could be suspended, even then it is not possible to stop a paribhāṣā to appear in such places when there is an occasion for it to do so. This is in consonance to the kāryakāla paribhāṣā and there is a rule “sarvadhātukārdhadhātukayoh” (7.3.84) which directs guna whenever särvadhātuka and ardhadhātuka suffixes follow. Now the guna for ardhadhātuka suffixes is prohibited by the rule “na dhatulopa ärdhadhātuke” (1.1.4). But for prohibiting guna we need a rule like “kniti ca”, for ‘yan’, which has been elided, is a myth. But there is a further objection that ‘kniti ca’ is not a paribhāṣā and it cannot be allowed to appear whenever there is an occasion for it in according to the maxim “karyakala-samjñā-paribhāṣam”, for that maxim applies only to paribhāṣā and “kniti ca” is not a paribhāṣā. The reply to such an objection is now being attempted in the Bhāṣya immediately following.
5. Since the rule “kniti ca” is not generally regarded as a paribhāṣā we have to find some other means of justifying the usage of ‘roraviti’. Three provisional answers are given in this connection: (i) There is a rule “yan aci ca” (2.4.74) which means, as we have seen, that before a vowel suffix, yan is elided. But before this rule there is another rule “bahulam chandasi” (2.4.73) by which ’lyap’ is elided optionally in Vedic usage. By force of the word ‘ca’ in ‘yan aci ca’ the word ‘bahula’ from the rule ‘bahulam chandasi’ is drawn in, and thus it may be said that ‘yan’ is elided in the Vedic usage. Thus the word ‘roraviti’ may be justified as a Vedic usage. (ii) The view that ‘roraviti’ should be supported as a Vedic usage, has to be adopted in the last extremity and for this reason, it is worth while to attempt another solution. With this idea in view the Bhāṣya says that as the prohibition of guna is due to the elision of na in ‘yan’, this prohibition is internal (antaranga) with reference to the occasion of guna that arises when subsequently the sarvadhātuka suffixes follow and therefore sarvadhatuka suffixes being null (asiddha) as being external (bahiranga), they do not exist and since they do not exist, there is no occasion of guna and no prohibition is required. (iii) In the 3rd interpretation it is suggested that the use of the word ‘ardhadhātuke’ in the preceding sutra “na dhatulopa ardhadhātuke”-it may well be regarded as out Fourth Ahnika 241 Why not in the case of ’laigavāyana’ and ‘kamayate’? (Here the Värttika intervenes)- in the case of taddhita suffixes and the root ‘kami’ (there is a prohibition of guna and vṛddhi) with reference to ‘ik’. (Bhāṣya:) The guna and vṛddhi are prohibited with reference to the vowels that come within the ‘ik’ group. But these do not belong to the ‘ik’ group, 172 (Here the Varttika intervenes)- the law being ‘knit’ all the adeśas prescribed in its place may be regarded to have the same status as law. (The Bhāṣya interprets it as follows). (Bhāṣya:) And all the ten suffixes technically called la-kara- and all that are prescribed in their place, should also be ‘knit’ as they have the same status as the law in whose place they have come.173 of point. For even without the word ‘ardhadhātuke’ we can have the sense of ‘bahuvṛihi’ in the word ‘dhatulopa’ as ‘dhator lopo yasmin’ and guna may be prohibited by virtue of the elision of ‘yan’ by the rule ’na dhatulopaḥ’. The idea is that the rule “na dhatulopa ärdhadhātuke” is spilt up by yoga-vibhāga-na dhatulopa being on one side and ardhadhatuke on the other. Dhatulopa is formed by a genitive compound. dhātu means a part of dhatu and dhatulopa means therefore ‘dhatvekadeśalopa’ that is why one part of the dhātu is elided as in the case of the elision of ‘yan’. Now, therefore, when there is such a ‘dhatvekadeśa-lopa’ if there is on account of that any prohibition of guna and vṛddhi, then that prohibition should be operative only when ardhadhātuka suffixes follow. But in ‘roraviti’ there is no ärdhadhātuka suffix and therefore the elision of ‘yan’ cannot lead to the prohibition of guna, and even if the prohibition was available by some other sutra, the scope of the operation of that rule will also be arrested, for the rule “na dhatulopa ārdhadhātuke” distinctly shows that guna will be prohibited only when ardhadhätuka suffixes follow. Therefore the operation of guna due to a sarvadhātuka suffix is not arrested, and we have the forms ‘rorute’ where there is prohibition of guna. The Bhāṣya regards it and such other case where yan is elided as being Vedic usage.
6. Nāgesa in commenting on the above Bhāṣya, says that though guna and vrddhi apply to ‘ik’ yet in laigavayana (ligor apatyam puman-ligu + phak by the rule “naḍādibhyaḥ phak’’ (4.1.99) the guna and vṛddhi apply, because though it is not ik-lakṣaṇa, it is ik-sthānika. The point is that the phak-suffix is added to the ligoḥ and not ligu. But in ligoh we have already o-kara and a visarga after it. Therefore, it is not an ik-lakṣaṇa word, it does not end with ik’. But since the transformation has occurred with reference to the ‘ik’, it may be regarded as ik-sthānika, for the phak is ultimately applied to ligu.
