Grammatical Riddles from Jain Works

Author(s): Nalini Balbir
Publisher: Z_Nirgranth_Aetihasik_Lekh_Samucchay_Part_1_002105.pdf and Nirgranth_Aetihasik_Lekh_Samucchay_Part_2
Catalog link: https://jainqq.org/explore/269046/1

Page #1

Grammatical Riddles from Jain works
Nalini Balbir
0. Broadly speaking, Indian riddles can be divided into two groups : First, those where the question invites a global answer corresponding to the meaning of the stanza taken as a whole are sometimes instances of fine muktaka-poetry. The technical term used to name representatives of this category is generally prahelikā. The second category includes all types of charades and clearly belongs to the field of śabda : there, the stanza is made of two or more independent questions, according to the pattern A, B, C, ... N. While the answers A, B, C ... are made of short words or monosyllables, the last one N consists in the sum of all the elements. The generic term used for this category is praśnottara, of which there are several varieties, depending, for instance, on the types of combinations or on the nature of the elements to be guessed. Like all fields of science in India, riddles have their śāstras too. The Vidagdhamukhamandana by Dharmadāsa (on which see below § 1.3) is the most famous of them. It extensively and systematically deals with the subject and considers all subdivisions of the two categories defined above.
Both types are tests of intelligence, but they require slightly different qualities from the questioner as well as from the discoverer. For praśnottarariddles, which are thus based on the de-composition of long sequences into small elements and are clearly meant to display knowledge, it is indispensable to possess erudition in mythology, poetic conventions, realia, etc. and full proficiency in all topics connected with the mastering of language. This means a perfect knowledge of lexicography in all its aspects, mainly of a synonymic kosa such as of Amarasimha and even more of the so-called Ekāksara- or Dvyakşara-kosas, since words consisting in one or two syllables are so widely used in this context. Needless to say,

Page #2

270
Nalini Balbir
Jambu-jyoti
an unbeatable familiarity with all the niceties of grammar, especially sandhi and morphology, is an utmost requirement2.
In this paper, I would like to study some riddles propounded by Jain munis of the past which have grammar as their central subject. This is humbly offered as a token of respect and profound admiration for a Jain monk who is an embodiment of the Jain tradition as well as of that makes a genuine pundit. Let us just remember that, apart from editing several works from the Jain Agamas as well as important Jain treatises like the Dvādaśāranayacakra, the Strīmuktiprakaraṇa, the Yogaśāstra, or the Pañcasūtraka, Muni Jambuvijayaji Maharaj has also to his credit a critical edition of the Sanskrit part of Hemacandra's grammar with its auto-commentary3, and is himself a capable writer of learned articles or introductions in Sanskrit besides Gujarati.
This is certainly not a coincidence if all the riddles which will follow are written in Sanskrit, even if they sometimes occur in works otherwise written in Prakrit (Jayasimhasūri's Dharmopadeśamālāvivaraṇa, Amradevasūri's Akhyānakamaṇikośa-vṛtti) or in the work of an author who can write in both languages (Jinavallabha, § 1.1). The study of a good deal of riddles found in Jain Prakrit narrative works clearly shows the strong impact of the Sanskrit tradition when it comes to this topic, and the feeling that Sanskrit is undoubtedly considered the language of scholarship and erudition par excellence prevails: besides the few cases of Prakrit works including riddles in Prakrit, there are those which, having Prakrit as their basic language, use more Sanskrit than Prakrit or exclusively turn to Sanskrit for riddles. On the other hand, even when praśnottaras are written in Prakrit, they very often appear as transpositions or translations of Sanskrit patterns, using, for instance, the same ekākṣara-words as their Sanskrit counterparts. In the case of grammatical riddles, the place of Sanskrit becomes even more important they all are in fact in Sanskrit, to the extent that there is no trace whatsoever of any reference to the rich tradition of Prakrit grammar and hardly any presence of Prakrit forms in our corpus.
What I call "grammatical riddles" refer to five cases:

---Praśnottaras of any variety containing grammatical questions (I); -Praśnottaras where the ultimate aim is to discover a form which is both a nominal and a verbal form (II);

Page #3

Grammatical Riddles from Jain works
271
-Praśnottaras where the ultimate aim is to discover a grammatical technical term (III);
--Praśnottaras where the ultimate aim is to discover a grammatical sūtra (IV);
--Praśnottaras concerned with knowledge of metrics (V), since chandas can be considered as an appendix of grammar.
1. Besides the corpus of Jain narrative works just mentioned, where riddle-verses generally form a small separate section and are exchanged as a pastime between a group of friends, members of a learned gosthi or a bride and bridegroom just after marriage, the sources used for the present investigation include three specialized texts :
1.1 The Praśnottara-sasti-ekašata (=JP) is from the pen of Jinavallabha, a prominent Svetāmbara acārya from the Kharatara-gaccha who lived in the last quarter of the 11th and the first of the 12th century and has many works to his credit.Jinavallabha's work is not a śästra because it completely lacks definitions. Except for the first mangala-verse dedicated to Pārsvanātha, the remaining 159 stanzas are riddle-verses, including the two last ones where the author gives the names of his spiritual masters and his own (vss. 159-160). Each of this riddle-verse is followed by the name of the variety it represents and by the answer. Since Jinavallabha's work is obviously meant to display excellence and cleverness in the topic, these mere indications are by no means sufficient to make the common reader able to understand the process of the riddle. Therefore a commentary is indispensable.
The printed text of the Prašnottara is found in a book where various small texts, mainly hymns, are collected : Śrīstotraratnakaradvitīyabhāgah satīkah. Śrī Yaśovijaya Jainasamskrta Pāthaśālā, Mehsana, vīra samvat 2440/A. D. 1914, p. 1a-33b. Each verse is followed by an anonymous Sanskrit avacūri (here : Ed.). Besides this edition, the following two manuscripts have been used for the present investigation and will be quoted whenever they are found to be illuminating :
-A1: manuscript No. 5000 kept in the L. D. Institute of Indology, Ahmedabad; 32 pages; tripātha-type; prsthamātrā-script; copied in sam. 1686 ( = A. D. 1629).

Page #4

272
Nalini Balbir
Jambu-jyoti
-A2 : manuscript No. 6198 kept in the L.D. Institute of Indology, Ahmedabad; 20 pages; pañcapātha-type?.
The identical commentary they contain, although apparently based on the avacūri found in the printed edition, is sometimes more detailed or clearer in its wording. It was written by a certain Punyasāgara in (V.]Sam. 1644 (= A.D. 1587).
The popularity of Jinavallabha's work is demonstrated by the rather larger number of available manuscripts, the fairly good number of commentaries it has given birth to, as also by the fact that some of his riddles have been borrowed by Jain writers (see below $ 3: JP 31).
1.2. The Praśnāvali ( = P), edited along with Jinavallabha's work (p. 55a-58b), is a short set of 15 Sanskrit riddle-verses followed by an avacūri. There is no mention of the author in the printed edition, but the text is usually ascribed to Municandra, a famous Jain author and commentator of the late 11th and early 12th century (he died in [V.]Sam. 1178 A.D. 1122). See further § 4 below.
1.3. Mahākavi Ajitasena's Alamkāracintamani (=AC) is a fullfledged work on poetics in Sanskrit verses. The system of exposition is what we expect in an alamkāraśāstra : definitions followed by illustrations. There has been some discussion about the date of its composition, which has been reasonably ascribed to the last quarter of the 15th century by Dr. Nemi Chandra Shastri, the editor of the text, and by A. N. Upadhye. All the available manuscripts, which are not very numerous, are written in Kannada script and housed in the libraries of Karnataka (Moodbidri); none seems to hail from North India. The second chapter (pariccheda), where the author states his wish to tackle the subject of sabdālamkāra (vs. 1), is in fact mostly devoted to a detailed treatment of prašnottaras and citrabandha. As is wellknown, the position of these two varieties of literary compositions is somewhat ambiguous: they are considered as attractive and striking because of their peculiarities and the great amount of virtuosity they require, but, at the same time, they are said to represent poetry of inferior quality (adhamakávya), especially by the strong advocates of the dhvani-theory.*
Ajitasena's work, coming rather late in the rich Indian tradition on poetics, is naturally indebted to the authors who have preceded him. It is
* Like Anandavardhana (c. 9th cent.). ---Editors.

Page #5

Grammatical Riddles from Jain works
273
interesting to note, for instance, that it shows several affinities with Vidyānātha's Pratāparudriya, a popular treatise in South India*. The Hindi introduction of the editor includes a detailed comparison of the AC with other Indian works on poetics (Natyaśāstra, alamkāra-chapters of the Agnipurāna, Bhāmaha, Dandin, Udbhața, Rudrata, Rājasekhara, Bhoja, Mammata, Vägbhata, Hemacandra, Visvanātha and Vijayavarni's Srngärārnavacandrikā) and presents a critical appraisal of the material included in the AC. As a Digambara Jain, Ajitasena of course draws inspiration from his own religious tradition and often quotes from the basic works or authors representing this cultural background (Adipurāna, Samantabhadra, etc.).
The second chapter on praśnottaras (not considered as such in the Hindi introduction) is no exception to this tendency. Ajitasena's sources are not here the above-mentioned alamkāraśāstras, where riddles, albeit present, are not the main topic, but the standard treatise on the subject, namely Dharmadāsa's Vidagdhamukhamandana (= VMM). Dharmadāsa's original Buddhist affiliation can be stated with some amount of certainty on the basis of internal evidence and there is no need to reject it as insufficient", but this has never been an obstacle to the diffusion of the VMM in broad circles. The Jains, who have shown considerable interest in all linguistic games and have never despised anything connected with śabdālaskāras, could not ignore the VMM : manuscripts of the VMM are numerous in Jain librariestoJinaprabhasūri (early 14th cent. A.D.), a Svetāmbara, is among those who commented upon it, and our Ajitasena's aim was obviously to make the VMM suitable for a Digambara Jain audience. Hence the following result : the verses giving the table of contents and the definition-verses, which are devoid of all sectarian character whatsoever, are almost identical to those of the VMM or, in case they are changed, show very minute differences, as if the author had had before him a copy of the original work and had deliberately introduced some small modifications like a plagiarist would do (see examples below § 4, 9 and 15). As for the illustrative verses, they are sometimes clearly based on the VMM (below 8 7 and 9) or at least inspired by it, they are simply different because
* If it has reference to Prataparudra, the last Kākatīya king of Tilangāna, its date of
composition may be some year in the first quarter of the 14th century.--Editors

Page #6

274
Nalini Balbir
Jambu-jyoti
the author has produced some verse of his own, or they are deliberately changed in order to give them a Digambara Jain colouring instead of a Buddhist one: a clear instance is provided by the riddle of the variety called tārkya, where the stanza is so built as to yield the answers vijñānavādinah in the VMM (2.58) andanekantavādah in the AC (2.59)". Whatever the case, a close comparison with the VMM always proves useful, especially when, more often than once, the present text of the AC (and its Hindi rendering) are unsatisfactory. One of the aims of this paper is to show that the second chapter of this work cannot be neglected by those who study riddles12.
(1)13
2. Grammatical riddles meant to test basic knowledge of the language are illustrated by the following stanza :
...kim vā rūpam tado jasi ?
śatr-caturthy-ekavaco bhavater iha kim bhavet ? (bhavate; Jayasimhasūri, Dharmopadeśamālā-vivarana p. 43).
... Or what is the form of tad in the nominative plural ? te
What would be the form for the root bhū in the present participle dative singular ? bhavate
This is one among many cases showing the all-pervading prevalence of Pāninian metalanguage, at use even in works of indisputable Jain origin, Here, the aim is not to investigate the familiarity with a specific grammatical school and its terminology, but only to ask a rather simple question which should not arouse any difficulty for anybody having a minimal training in indigenous Sanskrit grammar. As a matter of fact, the technical terms used are of an elementary type. The answer to the first grammatical question is to some extent dependent on the answer to the second one, since jas could theoretically refer to the feminine as well as to the neuter. But both answers have to be compatible, therefore täh/ tāni are easily ruled out.
Practice of language and knowledge of technical terminology go together in the following question :
*... pada-nipunaih pañcami kena vācyā ? (bhämäratasanatemanasi; JP 71)
By which (element) can the connoisseurs in the field of words express the ablative case ? - By tas (tasā, instr. sg.).

