Jain Journal 1996 04

Author(s): Jain Bhawan Publication
Publisher: Jain Bhawan Publication
Catalog link: https://jainqq.org/explore/520122/1

Page #1

ISSN 0021-4043 A QUARTERLY
ON
JAINOLOGY
VOL. XXX
No. 4 APRIL
1996
Jain JOUrnal
| | | | JAIN BHAWAN PUBLICATION

W ainelibrary.org

Page #2

Şaṭkhaṇḍāgama and Śauraseni Satya Ranjan Banerjee
A Primer of Sauraseni Richard Schmidt
Contents

101
125

Page #3

JAIN JOURNAL
Vol. XXX
No.4 April
1996
ȘAȚKHAŅDĀGAMA AND SAURASENI
SATYA RANJAN BANERJEE
1 Preamble
The discovery of the Satkhadāgama, Kasāyapaida, Mahā-bandha and some other similar works of the Digambara Jaina canonical texts sometime in late thirties and forties is a remarkable contribution to the field of Prakrit studies. These texts are important from the point of view of the Prakrit language. There has been a consensus of opinion that the two groups of Jains have two sets of canonical literatures. For the Švetāmbara the 45 Āgama texts as current among the community are in Ardhamāgadhi and the Digambara canonical texts as reflected in the Satkhanļāgama and the others mentioned above are in Sauraseni. Whether this conclusion is true or false is not our concern. The fact is that the languages of both these groups of canonical literature are not exactly the same. There are certain features which are exclusively found in one and rarely in the others. As far as the Svetāmbara canonical texts are concerned, it is the general belief that they are in Ardhamāgadhi. This does not mean that the Śvetāmbara texts are very near to Māgadhi as the name indicates, i.e. half Māgadhi and half the others, but, in general, the characteristic features of Prakrit, or say Māhārāștri, are also found there side by side with some other forms which are not regular in Māhārāstrī, but are available only in Ardhamāgadhi. But in the case of the Digambara texts again there are certain forms which are not found in Ardhamāgadhi texts, but are found exclusively in the Digambara canonical texts. However, one thing is very certain that whether it is Sauraseni or Ardhamāgadhi, or Māhārāștri or Māgādhi, there are major forms which are very common to all types of Prakrit, except a few phonological or morphological variations which are normally regarded as nothing but dialectal features. For example, the Sanskrit word krta has several forms in Prakrit and each form has a distinctive feature with regard to a particular dialect, e.g. kệta >kaa (Māh.), kaya (Amg), kida ($), kada (Mg). These variations are generally found in the respective dialects as enunciated by Prakrit grammarians. In case,

Page #4

JAIN JOURNAL: Vol-XXX, No. 4. April 1996
either in an Ardhamāgadhi or Śauraseni text kaḍais found, it should neither be regarded as a Śauraseni text, nor should it be treated as a sort of archaic Prakrit, rather it should be regarded as an editorial defect. This type of anomaly in an edited Prakrit text is profusely found, and as a result, we are at a loss to determine the language of a particular text. Let us explain this problem with particular reference to the Satkhanḍāgama.
102
2 Prakrit and its dialects
In ancient India the three languages were prominent-Sanskrit, Pali, and Prakrit. Of these three languages, Prakrit is considered as the language of the common people. Without going into controversy about the origin of Prakrit which is either from Skt. or from natural one, it can be said that it was the language which the masses used to speak for their communications. Once it is accepted as a spoken language, it is quite natural then that the same Pkt. should be spoken by different people at different places in a different way, as it is the general nature of a spoken language. Naturally Pkt. was also fused into different dialects used by different people at different places. The names of these languages were given in accordance with the name of the place. As a result Pkt. spoken in the area of Magadha is known as Māgadhi, and that spoken in the area of Surasena, is Śauraseni, and that also of Mahārāṣṭra is Māhārāṣṭrī. Besides these, some names which are given because of the type of people speaking that language are Paiśācī, Pracyā etc. However, in this way gradually the same language is perhaps divided into some dialects which bear some characteristic features which are peculiar to one dialect, but absent in others, even though some overlappings between the dialects are not uncommon. Naturally some of the features of common Pkt. are found in almost all the Pkt. dialects as recorded by the grammarians. So whatever dialects it might be some common features are bound to be found in almost all the dialects, e.g. Skt. sakala >sayala in Pkt. and the same form is found in Mah. Śau. and Mg. and also in Amg., even though sometimes sagala is also found in Amg. The latter form i.e. sagala when it is consistently used in Amg., we can consider it as one of the additional forms of Amg. and in this way some additional characteristic features are generally added to some respective dialects which are supposed to be absent in others. Otherwise, the grammarians could have written several Pkt. grammars for several Pkt. dialects. Of course, we know at a later stage one Ardhamāgadhi Vyakarana was written by someone, but this book does not mean that Amg. has something peculiar which cannot be justified by other grammarians. However, in a similar way, I do not know yet any

Page #5

BANERJEE : ȘAȚKHANDĀGAMA AND ŚAURASENÍ
103
grammar for Sau. Mg. or Paisāci, except that in modern times some have composed some grammars on each dialect of Pkt. based on the modern outlook of linguistics. When some of the dialects became very prominent either because of their literature or of the political supremacy of a particular language in a particular area, some dialects got a chance in the recorded documents of Indian history. I am at present going to refer to those texts where the Sau. is found as one of the dialects of Pkt. With this background in mind let me proceed to ransack the literature where the features of Sauraseni are mentioned.
3 Sources of Sauraseni
As far as we know the sources for the features of Sauraseni are the Prakrit grammarians, Sanskrit dramas and some other literature, and the studies of modern Pra.crit scholars on Sauraseni. It is a fact worth noting that for the characteristic features of Sauraseni, we will have to depend primarily on the Prakrit grammarians. These grammarians are Vararuci (5th/6th cent. A.D.), Hemacandra (10881172 A.D.), Puruşottama (13th cent. A.D.), Kramadiśvara (13th cent. A.D.), Trivikrama (14th cent. A.D.), Laksmidhara (15th cent. A.D.), Rāmasarmā (16th cent. A.D.) Märkandeya (17th cent. A.D.), and many others. As all these grammarians belong to different times and places, their views are to be analysed and judged from the historical point of view.
The specimens of the Sauraseni language can also be gathered from the Sanskrit dramas beginning from Ašvaghoșa (1st cent. A.D.) down to Rājasekhara (10th cent. A.D.).
The modern Prakrit scholars like Christian Lassen, Cowell, Pischel, Schmidt and many others have generally given the characteristic features of Saurasenī from the analysis of Sauraseni texts or passages. The fact that they have considered the text as a source of the Sauraseni passages is not generally questioned. But the point is how do we know that such and such are the features of Saurasenī? Here in this very context we should take the help of grammarians. The grammarians, from whatever sources it might be, have given some characteristic features of Sauraseni and other dialects or subdialects from which we definitely consider the features as genuine. Just as we have some traditions before us, so also the grammarians in those days had some authoritative texts before them. It is in this connection that we take the help of the grammarians for the features of respective dialects.
I think the Western Prakrit scholars have got the clue of this dialect