7. The point here is that since the ’tin’ vibhakti is ‘knit’ all the adeśas that take in its place are to be regarded as themselves also ‘knit by the Panini’s rule “sthanivad- adeśo’nalvidhau” (1.1.56), and therefore there should be prohibition of guna in ‘acinavam’ by the rule “kniti ca”. Kaiyata here comments that the anubandha of the la-kāra being called ‘it’ when it is elided, is not attached or associated with that which 242 Mahābhāṣya of Patanjali The operation directed for the ’nit’, shall take place, but not the operation that is due when ’nit follows. Should this be definitely stated? How can one understand it if unless it is definitely stated? It can be deducted from the fact that ‘yasut’ can be declared as ‘knit’, ‘yasut’ alone is unable to endow all la suffixes with a nit character and yet ‘yasut’ has been declared as ’nit’. The necessity for that is this- all such operations as should be done to a ’nit’, should be done but no such operation should be done as are due when a ’nit’ suffix follows.174 has taken its place and therefore the prohibition to sthanivadbhava as dictated by the phrase ‘analvidhau’ in the rule “sthanivad adeśo’nalvidhau” (1.1.56) will not apply. From the rule “sarvadhatukam apit’ (1.2.4)- the sarvadhātuka that is not ‘piť becomes like ‘knit’, the Bhāṣya infers that which is ‘pit’ is not ‘knit’, that which is ‘knit’ is not like ‘pit’. By adopting the principle of ‘yoga-vibhāga’ in the rule “sarvadhātukam’apit” either by “sthanivadbhava” or by being directly a ‘knit’ the prohibition of guna will apply to all la-kāras. Thus lat is an apit, being an apit it is a ‘knit- so also ’lut’, and ’lot’ and ‘Irt- and the rest are directly ‘kniť. Here the Varttika puts forward an answer of a provisional nature and says: the suggestion that the tense la being ‘knit’ all the adeśas would have the same status is obviated by the fact that the suffix ‘yasut’ has been declared as a myth. This the Bhāṣya interprets as follows: the fact that ‘yasuta la-suffix has been specifically mentioned by Panini, indicates the fact that adeśas occurring in the place of ‘knit’ suffixes would not themselves be regarded as ‘knit’. If this is so, why then “nityam nitah” and “itas ca”. There is a rule “yasut parasmaipadeṣudatto nitas ca” (3.4.103)- in the paras- maipada the augment ‘yasut’ will come in place of lin and this ‘yasut’ will have the udatta accent and be treated as ’nit’. Now if this sthanivadbhāva maxim be held in force, the yasuț occurring in place of ’lin’ will naturally be a ’nit’ and there will be no necessity for Panini to declare it specially as ’nit’. From such a declaration the natural implication is that there are other adešas prescribed in place of the tense la- kära should not be regarded as ’nit’. There is a rule “nityam nitah” (3.4.99) and it means that the sa of the first person of a ’nit’ tense la is elided. There is another rule “itas ca” (3.4.100) apacāva. It means the i associated with a nit la-kāra will always be elided in the parasmaipada -apacat, apakṣit. Now the point raised in the Bhasya is that these two sūtras-assume or indicate that Pāņini admits that there is tense la which should be regarded as ’nit’. But it should be noted that the contention is not relevant. No one denies that there are some tense la which have na as anubandha and these may naturally be regarded as nit. But this does not indicate that the adeśas coming in their place should be regarded as ’nit’ for in that case a lat being an apit, is ’nit’ and we should expect that the sa in pacāva should be elided by the rule “nityam nitah” and we should expect ‘pacāva’ instead of ‘pacāvaḥ’, as we find in lat.
8. The idea is that we do not deny that lan, lun, etc. are themselves nit but we hold that no ādeśas that are to occur in place of these nit lakaras, should themselves Fourth Ahnika 243 be regarded as nit and should be exposed to such restrictions on the words preceding them as are due when nit suffixes follow. When ‘yasut’ is made nit it induces the operations that take place when a nit suffix follows such as samprasarana etc. If yāsuṭ could have been regarded as nit on the ground that ‘yasut’ is an adeśa in place of lin which is nit by the sthanivadbhava maxim, then the declaration in the sutra that ‘yāsuṭ’ should be regarded as nit (3.4.103) would have been quite unnecessary. The main outline of the discussion in the rule “kniti ca” is as follows: ‘kniti’ in the sūtra is in the seventh case. The seventh case may have two different relevant meanings, one guided by the rule “tasmin niti nirdişte pūrvasya” (1.1.66)-it means that when anything in the sutra is in the seventh case, it implies that the consequential operation shall take place with reference to the vowel or consonant that immediately precedes as in the rule “iko yan aci’ (6.1.77)-dadhyodanam. There is another rule “yasya ca bhave bhavalakṣaṇam” (2.3.37). It means that when by the action of anything the action of any other thing is describable, then there will be seventh case with regard to the former. This is called nimitta saptami. Now, if the seventh case in ’niti’ is guided by the