Page #7

Grammatical Riddles from Jain works
275
Regarding phonetics, questions connected with the status and transformation of the visarga form a specific group :
*... kidrśaḥ syād visargaḥ ? (vyantarādivyastah; JP 27)
How would the visarga be? - It throws down the s (s-tah, sam sakāram tasyatīti s-tah, kvip, ct. mss. A1, A2, Ed.).
*... kam upaiti sarva-vidhibhir-jāto visargah kutah ? (P 11)
Where does the visarga go, produced from where through (the application) of all rules ?. (it goes) to r (coming) from r (ram rah).
*... kidrg nămi-parā visarga-vihatir-vidvan ? (P 11)
O learned man, how is the destruction of the visarga when it follows a vowel causing cerebralization ? - It gives (i.e. produces)r (ra-rā); cf. nāmi-paro ram, Kātantra 1.5.12, and the other rules of this section which all concern the treatment of the visarjaniya.
3. Questions where the point is to find a synonymical root are rather frequent in riddle-verses. The knowledge of dhātupāthas, a field which belongs both to grammar and to lexicography, is required for this purpose 14 Since there are generally several synonyms, the choice could be difficult. The selection of the adequate root and the elimination of the others are to be operated in view of all the questions making the riddle, because only one is able to fit in.
? (rāmāyane; Jayasimhasüri,
*... danammi ko vă dhău Dharmopadeśamála-vivarana p. 90)
...Or what is the root for 'to give'? - rā -: cf. rā dāne, Pān. Dhātup. 2.48; Palsule 1955, 180 s.v. dāne.
*... viratau ko dhātuh ? (malayamarutah; JP 31)15
... What is the root for 'to cease' ? - yam -; cf. Palsule 1955, 193 s.v. viratau, the only source mentioned being Vopadeva.
* bhuktyarthateha vihitā katamasya dhātoh? (visadapañcamah; JP 53)
For which root has the meaning 'to eat' been prescribed ? - camah, "for the root cam-": cf. Palsuie 1955, 166 s.v. adane.

Page #8

276
Nalini Balbir
Jambū-jyoti
*... prayoga-nipunaih kah sabda-dhātuh smrtah? (madguravojinesvarasürayah; JP 159)
... Which is the root for 'to emit a sound that the people versed in the correct use of language have taught ? - ras : cf. Palsule 1955, 195 s.y. sabde.
*ko dhātur gamane ? (P 7)
... What is the root for 'to go'? - mimu : cf. ama drama hamma mimr gam gatau quoted by the avacuri on P = Pān Dhātup. 1.493-496; compare Hemac. Dhātup 1.392-396 : ama drama hamma mimr gamlrm gatau. Here the choice has to be compatible with the last question of the stanza vidito vargäntya-varnaś ca kah? which directs towards a root beginning with the letter m
* ko dhatur yada resane ? (P 11)
Say, what is the root for 'to yell'? - ri : cf. ri resane ri-dhātuh of the avacūri, reminding of Pān. Dhātup. 9.30 ri gati-resanayoh (and Hemac. Dhātup. 3. 18 rims gati-resanayoh); Palsule 1955, 190.
In our next two examples, the standard straightforward wording is replaced by something less expected, but quite usual in the context of riddles :
*... atho khādad-grhītāvadat : "kidrg bhati saro 'rhataś ca sadanam ?"...(jinavallabhena; JP 160)
And the root 'to eat' together with the root 'to take' said : "How does a lake look beautiful ? And how does a Jain temple look beautiful"...
As a matter of fact, in riddles anybody or anything, whether an object, an abstract entity, or even a word, can be the grammatical subject of a verb meaning 'to speak'. This has an impact on the reply, which will in turn be addressed to these speakers, now in the vocative form. This is another way to demonstrate one's own ability in the field of grammar by building vocative forms of monosyllabic words or artificial compounds. Thus here, the answer will be :
Root 'o eat' together with root 'to take' ! adorned with lotuses (does a lake look beautiful) and full of Jinas (does a Jain temple look beautiful) : al-la, vana-ji (i.e. secondary derivative with the possessive suffix -in from vana-ja “otus"), jina-vat.

Page #9

Grammatical Riddles from Jain works
277
Al-la is the vocative of the dvandva formed by ad- to eat and lā - 'to take' (cf. Palsule 1955, 168 s.v. adāne) 16.
* ko ... krīdārtha-dhātur vadet "ko'stavimśatayā prasiddhim agamad varnah kaka-lekhake ?" (P.12)
Who, being a root meaning 'to play', would say "Which is the letter of the consonant chart that is known to be the 28th ?" - Root 'to play'! the letter l.
The last question stated in a rather simple way points to the root las, recorded in the Dhātupāthas (cf. lasa ślesana-kridanayoh, Pān. Dhātup. 1.746 = Hemac. Dhātup. 1.543; Palsule 1955, 172 and 196). The root which was the speaker in the question has to take the form of the vocative when it becomes the addressee in the answer (las lasatīti kvip prathamă si, vyañjanác ca si-lopah, he lah, ct.); lah is also the nominative form for the la-kára.
4. Such an elementary prerequisite as the knowledge of the alphabet and of the Sanskrit phonemes has been variously applied in the context of riddle-making.
First, the varnas are useful because they provide convenient and obvious monosyllables for praśnottaras :
* tavargge pañcamaḥ ko vā? (kumjarena; Jayasimhasūri, Dharmopadesamālā-vivarana p. 43)
What is the fifth letter in the group of cerebrals? - na
.*... khăntam brāhi kim nāntam icchasi ? (īhālamkārasangatam; Bhavadeva, Pārsvanāthacarita 2.224)
Tell the letter which comes after kha (and) which one do you require coming after na ?- ga (and) tam.
In the 15 verses of the Praśnāvali, the last question of the riddle invariably concerns a letter (varna), more precisely a consonant or a class of consonants (varga). It can be expressed very simply under the form: What is the third consonant (vs. 2), the twentieth (vs. 4), the twenty-first (vs. 5), the twenty-sixth (vss. 8-9), the twenty-seventh (vss. 10-11), the twenty-eighth (vs. 12; above), the thirtieth (vs. 15), meaning that all answers will contain respectively ga, na, pa, ya, ra, la, and śa. A short descriptive statement can replace the serial number : the questions 'What is the last phoneme of the last group ?' (vs. 7) and 'What is the last phoneme

Page #10

278
Nalini Balbir
Jambu-jyoti
among the semi-vowels?? ?' (vss. 13-14) respectively mean words including ma and va. When they occur, indirect statements based on numerical equivalences are not too difficult to grasp : 'What is the phoneme in the sequence of phonemes and at the same time the label of a group of consonants which corresponds to the digits of the moon ?' (vs. 3) refers to ta, the 16th consonant, and 'What is the letter in the chart of consonants (ka-ka-lekhake) that has got the (same nunber as the) number of Jinas ?' (vs. 6) refers to bha, the 24th consonant. The point is then to use this information as a hint to answer the other questions of the stanza, which are much more tricky, to find out ekāksaras or dvyaksaras containing the given consonant (e.g., mama, ami, mimu, mu, ume, amai, mas, ama, aumam and mas are the words answering the different questions put in vs. 7).
Jinavallabha combines the use of the alphabet as well as lexicographical matter in a sophisticated manner which is peculiar to his work. The first stage is to produce new synonyms for words offered in the question by adding or suppressing syllables in a given phonetic sequence (dhvani, sabda, pada, varnāli). Then comes the real answer to the riddle, which consists in making a bahuvrihi-compound describing how the starting phonetic sequence has been modified. There are 16 such instances which can be classified as follows according to the pattern they exhibit :
The starting point is a word, i.e. a meaningful phonetic sequence which will acquire a different meaning suiting the question once it has been submitted to modification. The position within the alphabet of the letter to be discovered is described in a rather indirect fashion :
* jāmātaram samākhyāti kidršo vathara-dhvanih? (agadaśamah; JP 35)
How does the word 'stupid' denote a bridegroom ?--Once it is deprived of the tenth letter, starting with ga (a-ga-daśamah), i.e. tha : vathara + vara.
* aśiva-dhvanir ākhyāti tiryag-bhedam ca kidrśah ? (aparäjayah; JP 15) How does the word 'inauspicious' denote a kind of animal ?.Once the vowel coming after a is removed (a-paraj-ayah, where ac is the pratyähära for a vowel in Pānini's grammar), i.e. i : aśiva → aśva 18.
* brüte sipha-dhvanir atha Sriyam atra kidrk? (visadapañcamaḥ; JP 53)