Page #6

104
JAIN JOURNAL: Vol-XXX, No. 4. April 1996
from the Sanskrit dramaturgical texts. As early as 3rd cent A.D. Bharata had mentioned the names of the languages to be spoken by such and such characters of the Sanskrit dramas. And as the ladies and the Viduṣaka of the Sanskrit dramas will speak Sauraseni, the passages of Viduṣaka and the ladies of the Sanskrit dramas are considered as Sauraseni. Naturally, the type of Prakrit found in their speeches are generally analysed as the language of Śauraseni. These passages, of course, do not go against what the Prakrit grammarians say in their respective treatises. As a result though their studies on the Sauraseni language is very elaborate, it does not practically overstate the case in point. What I generally mean by this is that though the modern scholars, except, perhaps, Pischel, have not consulted the Prakrit grammarians, but they, on the contrary, follow the Prakrit grammarians indirectly.
It is to be noted that Prakrit is the generic name for common man's language which is obviously different from Sanskrit or other languages at that time. Prakrit grammarians have generally given first the features of Prakrit. This Prakrit is more or less equivalent to Mahārāṣṭri. Except a few grammarians who have mentioned that they are going to describe the characteristic features of Māhārāṣṭrī (atha Mahārāṣṭri), almost all the grammarians have never mentioned that they are describing the features of Mahārāṣṭri, yet we do not see any difference between the features of the Mahārāṣṭrī language on the one hand and Prakrit, on the other. As a result we generally come to this conclusion that Prakrit is the common generic name of the language and Māhārāṣṭrī would be the language par excellence. When that equation is accepted then the features which are not common to Māhārāṣṭrī, are considered as dialects, such as, Šauraseni, Māgadhi, Paiśāci and others. These dialectal features are also recorded by the Prakrit grammarians under different headings, sometimes, inter alia, in describing Māhārāṣṭri or Prakrit. The name Māhārāṣṭri is not very old. In Bharata's Natyaśāstra the name Māhārāṣṭrī is not found, rather the name Dākṣinātyā was used. Probably in the 3rd A.D. or before that, Mahārāṣṭra was regarded as Dākṣinātya, modern Deccan (cf Deccan Queen) rather than Mahārāṣṭra. But the name Dākṣinātya was changed perhaps by the time of Dandin (7th c. A.D.) who in his Kavyadarsa (1.33) has mentioned that the poems written in Māhārāṣṭri is the best (mahārāṣṭrāśrayām bhāṣām prakṛṣṭam prākṛtam viduḥ). Perhaps from that time onwards Māhārāṣṭri became popular and got a place in the ancient languages of Prakrit. However, the basic point is that the grammarians have recorded the characteristic features of such and such Prakrit which should not be ignored when we consider the features of subdialects of Prakrit. Naturally our starting point of

Page #7

BANERJEE : ȘAȚKHAŅDĀGAMA AND SAURASENI
105
getting the features of Sauraseni would be the Prakrit grammarians on the one hand and the modern Western Prakrit scholars, on the other.
While considering the language of the Șatkhaņdāgama we shall see how many of the features given by the grammarians are found in the text. And as such we shall be discussing whether the texts edited by scholars are on a par with the Prakrit grammarians. Here in this connection a problem is raised which is very fundamental and at the same time important for the language of the texts. The problem is how far the Prakrit grammarians are to be accepted against the manuscript of a particular text. The problem is obviously a difficult one to settle at this stage of our knowledge, yet let me start the problem afresh.
4 Śaurasenī Literature
Before considering the language of the Șațkhanţāgama, it will not be out of place here, I suppose, to give a brief survey of the Śauraseni literature considering that the Şațkhandāgama is written in that language.
The starting point of Prakrit, as it is, at least, historically recorded, is reckoned as the time of Lord Mahāvīra who is supposed to have flourished in the 6th /7th c. B.C. It is said that Mahāvīra preached his doctrine in the then common man's language which was later on known as Ardhamāgadhi. In the Samavāyānga-sūtra it is said that Mahāvira preached his doctrine in the then common man's language which was Ardhamāgadhi for the understanding of the masses. As a result, when the doctrines of Mahāvīra were codified by the 5th cent. A.D. by Devardhigani Kşamā-Śramaņa at Valabhi, the language as represented in the 45 Āgama texts, is, therefore, considered as Ardhamāgadhi on the basis of the statement found in the Samavāyānga-sutra. After the establishment of the two schisms of the Jains, i.e. Śvetāmbara and Digambara, it was found that these 45 Āgamas were only accepted by the Svetāmbaras, while the Digambaras have doubts about the contents of these texts. To them the codification of Mahāvīra's doctrine is embalmed in the Drstivāda which is, of course, lost. Anyway when in 1939 the Şațkhandāgama was published, it was immediately reckoned that this text was the text of the Digambara canon. And subsequently some other texts of the Digambara canon were also discovered and published. It was also said by the learned editors that the language of all these Digambara canonical texts was in Sauraseni, because some of the features of Sau. are found there. It was also decided that in the 1st cent. A.D the

Page #8

106
JAIN JOURNAL: Vol-XXX, No. 4. April 1996
Digambara text was first codified and then was lost, and some of these lost texts are nothing but the present Saṭkhaṇḍāgama, Kaṣāyapāhuḍa, Mahābandha, and others. These are regarded as Digambara Jaina canonical texts and these texts are the earliest ones which are discovered, of course, very lately.
After these Digambar a canonical texts the most important writer of non-canonical text is Kundakunda whose works, such as, Samayasara, Niyamasāra, Pravacanasāra, Pañcāstikāya and so on are in Śauraseni. And the date of Kundakunda varies from the 1st to the 3rd c. A.D., some even have gone down to 5th c. A.D, but not beyond that. It is not quite certain that in the 1st c. Śau. was predominant or not, but if the Digambara canonical texts were codified in the 1st c. A.D. then it was quite a good guess to supose that in the 1st c. A.D. Śau. was current.
To the first centuries of the Christian era also belonged Vaṭṭakera and Karttikeya Svāmi. Vaṭṭakera wrote his Mulācāra and Trivamṇācāra in Prakrit, or say, in Sauraseni, and so also the Kaṭṭigeyāņupekhā of Kārttikeya Svāmi. According to Pischel (§ 21) these works are written in Jaina Śauraseni.
It is also believed that the Tiloyapannatti of Yati Vṛṣabhācārya (bet. 6-8 centuries A.D.) is also written in Sauraseni.
Though the date is not certain, yet it is said that the BhagavatiĀrādhana of Śivarya containing about 2170 (or 2166) verses, a pretty lengthy text, is also written in Prakrit or say, in Jaina Sauraseni. "The Prakrit dialect shows", says A.N. Upadhye in his Bṛhat-kathākoşa, "close affinities with the Ardhamāgadhi canon on the one hand and with the works of Kundakunda etc. on the other; and the commentaries explain certain queer forms as ārṣa" (p.55). As regards the date of Śivarya or the Bhagavati-Ārādhanā, Upadhye further says that "the Bh. A. belongs to the earliest stratum of the Pro-canon of the Digambaras consisting of the works of Vaṭṭakera, Kundakunda etc. It is quite likely that Śivarya might be senior even to Kundakunda, but we have to await further researches." (p. 55).
Besides these books, there are some works where the specimens of Sauraseni are found. In the 8th century, in the Samaraicca-kahā of Haribhadra Suri (bet. 705 and 775 A.D.) there are some specimens of Sauraseni. "The "Samaraicca-kaha is written in prose with inserted verse passages of varying length (usually in the Arya metre). The language is Jaina Māhārāṣṭri. In the verses it does not differ from the