second rule then nit suffix will be regarded as a nimitta or cause for the prohibition of guna and there is no such condition that this prohibition will take effect only with reference to the immediately proximate ‘ik’. If the meaning of the seventh case however be determined by the first rule (1.1.66) then the prohibition should effect only the immediately proximate vowel preceding. The first stand taken in the Varttika and the Bhasya is that the seventh case in ‘kniti’ should be regarded as a nimita saptami according to the second rule quoted above (2.3.37), for otherwise one cannot provide for the prohibition of guna in the penultimate vowel and also satisfactorily solve the word ‘roraviti’. Provisional objections are raised against this view. It is suggested that the seventh case of ‘kniti’ might be taken in the first meaning (1.1.66) and it may be held that the prohibition would apply to wherever there is scope whether the vowel be ultimate or penultimate. But then the meaning according to 1.1.66, definitely bars the operation with reference to the vowel that is not immedi- ately proximate to the ‘knit’ suffix and therefore the prohibition cannot apply to the penultimate. It is again suggested that if the second meaning is accepted then verbs with a following sa or can cannot be solved. The reply given to this is that in those cases guna does not take place because the prescription of ‘iyan’ and ‘uvan’ is internal or antaranga. Then another attempt at justifying the nimitta view in the seventh case is made with reference to the words hato, hathaḥ, upoyate, but this objection also is overriden. In further explaining the rule it is said that the prohibition of guna would apply to the ik of a vikarana or a dhātu in immediate proximity. But the problem regarding affecting the penultimate by the prohibition of guna still remained inexplicable unless the seventh case is taken in the sense of nimitta. But all these objections are brushed aside for the suffix ‘knu’ has been made ‘knit’ only with a view to arrest the guna of penultimate vowel. We here find that it is unnecessay to ascribe the second meaning of nimitta saptami to ‘kniti’. Its meaning therefore is that which is controlled by 1.1.66. The second meaning of nimitta saptami is not also necessary for the formation of the word ‘roraviti’, for this word may be regarded as being of Vedic usage. Apart from 244 Panini. Mahābhāṣya of Patanjali didhivevītām 1.1.6. In the case of the roots ‘didhi’ and ‘vevi’ and the augment ‘it’, there will be no guna or vṛddhi. (The roots ‘didhi’ and ‘vevi’ are mostly used in the Vedas. ‘Didhi’ means ’to reflect’, ’to meditate’, ’to shine’ and ‘vevi’ means ’to go’, ’to permit’, ’to derive’, ’to love’, etc. as didhite’, ‘vevite’. Bhāṣya. Värttika. Bhāṣya. kimartham idam ucyate? guna-vṛddhi mā bhūtām iti. ‘ādīdhyanam ādidhyaka’, ‘avevyanam āvevyakaḥ’. ayam yogaḥ śakyo’ kartum. katham ? didhivevyoś chandoviṣayatvad dṛṣṭā- nuvidhitvāc ca chandasaś chandasy adidhed adidhayur iti ca gunadarśanād apratiṣedhaḥ. ‘didhi-vevyoś chandoviṣayatvāt’. didhi- vevyau chandoviṣayau. dṛṣṭānuvidhitvāc chandasaḥ. dṛṣṭānuvidhiś ca chandasi bhavati. ‘adidhed adidhayur’ iti ca gunasya darśanad apratiṣedhaḥ. anarthakaḥ pratiṣedho’pratiṣedhaḥ. “prajapatir vai yat kimcana manasādīdhet”, “hotrāya vṛtaḥ krpayann adidhet (RV. 10.98.7), “adidhayur dāśarājñe vṛtāsaḥ” (RV. 7.33.5). it, other solutions are also provisionally proposed with reference to some of the fundamental problems. These provisional replies are not, however, satisfactory. The conclusion then is that the prohibition of guna before a knit suffix applies to an ultimate or a penultimate vowel, as the case may be. Another point raised here is the question as to whether the la-kāra suffixes prescribed in the case of la-kara should be regarded as ’nit’ and thereby all guna operations should be suspended whether they belong to the ultimate vowel or the penultimate. But the answer is that for the fact that ‘yasut’ has been made a ’nit’ it follows that other instances of la-kāra should not be regarded as ’nit’ in the sense that when they follow there is no guna-vrddhi with reference to the immediately proximate vowels or penultimate vowels. No specific declaration need be made for it as it is deducible from the fact that ‘yasut ‘has been made nit by Panini (3.4.103). Värttika. Bhāṣya. Fourth Ahnika 245 bhaved idam yuktam udāharaṇam- ‘adidhed’ iti. idam tvayuktam - ‘adidhayur’ iti. ayam jusi guṇaḥ pratiṣedha-viṣaya āra- bhyate, sa yathaiva “kniti na” ity etam pratiṣedham badhate. evam imam api bädheta. naișa dosah. jusi gunah pratiṣedhaviṣaya ārabhyamāņas tulyajātīyam pratiṣedham badhate. kaś ca tulyajātīyaḥ pratiṣedhaḥ ? yaḥ pratyayasrayah. prakṛtyāśrayaś câyam. athava-yena nâprapte tasya badhanam bhavati. na câprāpte “kniti na” ity etasmin pratiședhe jusi guņa ārabhyate. asmin punaḥ prapte câprāpte ca. yadi tarhy ayam yogo nârabhyate, katham ‘didhyad’ iti? didhyad iti ca syan-vyatyayena. didhyad iti ca syan vatyayena bhaviṣyati. itas câpi grahanam sakyam akartum. ‘katham-akaniṣam’, ‘araniṣam’, ‘kaṇitā śvaḥ,’ ‘ranita śvah’, iti? “ardhadhātukasyed valādeḥ” (Pā. 7.2.35) ity atra ‘id’ ity anuvartamāne punar id- grahaṇasya prayojanam -‘id id eva yathā syad, yad anyat prāpnoti tan ma bhūd’ iti. kim cânyat präpnoti? guṇaḥ. yadi niyamaḥ kriyate, ‘pipathiṣater apra- tyayaḥ pipathiḥ’-dirghatvam na prāpnoti. naiṣa doṣaḥ, āngam yat kāryam tan niyamyate, na caitad āngam. athava-asiddham dirghatvam tasyâ- siddhatvan niyamo na bhaviṣyati. (Bhāṣya:) Why it is this sūtra made? So that there may not be any guna or vṛddhi -ādidhyanam, ādidhyakaḥ, āvevyanam, āvevyakaḥ. 