Page #11

Grammatical Riddles from Jain works
279
How does the word 'fibrous root' denote Laksmi?-Once it is deprived of śa and of the fifth letter, starting with da (vi-sa-da-pañcamah), i.e. pha : sipha > ita > Ya, one of the several synonymic monosyllables for Laksmī.
The same process is applied to a word which is partly meaningful and partly made of a sequence of phonemes created ad hoc :
* kidrg bhūmiśubhāsa-sabda iha bho viśrambha-vācī bhavet ? (atanavamadaśamah; JP 55)
How does the word bhumiśubhāsa denote 'trust'?-Once it is deprived of the ninth and the tenth letters, starting with ta (a-ta-nava-dasamah), i.e. bha and ma : bhümiśubhāsa > *uiśuasa > visvāsa.
* nayanagati - padaṁ kidrk pujayatiti artham abhidhatte ? (apaprathamangajam; JP 85)
How does the word nayanagati express the meaning 'to worship'?Once it is deprived of the first letter before pa (i.e. na) and once it has ja instead of ga (a-pa-prathamam ga-jam): nayanagati → yajati.
In several cases, the starting point is clearly a mere phoneme or a phonetic sequence without any meaning. A rather simple instance involving only one type of modification, namely an addition, is :
*... atha silpi-šikya-dehān udāharati kā-dhvanir atra kidrk ? (jinadantarucayah; JP 2)
How does the sound kā denote 'a craftsman', 'a rope', 'a body'?-- Once it ends (respectively) with ru, ca or ya (anta-ru-ca-yah) : kā kāru, kāca, käya.
An illustration for modification through removal of a syllable is :
* nasthātavyańca-śabdo'yam pradosan kidrśaḥ ? (vitathavacanah; JP 62)
How does the word 'and it should not be stated' denote evening ?' — Once it is deprived of ta, tha, va, ca and na (vi-ta-tha-va-ca-nah) : nasthātavyamca → sāyam.
In the following riddle, the syntactic pattern is not a bahuvrīhi as is usual, and instead of forming a long compound, the answer is extremely short, leading to an effect of surprise, which is also a part of riddle-games :

Page #12

280
Nalini Balbir
Jambu-jyoti
* adivāri-sabdo vā kais tyaktah prāha grha-deśam ? vaih; JP 52)
Of which elements should the word adivārī be deprived to denote a part of a house ? - Ofi, i and a (i-1-a in the instrumental plural > yaih) : adiväri → dvār, a door'.
When the synonyms which are to be found out are not ordinary words, but typical ekāksaras, the effect of surprise and the pleasure of the discovery are even greater :
* asumeti-padam kidrk Kämam Laksmim ca bodhayati ? (avatamasam; JP 7)
How does the sequence asumeti refer to Kāma and Lakşmi in the vocative ? - Once it is deprived of the letters va (i.e. u and a), ta, ma and sa (a-v-a-ta-ma-sam) : asumeti → E, vocative of I, a frequent synonym of Käma, and 1, vocative of l, a frequent synonym of Laksmī. Both these monosyllables are well recorded in Ekāksarakosas and widely used by Jinavallabha in his Praśnottara.
On the other hand, there are riddles of this type implying two types of modifications, an addition combined with a suppression, in which case the answer can be expressed through two bahuvrihi-compounds :
* Hara-nikara-patha-svah-srsti-vāci narnaga-padam kidrg ? (bhavamāsvasādiśastanam; JP 16)
How does the word narnaga designate 'Siva', 'a collection', 'a path', 'the sky', 'creation'?—Once it starts (respectively) with bha, va, mā, sva or sa and is deprived of na (bha-va-mă-sva-sadi sasta-nam): narnaga Bharga, varga, marga, svarga and sarga.
* vadati javina-sabdah kidrśah sat-kavindrah
kathayata jana-śünyam kajjalam bhartsanaṁ ca ? (vyantarādivyastah; JP 27)
Good poets tell : how does the word javina denote 'deserted by people', 'collyrium' and 'threat' ?-Once it starts with vi, am or tar, and once vi has been thrown away (vy-am-tar-ādi-vy-astah) : javina- vijana, añjana, tarjana.
* ... kidsk/pūta-vāta-paritāpa-mlecchopăsti-nuti-grha-krīdā
homa-viśva-vegavato jalpati pavadanada-padam ? (pāpadayasestabhadehabhujādivipadam; JP 34).

Page #13

Grammatical Riddles from Jain works
How does the word pavadanada denote 'fire', 'wind', 'scorching heat', 'barbarian', 'worship', 'praise', 'house', 'play', 'oblation', 'universe', 'quick'? Once it starts with pā, pa, da, ya, se, sta, bha, de, ha, bhu or ja and once it is deprived of pa and da (pa-pa-da-ya-se-sta-bha-de-ha-bhu-jādi vi-padam): pavadanada → pāvana, pavana, davana, yavana, sevana, stavana, bhavana, devana, havana, bhuvana and javana.
s
tvamaladaya-padaṁ vā āśrayābhāva-mūrcchā kataka-nagaviseṣān kīdṛg āmantrayeta ? (avipravamādyatvamadam;
281
JP 39)
How would the wordtvamaladaya refer to ‘abode', 'inexistence', 'fainting', 'bracelet' and a specific mountain in the vocative?-Once it starts withā, vi, pra, va orma and once it is deprived oftva, ma and da (a-vi-pra-va-mādy a-tva-madam) :tvamaladaya→→ ālaya, vilaya, pralaya, valaya and Malaya.
*
ura-sabdaḥ kalyāṇada-bala-hima-śṛngān vadati kīdṛk ? (adisyantaravaviśikhānuḥ; JP 11)
How does the word 'sheep' denote 'bestowing prosperity', 'army', 'cold' and 'summit' ?-Once it starts with si, has in the middle va, vi, si or kha, and once it is deprived of u (adi-si-antara-va-vi-si-kha-an-uḥ): ura → śivara (ra, agent noun built on the root rā 'to give'), sivira 'military camp', sisira and sikhara.
Finally, substitution of syllables is also one process at work. It can be expressed either through the pregnant use of the locative case in conformity with the process applied in the grammatical metalanguage or through a compound (va-na, JP 54, below; and ga-jam, JP 85, see above) :
* ravaravaka-varṇālī kīdṛg bravīti gatāratim? (aparāvaṇā; JP 54)
How does the sequence ravaravaka denote somebody who has no regret? Once it is preceded by a and once it has ṇā instead of va (a-parā va-ņā): ravaravaka → a-raṇaraṇaka.
*paścādudbhava-jānusambhava-narān daityāntya-daṁṣṭrāngajān
Mandam ca kramaso muja-dhvanir agāt kīdṛk kva kasmin sati? (ajamadahatabhäpūrvomene; JP 22)
How does the word muja denote respectively 'one who is born after', 'coming from the knee', 'man', 'demon', 'tusk', 'born from one's own body', 'Saturn', what being where ?-Once it starts with a, jā, ma, da, ha, ta, or

Page #14

282
Nalini Balbir
Jambu-jyoti
bhä, when there is na instead of ma (a-jā-ma-da-ha-ta-bhā-pūrvo me ne): muja - anuja, jānuja, manuja, danuja, hanuja, tanuja and Bhānuja!.
In all these cases, whatever the process at work, the analysis of the compound which forms the answer in the smallest possible elements looks both artificial and skillful.
(II)
5. We now come to varieties of grammatical praśnottaras which are defined and illustrated in the relevant śāstras. In the so-called nămākhyātajāti, the same sequence of phonemes is submitted to a twofold analysis and can be understood both as a noun (substantive, adjective, indeclinable) and as a verb. Mahākavi Ajitasena states :
sup-tin-anta-prabhedena suyogitvăd dvidhottaram ekam eva bhaved yatra tan nāmākhyātam ucyate (AC 2.55)
When an answer, which is twofold because of a good grammatical connection, corresponding respectively to an inflected nominal form and to an inflected verb form, is the same, the riddle is called nāmākhyāta, "noun (and) verb."
This verse obviously echoes Dharmadāsa's definition; both the syntactic pattern and the vocabulary are identical, except for one word :
ekam evottaram yatra suslistatvad dvidhā bhavet sup-tin-anta-prabhedena nāmākhyātam tad ucyate (VMM 2.37)
When the same answer becomes twofold because of a good double meaning, corresponding respectively to an inflected nomimai form and to an inflected verb form, the riddle is called nāmākhyāta, "noun (and) verb."
In the VMM, this variety is illustrated through 18 examples, a fairly large number compared to what we get for other types. Three groups can be sorted out depending on the pattern at work :
5.1. Q1 + Q2 ... + x + A + A, ... + A as nominal forms and Q. → A, A, ... + A, as a verbal form. E.g.,
padam anantara-vāci kim isyate ? kapi-patir vijayi nanu kidrśah ? para-gunaṁ gaditum gata-matsarāh kuruta kim satatam bhuvi sajjanah ? (VMM 2.39)

Page #15

Grammatical Riddles from Jain works
283
Which word expressing the absence of interval is desired ?-In immediate succession anu
And how can the Lord of Monkeys be victorious ?_With Rāmah saRāmah
O good people who are free from envy, what do you always do on the earth in order to tell others' virtues ? _We follow (them) anusarāmah
This pattern is exhibited in 13 cases :
a-bhi bhavati / abhibhavati (2.38) - anuja grhe / anujagrhe (2.40) - samā dadhi re / samādadhire (2.41) - avamam tårah / avamantārah (2.42) - a-nu mātā ase / anumātāse (2.43) - dāsya mahe / dāsyāmahe (2.44) - prayāsī ati / prayāsyati (2.45) - santi a-jatu / santyajatu (2.46) - ali lavam alīlavam (2.48)--samaram jayam / samaranjayam (2.51) - parā ājaye mahi parājayemahi (2.52) - vină udaye yam / vinodayeyam (2.53).
Whereas some of the verbal forms to be discovered are rather simple, some others are not so common in the ordinary language and require a full knowledge of grammar.
5.2. The functioning of the second pattern, represented only by two instances, is basically the same, but instead of forming only a verbal form one of the answers is a syntagm or a full sentence :
nadi iyatām / na dīyatām (2.47) - mānava nagäḥ / mä anavana gāḥ (2.49).
5.3. In the last pattern, the stanza is made of two questions only. The first one produces a noun, the second one a verb : Q, →A as a noun and Q, → A as a verb. The difference is that here the answer is not progressively built with the help of two or more elements which are piled up as in the preceding cases. E.g.,
kim akaravam aham Harir mahīdhram sva-bhuja-balena gavāṁ hitam vidhitsuh ? priyatama-vadanena piyate kah
parinata-bimba-phalopamah priyāyāh ? (VMM 2.50) I, Hari, what did I do to a mountain with the strength of my own arms, as I wanted to do something useful for the cows ?—You held (the Govardhana) adharah

Page #16

284
Nalini Balbir
Jambu-jyoti
In a beloved one what is the thing resembling a ripe bimba-fruit that is drunk by the mouth of a lover ?The lip adharah
The two other such instances are : sa-mada-āsyah/samadāsyah (2.54) and a-hāsyatām/ ahāsyatām (2.55).
6. In Jinavallabha's Praśnottara there are three instances which come under the caption nāmākhyāta-jāti according to the commentary of mss Al and A2, while the avacūri going with the printed edition gives them the more general label dvir-gata-jāti (for JP 81, 83) or vyasta-samasta-jāti (for JP 66). The first two correspond to the definition given above and are identical with the last pattern described for the VMM.
* prcchāmi jala-nidhir "aham kim akaravam sapadi sasadharābhyudaye ?"
alam udyamaih sukstinām ity ukte kidrśaḥ kah syāt? (JP 81)
I, the ocean, I ask : "What did I immediately do at moonrise ?"_You gleamed20 samudalasah
When he is told : "Enough exertion for good people !", who would be how ?-The idle fellow (would be) full of joy sa-mud alasah
* rataye kim akurvātām parasparam dampati ciran militau ? mokşa-patha-prasthita-matih pariharati ca kidršīm janatām ? (JP 83)
What did a couple do mutually concerning love once they met after a long time ?-Both of them rushed (to make love) atattvaratām
What type of crowd does the one who has engaged himself on the path of Liberation avoid ?-(The crowd) who finds pleasure in what goes against the (true) principles a-tattva-ratām
In this verse, the discrepancy between the tense used in the question (the imperfect) and the one used in the answer (aorist) is worth noting. It is unusual in the context of riddles, but easily explainable within classical Sanskrit where both tenses are more or less freely exchanged.
The next instance shows how basic patterns are liable to variations. Instead of one verbal form, there are two here, which are then combined to make a rather sophisticated compound :
* tanvi ! tvam netra-türņodgata -madana-sarākära-cañcat-katāksair laksyi-krtya smarārtān sapadi kim akaroh subhru ! tīksnair abhīksnam ?