Page #9

BANERJEE : ŞAȚKHANDĀGAMA AND SAURASENI
107
dialect used generally in Jaina-Prākrit, in the prose it is mingled here and there with peculiarities of Śauraseni (History of Indian Literature, Vol-II, p. 525).
At a much later period, there are some works written in many Prakrit dialects. Dharmavardhana's (about 1200 A.D.) Şadbhāsā. nirmita-Pārsva-jina-stavana and Jinapadma's (1325-1344 A.D) Şaạbhāṣā-vibhūşita-Santinātha-stavana are poems written in six languages of which Śauraseni is one, besides Sanskrit, Māhārāștri, Māgadhi, Paiśāci and Apabhramsa.
The Sau. literature that we find next is the Skt. dramas. All the Skt. dramatists have given some passages in Sau. representing some types of characters according to the Skt. dramaturgy. Beginning from Ašvaghosa (1st c. A.D.) or Bhās 1 (1st c. A.D), Śūdraka (2nd c. A.D), Kālidāsa (5th c. A.D) and many others have given many Sau. passages in their respective dramas and these passages are good imples of the sau. language and most of the Sau. features are generally found in these Skt. dramas. These dramas came down upto 10th c. A.D. The only Pkt. drama written in Śau. is the Karpūramanjarī whose language is a controversial one.
Beyond these three categories we do not have any direct evidence of Sau. literature. But, on the contrary, we find some of the language specimens in some other contexts also. In the Inscriptional Prakrits beginning from Asoka (3rd cent. B.C) down to the 5th c. A.D. or even later than that, some of the characteristic features of Sau. are, of course, found in the Ins. Pkt., but these features do not mean that those inscriptions are in Sau. The point is that a literature written entirely in Śau. is practically not found after Kundakunda, but for the study of the sau. ig. some inscriptional and dramatic Pkts. can be consulted besides Kundakunda and the Digambara canonical literature.
In this connection it should be mentioned that there is a difference between the dramatic Sau. and the canonical one. In most of the canonical literature the uniformity of the Sau. Ig. is not maintained. As a result, those irregularities are considered as archaic forms and naturally they justify them by calling them the oldest specimens of canonical language. Anyway for my present study this much is sufficient to say that before considering the language of the Şatkhandāgama, let us see how many of the Sau. characters are preserved in the Șatkhanţāgama.
Before discussing any other linguistic problem, it is better to give some of the linguistic features of the Şatkhandāgama to see in which dialect it is written, or whether it is written in Sauraseni or not.

Page #10

108
JAIN JOURNAL: Vol-XXX, No. 4. April 1996
5 The Features of Sauraseni
The characteristic features of Sauraseni as given by Hemacandra and other Prakrit grammarians mentioned above are generally considered as features of Śauraseni. This does not mean that the features which are not found in Prakrit grammars should not be considered at all as features of Śauraseni. Every dialect has some peculiar features which are absent in the others. Most of the featues of Mahārāṣṭrī or Prakrit being common to all types of Prakrit dialects, will also be considered as features of that dialect. Whether such and such features are features of such and such dialects depends solely on the basis of Prakrit grammars. As a result, the Prakrit grammarians have only described those features which are very distinctive in a particular dialect. Here in this dissertation I will discuss the features of Śauraseni as found in the Prakrit grammars of Vararuci, Hemacandra, Purusottama and others to see whether these features are found in the Satkhandagama.
Phonology
(i) Intervocalic -t- changes into -d-.
All the Prakrit grammarians have said that the intervocalic nonconjunet -t- becomes -d- in Śauraseni; in other words, medially dis allowed in Śauraseni, e.g. tataḥ> tado (S), yataḥ > jado (S); tātaḥ > tādo (Ś). In this way examples can be multiplied.
Normally this feature of Śauraseni is found in almost all the Śauraśeni passages of the Sanskrit dramas. But there are some supposed Śauraseni texts where this rule is violated, i.e. instead of tado we also come across tao, and this type of example is a disturbing one for which the Śauraseni passages are mutilated.

It should be noted that the basic difference between Śauraseni and Mahārāṣṭri with regard to this particular point is that this intervocalic -d- in Śauraseni is never elided, whereas in Māhārāṣṭri it is generally elided. The elision of intervocalic -d- in Śau. should generally be regarded as wrong editing. In most of the cases this type of wrong editing has forced scholars to believe that even in Śauraseni this intervocalic -d- is often elided. That this intervocalic -d- is never elided has some evidence also from the grammarian Hemacandra. While describing the ablative declensional endings of the Māhāraṣṭri Prakrit, Hemacandra has made the sutra-nases-tto-do-du-hi-hintolukaḥ (Hc.III.8.) and in the vṛtti of this sūtra he has made a remark that 'dakāra-karaṇam bhāṣāntarārtham' which simply means that the inclusion of -d- in the sūtra [of Māhārāṣṭrī] is to indicate that

Page #11

BANERJEE : ȘAȚKHANDĀGAMA AND SAURASENI
109
this -d- is retained in some dialects (say, in Saurasenī). And this simple instance of Hemacandra is sufficient enough to prove that wherever in any sūtra of Hemacandra we come across a rule of intervocalic-d-, that-d- must not be elided in a dialect, say, Sauraseni. In all these cases the evidence of manuscripts is not a strong proof to accept the reading in Sauraseni with the elision of -d-. So if any text is full of this type of readings (where this type of elision is found) that text should be considered as erroneous, because this is again principle of grammar or linguistics.
In the Șatkhandāgama both the elision and retention of intervocalic -d- are found. ii) Intervocalic -th-, -dh-changing into -dh-.
In a similar way another phonological feature which is strong enough to make Sauraseni distinct from the others, is the sound dh.
iraseni, the intervocalic -dh-is retained. This is not elided, nor is it reduced to -h- as is done in the case of Māhārāştri. Not only that, even the -th- is changed to -dh- in Sauraseni cf. tāthā>tadhā (%); kathayati>kadhedi (%) and so on.
This rule of Saurasenī is very often violated. In most of the editions of Sauraseni texts, if not all, this sort of confusion is regularly found. In this regard scholars are divided into two : one group of scholars thinks that the retention of -dh- or -th-> -dh- is not as such a very distinctive feature of Sauraseni. To them -dh- and -h- are mutually interchangeable. As a result in most of the texts where Sauraseni passages are found this type of interchanges between -dh- and -hare also found. The other group is of opinion that the retention of -dhand the change of -th- into -dh- are distinctive features of Saurasenī, and as such, this should be retained in the text. As a result, this group of scholars wants to correct the Sauraseni passages where -his retained by changing it into -dh-.
In the works of Kundakunda and even in the Satkhandāgama this phenomenon is found. And naturally scholars do not categorically say that this is not possible in Sauraseni. Though it is difficult to remove this sort of traditional belief, scholastic approach should be free from any kind of dogma.
If we look at the problem historically we can say that this retention of-dh- and the change of -h' or -th-into-dh- is definitely a distinctive feature in Sauraseni. The country Surasena was situated in the midland (Madhyapradeśa) which is regarded as modern Mathura. In