246 Mahābhāṣya of Patanjali This sūtra could have been avoided. How? (Here the Värttika intervenes)-‘didhi’ and ‘vevi’ being Vedic usages and usages from there having been current in language in the Vedic usages, we have adidhet and adidhayuh which show that they are affected by guna and therefore the prohibition of guna does not apply. (The Bhāṣya explains it as follows)-‘didhi’ and ‘vevi being found only in the Vedas, these (‘didhi’ and ‘vevi’) are Vedic incidents. The Vedas can use only words that are current. In the Vedas the current words are accepted as current and used. In ‘adidhet’ and ‘adidhayuh’ guna being found, it (guna) cannot be prohibited. A useless prohibition is no prohi- bition. Prajapati medicated upon (adidhet) whatsoever in his mind being nominated for the position of a hotṛ. Through kindness he reflected (adidhet). Those who were nominated, reflected (adīdhayuh).175 The example ‘adidhet’ is a proper one, but the example ‘adidhayuḥ’ is not so. This guna in the presence of a following ‘yus’ is an object of prohibition and just as ‘kniti ca’ prohibits guna, so this sutra also prohibits guna in that place. This is no fault. The guna in ‘yus’ is the object of prohibition and this prohibition, as it works, contradicts another prohibi- tion of its own class. What is the prohibition of the same class? That which depends upon the suffix. But this prohibition is with reference to the original dhātus. 176 Or, rather, that which has a general rule is contradicted. If the rule ‘kniti ca’ had not exercised its prohibitive control, the rule ‘yusi gunah’ would
9. The word ‘drṣṭānuvidhitvac ca chandasah’ as used in the Bhasya deserves a little comment. Didhi and vevi have currency in ordinary language and though in the ordinary language the words that are actually seen current in usage, but which are grammatically correct and correctness of which we can assume by their use by master minds may be re-accepted in language. But these forms such as adidhyanam as illustrated before, are not actually found in the Vedas. The Vedic forms are ‘ādidhet’, ‘ādīdhayuḥ’ where guna has been commended. The prohibition of guna in case of ‘didhi cannot therefore be accepted. If guna did not take place, the forms might be ‘adidhyanam’, ‘adidhyakah’. Such forms are not found current in the Vedas. We find guna in the Vedas and therefore the universal prohibition of guna by the present rule is not justificable. Adidhet is the lar form of didhi, and ‘adidhayuḥ’ is formed by change ‘jhi’ into ‘yus’ by the rule “jakṣityadayaḥ ṣat” (6.1.6)-The root ‘jaksa’ and six other roots are regarded as abhyasta.
10. There is a rule “jusi ca” (7.3.83) which means that when the ‘yus’ suffix follows the iganta part of a root, is affected by guna, as ‘ajuhavuḥ’, ‘avibhayuḥ’. In the formation of the word ‘adidhayuh’ there is guna by this rule “jusi gunah”. The sūtra “didhi-veviṭām” should have opposed this guna by “jusi ca”, but the present sūtra “didhi-veviṭām” is dependent on the nature of the original roots ‘didhi’ and ‘vevi’ and is not dependent upon the suffix that followed them. Fourth Ahnika 247 not have been introduced. Thus, there being the general prohibition by “kniti ca”, there is the special rule, “jasi ca”. But if thus, this rule was not introduced, how to solve the form ‘didhyat’ (didhi- let ’tip’)? Then by the rule “itaś ca lopaḥ parasmaipade” (3.4.97) the i-kāra in didhi is elided. (Here the Värttika intervenes)-didhyat has been formed by the irregularity of ‘syan’. (Bhasya:) ‘Didhyat has been formed by the irregularity of ‘syan’.177 The ‘it’ in the sutra is also unnecessary. How so? ‘akaniṣam’, ‘araniṣam’, ‘kanita śvah’, ‘ranita svaḥ.‘178 In the rule “ardhadhātukasya id valādeḥ” (7.2.35; when a suffix of the ärdhadhātuka class with any of the alphabets contained in the group ‘val’ follows, there should come the augment ‘it’). ‘It was available by continuity or adhikara and there was no necessity for including the word ‘it’ in the sutra 7.2.35, had it not been for the fact that it indicates that this ‘it will remain unchanged-whatever else might be prescribed in its place, should not take place. The adhikara of ‘it’ (7.2.35) may be traced from the rule “nedvaśi krti” (7.2.8). If a restriction is made when we have ‘pipathiḥ’ from ‘pipathiṣati’ without a visible suffix, the ‘it’ ought not to have been lengthened.179
11. Didhi-syan + tip-the ‘Syan’ being ’nit’, there is no guna by “kniti ca”. By the rule “yivarṇayor didhi-vevyoh” (7.4.53; when ya and i follow, the i of didhi and vevi, they are elided). The i of didhi is elided and thus we have didhyat.