Page #17

Grammatical Riddles from Jain works
285
kim kurvāte bhaväbdhim su-muni-vitaranād dāyaka-śrävakau drāk ?
sraddhāluh prāpta-mantrady-ucita -vidhi-parah prāyasah kīdršah syāt? (JP 66)
O fair one ! Having taken as your target those who suffer from love, what did you do immediately and repeatedly, lovely woman ! with your moving and sharp sideglances resembling the arrows of Kāma quickly discharged by your eyes ?-1 pierced (them) avidhyam
What do the one who gives and the Jain layman quickly do regarding the ocean of rebirth when a good monk makes them cross?
-They both cross taratah
Generally how will be a faithful fellow who has received mantras and is fully devoted to the relevant rules ?—He will find pleasure in putting an end to what goes against the rule. a-vidhy-anta-ratah
7. Ajitasena's AC has two instances of the namākhyāta-variety. Both are clear representatives of the pattern analysed in 8 5.1.: questions 1 and 2 lead to the answer as a noun, question 3 to the answer as a verbal form. But, as is often the case, the text needs emendation :
* sevitä vihvalam kartun kā kşamā suciram ghatah nămbho dharati kidrksam sastram kurutha dhidhanah ? (AC 2.56) surāmah. surā, āmah. surāmah dadmah.
The above investigation of the illustrative verses found in VMM and JP should have made clear that namākhyāta-riddles, like all other praśnottaras, are built on strict grammatical patterns : a question including kim means a substantive in the answer (with the same grammatical gender as the interrogative pronoun), a question containing kīdrk or the like implies an adjective, simple or compound, in the answer, a question with a verb in the second person implies a verb in the first person in the reply, etc. Sticking to these facts is necessary and proves specially rewarding when it comes to understand texts which are badly transmitted. This holds true for the second line of the above verse which should be read :
nămbho dharati kidrk? kim21 śāstram kurutha dhīdhanah ? The whole verse can then be translated :
What is the thing, the addiction of which can make one confused ? --Alcohol surā

Page #18

286
Nalini Balbir
Jambū-jyoti
Which type of pot cannot retain water for long-An uncooked one amah
O intelligent fellows, what do you do concerning the treatise ? -We give (them) well (?; cf. rā dāne, Dhātup.) su-rāmaḥ Ajitasena's second instance reads : * yaminipratimāyoge kidrśam yatināṁ kulam ? kam vandante surā nityam ? kamar kim akarot sudhih ? (AC 2.57)
abhyabhavam. abhi bhayarahitam. abhavam saṁsārahinajanam. abhyabhavam niräkaromi sma.
Here again, a simple correction would suit the grammatical pattern of the riddle better. The answer as a verbal form is abhyabhavam, 1st person of the imperfect. The corresponding question should then normally include a second person : akaroh instead of akarot.
How is a group of Digambara monks when practising käyotsarga in the night ?-Without fear a-bhi
Who is it that the gods always respect ?-The one who is free from rebirth a-bhavam O wise man, what did you do with desire ?-I overcame (it)
abhyabhavam In this case, a comparison with the VMM further shows that Ajitasena has drawn inspiration from this work not only for definitions but also for illustrative verses, transforming what is found in VMM 2.38 in order to suit an ascetic context :
samara-sirasi sainyar kidrsam durniväram ? vigata-ghana-nisīthe kidrse vyomni sobhā ? kam api vidhi-vašena prăpya yogyābhimānam
jagad akhilam anindyam durjanah kim karoti ? What type of army is difficult to attack at the front of a battle ? --The one who has no fear a-bhi In a night without clouds in which type of sky is there light? -In (a sky) with stars bhavati
When, by chance, he has got some reason to be arrogant, what does a bad fellow do with the whole world of unblamable people ?-He humiliates (them) abhibhavati

Page #19

Grammatical Riddles from Jain works
(III)
8. In the preceding category, the words to be guessed belong to the language used in performance. But several grammatical riddles are also meant to test the proficiency in the field of metalanguage, when the question and/or the answer involves a pratyāhāra (§ 8.1) a kṛt or taddhita affix (§ 8.2) or a grammatical technical term (§ 9). As will be seen, in all instances but one (§ 8.2) the standard is Panini's metalanguage: 8.1. * agre gamyeta kena? (halāsamstarasamsarayetaḥ; JP 36) What is it that should go in front ?-The consonant (hala, instr. sg.) Hal is the well-known pratyāhāra for referring to the consonants. As for the respective importance of consonants and vowels in the linguistic system, the commentary (mss A1, A2) states: svarāt pradhanam vyañjanaṁ bhavati, a belief already handed down in the older tradition: cf. vyañjanaṁ svarāngam (Taittirīya Prātiśākhya I, 6 quoted by Abhyankar 1961 under vyañjana).
l
pratyāhāra-viśeśā vadanti: "Nandi nigadyate kīdṛk ?" (ajagaṇaḥ,
Certain pratyāhāras say: "How can Nandin be called?
-O ac, ak and an, (he can be called) an attendant of the Unborn (=
Śiva).
JP 76)
Here the sequence ajagaṇaḥ is understood in two different ways: first as aj-ag-aṇaḥ, which forms the vocative of a dvandva compound made of the three Pāṇinian pratyāhāras referring respectively to the vowels (ac), to the vowels as such (a/a, i/i, u/u, r and I: ak), to vowels and semi-vowels taken together (an), then as Aja-gaṇaḥ (nom. masc. sg.).
8.2. * syad utaḥ kena vṛddhiḥ? (virājñāvinudatipāpam; JP 6) By what is the vṛddhi of u caused?—By tip (tipā).
287
Tip is Panini's marker for the personal ending of the Parasmaipada 3rd person singular. The commentary of JP explains: tip-pratyayena 'uto vṛddhir' ity ādinā nautity adau yathā vṛddhir bhavati, and thus quotes the beginning of Pān 7.3.89 uto vṛddhir luki hali, “A root ending in u which has no Present characteristic gets vṛddhi before a pit särvadhātuka affix beginning with a consonant"22.
guna-vṛddhi vāj-jhalau kasya ? (syuh; JP 148)
*

Page #20

288
Nalini Balbir
Jambū-jyoti
Which (root) has a guna and a vrddhi which are both vowels and consonants ? - The root r (uh, gen. of this root).
The reason is that the guna of this root is ar (vowel + consonant) and its yrddhi is ar (idem), as the commentary explains : ar-är-laksana-gunavrddhi ac-halau svara-vyañjana-rūpe kasya dhātoh syātām ? uh r-kārasya rddhi r-to dur. ur ādese dity-antya-hrasvādeh r-lope un iti rupa-siddhih.
The following instance is fairly sophisticated : * "nirdambheti yad arthatah pranigaded rūpam vi-purvāc ca tat
mināteh kam apeksya jāyata ?", iti ktvä-pratyayah prcchati (bhavadyavādeśam; JP 98)
The affix ktvā asks: "What needs to be taken in consideration in order that the root 'to hurt', preceded by vi, takes the meaning ‘not deceitful' ? Having the substitute yap instead of you : bhavad-yab-ādeśam 23.
As a matter of fact, when the root mi, mināti is used as a simplex, it will form its absolutive with the suffix - tvā (ktva in Pāṇini's language), but when it is preceded by the prefix vi, its correct absolutive will be vi-maya (see Pān. 6.1.50 for the substitution of a in this situation) and will thus use the suffix -ya (krt-affix) which Panini calls lyap (cf.7.1.37 samāse 'nañ-pūrve ktvo lyap), and which Jinavallabha calls yap (> yab through sandhi)24 In doing this, he here follows the metalanguage of Hemacandra's grammar where the counterpart of Pānini's sūtra reads : a-nañah ktvo yap (3.2.154).
9. A variety of riddle where the answer should consist in a grammatical technical term is recorded in the specialized treatises (AC and VMM) and is named śābdiya-jāti. Here again, Ajitasena's AC follows Dharmadāsa's VMM and, like his predecessor, he considers the definition of this variety along with the definition of three other varieties involving knowledge of specific fields, namely tarkya-jäti "philosophical" (see above $ 1.3 end), sautra-jāti (below (V) and sāstraja-jäti "scientific" (AC 2.58 and VMM 2.56). But these types were probably considered obvious, or best understood through concrete examples, since there is hardly any definition (iñeyam.../ śābdīyam śabda-samjñabhih...., VMM 2.56;...sabdād udbhavam ... sābdam ..., AC 2.58).
The illustrative verse of the AC as it is edited reads as follows:
na slāgh[y]ate munih kasmai? sub-antam kim nigadyatām ?