Page #12

110
JAIN JOURNAL : Vol-XXX, No. 4. April 1996
this particular area the Vedic language had great influence and in Vedic language -dh- is retained in many places where in classical Sanskrit -h- is found, e.g. classical iha > Ved. idha > Ś. idha, Cl. kathayati > Ved.* kadhayati > S. kadhedi, > As kadhati and so on. Though this retention of -dh- in an intervocalic position in Asokan inscriptions situated at different places is not uniform in this particular feature, we can only assume that this Vedic feature which was retained in Sauraseni dialect has a sort of pan-Indian development even in later inscriptional Prakrits. This intervocalic -dh- is also found in many inscriptions other than those situated very near to Mathura. And in the early Sanskrit dramas, particularly in the Mrcchakatika and Bhāsa, this feature is prominent. Obviously, the later the Sanskrit dramas the better the preservation of this feature. This short survey of the retention of -dh- in ancient Surasena land is sufficient to say that the retention of intervocalic -dh- in Sauraseni is one of the dominant features. At a very later stage there was a controversy with regard to the language of a Prakrit drama Karpūramañjari by name. Let me discuss the problem of this drama in this particular case.
The language of the Karpūramañjari is a controversial one. Some say that it is entirely written in Sauraseni, Rājasekhara being a man of Śūrasena, it was quite possible for him that his drama should bear the stamp of the Sauraseni language. The other type of scholars thinks that the prose passages of the Karpūramañjari are written in Sauraseni whereas the verses are in Māhārāstri. The main pioneers of this problem are the two scholars-Manmohan Ghosh and Sten Konow. The edition of Sten Konow's Karpūramañjari was published in 1902 in the Harvard Oriental Series. Though the edition is excellent, full of copious variant readings, his edition bears two types of language, one is Sauraseni and the other one is Māhārāstri. Sten Konow has retained the Māhārāştri features in verses and the Sauraseni features in prose passages, sometimes against the evidence of the manuscripts. Sten Konow has consulted several manuscripts, about 10, of which W manuscript represents the Sauraseni features even in verses. Sten Konow has not accepted them in the main body of the text, e.g. the very first verse of the Karpūramañjari is given in Māhārāştri even though his W manuscript has Sauraseni features as well. e.g. Sten Konow reads bhaddam hou where W manuscript has bhodu and Y manuscript has hodu. Though hodu is a hyper-prakritaisation, the bhodu reading is supported by the Prakrit grammarians. S. Konow has not accepted that, but he has given these variants in the footnotes. This has been done throughout his text. The loss of intervocalic -t- is found in the verses where it should be -d- according to the

Page #13

BANERJEE : ŞAȚKHAŅDĀGAMA AND ŚAURASENI
111
prescription of Sauraseni. This edition had some influence for a long time till the appearance of Manmohan Ghosh's edition in 1938 published by the University of Calcutta.
Ghosh, on the contrary, has given the features of Saurasenī both in verses and in prose passages. In his opinion, Rājasekhara must have written the Karpūramañjari in Sauraseni. As a result, he has corrected all the Māhārāstri features of the verses into Sauraseni, of course, in the support of new manuscript designated by him as A which he thinks is akin to Sten Konow's W. On the strength of these two supporting statements, Ghosh has edited the Karpūramañjari all through in Sauraseni. In this connection we may note that even though Ghosh has a bold step over Sten Konow to make his text full of Sauraseni, he could not maintain his gusto till the end. Almost in the 4th act and in part of the 3rd, he could not maintain the Sauraseni features even in the verses. Whether Ghosh's unique manuscript was mutilated at the end or whether the manuscirpt is incomplete for this incongruity, is not easy to ascertain at this moment, particularly when Ghosh himself is not explicit in this regard. Be that as it may, one main point of difference between the two may be that Sten Konow was perhaps goaded by the influence of Sanskrit dramaturgy which says that the songs in Prakrit should be in Mahārāstri, the idea which Sten Konow might have extended to the verses as well. Whatever may be the reason between the two, it is pretty certain that scholars who follow either edition will be guided by the reading of that edition, and therefore, there cannot be any reconciliation among the scholars. But one thing is sure that whether it is Māhārāstri or Sauraseni the characteristic features of both the languages are maintained throughout. There is no intermixture between Sauraseni and Māhāraştri in the same passage.
But the edition of Kundakunda who is supposed to have written in Sauraseni is not uniform as far as the editions are concerned. Both-dh- and -h-are found in any one of the editions of Kundakunda; and not only this, even in the same verse this dh- and -h- are interchanged, even though the manuscripts do not always support it. As a result, what happens scholars who are not familiar with the Prakrit grammars or even the linguistic principles maintain that both are found or possible in Saurasenī. The same confusion is also found in the Satkhandāgama. I do not know whether the manuscript contains this confusion or the editor himself is responsible for this confusion. In either case, in my opinion, it should be rectified or corrected in accordance with the rules of grammar or linguistics.