12. The root ‘kan’ as in kanati, means ’to become small’, ’to cry in distress’ or ’to approach’. The root ‘ran’ as in ranati, raraṇa, means ’to rejoice’, “to praise’, ’to go’. akanisam kan+am (lun) = a+kan+ sic+ am=a+kan+i+sic+ am=akaṇişam. Here the guna of i was expected by the rule “pugantalaghûpadhasya ca” (7.8.86). It is suggested that the guna of this it as expected by the rule ‘puganta’ etc., is suspended by the present rule. This is the reason why the Bhasya says that if this rule was not composed by Panini, how could we have prohibited the guna of it in akaniṣam? Here a provisional reply will now be suggested by the Bhasya that follows.
13. We have the verb ‘pipathiṣati’ from path+san and then we add ‘kvip’ suffix to it. There is the rule “ato lopaḥ” (6.4.48; when an ardhadhātuka suffix follows, the a of the preceding verb is elided). Thus we have pipathiṣa + kvip= pipathiş (by the rule 6.4.48) the ultimate a is elided. Now, we have a rule ‘sasajuṣo ruh’ (8.2.66; a pada ending in sa and the word ‘sajus’ will have ra in place of their sa). We have another rule “in koh” (8.3.57; it means that the sa following the vowels included in the group ‘in’ or the consonants in the group ku, will become cerebralised). But since the rule “in koh” is a rule later than the rule “sasajuso ruh” and since they are both within the tripādi, i.e. the last three sections of Panini’s work, work, the rule, “in koh” is to be regarded as null or asiddha by the rule, “purvatrâsiddham” (8.2.1) and therefore the 248 Mahābhāṣya of Patanjali (Another provisional reply is suggested in the Bhāṣya that follows)— This is no criticism. It is the operation regarding an anga that is restricted. But this is not an anga operation.180 (The final suggestion on the point will be attempted in the Bhāṣya that follows) Or, rather, the lengthening is nullified (asiddha); that being nullified, there will be no restriction. 181 General review Both the Varttika and the Bhāṣya think that the rule “didhivevīṭām” could well have been dropped by Panini. The roots ‘didhi’ and ‘vevi’ are found in the Vedas only and they are found to undergo guna. The rule was not also necessary for the prohibition of the ‘it’, as the rule “ardhadha- tukasyed valadeh” indicates that this ‘it’ should not undergo guna. Kambala+da+ka-in the case of äkäränta dhātus with ka suffix there does not occur any guna of the ākāra. By associating the k-it suffix in case of ākārānta dhātus, provision has been made to render the ākāra of the rule “sasajuso ruh” applies and we have pipathir and we have thus pipathiḥ with the transformation of ra into visarga (:). But the objection of the Bhāṣya is that if by the restriction suggested in the previous section that an ‘it’ enjoined by the rule “ardhadhātukasyed valadeḥ” (7.2.35), should remain ‘it’ and should not undergo any change, with what reason could we introduce a change in this ‘it’ by lengthening the ‘it’ in pipathiḥ?
14. The point raised here is that the rule “ardhadhatukasyeḍ valadeḥ” enjoins the restriction that whatever may be in other cases, this ‘it’ will not be changed. But this operation or injunction words off any change or ‘it’ as an anga operation. But the rule “rvor upadhāyā dirgha ikah” (8.2.76; a pada or a dhatu ending in ra or va will have its vowel of the ‘ik’ group lengthened) does not prescribe a change of the anga type and therefore such a change of the ‘it’ was not contemplated in the rule “ardhadhätukasyed valadeḥ” which asserted in an indicatory manner that in no anga operation this ‘it’ shall be changed. It should be noted in this connection that the a that was elided by the rule “ato lopah” (6.4.48) forming the word ‘pipathiḥ’, for the maxim is that, that which is elided in connection with ‘kvip’ cannot be regarded as sthanivat.