Page #21

Grammatical Riddles from Jain works
289
a-kārädy-anubandhänām dhātūnām nāma kim vada ? (AC 2.61)
parasmaipadam. munih parasmai na slaghate sva-gunadhikam dharmam na jñāpayati, api tu sva-nindām para-prasamsam ca karotīty arthah. śābda-jatiḥ [Ed. śābdījātih].
Who is it that the ascetic does not praise ?--An other person parasmai How is an inflected nominal stem called ?-A word padam
Tell, which name is given to the roots having the anubandha a, etc. ?-- parasmaipadam
Now, the wording of the first and the third questions and their respective answers are obviously unfitting. The explanation of the first answer given by the author is rather tortuous, unconvincing and contradicts the answer parasmai. Further, parasmaipadam cannot apply to roots possessing the anubandhaa, on the contrary, this vowel normally "indicates the placing of the Atmanepada affixes after them, if it be uttered as anudātta and of affixes of both the padas if uttered svarita25.”
Once again, a look at the VMM helps to restore the proper text because it immediately appears that Ajitasena borrowed from it. In all editions and manuscripts we read :
na ślāghate khalah kasmai ? sup-tin-antam kim ucyate ? lädeśānām navānam ca tinam kir nāma kathyatām ? (VMM 2.61)
Who is it that the wicked fellow does not praise ?-An other person parasmai
How is an inflected nominal stem called ?-A word padam
What is the name given to the nine verbal endings having la as their substitute ? parasmaipadam
The contents of the first question and its answer are in agreement with common sense of observation, and the answer to the third question is in agreement with grammatical theory:lah parasmaipadam (Pān. 1.4.99)26. Now, on the other hand a part of the manuscript tradition, mirrored in the common Indian editions of the VMM. also adds another stanza 27.
satatań ślāghate kasmai nico ? bhuvi kim uttamam ? kartary api rucādināṁ dhātūnām kim padam bhavet? (VMM 2.62) Who is it that the mean fellow always praise ?-Himself ātmane What is the best thing on earth ?--A place padam

Page #22

290
Nalini Balbir
Jambu-jyoti
What would be the word (i.e., verbal affix) for the roots to shine' and others, even if there is an agent ? ātmanepadam
There are good reasons to think that this verse is an interpolation favoured by concatenation with the preceding one, the existence of an identical pattern in the first part of both verses, and the feeling that parasmaipadam could not go without ātmanepadam 28. But, whatever the textual history, it seems clear that the author of the AC had before him a manuscript of the VMM containing the two stanzas and that this situation produced confusion and overlapping. This is evidenced by the manuscript tradition of the AC itself : pada c has a variant reading which is nkārādyanubandhānam (instead of akarādy-anubandhānam) and the commentary too :ātmanepadam. ātmane na ślāghate munih sva-ślāghāṁ na karotity arthah. In short, the only way to make Ajitasena's verse agree both with common sense and grammar is to suppose that the answer is to be read as åtmanepadam :
Who is it that the ascetic does not praise ?-Himself atmane
... Tell what is the name to be given to the roots having the anubandha ? ātmanepadam
(IV)
10. Like the sabdiya-jāti (§ 9), the sautra-jāti is illustrated but not defined. It is probably not a result of mere chance that all examples of this variety (whether in the VMM or elsewhere) exclusively refer to the field of grammar: sūtra-style and grammatical teaching are traditionally considered to go together. The structural pattern of such riddles can be described on the basis of the two instances available in the VMM: as usual in praśnottaras, the global answer is progressively formed by piling up several units corresponding to the same number of questions (here three), whereas the last question concerns the sūtra itself and gives a hint as to the treatise where it is to be found :
yamah, agam, dhane / Panini-sūtram ca kidrksam ?→yamo gandhane, Pẫn. 1.2.15(VMM 2.59)
višesanam, ekā, arthena / sūtram Candrasya kidrksam ?—visesanam ekārthena, Cāndra-vyak. 2.2.18 (VMM 2.60).

Page #23

Grammatical Riddles from Jain works
11. Riddles of this kind available in our Jain corpus refer to a variety of works. To some extent they are of interest for the history of grammatical tradition and for an appraisal of the spread of grammatical schools in Medieval India.
291
11.1. Jinavallabha's Praśnottara devotes two rather elaborate stanzas to the sautra-jāti, but the commentators' silence about the quoted sūtras and their sources is noteworthy. In the first stanza the pattern is different from the instances of the VMM because it represents a mixed type, called by the author trih-samasta-sütrottara-jāti. Instead of being reached at through the piling up of micro-elements, the sequence forming the answer is taken as a whole and can be understood in three different ways:
sva-janaḥ prcchati, "jainair aghasya kaḥ kutra kīdṛśe kathitaḥ ?" kathayata vaiyākaraṇāḥ sūtraṁ Kätyāyanīyaṁ kim? (JP 64)
One's own kinsman asks: "What have Jains said of sin regarding which thing of which type ?"--Kinsman, bondage regarding the means of activity which is a very big fight bandho, bandho 'dhikarane 'dhika-rane
Tell, grammarians, what is Katyāyana's sūtra ? bandho 'dhikarane2
If we except the vocative and the rather artificial analysis of adhikarane as adhika-rane, a compound of rare occurrence (even in the Poona Sanskrit dictionary), the first part states fundamentals of the Jain doctrine with the relevant technical terminology: karmic bondage (bandha) is one of the fundamental truths (tattvas, cf. Tattvārthasūtra 1.4); as for the term adhikaraṇa, for which the standard passage is Tattvärthasutra 6.8 : adhikaraṇam jīvājīvāḥ (developed in the subsequent sūtras), its first meaning 'basis or means (of any activity)" applies here as well. However, the fact that all activity is viewed as generating particles of karman in the individual soul has led to a restriction of meaning; hence the explanation of adhikaraṇa as "a means, implement or weapon of karmic bondage" and the translation of TS 6.8 as "The instruments of long-term karmic inflow are both sentient and non-sentient entities"." This traditional semantic equivalence accounts for the gloss adhikaraṇe pāpa-vyāpāre of the commentaries on JP 22 (quoted in note 29).
On bandho 'dhikarane as a grammatical sutra, all the commentaries are silent. The identification of the Katyāyana mentioned by Jinavallabha

Page #24

292
Nalini Balbir
Jambū-jyoti
with Panini's vārttikakära can be easily ruled out because the corresponding rule in Panini's grammar is not a varttika and has a word-order different from what we have here : adhikarane bandhah (3.4.41), "(The affix 'namull comes) after the verb 'bandh' (to bind) when (a word expressing) location (is in construction with it).” In the context of riddles and grammar, where precision and strictness are fundamental, such a discrepancy is of importance. It is very likely, however, that our Katyāyana is the grammarian also known as Vararuci "to whom some works on Prakrit and Kātantra grammar are ascribed32." As a matter of fact, bandho'dhikarane is a sūtra that occurs in the fourth chapter of the Kātantra which is devoted to the discussion of krt-affixes (4.6.25)33, a problematic chapter generally
n to the original sūtrapātha and explicitly taken by the commentator Durgasimha to be the work of Kātyāyana :
vrksādivad ami rudhah krtinā na krtah krtah,
Kātyāyanena te srstā vibuddhi-pratibuddhaye34. Jinavallabha is in agreement with this traditional opinion. His riddle is one of the hints indicating the vast circulation of the Kātantra in Medieval India, also among Jain circles in Gujarat (see further 12).
11.2. In his second instance of the sautra-jäti Jinavallabha resorts to the pattern expected for this kind of riddle, but makes it sophisticated, hence savoury, through the use of uncommon words to be found in the answers. As a matter of fact, his purpose is not to write a handbook, as Dharmadāsa did with the VMM, but to display his virtuosity :
jantuh kascana vakti, "kā kva ramate?" procuh kacán kidrsan ? brahmadi-trayam atra kah krśayati ? kved-agamah syäj janen ? kim vānukta-samuccaye padam ? atho dhātuś ca ko bhartsane ? kim sūtram sudhiyo 'dhyagisata tathā Visrāntavidyadharāh ? (JP 122) sütrottara-jätih : jhasyekācovaśaḥsdhvoścabhas. A certain being says: "Which lady finds pleasure where ?" -Fish, Lakşmi (finds pleasure) in Visnu jhasi, 1 E > jhasye How do they describe the hair ? As "worshipping the head" kācah
Who is it that makes thin the three (Brahma, etc.)?—The one who makes thin Brahma, Śiva and Visnu va-śaḥ
Before what would there be the accrement it in the root 'to be born' ? -Before s and dhy s-dhvoh

Page #25

Grammatical Riddles from Jain works
293
And which is the word used in the meaning of accumulation though not actually stated ?- "And" ca
And which is the root that is used in the meaning “to insult” ?–The root bhas" bhas
And which is the sūtra that the clever Viśrāntavidyadhara taught ?-- The sūtra jhasy ekāco basah sdhvoś ca bhas
The lexicographical niceties of this stanza require some explanation : jhasi is the vocative of the feminine word jhasi; i for Lakşmi and A (here in the locative E) for Visnu are some of the ekáksara-words Jinavallabha likes best; kācah is the accusative plural of an artificial tatpuruşa compound of etymological value made of ka ("head") + añ)c (root noun meaning "which worships", cf. añc pūjane, Dhātup.) > kāc; va-śah is the nominative masculine singular of a tatpuruṣa compound made of U-U-A ( > va through sandhi) + sah (cf. so tanūkarane, Dhātup.), where U is an ekakşara-designation for Brahma and then for Śiva. One could also consider that the designation for Śiva is Ū (also found in Ekāksarakosas with this meaning).
Further, the stanza is also a small collection of individual grammatical questions of the types we saw earlier scattered in different verses :
—there is a question on a synonymic root : bhas bhartsane, see for instance Hemacandra, Dhātup. 1.521; Palsule 1955, 187.
— The statement defining ca is well-known from the Kāśikā : ca-kāro 'nukta-samuccayārthah (Kaś, on 2.4.18, etc.: Abhyankar 1961 p. 21; F.Kielhorn, "Notes on the Mahābhāsya : 7. Some devices of Indian grammarians”, Ind. Antiquary 1887, p. 251 = Kleine Schriften, ed. W. Rau. Wiesbaden, 1969, p. 240).
Finally, one is asked a question concerning grammatical metalanguage and requiring familiarity with Pānini's sūtra 7.2.78 : (iţ, implied by anuvrtti from 7.2.66) ida-janor dhve ca, where ca implies the recurrence of se from the sütra just coming before :
"(i is added before se (including the substitute sva]) as well as before dhve [including the substitute dhvam) in roots to praise' and 'to be born 35."
Like in other instances of the sautra-jāti, the question about the sūtra is the last of the riddle. The source is indicated, but is hardly palpable because

Page #26

294
Nalini Balbir
Jambū-jyoti
the corresponding text has not been handed down to us. The name Visräntavidyadhara is indeed known from references or quotations found in the indigenous tradition, but there have been some discussions on whether to understand it as referring to a work or to an author36. The syntax of our stanza makes it clear that Jinavallabha thinks about a teacher. But it is equally clear that in the following verse of the section of the Prabhāvakacarita devoted to Mallavädin
sabda-śāstre ca Viśrantavidyadhara-varābhidhe nyāsam cakre 'padhi-vrnda-bodhanāya sphutärthakam
(chap. 10, vs.38) it refers to a work (authored by a certain Vāmana, as other references show), and to the commentary written on it by the famous author of the Dvādasaranayacakra. Anyway this question is probably not so relevant, and it could well be that the rather expressive name Viśrantavidyadhara (= V.) designates both the work and its author, as when one says "the MonierWilliams". If the statement that Mallavādin (between sam. 400 and sam. 600)* commented it is true, it means that (the) V. is quite early.
The original source of the sūtra may be problematic, but the sutra itself
jhașy ekāco basah sdhvoś ca bhas
is well-known in wordings which are very close to each other in the following grammars :
* ekāco baśo bhas jħas-antasya sdhvoh (Pān. 8.2.37, with ante ca through anuvrtti from 8.2.29)
* jhasa ekācah sdhvor baśo bhas (Cāndra-vyāk. 6.3.69) * ekāco baso bhas jhasah sdhvoh (Jainendra-vyāk. 5.3.54) * baso bhas jhaṣaḥ sdhvoś caikācah pratyaye (Sākatāyana 1.2.76).
This shows that V. belongs to the Pāṇinian tradition and uses the same vocabulary and techniques as the great teacher of Sanskrit grammar. The use of ca both in V. and sākatāyana may mean that the sequence of
* Since he mentions Dinnāga (c. A. D. 480-560 or a few decades earlier), and also quotes from the Avasyaka-niryukti, he is best dated to the latter half of the sixth century A. D. - Editors.