Page #14

112
JAIN JOURNAL: Vol-XXX, No. 4. April 1996
(iii) Treatment of ks in Sauraseni.
Among the conjuncts the development of Sanskrit ks seems to be a little peculiar to Sauraseni. According to Hemacandra (II.3) ks has three developments in Prakrit: ks> kkh; ks> cch; and ks> jjh. Among these three developments modern scholars normally think that the development of ks into kkh is one of characteristic features of Śauraseni, e.g. pakṣa> pakkha. But initially kkh is not doubled, e.g. kṣamā > khamā.
But as far as the other developments are concerned they are generally sporadic. There is hardly any definite line of demarcation where ks should be cch, e.g. makṣikā > macchia > Beng. māchi. This development seems to be quite regular in the case of eastern region. And the same word is developed into kkh in the western region, e.g. makṣikā> makkhiā> mākhiin Marathi. Because kṣ develops into kkh in the western region, it is normally suggested that the development of ks into kkh is normally found in the western region and that into cch in the eastern region. And as a result, ks changing into kkh must be one of the dominant features of Mahārāṣṭrī rather than Sauraseni. But Pischel and Woolner think that kṣ> kkh is a feature of Śauraseni. In that case it can be said that Māhārāṣṭri has borrowed this feature from Śauraseni.
But the development of ks into jjh does not seem to be a very regular one, e.g. in the case ojjhara < Skt. avakṣara is a glaring instance of Māhārāṣṭri. So also kṣiņam > jhiņam. In the Kasayapāhuḍa the forms jhiņamajhiņam occurs in the passage-payaḍie mohanijjā. vihatti taha ṭṭhidie aṇubhāge ukkassamaņukkassam jhiņamajhiņam ca ṭṭhidiyam vā (verse 22). In the Jayadhavalāṭīkā the form jhiņam is not explained. This development is mostly found in the Māhārāṣṭri dialect. In fact the examples of this development are not very many.
Coming back to the development of ks into kkh we can say that this development is equally found both in Mahārāṣṭrī and Sauraseni, with this exception that Māhārāṣṭrī's other two developments (cch and jjh) are not found in Śauraseni, at least, in the Śauraseni passages of Sanskrit dramas. Whether kkh was a regular development either in Śauraseni or in Mahārāṣṭri is a matter of speculation. Supposing this development is in either language, we can say that one has borrowed this development from the other, and as in most of the cases Śauraseni passages have only one development, i.e. kkh, scholars might have accepted this development of ks into kkh as a feature of Śauraseni. So, Śauraseni, being in the midland, has influenced both

Page #15

BANERJEE : ŞAṬKHANDAGAMA AND SAURASENI
Māhārāṣṭri on the one hand and Magadhi on the other. We also find kkh in Magadhi.
To sum up from the above discussions, we can say that kkh is the normal development of ks, while the other two developments (i.e. cch and jjh) are sporadic and are also confined to Māhārāṣṭrī only. In the case of Māgadhi either is possible, though kkh is the dominant
one.
113
Coming back to the Satkhanḍāgama it is found that this book has only kkh and not the other one.
(iv) The retention of bh in the root bhu in Sauraseni.
The root √bhū has several forms in Prakrit, depending on the dialects. In Mahārāṣṭrī the root √bhū > √hū and then √ho (a gunated form) to which the personal terminations are added, making it hoi, hosi, homi etc. Though initially bh does not become h as per rules of aspiration in Prakrit, in the case of √bhū it is an exception. But in Śauraseni this bh of the original Sanskrit root is not changed into h as in Māhārāṣṭrī. So the Sauraseni forms will be bhodi, bhosi, and bhomi. These forms of Māhārāṣṭrī and Śauraseni are very regular. In the case of Śauraseni, Hemacandra (IV.269) has optionally accepted hodi, hosi, homi as well. This form i.e. bh>h in Sauraseni is not regular even though we find this usage in literature. There are other two forms of the root √bhu which are also accepted both in Māhārāṣṭri and Śauraseni, e.g. from Skt. bhavati we have havaï in Māhārāṣṭri and bhavadi and havadi (HC. IV. 269) in Śauraseni. There are other forms as given by Hemacandra both in Mah. and Śau. as Mãh. huvaï, Śau. huvadi, so also bhuvaï in Mah., bhuvadi and huvadi in Sau. To speak frankly the base bhava or bhuva or hava and huva are of later origin than hu or ho. We can call these forms hyper-prakritisation directly from Sanskrit, or in other words, it is Sanskrit based form in Prakrit. The reasons of the development of these forms are not known to us, except that they are developed out of metrical necessity, and so we find these bhuva or huva, bhava or hava used mostly in poetry rather than in prose. Now if a manuscript reads bhuvadi or huvadi, bhavadi or havadi and its corresponding mansuscript reading is bhodi or hodi, I personally prefer the latter readings to the earlier ones, provided it does not disturb the metrical necessity.
In Kundakunda's works we come across more bhuva or bhava, huva and hava, rather than bhodi. The form hodi is, of course, also found there. In the Niryukti the forms havaï and bhavaï are more frequent than bhuvaï and huvaï. To my mind it appears that these

Page #16

114
JAIN JOURNAL: Vol-XXX, No. 4. April 1996
forms are not very much regular in Pkt. they are the later tendencies and these have crept into the manuscripts by the copyists who were perhaps influenced by their knowledge of Sanskrit as it is done today.
In Skt. dramas, of course, the form bhodi is more frequent than hodi, even though they are more or less used side by side.
In the Satkhaṇḍāgama we do not come across very often bhodi or hodi, but in the commentary, the form bhodi is more conspicuous than the form hodi. However, unless a statistical account is made we are not sure whether bhodi supersedes the hodi. In the Kasayapāhuḍa the forms hodi, havadi and homti (pl) are found. To me it seems that they are not regular forms in Sauraseni.
Here perhaps a comment on the forms of hodi, bhuvadi, havadi is necessary. The inclusion of these forms by Hemacandra seems to be historical rather than a genuine characteristic feature of Śau. We have come to this conclusion on the basis of the fact that these forms are also
available in Mg. Śau. being a language in the midland, it has perhaps influenced Mg. as well. While, on the contrary, these forms have come down to us from the influence of Mah. as well. As Hemacandra has made a special sūtra for the root √bhū which becomes bha (bhuvo bhaḥ IV. 269) in Śau., it appears that if that is not the dominant feature of Śau., Hemacandra could not have made this sūtra at all. The optional forms based on hu is in a sense a loan from Mah. However, my main intention of mentioning this is to emphasize this point that the genuine Śau. form of the root √bhū would be bho with usual suffixes. As in the Satkhandagama as well as its Dhavala the occurrence of the form bho, bhodi etc. is found side by side with ho, hodi, the Satkhanḍāgama and Dhavala have also the features of Sauraseni.
That in the initial position Śau. retains the sound bh is also proved by the fact that the words bhavat and bhagavat also become bhavam and bhagavam in Śau (bhavad-bhagavatoḥ, IV. 265) where Hemacandra has not suggested any other alternative forms with havam as in the case with the root √bhū. So all these points lead us to this conclusion that in Śau. the retention of initial bh, particularly in the root √bhū, is more logical and historical than the other forms with hu or ho.
(v) Changes of ry to yy
According to Hemacandra (IV.266), in Śau. ry is optionally changed to yy, and naturally the usual change of ry to jj is also found in Śau. e.g. ārya ayya, also ajja, kārya> kayya, kajja, sūrya suyya, sujja.