15. It was suggested that the rule “ardhatukasyed valādeḥ” (7.2.35; implied that this ‘it’ would not be changed and it was then objected: how we could lengthen the i by the rule “rvor upadhaya” etc. (8.2.76)? The reply is that the rule “rvor upadhāyā” etc. (8.2.76) is a rule in the tripädi and hence its operation is non-existent (asiddha) in the eye of the rule “ardhadhātukasyed valadeḥ” (7.2.35). This rule “ardhadhātukasya” etc. does not know that its injunction regarding the inviolability of ‘if’ was trans- gressed by the rule “rvor upadhaya” etc. in the tripādi. Fourth Ahnika 249 dhatu into akāra. This shows that guna does not occur in such cases and hence the sutra. Similar is the case with “janer dah”. (i) puganta-laghûpadhasya eva sarvadhatukârdhadhatukayoh (ii) puganta-laghûpadhasya sarvadhātukārdhadhātukayor eva. Panini. halo’nantarāḥ samyogaḥ 1.1.7. Consonants unintervened by a different class of vowels are together technically called conjunctions (samyoga) Bhāṣya. anantara iti. katham idam vijñāyate- avidyamānam antaram yeṣām iti, ahosvid avidyamānā antaraiṣām iti? (The word ‘anantarah’ is now picked up for interpretation and a question is now raised as to the meaning of the word). How should we understand the meaning of the word - is it, those of which there is no intervening time or is it those of which there is nothing intervening. 182 Bhāṣya. kim cataḥ ? yadi vijñāyate-avidyamānam antaram yeṣām iti. avagrahe samyoga-samjñā na prāpnoti- “apsvityap’ su” (RV. 1.23.19) iti. vidyate hy atrântaram. atha vijñāyate-avidyamānā antaraiṣām- iti na doṣo bhavati. What of that? If it is understood as those of which there is no hole (the first meaning), in that case the term ‘samyoga’ or conjunction could not be applicable to the avagraha (for meaning see below), as in “apsvityap’su”. There is a hole here. If it is understood as those of which there is no intervening letter, there is no harm.183
16. The two alternative meanings suggested are as follows: (1) In the first meaning ‘antara’ means a hole, i.e. moment devoid of any letter. In the first meaning ‘anantara’ is formed by a bahuvrihi compound with nañ particle. ‘anantara’ thus means holeless (niśchidra). “halo’nantaraḥ” thus means consonants in the pronounciation of which there is no hole or lapse of time. (2) In the second meaning, the word is ‘antara’ and not ‘antara’. ‘anantara’ means - in the middle. “halo’nantaraḥ” thus means those consonants that have no letter in between them.
17. Avagraha’ means the unit of time of the length of half measure or ardhamātrā and ordinarily the period of time required in passing from one alphabet sound to250 Bhāṣya. Mahabhāṣya of Patañjali yatha na doṣas tatha’stu. athavā punarastu-avidyamānam antaram yeṣām iti. nanu coktam - avagrahe sam- yoga-samjñā na prāpnoti- “apsvityap’su” iti. vidyate hy atrânaram iti. naiva doṣaḥ, na prayojanam. Then let it be that in which there is no harm of rather we may also accept the first interpretation as those of which there is no hole. But it has been said that in such a case the term ‘samyoga’ should not be applicable to an avagraha as in ‘apsvityap’ su’, there is a hole here.184 another alphabet sound or varna is called half measure, ardhamātrā. When two consonants are pronounced one succeeding the other, there is still the delay of half measure as determining the period of succession. But there may be a delay of an additional half measure in the case of pluta accent as in “guroranṛta” or “apsvityap’su”. Thus in such cases there would be one measure of time and consequently the term ‘samyoga’ would not be applicable. This interpretation is rejected.