Page #27

Grammatical Riddles from Jain works
295
sūtras was similar in both works and implies that the mention padānte, which has to be applied through recurrence from 1.2.67 in Śākatāyana, is also valid in V. The wording of V. can thus be rendered : "Before the voiced aspirate consonant (has = jha, bha, gha, dha and dha) of a monosyllabic root, instead of the non aspirate voiced consonant (baś = ba, ga, da and da), there will be the corresponding voiced aspirate consonant (bhas = bha, gha, dha and dha), also in front of s and dhy (i.e. at the end of a word as well as in front ofs and dhv)". The process described refers to forms such as bhotsyate or abhuddhvam from root budh, or to cases like godhuk37. It is of course purposefully that Jinavallabha selected for a riddle a sūtra of such a high technical level where all the words belong to the grammatical metalanguage, being either pratyāhāras (bas, bhas, jhas)—and not very common ones—, pratyāhāra-based terms (ekāc) or endings (s-dhvon).
12. In our body of narrative works, the only instance of sautra-jāti is found in Amradevasuri's Akhyanakamanikosavrtti (dated sam. 1190 = 1134 A.D.). A group of young men and a lady exchange riddles. All of them are rather complicated or even obscure. When the type illustrated by the stanza is indicated, it is a useful hint, as in the present case :
jai evam suttaṁ ciya maha panhottaram imam sunaha In this case, listen to my riddle, which is of the sutra-type.
After this introductory Prakrit sentence, comes the riddle, in Sanskrit, like all those which are included in this text (except for one which is told by the lady of the group).
ká saukhyaikanibandhanam tri-bhuvane ? kesāṁ mahad gauravam ? nīrug vakti, "janasya tättvika-ripuh kah ? kam ca babhrur dvişan ?" sangho vakti, "Sudarśanam vada kare kesām ?" svarāgryo 'bhidhāt "ko varno na-paro vitundati ca?38kim sūtraṁ purasyānañah ?
(Akhyānakamaņikośavrtti chap. 36, vss. 243-244)
As it often happens, the puzzled reaction of the audience is indicated, and this serves to enhance the intelligence of the hero who alone is able to find the answer. This becomes possible for him only when he has again taken into consideration the variety of the riddle which has been proposed to him :

Page #28

296
Nalini Balbir
Jambū-jyoti
navaram vinnāyam imam mai-sāyara-santiyam na kenāvi panhottara-gaya-suttam muniyam kumarena tam tu imam (vs. 245)
The solution then comes as ahimsarthānām ajvaren, a sequence which is not part of the verse and which is arrived at as follows:
What is the only basis for happiness in the three worlds ?
-Non violence ahimsa What are the things one is very proud of ?-Richesarthanam
A man keeping good health says : "Who is the fundamental enemy of a human being ?- you who have no fever, it is love ajvara, Ih
Who is the one that the mungoose hates ?—The snake ahim
A group of people says: "Tell, in the hands of which beings is Sudarśana carried) ?"-O group, (in) Vişnu's hands sārtha, Anām → sārthānām
The best vowel said: "Which phoneme acts with regard to n?"
- best vowel, (it is) i aj-vara (voc.), ih39
And then, once the privative a standing in front is removed what is the sütra? himsārthānām ajvareh
The process to be applied in order to come to the solution is identical to what was seen in the VMM or in Jinavallabha's riddles. Like in the preceding case (8 11.2), the sūtra represents a sort of climax of the riddle but is not the only grammatical item to discover. Here, the question just before the last one is rather technical and refers to Pān 7.1.58 : id-ito num dhātoh, "n is inserted after the last vowel of a root ending with mute i (in the Dhātupātha)".
As to the final answer which forms a sütra, the difference is that here no hint is given about the grammatical text where it can be traced. Yet the source can be identified as Katantra 2.4.4040, a sūtra concerning the use of the genitive for referring to the object : sasthi is implied and forms a recurrent term to be taken from 2.4.37, karmani comes from 2.4.38: "(The genitive is used for the object) in roots meaning to 'inflict pain', (but) not for the root jvar in the causative.” Examples given in the commentaries*l show that the genitive can alternate with the accusative and is not compulsory (anityam, Durgasimha) except for jvar. Verbs chosen are (caurasya/cauram) prahanti, nihanti, pranihanti, utkrāthayati, pinasti, rujati, āmayati, ujjāsayati, unnātayati. This shows that Kätantra's sūtra is in fact a blend of two sutras which are kept independent by Pānini : rujārthanam

Page #29

Grammatical Riddles from Jain works
bhāva-vacanānāṁ ajvareḥ (2.3.54), "(The genitive is used for the object) for roots meaning 'to pain' when they express a condition (i.e., when the subject is an abstract noun), except for the causative verb jvaraya" and jāsiniprahaṇa-nāṭa-kratha-piṣāṁ hiṁsāyām (2.3.56), "The genitive is used for the object) for the roots 'to strike' in the causative, 'to kill' (when prefixed by) ni and/or pra, for nat in the causative, krath in the causative and pis, when meaning violence42."
A reference to the Katantra in a Jain work should not be too surprising : manuscripts of Durgasimha's vṛtti are housed in Jain libraries, Jain commentators, whether Śvetambara or Digambara, also applied their scholarship to the explanation of this treatise, whereas, on the other hand, various literateurs took it as a basis for sophisticated poetic exercises; last but not least the Katantra is also known to have had some impact on Hemacandra's grammar.
297
13. The last instance of a sautra-jāti riddle is provided by Ajitasena's Alamkāracintamani :
uktasya nuḥ paramṛṣṭau kaḥ śabdo ? bheda-vāci kim
avyayam? kena nătoși ? sūtram kim Prakriya-sthitam ? (AC 2.60) sahārthena. sautra-jātiḥ.
What is the word used in referring to a masculine already mentioned ? -"He" Sa(h)
What is the indeclinable expressing differentiation ?-Ha ha43 What is it that does not bring satisfaction ?-Money arthena What is the sutra found in the Prakriya ?-"With (words) having the meaning 'along with'" sahārthena
On the basis of the general pattern of the sautra-jāti riddles, where the last question always refer to a grammatical work, it can safely be assumed that Prakriya designates some treatise of this type. Since no work having only this title seems to be known, Prakriya is probably an abbreviation. There are then two possibilities:
(i) Prakriya refers to the Prakriyasaṁgraha, a recast of Śākaṭāyana's Śabdānusasana by Abhayacandrasūri, "a commentary of the Kaumudi - type", or even, by extension, to śākaṭāyana's work in Abhayacandra's recension, which is possible since a sütra of the wording sahārthena

Page #30

298
Nalini Balbir
Jambu-jyoti
prescribing the use of the instrumental case (trtīyā) “(with words) having the meaning "with" is taught there :
sahärthena (1.3.129 = sūtra 20 in the vibhakti-section of Abhayacandrasūri's commentary)
c. sahārthena yuktát trtīyā bhavati. putrena sahāgatah putrena saha sthulah, putrena sardham, putrena sākam, putrena samam, putrena satrā, putrenāmā bhunkte45
(ii) Prakriya designates one of the commentaries on the Jainendravyākarana, for instance Śrutakīrti's Pañcavastuprakriyā or the Sabdarnavaprakriyā based on Somadeva's Sabdārnavacandrikā. This is not impossible because the sūtra sahārthena is also attested in the Jainendra 1.4.30.
The solution could come from another verse of the AC, identified by the editor of the text as "a mangala-verse of Sākatāyana's Prakriyā", but this verse is unfortunately not traceable in any of the sources consulted by me46
As for the sūtra itself, a comparison of śāk, with the corresponding aphorism in Panini saha-yukte 'pradhane (2.3.19) confirms F. Kielhorn's conclusions regarding the relation of the two works : Śākațāyana's "pathological concern for economizing the number of syllables", his way of incorporating material from the commentaries (Katyāyana or Patañjali), but, at the same time, his manner of simplifying the matter by removing a specification which he probably considered as having minimal importance (apradhane). Similarly, a comparison with the corresponding sūtras in the Candravyākarana and in Hemacandra's work shows Śåkatāyana's indebtedness to Candragomin (where sahārthena is sūtra 2.1.65) and Hemacandra's indebtedness to sākatāyana (Hemac. 2.2.45 has sahārthe)47
(V)
14. Although less frequently, metrics is like grammar a field of knowledge referred to in riddles. In the following instance
... rucirā kā satäm vrtta-jätih ? (JP 69)
Which is the kind of metre that connoisseurs find beautiful ?, the selection of the right answer among many possibilities, namely mālinī, is determined by the next question.

Page #31

Grammatical Riddles from Jain works
299
ko vā diksu prasarati sada kantha-kandāt Purāreh ? (JP 69)
And what is it that always spreads in all directions from the place of Siva's throat ?
where the answer should be nilimā, i.e. the same word read in a reverse order so that it fits with the type of riddle here illustrated, namely the manthānantara-jāti48.
In the question
amantasu anta-gurur ... (sasankam; Dhaneśvara, Surasundaricariya 16.53)
Call the one that ends with a long (syllable),
the point is to test basic knowledge of the technical terminology relating to metrical science. The answer is sa, vocative of the technical term sa referring to a gana of the form vu–, and immediately comes to the mind of those who remember the relevant samjñasutras or kārikās beginning any chandahśāstra49
15. The variety corresponding to the sābdīyajāti of grammar (above S 9) is termed vrttanamajāti. There is no illustration of it in any of the Jain narrative woks considered in our investigation. But it is both defined and illustrated in the second chapter of Mahākavi Ajitasena's Alarkāracintamani. Once more, Ajitasena's definition appears as a clear rewording of Dharmadāsa's Vidagdhamukhamandana. Compare
vrtta-nama bhavet praśna-vrttanāmottarād hi yat (AC 2.52cd)
It would be a yrttanaman, because it has as an answer the name of the metre of the question
and
(yatra) vrttanāmottaraṁ prstam bhavet tad vrttanāmakam (VMM 2.34cd)
(Where) the question has as its answer the name of the metre it would be a vrttanamaka.
However, Ajitasena's awkward formulation includes an element which is not explicitly mentioned by Dharmadāsa : not only does the final answer give the name of a metre, but the riddie-stanza should itself be composed "by the way of mudrā" in this very metre which is to be guessed so. A look at