Page #17

BANERJEE SAṬKHAṆDĀGAMA AND SAURASENI
A careful perusal of all these examples will immediately reveal the fact that the change of ry to yy cannot be a pure Śau. form. As j is changed into y in Mg. (Hc. IV. 292) and therefore ry is also assimilated to yy in Mg. whereas the change of y intoj is found in Mah. as well as in Sau. So it seems that this character of Sau., i.e. ry changing into yy is a loan from Mg., because that is one of the dominant characters of Mg. So Sanskrit ārya > Māh. Śau. ajja, and Mg. ayya. Though Hemacandra has given this rule in the case of Sau. this is in fact a Mg. rule as is known by his sūtra (Hc. IV. 292). The picture is given below:
In Pkt., or for that matter in Māh, the initial y is changed into palatal j (Hc. I. 245) e.g. yadi > Pkt. jaï; yama jama and so on. But when it is in the middle position this ya is generally elided in Mah. (Hc. I. 177) and is again replaced by ya-śruti. (Hc. I. 180), e.g. samaya> samaa samaya.
115
Apparently it seems that perhaps medially -ya- is not elided; but that is not the case. In Śau. also the same form occurs. But in Mg. this -ya- is not as such elided, but is retained. This is mostly evident in the case of conjuncts. In conjuncts ry is normally assimilated to jja (<* rj). e.g. ārya> ajja; kārya> kajja. The same form is also found in Śau. But in Mg. as there is no j, it is assimilated to yy, e.g. ārya> ayya; kārya kayya. Naturally there is no scope of ry> yy in Śau. unless we accept that it is a loan from Mg.
In the Sau. passages of Skt. drama the assimilated form jj of ry is found. Whereas in the Mg. passages yy is found. In the works of Kundakunda the form jj is found, and the form yy should be regarded as badly edited texts in Sauraseni.
In the commentary on the Saṭkhaṇḍāgama the forms are mostly found with jj and not really with yy, unless it is mistakenly incorporated into the text.
Morphology
(vi) Declension
- ablative singular in du/do.
In the case of declension, Śau. has not got as such any peculiarity except in the abl. sg. of a- base. In this particular case the retention of a is normally found, i.e. from Skt. narāt we have narādu or narado, while narāuand narão will be in Mah. This peculiarity can be justified by the general rule that in Sau. the intervocalic -d- is retained. And hence Hemacandra has made a sūtra of the abl. sg. as nases tto do du hi hinto lukaḥ (III. 8) where in his vṛtti he has said dakāra-karaṇam

Page #18

JAIN JOURNAL: Vol-XXX, No. 4. April 1996
bhāṣāntarārtham and by bhāṣā he, of course, primarily means Śau. and perhaps secondarily Mg. where also the same form occurs.
116
However, this feature as such occurs in the Saṭkhanḍāgama and is in general governed by the general rule that the intervocalic -t-> - d- in Śauraseni.
(vii) Vocative in a or am.
But in the case of Skt. words ending in -n, in Śau. this -n- becomes optionally -ā in the voc. sg. (Hc. IV. 263), e.g. Skt. bho kañcukin> Sau. bho kancuia; bho sukhin> Sau. bho suhia; this one seems to be a special rule for Sau. in voc. sg. of words ending in -n. In other case n is dropped in the words ending in -n e.g. Skt. bho tapasvi >Śau. bho tavassi; so also bho manassi. This means that the general Mah. form of voc. sg. of words ending in -n is used in Śau. but the present rule ending in -ā in the voc. sg. is a special rule for Śau.
There is no example of this type of rule in the main body of the Satkhanḍāgama, because there is no such scope in the work. But in the case of words ending in -n or a- base the anusvära takes place in place of -n in the voc. in Sau. (Hc.IV.264), e.g. Skt. bho rājan Śau. bho rāyam, similarly bho vijayavarman > Sau. bho viayavammam; sukarman >sukammam. In a similar way we have examples in literature like bho bhagavan kusumāyudha> bho bhayavam kusumāyuha; and so also bhayavam tittham pavatteha.
In all these examples the words ending in -an> -am in the voc. do not seem to be a very peculiar feature of Sau., even though grammarians admit that bhagavan could have been equally bhaava also. This shows the influence of Mah. on Sau. However, as far as the Satkhaṇḍāgama is concerned there is no such example in the work. (viii) Pronominal declension - tasmāt > tā.
In the pronominal declension according to the grammarians (Hc.IV.278) the Skt. form tasmāt>tā in Śau. e.g. tā jāva pavisāmi; tā alam ediņā māņeṇa and so on.
In the sutra of the Satkhanḍāgama the form tā from tasmāt is not met with.
(ix) Future in -ssi.
Apart from this, the grammarians (Hc. IV. 275) say that in Sau. the ending -ssi is added in the future tense instead of -hi as in the case of Mah. eg. bhavissidi, karissidi and gacchissidi.

Page #19

BANERJEE ȘATKHANDAGAMA AND SAURASENİ
It is a fact worth noting that according to Hemacandra (III. 167) the -ssa is used in Mah. and this is an extension to Śau. which shows that the other forms of Mah. is not possible in Sau. If this statement of Hemacandra is correct we can say that the special rule -ssi for Śau. must be compulsory and peculiar to the Sau. language.
117
We could not find any example of the future form of Śau. with this -ssi in the Satkhanḍāgama.
(x)The formation of gerundial suffix-ktvā
In the formation of the gerundial suffix, Śauraseni has again some peculiarities which are different from others. According to the grammarians (Hc.IV.271) the gerundial suffix in Śauraseni, is -iya, and -dūņa. Of these two the suffix-dūņa is pure Śauraseni, while the suffix-iya is a loan from Mah. As the suffix -iya does not contain any -d, it is possible that it could be used in Śau. as well, e.g. √bhū > bhavia, bhodūņa, and havia, hodina, so also path> paḍhidūņa, padhia; √ram> randūna, ramia. A glance at these examples will reveal the fact that as Śau. retains -d-intervocally the forms with -dūņa should be a recognised one for Śau. while the others are Mah. or Pkt. in general. There could be another formation of gerundial form both in Śau. and Mäh. This form is nothing but the assimilation as -tvā > -ttā. This shows that the gerundial forms with -d is more important for Śau.
than the other forms. So the roots mentioned above have other forms as well, say bhottä, hottā, paḍhitta and rantā.
But in the case of √kṛ and √gam we have some special forms (Hc. IV. 272) besides the suffixes mentioned above. From √k we have karidūņa and karia, gacchidūņa/gamidūņa, and gacchia/gamia. Besides these forms we have another suffix with -ua. So we have from
Vk kaḍua and from √gam> gaḍua. The Pkt. grammarians have called all these forms a special case in Śau. To my mind it appears that kaḍua and gaḍua are forms of Mg. as Mg. √kṛ+kta kaḍa and √gam+kta > gaḍa. So also √mr+-kta > maḍa. (Vara XI.15) And these forms are not used in Śau. Naturally it appears that the gerundial suffix -ua is added to these Mg. forms of √kr, Vgam, and √mr. So it appears to me that the Śau. forms of these kaḍua and gaḍua are nothing but transferred to Śau. from Mg. and just as we have kaḍua in Mg. and Śau., so also kāūņa or kariūna in Mah.
So if a text contains the forms kaḍua and gaḍua there is no reason to consider them as special case of Sau. We could say only this much, as it does not disturb the language, there is no harm in accepting or removing these forms at the time of collation of manuscripts. In most