18. A few points have to be first noted before we can understand the meaning of the above version. There is a sutra “pratyabhivade’sūdre” (8.2.83.)-At the time of counter-greeting on the part of a teacher-the ti part of the name should be both pluta and udatta ’ti’ means the part of a word beginning with the last vowel “aco’ntyādi ti” (1.1.64). There is another rule “samyoge guru” (1.4.11.), i.e. when a conjunction or a samyoga follows the vowel-part preceding it is called guru. There is another rule “guror anṛto ’nantyasyapy ekaikasya prācām” (8.2.86). In the case of counter-greeting on the part of a guru and not referring to a sudra, the pluta accent should be modified by the present rule to the extent that vowels not being a rand being non-ultimate as well as guru or ending in ‘api’ should have the pluta in the successive part of the ti, optionally. Thus the word ‘Devadatta’ can be made pluta in the following way- De3vadatta/ Deva3datta/ Devadatta3. But in ‘Kṛṣṇamitra’ where there is a krwe have only two-Kṛṣṇa3mitra/ Kṛṣṇamitra3. Now in the phrase “apsvityap’su” ‘a’ would be guru vowel as we would expect here pluta by the rule “guroranṛta” etc. But there is no pluta here, and this would ultimately lead to the view that the ‘su’ in ‘apsu’ is not a samyoga. But if this is so why then again accept such a view. The reply is that the samjñā-s are made for particular utility and for removing certain criticism. But the word “apsvityap’su” means ‘in the water in the water’. Pluta is possible only in the case of calling at a distance-durad hūte; and therefore there is no occasion for a pluta accent here, for ‘apsvityap’su’ is not a name called from a distance and therefore there is nothing to be afraid of that in such cases there cannot be any pluta and that this would lead to a miscarriage of our definition of samyoga as holeless. INDEX a ai un (Śivasūtra 1) 65 akarasya vivṛtopadeśa ākāragrahaṇârthaḥ (Va.) 65 agrahanam cen nuḍvidhiladeśavināmeṣu ṛkāragrahaṇam (Va.) 110 ați coktam (Va.) 181 atapara eva igghrasvadeśe (Va.) 96 atha śabdânuśāsanam (Pa.) 1 anākṛtiḥ (Va.) 146 anākṛtiḥ samjñā, ākṛtimantaḥ samjñinaḥ (Va.) 148 anarambho vā (Va.) 216 anukaranam śistâśiştâpratisiddheṣu yatha laukika-vaidikeṣu (Va.) 87 anubandhakaranârthaś ca (Va.) 57 anubandha-sankarastu (Va.) 73 anyatarartham punarvacanam (Va.) 185 anyatara-sahavacanāt samudaye samjñāprasangaḥ (Va.) 158 aprayukte dirgha-satravat (Va.) 44 aprayogaḥ prayoganyatvāt (Va.) 43 abhyupayo vā’paśabda-jñānam sabdajñāne (Va.) 48 ayogavāhānām atsu, natvam (Va.) 133 allopasya sthanivattvāt (Va.) 216 astyaprayuktaḥ (Va.) 41 astyaprayukta iti cennârthe śabdaprayogāt (Va.) 43 ā ākārasya tapara-karanam savarnârtham (Va.) 161 akṛtigrahaṇāt siddham (Va.) 81 akṛtyupadeśāt siddham (Va.) 57 ākṛtyupadeśāt siddham iti cet samvṛtādinām pratiṣedhaḥ (Va.) 57 ācāre niyama iti ced yājñe karmani sa niyamo’nyatrâniamaḥ (Va.) 49 ācāre niyamah (Va.) 46 ācāryācārāt samjñāsiddhiḥ (Va.) 145 adityavat syuḥ (Va.) 80 anyabhavyam tu kāla-śabda-vyavāyāt (Vā.) 78 āvartinyaḥ samjñāḥ (Vā.) 148 252 Mahābhāṣya of Patanjali i ik-prakaraṇān numlope vṛddhiḥ (Va.) 215 iko gunavṛddhi (Pā. 1.1.3) 167 ig-grahaṇam at-sandhyakṣara-vyañjananivṛttyartham (Va.) 168 inmātrasyeti cej jusisārvadhātukārdhadhātuka-hrasvādyor guneṣvanantya- pratiṣedhaḥ (Va.) 176 itaratha hy asampratyayo yathā loke (Va.) 143 inârtham câpi (Va.) 181 iṣṭa-buddhyarthaś ceti ced udattânudātta-svaritânunāsika-dirgha-plutānām upadeśaḥ (Va.) 57 u uktam śeṣe (Va.) 217 uttve ca (Va.) 124 upadha-roravityartham (Va.) 228 ubhaya-samjñāny api cchandāmsi (Va.) 136 uran, rapare ca (Va.)185 r!k (śivasūtra 2) 86 ! 1-kāropadeśo yadṛccha-śaktijānukaraṇa-plutyadyarthaḥ (Va.) 86 e e on (śivasutra 3) 96 ekatvad akārasya siddham (Va.) 73 ekadeśa-vikṛtasyânanyatvāt plutyādayaḥ (Va.) 89 ekayogo vā. (Va.) 173 ekavarnavac ca (Va.) 109 ekājanekāj-graheņeṣu cânupapattiḥ (Va.) 73 ekājanekāj-grahaṇeṣu cavṛttisankhyānāt (Vā.) 77 ekādeśe dirghagrahaṇam (Va.) 97 ai ai auc (Śivasūtra 4) 96 Index ka 253 kimartham sastram iti cen nivartakatvāt siddham (Va.) 157 kūpa-khānakavad vā (Va.) 49 kniti ca (Pa. 1.1.5.) 