Page #32

300
Nalini Balbir
Jambu-jyoti
the three illustrative verses available both in the VMM (2 verses) and in the AC (1 verse) shows that it is actually the case :
Metre of VMM 2.35 śikharini / Answer to the riddle : śikharini (< śikharini + 1)
Metre of VMM 2.36 mâlini / Answer to the riddle : mālini (< mă + alini) 51
Metre of AC 2.54 indramālā (see below) / Answer to the riddle : indramālā.
But Ajitasena's definition has the advantage of making this feature a structural element specific to the variety and not a matter of chance. It is both an additional challenge for the one who asks the riddle and a hint for the one who is to answer it, provided he is decently trained in recognizing the metre of a stanza read to him.
Ajitasena's example reads :
sambodhanam kim suralokanāthe ? bhramad-dvirephā surabhi-sphuta kā ?
Kā yäti nākāj Jinapujanartham ? vrttam kim ābrūhy upajāti-laksma? Which word is used to address the lord of the world of gods ? Indra ! What is it that is famous for its fragrance, where bees turn around ?
A garland of flowers-mālā
Who comes from the sky to worship the Jinas?.--A line of Indras Indramålā
What is the metre that has the characteristic of an upajāti? indramālā
The pattern of the riddle-verse agrees with what we find in the VMM : in the vrttanāmajāti, the first two or three questions can relate to any topic, but the last one always gives an indication about the structural pattern of the metre to be guessed (... a metre having nine light and eight heavy syllables for the śikharini in VMM 2.35, ... a metre having the same number of light syllables as the number of mountains [ = 8 ) and the same number of heavy syllables as the number of oceans [ = 7 ] for the mâlini in VMM 2.36). Here also the metrical pattern of the stanza fits with the hint given about the metre : three indravajrapadas (a, cand d) and one upendravajrāpada (b) makes an upajāti, since this syllabic metre of 11 syllables is precisely made of the combinations of these two varieties, in any

Page #33

Grammatical Riddles from Jain works
possible type of mixture. As for the word indramālā, it is probably not by chance that Ajitasena makes use of it. This synonymic designation of upajäti is not attested in all metrical treatises, but precisely in three works which come from South India, as Mahākavi Ajitasena himself, and which, for two of them, have Jain authors$2.
(i) The earliest reference is found in the Ratnamañjūṣā, an anonymous Jain work on Sanskrit metrics which is one of the oldest existing Indian chandaḥsāstras 53:
tristubh (5.24)
indravajra śare (5.25)
upendravajra șare (5.26)
indramälä dvayam (5.27); ct. yadīndravajrāupendravajre sahaikasmin sloke bhavataḥ, bhavati indramālā nāma54.
(ii) The wording of Janāśrayi's Chandoviciti (end of 6th cent. A.D.) recalls the Ratnamañjūṣā, although the technical designations of the ganas used by this author are peculiar to him:
Variety No. 1
Variety No. 2
Variety No. 3
Variety No. 4
Variety No. 5

indravajra bejr (4.34)
upendravajra kejr (4.35)
ubhaya-miśrendramālā (4.36)55.
The author goes a step further, stating that there are fourteen different varieties of indramālā (sā caturdaśa-bhedā, 4.37), as other metricians also do56. But he is one of the rare who provides illustrative stanzas for these varieties, namely twelve of them (the two remaining ones, the akhyānikā and the viparītākhyānikā, which he has already treated separately in 3.7 and 3.8, are not repeated) :
301
Indravajrapada(s)
ab
cd
ad
bc
a

Upendravajrapada(s)
cd
ab
bc
ad
bcd

Page #34

302
Nalini Balbir
Jambū-jyoti
b
acd
C
abd
Variety No. 6 Variety No. 7 Variety No. 8 Variety No. 9
abc
bcd
acd
Variety No. 10 (corresponds to the pattern of the above verse of AC)
abd
Variety No. 11 (= Jánaktharana 1.37)
abc
Variety No. 12 + (Akhyānikā
ac
bd
Viparītākhyānikā
bd
ac)
(iii) Finally, Jayakirti, a Digambara Jain from Karnataka who lived about 1000 A.D., gives the following definition in his versified Chando'nuśasana edited on the basis of a unique palm-leaf manuscript from Jaisalmer :
upendravajrā-sphurad indravajrāpādau vimiśrau yadi tau bhavetām nänā-vikalpair upajātir esă prakalpitāryaih kvacid indramālā (2.117)57.
Because of common sectarian affiliation, this could well have been the book which was the source used by Mahākavi Ajitasena.
The above investigation should have contributed to make clear that Jain authors are extremely fond of all types of games connected with the manipulation of language and that their proficiency in the fields of knowledge pertaining to language (grammar, lexicography and metrics) is of a very high level. For them, as for other Indians, erudition is not boring, on the contrary it serves the games of intelligence and as such is valuable.

Page #35

Grammatical Riddles from Jain works
303
Annotations : 1. For a convenient list of such dictionaries see C. Vogel, Indian Lexicography.
Wiesbaden, 1979, p. 369-371. 2. This has rightly been underlined by L. Sternbach, Indian Riddles. A forgotten chapter
in the history of Sanskrit Literature. Hoshiarpur, 1975, $ 132ff. 3. Śri Siddhahemacandra Sabdanuśāsana. With the Auto-commentary Laghuvrtti.
Patan, 1994. 4. Representatives of the first category are Uddyotanasūri's Kuvalayamālā, Maheśvarasūri's Nanapancamikahā, śāntisūri's Puhaicandacariya, Dhaneśvarasūri's Surasundaricariya or an anonymous Jinadatta-ākhyāna. Second category:
Dharmopadeśamälāvivarana Sk.: 3 Pk.: 2 Manipaticarita Sk.: 2 Pk.: 3 Akhyānakamanikośavrtti Sk.: 8 Pk.: 1
Kahārayanakosa Sk.: 4 Pk.: 2 Third category : e.g. Samarāiccakahā with 4 riddies, all in Sanskrit. This is just a sample of references. More details about the corpus will be given in an extensive
study on riddles found in Jain narrative works (in preparation). 5. Partly for lack of space, riddles belonging to the category where a verbal form
(kriya) or a nominal form corresponding to any of the grammatical cases (kartro, karma", karana', etc.) is concealed (gupta) in a given verse and is to be detected, are left out. See VMM 4.33ff., or Kuvalayamālā and Samarāiccakahā for examples
in Jain works. 6. Mentioned in L. Sternbach's monograph Indian Riddles $ 120, but with more
information about the author's career than about his work itself. 7. Description and extracts from the praśastis in Catalogue of Sanskrit and Prakrit
Manuscripts. Muniraja Shri Punyavijayaji's Collection. Ahmedabad, 1968,
p. 432, Serial Nos. 3384 and 3385. 8. See New Catalogus Catalogorum under Praśnávali and the list of Municandra's
works in e.g., H. R. Kapadia, introduction to Haribhadra's Anekantajayapatākā
(commented upon by Municandra). Baroda, 1940 (G.O.S. 88), p. XXX. 9. Available evidence is carefully examined by M. Kraatz, VMM 1968, vol. I,
p.XI-XIV, and his conclusion is very cautious (perhaps excessively so). 10. See for instance H. D. Velankar, Jinaratnakosa. Poona, 1944, p. 355; Pandit
Ambālāl Pre. Shah, Jain Sähitya kā Brhad Itihās, vol.5: Läksanik Sahitya. Varanasi, 1969, p. 127-129.

Page #36

304
Nalini Balbir
Jambu-jyoti
11. See also below § 13. Other instances of riddles providing answers of Jain or
Digambara Jain colouring are : Tirthakarah (AC 2.11), Akalankah (2.13),
Vidyanandah (2.109), paramesthi (2.110). 12. It is not mentioned in Sternbach's monograph. 13. In this section and the next, only the parts of the verses which contain grammatical
riddles will be quoted and considered. The verbal sequence forming the answer to the full riddle will be given because it serves as a hint to direct the reader in cases where there could be several possibilities, but it will not be extensively explained : the more questions, the more problems, a detailed discussion of which would lead
to unnecessary digressions... 14. For a modern reader, the “alphabetical Index of meanings with corresponding
roots (also arranged alphabetically) as given in the different Dhātupathas" found
in Palsule 1955 is a perfect tool. 15. This riddle and another one (JP 33) are also found in Devabhadrasūri's
Kahārayanakosa (p. 280a). 16. Ct.: ada psä bhakşane attiti at kvip adam lāti dadätiti allas tat-sambodham he 'lla
(mss A1, A2). 17. Antasthāsv antya-varnaś ca kah prasiddho mahitale ? 18. Ct.: a-kārāt paro 'c i-kārah aparāc tasyāyah ksayah (Ed.). 19. Ct.: atrottaram : aś ca jaś ca maś ca das ca has ca taś ca bhāś ca ajāmadahatabhas
te púrve yasya mah ajāmadahatabhā-purvah tathā me muja-sabda-sambundhi ma-kāre krte ne na-kare krte sati etāvată muja-sthane nuja iti jāte adau ajadisu varneșu datteșu yathākramam anuja 1 jānuja 2 manuja 3 danuja 4 hanuja 5 tanuja 6 bhānuja 7 iti śabdah bhavanti. Hanujah antyadamstra bhanujah sanih,
śesā spastā eva (mss A1, A2). 20. Ct.: tvaṁ samudalasah samullasitavān athavā sam sāmastyena ut prābalyena alasah
sabditavān tus hras las sabde las ity asya rupam; but las with this meaning is not
recorded in Palsule s.v. śabde. 21. This is perhaps what is actually meant by the reading kidrvikam found in two mss.
of the AC. 22. Translations of Pāņ. are based on S. C. Vasu, The Ashtādhyāyī of Pāṇini, 1891,
Reprint Motilal Banarsidass, 1962 and O. Böhtlingk, Panini's Grammatik. Leipzig,
1887. 23. Ct.: krvā-pratyayar praty uttara-dātā vakti : bhavatas tava krvā-pratyayasya sthane
yo 'sau yap ity adeśo bhavad-yab-adeśas tam apeksya vimāyeti bhavati. ayam atra