Page #20

118
JAIN JOURNAL : Vol-XXX, No. 4. April 1996
of the Sanskrit dramas the forms kadua and gaļua are found in the Mg. passages, whereas in pure Sau. text karidūņa, gamidūņa or gacchidūņa are found.
In the case of the Şațkhandāgama the form does not occur in the sūtra of the text. But in the commentary Dhavalā the forms used are with dūņa rather than with -ia. However, if the forms with da is found, it is, of course, a sau. dialect, but other than this should be regarded as common forms of Pkt. (xi) Indeclinables
In some indeclinables Šau, has some peculiarities. We are giving these peculiarities on the basis of Hemacandra. e.g.
idānimo dāņim (IV. 277).
In Sau. the word idānim (now)> dāņim, e.g. añantara-karaniyam dāņiņāņavedu ayyo.
These examples of this sūtra can be found only in the case of dramas. It is difficult to get an example of dāņim in the Şațkhandāgama as it is a philosophical text.
evārthe yyeva (IV. 280).
In the sense of eva meaning 'indeed', 'in reality' the particle eva >yyeva in Sau. eg. mama yyeva bambhanassa, so yyeva eso.
These examples of this sūtra can also be found in Skt. dramas rather than in the Satkhaņdāgama.
ņam nanvarthe (IV. 283).
In Śau. nam is used in the sense of nanu i.e in the sense of doubt or question etc., e.g. ņam aphalodayā, nam ayyamissehim pudhamam yyeva āņattań, ņam bhavam me aggado caladi.
The examples given by Hemacandra are from the Skt. dramas mainly from the Abhijñāna-sakuntalam and Mscchakațikam.
It is a fact worth noting that in Jaina canonical literature, particularly in the prose portion, this ņam is used as a sort of emphatic, eg. jayā ņam, tayā ņam, etc. But in the Șatkhandāgama, particularly in the main sūtras of the text, this nam is not used, but in the commentary Dhavalā it is found profusely. (xii) Interjections.
In the case of interjections there are some words which are peculiar

Page #21

BANERJEE : ŞAȚKHAŅDĀGAMA AND SAURASENI
119
to Śau. Hemacandra in his Pkt. grammar has mentioned some of them. e.g.
hanje cețyāhvāne (IV. 281).
In Śau. (as used in the Skt. dramas) hanje is used in calling a maid-servant (ceți). e.g. hañje cadurike.
As it is generally used in Skt. dramas, its usage in the Şațkhandāgama is not as such possible.
himāṇahe vismaya-nirvede (IV. 282)
In Sau. the word himānahe is used in order to indicate a surprise or dejection. e.g. himāṇahe jīvanta-vacchā me jaņaņi (in surprise); himāṇahe palissantā hage edena niyabidhiņo duvva vasideņa (in dejection).
ammahe harșe (IV. 284)
In Śau. the word ammahe is used in the sense of delight. e.g. ammahe eāe summilāe supaligaļhido bhavam.
hi hi vidūşakasya (IV. 285)
In Sau. in order to indicate the delight of the Vidūşaka the particle hihi is used, e.g. hihi bho sampannā manoradhā piya-vayassassa.
In these two above cases it is seen that the examples of these two sūtras can only be found in the Skt. dramas. Naturally it is difficult to get any example of these sūtras from the Satkhaņdāgama.
In the end we may say that the above analysis has given us this impression that whether the language of the Şațkhandāgama is Śauraseni or not, it is mixed up with other Prakrit dialects, and this is also said by Hiralal Jain in the Introduction to his edition of the Şațkhaņdāgama: "The present work (i.e. Şațkhaņdāgama) consists of the original sūtras, the commentary of Virasena called Dhavalā and the various quotations given by the commentator from writings of his predecessors. The language of the sūtras is Prakrit and so also of the most of the quoted Gāthās. The prose of Virasena is Prakrit alternating with Sanskrit. In the present portion Sanskrit predominates, being three times as much as Prakrit. This condition of the whole text clearly reflects the comparative position of Prakrit and Sanskrit in the Digambara Jain literature of the south. The most ancient literature was all in Prakrit as shown by the sūtras and their first reputed commentary Parikarma as well as all the other works of Kundakunda,

Page #22

120
JAIN JOURNAL: Vol-XXX, No. 4. April 1996
and also by the preponderance of Prakrit verses quoted in the Dhavalā. But about the time of Virsena the tables had turned against Prakrit, and Sanskrit had got the upperhand as revealed by the present portion of Dhavalā as well as its contemporary literature."
"The Prakrit of the sutras, the Gāthās as well as of the commentary, is Sauraseni influenced by the older Ardhamāgadhi on the one hand, and the Mahārāṣṭri on the other; and this is exactly the nature of the language called Jain Sauraseni by Dr Pischel and subsequent writers. It is, however, only a very small fraction of the whole text that has now been edited critically so far as was possible with the available material."
6 Manuscript vs Prakrit grammarians
It is at this point a controversy is raised with regard to the problem of manuscript vs Prakrit grammarians. From the time the Europeans started editing Prakrit texts, this question got its prominence. As far as I know T. Bloch in his Vararuci und Hemacandra (Gütersloh, 1893), was the first scholar who raised this question while discussing some of the readings of the sutras of Vararuci and Hemacandra. In his opinion the reading of the manuscript should be given priority despite the fact that it goes sometimes against the grammarians. He further says that there has always been a historicity in the different manuscripts of the same book which might reflect earlier usage of a particular reading preserved fairly in a chronological order. Naturally he has given priority to certain readings of Vararuci which he thinks are older specimens of the language than Hemacandra. Pischel, on the other hand, is of different opinion. In his book Grammatik der Prakrit Sprachen (Stuttgart, 1900, § 42) he has attached much more value to the grammarians than the manuscripts. He has not ignored the readings of manuscripts, but what he says is that the manuscripts should be corrected, if necessary, in accordance with the rules of Prakrit grammar. The glaring instance of his theory is the Pkt passages of Kalidasa's Sakuntala which he has corrected in accordance with the rules of Hemacandra. This is greatly seen in the 6th act of the Śakuntala particularly in the Magadhi passages of the fisherman. As the manuscripts of that passage is corrupt and not inconformity with the rules of Prakrit grammar, he has corrected them in accordance with the rules of Hemacandra's Prakrit grammar. In short, Pischel has not totally ignored the manuscripts, but his emphasis is on the correctness of the text. At a much later time E.B. Clark in his article Magadhi and Ardhamāgadhi published in the JAOS, 44, 1924, pp. 81-121, has reiterated this problem. But he could not find out any