227 kniti pratiṣedhe tannimitta-grahaṇam upadha-roravityartham (Va.) 228 ga gargādi-bidādi-pāthāt samvṛtādinām nivṛttiḥ (Va.) 58 chandovat sūtram (Va.) 136 ch ja jñāne dharma iti cet tathâdharmaḥ (Va.) 46 tac ca mrjyartham (Va.) 180 ta tatrânuvṛtti-nirdeśe savarnagrahaṇam anaṇtvāt (Va.) 73 taddhita-kāmyor ik-prakaraṇāt (Vā.) 232 tadvacca tapara-karaṇam (Va.) 81 tasmād iglakṣaṇā vṛddliḥ (Va.) 184 tasya vivṛtopadeśad anyatrâpi vivṛtopadeśaḥ savarṇagrahanârthaḥ (Va.) 68 tulyarupe samyoge dvivyañjanavidhiḥ (Va.) 112 da didhiveviṭām (Pa. 1.1.6) 224 didhivevyoś chandoviṣayatvad dṛṣtānuvidhitvāc ca chandasaś chandasy adidhed adidhayur iti ca guṇadarśanad apratiṣedhaḥ (Va.) 244 didhyad iti ca śyan-vyatyayena (Va.) 245 dirgha-pluta-vacane ca samvṛtanivṛttyarthaḥ (Va.) 71 dirghe hrasva-vidhi-pratiṣedhaḥ (Vā.) 107 dravyavac copacārāḥ (Va.) 73 na na dhatulopa ardhadhātuke (Pa. 1.1.4) 213 nâvyapavṛktasya avayave tad vidhir yatha dravyeṣu (Va.) 109 254 Mahābhāṣya of Patanjali nipātanat syadādiṣu (Va.) 215 niyamād iko guṇa-vrddhi bhavato vipratiṣedhena (Va.) 177 numlopa-srivyanubandha-lope pratiṣedhartham (Va.) 214 nyayyabhāvāt kalpanam samjñādiṣu (Vā.) 87 pa puganta-laghûpadha-grahaṇam anantya-niyamartham (Vā.) 176 pūrvoccāritaḥ samjñi paroccāritā samjñā (Vā.) 148 pūrvopadese kittvakseḍ-vidhayo jhal-grahaṇāni ca (Va.) 125 pūrvopadese kittva-pratiṣedham vyalopa-vacanam ca (Va.) 129 pratyayāśrayatvad anyatra siddham (Va.) 215 pratyavayavam ca vakyaparisamāpteḥ (Va.) 158 pratyapatti-vacanam (Va.) 58 pratyekam vṛddhi-guna-samjñā-vacanam (Vā.) prayoge sarvalokasya (Va.) 46 prasaraṇe ca (Va.) 185 proktādayaś ca taddhitaḥ (Va.) 51 pluta-samjñā ca (Va.) 96 plutāvaica idutau (Va.) 112 plutyādiṣu aj-vidhiḥ (Va.) 96 bhave ca taddhitaḥ (Va.) 51 bhedakatvad guṇasya (Va.) 161 bhedakatvad svarasya (Va.) 161 bha ma maṇḍukagatayo’dhikārāḥ (Va.) 173 mrjyartham iti ced yoga-vibhāgāt siddham (Va.) 181 ya yanyakkyava-lope pratiṣedhaḥ (Va.) 214 yatha laukika-vaidikeṣu (Va.) 39, 145 yugapac ca deśa-pṛthaktvadarśanat (Va.) 78 Index 255 raki jyaḥ samprasaraṇam (Va.) 215 ravat pratiṣedhāc ca (Va.) 90 rephasya paropadese 125 ra anunāsikadvirvacana-parasavarna-pratiṣedhaḥ (Vā.) rūpa-sāmānyād vā (Vā.) 81 la lakārasya nittvād ādeśeṣu sthānivadbhāva-prasangaḥ (Va.) 232 lakārasya nittvād adeseṣu sthānivadbhāva-prasanga iti ced yāsuṭo nidvacanāt siddham (Va.) 232 lakṣya-lakṣaṇe vyākaraṇam (Va.) 54 lingârthā tu pratyāpattiḥ (Va.) 58 lingena vā (Va.) 146 lokataḥ (Va.) 37 lokato’ rthaprayukte śāstreņa dharma-niyamaḥ (Va.) 38 va varṇaikadeśā varṇa-grahaṇena cet sandhyakṣare samānākṣaravidhi- pratiṣedhaḥ (Va.) 107 viṣayârtham punarvacanam (Va.) 185 vṛtti-samaväyârtha upadeśaḥ (Va.) 56 vṛddhi-guṇāv alo’ntyasyeti cen midhimrji-puganto-laghûpadharcchi-dṛśi- kṣipra-kṣudreṣv ig-grahaṇam (Va.) 174 vṛddhi-grahanam uttarartham (Va.) 180 vṛddhipratiṣedhânupapattis tvik-prakaraṇāt [tasmād iglakṣaṇā vṛddhiḥ] (Va.) 181 vrddhir ādaic (Pa. 1.1.1.) 135 vṛddhirādaij iti mangalârtham (Va.) 148 sa śakunivat syuḥ (Va.) 80 śacañantasyântaranga-lakṣaṇatvāt [siddham] (Va.) 230 sabdapratipattiḥ (Va.) 50 sabde lud-arthah (Va.) 51 śarşu jaśbhāvaṣatve (Va.) 133 śāstra-pūrvake praoge’ bhyudas tat tulyam veda-sabdena (Va.) 47 256 Mahābhāṣya of Patanjali şa ṣaṣṭhyāḥ sthāneyogatvad ig-nivṛttiḥ (Va.) 185 sa samjñā-vidhāne niyamaḥ (Va.) 173 samjñādhikāraḥ samjñā-sampratyayârthaḥ (Va.) 143 samjñā-samjñyasamdehaś ca (Va.) 144, 146 sato vṛddhyādiṣu samjñā-bhāvāt tadāśraya itaretarāśrayatvad aprasiddhiḥ (Va.) 149 sandhyakṣare vivṛtatvāt (Va.) 110 sandhyakṣareşu taparopadeśaś cet taparoccaraṇam (Va.) 96 sambandham anuvartiṣyate (Va.) 173 samudaya-sādhutvārtham tu gargādi-bidādi-pāṭhaḥ (Va.) 58 sarvādeśaprasangaś cânigantasya (Va.) 175 sarve deśāntare (Va.) 44 siddham tu nityaśabdatvāt (Va.) 150 siddham tu sasthyadhikare vacanāt (Va.) 185 siddhe śabdartha-sambandhe (Va.) 31 sūtre vyākaraṇe şaşthy artho’nupapannaḥ (Va.) 50 ha hakārāsya paropadese ad-grahaneṣu hagrahaṇam (Va.) 124 ha ya va rat (śivasūtra 5) 124 halo’nantarah samyogah (Pa 1.1.7.) 249 halgrahaneṣu ca (Va.) 81 ICPR SERIES IN CONTEMPORARY INDIAN PHILOSOPHY General Editor: R. Sundara Rajan Language, Knowledge and Ontology KALIKRISHNA BANERJEE Rs 180 The Art of the Conceptual-Explorations in a Conceptual Maze Over Three Decades DAYA KRISHNA Rs 200 Karma, Causation and Retributive Morality: Conceptual Essays in Ethics and Metaethics RAJENDRA PRASAD Rs 250 The Philosophy of Nikunja Vihari Banerjee EDITED BY MARGARET CHATTERJEE Rs 150 Philosophy and Religion: Essays in Interpretation J.L. MEHTA Rs 175 The Philosophy of J. N. Mohanty EDITED BY DAYA KRISHNA AND K.L. SHARMA Rs 170 Rs 160 ISBN: 81-215-0530-3