Page #37

Grammatical Riddles from Jain works
305
bhāvah: yadrśam nirdambha-śabdenarthato rūpam abhidhiyate tādrśam vipürvăn mināteh krvā pratyayasya yavādese sati, tatha hi nirdambha-sabdena nirmāya ity
ucyate anenāpi vimāyeti sa evārthah. 24. The difference between ba andva is never relevant in the context of riddles. In this
answer one should read ba, while in the two other answers of the same riddlestanza one should understand va (respectively bhavad yavā(h) deśam and Bhava,
dya vāde 'sam). 25. Abhyankar 1961, p. 63-64. 26. See further Abhyankar 1961 p. 304 and Kraatz's commentary on VMM 2.61 (vol.
2 p. 191). 27. For instance, in the Berlin manuscript "Ms. or. fol. 1034" (Weber, Verzeichnis II,
I p. 285 No. 1727) it is added in the margin by another hand. 28. See M. Kraatz's observations in his commentary of VMM (vol. 2 p. 192). 29. Ct.: he bandho svajana adhikarane pāpa-vyāpāre bandhah kathitaḥ, kidsse
adhikarane 'dhikaranam yatra tat tasminn adhikarane... (mss A1, A2); afterpapavyāpāre the avacūri of the printed edition reads: kimviśiste ? adhikarane adhikar
ranam samgrāmam yatra tat tathā tasminn adhika-rane. 30. See Poona Dictionary p. 1551, meaning 6. 31. See respectively Pt. Sukhlalji's Commentary on the Tattvārtha Sutra, Ahmedabad,
1974 (L. D. Series 44), p. 239 and N. Tatia's translation of TS, That Which Is.
London, New York: Harper Collins, 1994, p. 154. 32. Abhyankar 1961, p. 109. 33. = Pujyapāda Devanandin's Jainendra-vyākarana 2.4.28 (Ed. with the Jainendra
mahāvștti of Āc. Abhayanandi. Ed. Pandit Shambhu Nath Tripathi. Kashi, Bharatiya Jnanapith, 1956 (Jnanapitha Moorti Devi Jain Grantha Mala, Sanskrit Grantha
No. 17]). 34. Kātantra, ed. J. Eggeling, Calcutta 1874 (Bibliotheca indica 297-298), p. 299.
Quoted by S. K. Belvalkar, An account of the different existing systems of Sanskrit grammar. Poona, 1915, p. 27 n. 2; see also p. 84-85. - Kätantra's recension commented upon by the Digambara Bhāvasena Traividya (see below n. 41) does not include the sūtra under consideration. It ends with guros ca nisthāyām setah =
4.5.81 in the recension commented upon by Durgasimha and edited by J. Eggeling. 35. On the part played by ca in "offering an abbreviative interpretation" in this sequence
of sūtras see S. D. Joshi and Saroja Bhate, “The Role of the Particle ca in

Page #38

306
Nalini Balbir
Jambū-jyoti
the interpretation of the Astādhyāyi': Proceedings of the International Seminar on Pānini. University of Poona. Pune, 1983 (CASSE, 9), p. 180-181. - The corresponding sūtra in the Jainendra-vyākarana does not apply to the root jan: idah s-dhve
(5.1.136). 36. Some information on V. can be found in the following sources : Abhyankar 1961,
s.v. Višrāntavidyādhara, with the quotation of a popular verse (p.336); F. Kielhorn, “Indragomin and other grammarians", Indian Antiquary 15 (1886), p. 182 and n. 4(= Kleine Schriften, p. 243): quotations from Hemacandra's Nyasa on his own Sanskrit grammar, and references to the Ganaratnamahodadhi; H. R. Kapadia, Jaina Samskrta Sahitya no Itihasa Vol. 1. Baroda, 1956, p. 23-24 of the second
chapter: Convenient collection of whatever information is available. 37. TheAmoghavrtti on sākatāyana (ed. by Pt. Shambhu Nath Tripathi Kashi, Bharatiya
Jnanpitha, 1971 (Inanapitha Moorti Devi Jain Grantha Mala, Sanskrit Grantha No. 37) quotes both verbal and nominal forms, whereas Abhayacandrasūri's commentary
(ed. by G. Oppert. Madras, 1893) takes instances from the stemgo-duh-only. 38. Correction suggested by Prof. P. S. Filliozat (see also below). Ed. vitundati na is
probably wrong. The hints provided by the Indian editor for this part of the riddle
are not helpful at all. It seems he missed the point. 39. I am extremely thankful to Prof. P. S. Filliozat, without whom I was at a complete
loss to understand this question and its answer. His help has also proved invaluable
for solving a few other points connected with this paper. 40. See n. 2 p. 297 in the edition of the Akhyanakamanikośavrtti. 41. Commentary (vrtti) by Durgasimha in ed. J. Eggeling. Calcutta, 1874 (Bibliotheca
Indica 297-298), translated in B. Liebich, Zur Einführung in die einheimische Sprachwissenschaft. I. Das Kätantra. Heidelberg, 1919, p. 48; commentary (called Rupamālāvrtti) by the Digambara writer Bhāvasena Traividya, ed. with Hindi translation by Ganinī Āryika Śrī Jñanamatī mātāji. Hastinapur, 2nd ed. 1992. In
this edition, the sūtras have a continuous numbering. Our sūtra is No. 411. 42. Similar wording in the Jainendra-vyāk.: rujarthasya bhāva-vācino 'jvari-santäpyoh
(1.4.61) and jāsa-niprahana-nata-krātha-pisan himsāyam (1.4.63); in säkatāyana : rujo 'jvarisantāpyoh kartari bhāve (1.3.113) and himsāyām jāsanātakrathapisanipradhnam (1.3.114); in Hemacandra : rujārthasyajvarisantăper bhāve karttari (2.2.13), jāsa-nata-krātha-piso hiṁsāyām (2.2.14) and ni-prebhyo ghnah (2.2.15). There does not seern to be anything comparable to these sūtras in the Candra-vyäk.

Page #39

Grammatical Riddles from Jain works
307
43. Cf. ha vinirgrahe according to Medini, quoted in the Amaravyakhyāsudhā
(information provided by Prof. P. S. Filliozat). 44. R. Birwé, introduction to sākatāyana's grammar (e.d. by Pt. Shambhu Nath
Tripathi. Kashi, 1971), p. 3. 45. This is Abhayacandrasūri's commentary, ed. by G. Oppert. Madras, 1893, p. 75;
the Amoghavrtti (ed. by Shambhu Nath Tripathi. Kashi, 1971) is more extensive. 46. The verse in point, clearly of Jain origin, is :
jayati jagadisa-mastaka-mani-kirana-kalapa-kalpitărgha-nidhi Jina-carana-kamala-yugalam ganadhara-gananiya-nakha-keśarakam (AC 5.301), identified by the editor as "sākatāyana Prakriya kā mangala padya" on p. 365. What the editor exactly means is not clear to me, since neither Abhayacandrasuri
nor the Amoghavrtti include this verse as their mangalas. 47. See F. Kielhorn, “On the grammar of sākatāyana", Indian Antiquary 16 (1887),
24-28 and "Die śākatayana-Grammatik", Göttinger Nachrichten 1894, 1-14 = Kleine Schriften, p. 246-250 and 276-289 (“jenes fast krankhafte Streben nach
möglichster Kürze", p. 285). 48. This variety of riddle is akin tocitrakāvya and can be best represented as a cross,
the centre of which is occupied by a syllable common to all the answers of the riddle. There is then one horizontal answer which can be read in both directions, a vertical answer which can also be read in both directions. A circular reading ending in the centre makes the last answer. E.g., in the present case :
na
ma
li
ni
na
1) malini/ nilima; 2) nalina / Nalina; 3) mánanīnam āli. There are several such instances in JP (32, 74, 143, 145, 152, 154), but this variety does not seem to have
been very popular outside this work, and the VMM does not seem to know of it. 49. See e.g., Kedārabhatta's Vrttaratnākara 1.8, Hemacandra, Chando'nušāsana 1.2. 50. See M. Kraatz's translation of VMM : "(Wenn) ein Rätselvers als Antwort den
Namen (seines) Versmaßes hat - das ist vrttanamaka benannt nach dem Versma". 51. Let us note, by the way, that this riddle is quored with its source in the Pali
grammar of the Burmese monk Aggavamsa from the 12th century : tathā hi Vidaddhamukhamandanatikāyām mālini ti padass' attham vadatā 'ma vuccati Lakkhi, alini ti bhamari' ti vuttam (Saddaniti. Ed. H. Smith. Lund, 1928, p. 244, lines 19-21).

Page #40

308
Nalini Balbir
Jambu-jyoti
52. The matter is treated here in some details because all the relevant books are not
easily available, and these technicalities are not very well known. See further note
54.
53. On its importance in the development of metrical science see C. B. Tripathi,
“Ratnamañjūsā and 'Chandoviciti”: Beiträge zur Indienforschung (E. Waldschmidt
zum 80. Geburtstag gewidmet), ed. H. Härtel. Berlin, 1977, p. 549-560. 54. Ratnamañjūsā with Bhāsya by some unknown Jain authors on Sanskrit prosody.
Ed. with A Critical Introduction and Notes by Prof. H. D. Velankar. Kashi, Bharatiya Jnanapith, 1949 (Jnanapitha Moorti Devi Jain Granthamala, Sanskrit Grantha No. 5) : "Indramālā : Another name of the Upajāti: Mandāramarandacampü (Kāvyamālā ed. 1895, p. 8 line 7) calls it upendramālā; but every other writer that I know of calls it upajăti". (Velankar's note p. 59, to be revised in the light of
the material collected here). 55. Jánaśrayi Chandoviciti. Published by the Curator, The University Manuscripts
Library. Trivandrum, 1949 (Trivandrum Sanskrit Series No. 163). 56. See e.g., Prakrtapaingala 2.119 and 2.121 for the fourteen designations of
the varieties of upajätis : one of them is called mālā, a word which is anyway quite common in names of metres (cf. vidyunmālā, ūrmimālā, candramālā, etc.).; Hemacandra's Chando'nuśāsana 2.156 : etayoh parayoś ca saṁkara upajātis
caturdaśadha. 57. Edited in Jayadāman. A collection of ancient texts on Sanskrit Prosody and A
Classified List of Sanskrit Metres with an Alphabetical Index. Ed. by H. D. Velankar. Bombay, 1949.

Page #41
Grammatical Riddles from Jain works 309 ABBREVIATIONS Abhyankar 1961 = K. V. Abhyankar, A Dictionary of Sanskrit Grammar. Baroda, 1961 (G. O. S. 134). AC = Alamkaracintamani of Mahakavi Ajitasena. Ed. by Dr. Nemi Chandra Shastri. Delhi : Bharatiya Jnanapitha, 1973 (Jnanapitha Moorti Devi Grantha Mala : Sanskrit Grantha No. 43); see $ 1.3. JP = Jinavallabha's Prasnottara-sasti-ekasata (see $ 1.1). P = Municandra's Prasnavali (see $ 1.2). Palsule 1955 = G. K. Palsute, A Concordance of Sanskrit Dhatupathas (with Index of meanings). Poona, 1955 (Deccan College Dissertation Series 14). V. = Visrantavidyadhara (see $ 11.2). VMM = Dharmadasa's Vidagdhamukhamandana. The editions used are : * Sri Dharmadasasuripranitam svopajnavyakhyasamalankrtam idam Panasikaropahva-Laksmanatmaja-Vasudevasarmana samsodhitam. Bombay, 1905. * Vidagdhamukhamandanam of Dharmadasa Suri. With 'Chandrakala Commentary and Hindi translation. Editor and Translator Acharya sesaraja Sarma. Varanasi-Delhi : Chaukhambha Orientalia, 1984. * Das Vidagdhamukhamandana des Dharmadasa (Ein Lehrbuch der Ratselkunde) 1. und 2. Kapitel. Inaugural dissertation ... Philipps-Universitat zu Marburg vorgelegt von Martin Kraatz. Marburg/Lahn, 1968, 2 vols. 000