Page #23

BANERJEE : ŞAȚKHAŅDĀGAMA AND ŚAURASENI
121
ways and means by which a solution could be found. However, the problem has been going on. And again almost after a century I have discussed the problem anew. In one of my articles Prakrit Textual Criticism published in Jain Journal, January, No. 3 1988, pp.87-97, what is noted there is being summarised here for the benefit of the readers.
“The basic problem is whether the grammarians or the manuscripts are to be followed. It is not easy to answer this question, particularly when most of the scholars think that any kind of linguistic phenomenon is possible in Prakrit. Perhaps, under the tacit influence of this so-called idea, some of the Prakrit forms have been incorporated in some editions which sometimes baffle and betray some of the basic notions of the Prakrit language including dialects as enunciated by Prakrit grammarians. It is true that Prakrit grammarians are not very old, and most of the authors belonged to a time when the Prakrit language was almost stereotyped like Sanskrit. As a result, the Prakrit features as embalmed and treasured up by the grammarians vary from author to author, except a few general forms which are common to all; the variations are such that it is difficult to follow any particular reading from the manuscripts. The copyists are not always learned, more so, they may not have any knowledge or a very limited knowledge of the language, and hence every possibility of making mistakes. The phonetics of the language is not always regular. Sometimes the copyists add something to the manuscripts to improve upon the text”.
"It is my personal feeling that some sorts of emendations are necessary to edit a Prakrit text --if the manuscripts of a particular text do not help us much." It goes without saying that the original readings of the text must be recorded as variants in the foot-notes. “It is, indeed, very difficult to form direct cut and dried principles for this purpose, unless we base our arguments on some principles by which grammarians are involved in the matter."
The main reason for which I have discussed the problem here is to say that if the readings of a Prakrit text do not always conform to the rules of grammar as laid down by the Prakrit grammarians, it is then very difficult for us to follow any definite rules for a particular Prakrit text. Moreover, most of the Prakrit texts, if I am permitted to say so, are corrupt from the point of view of grammatical rules. More
is a question of Prakrit dialect. So is the case with the language of the Satkhandāgama and other similar Prakrit texts like Kasāyapāhuda, Mahābandha, Tiloyapannattī, Kundakunda's works and so on. In the Kasāyapāhuda, the Sanskrit word prathamārdhasya

Page #24

122
JAIN JOURNAL: Vol-XXX, No. 4. April 1996
has two Prakrit forms as paḍhamaddhassa and padhamatthe in a few pages gap. Which one is to be considered as the Sauraseni form? The paḍhamatthe can be derived from prathamarthe, but it has a different reading padhamaddhe. What I want to emphasize is that sometimes in the edited texts we might often get a wrong or irregular form of a language, e.g. Skt. kartavyā> Ś. kāyavvā, Skt. nikṣitavyā> Ś. nikkhiyavvā, and as far as Sauraseni is concerned, these forms are not regular, they should be corrected as kādavvā and nikkhidavvā.
The reading arihantāṇam is a puzzling one. It has three readings; arahantāṇam, arihantānam and aruhantānam. In the Hātigumphā Inscription of Khāravela at Udayagiri Hill in Bhubaneshwar belonging to the second or first century B.C., the reading arahantāṇam (namo arahamtāṇam namo sava-sidhāṇam) is found. This is, perhaps, the earliest written document where the reading arahantānam is available. Then the reading ariha- was developed in Prakrit from Sanskrit arha (cf. Hemacandra, rha-sri-hri-kṛtsna-kriya-diṣṭyāsvit, II. 104). So also we have the reading arihantāṇam (gen. pl); and this reading is mostly found in almost all the Jaina texts, canonical and non-canonical. Then, in course of time, before Hemacandra (1088-1172 A.D.), the reading aruha-was also developed (cf. Hemacandra, uccārhati, II, 111). I have seen this reading in the foot-note of the recently published text of Bhagavati-joḍa (1981) by the Jain Vishva Bharati Institute. Thus far is all right as far as the development of the Sanskrit word arha in Prakrit is concerned.
But the difficulty is-in some texts published by the same Institute or edited by the same scholar, the two readings - arahantānam and arihantānam are found. For example, in the Satkhandagama edited by Hiralal Jain the reading arihantāṇam is accepted, but in the Somasundara-cariu of Nayanandi (samvat 1100) edited by the same scholar, Hiralal Jain (1970, Vaiśāli), the reading of the namokkāramantra is arahantāṇam and not the other one. It has no variant in the foot-notes. If we consider his reading as correct, then we can say that the arahanta- reading was also prevalent in the samvat 1100. In the Bhagavati-joḍa of Jayācārya, published by the Jain Vishva Bhārati (1st edn 1981) we have two readings side by side. In the Prakrit namokkāra-mantra, the reading arahantāṇam is found, but in the duha (58) the reading is namo-arihantānam (parameṣṭhi pamcaka 'namo-arihantāṇam' āda). In the foot-note of this verse, the editor has mentioned that according to the vṛttikara, there are various forms of arahanta of which arihantānam and aruhantāṇam are variants. Again in duha (61) the reading arahantā is given. What I want to emphasize from all these discussions is that before editing a Prakrit text, a

Page #25

BANERJEE ȘATKHANDAGAMA AND SAURASENİ
conscious editor should formulate some editorial disciplines, so that there could be a consistant and coherent readings of a Prakrit text, no matter what the manuscripts say.
7 Conclusion
From the above survey it appears that some of the features of Pkt. are restricted to Ŝau. This means that the rest of the features of the language are to be collected from Mah. or Pkt. in general. Hence, all the grammarians have said that the rest of the features of Śau. are to be collected from Pkt. in general. (seṣam Prākṛtavat Hc. IV.286, seṣam Mahārāṣṭrivat, Vara XII.32). Naturally some of the features are common to both Mah. and Śau. It is in this connection we see some of the lapses in the Sau. lg. and it is in this connection we have some misunderstandings in describing the characteristic features of Śau. Hence we find confusions between t and d, between h and dh, and between Mah. and Mg. where the influence of both the languages is possible on Śau. The corruption in the manuscripts is another reason which also puzzles us in order to determine the features of Sau. As the grammarians have restricted some features exclusively to Śau. it should be our duty to adhere to those rules which are exclusively meant for Śau.; the other features being the same with Mah. we have no alternative to relying on the grammarians. In this particular case manuscript may not be a dependable source for detecting the features of Sau. As most of the scholars do not pay much attention to the fact that the Sau. Ig. has some distinctive features which are basically the retention of intervocalic d and dh, and at least, these 2 features are exclusively meant for Śau. even though manuscripts sometimes betray this conception. If we could have got the autographed manuscript of the author we could have thought of this matter in a different way. As the copying of the manuscripts has been handed down to us from generation to generation the reliability of the lg. is vitiated rather than copying correctly the manuscript from the other. This fact should not be ignored when a Pkt. text is edited from manuscripts. By this process, perhaps, somebody might say that this means the mutilation of the manuscript, but it should be considered at the same time that the wrong representation of the lg. does not lead us to the correct understanding of the text. Grammarians normally register some facts of the lg. which the writers of a particular book is supposed to know. When the lg. is current and spoken there might be some words which are not included in the books of grammar, but those can be judged from the linguistic point of view, whether that sort of form is possible or not, e.g. the Skt. word padma has at least 3 forms in Pkt. paduma (Śau), paüma (Mah.), and pomma (Amg). Of these 3 forms, both

123

Page #26

124
JAIN JOURNAL : Vol-XXX, No. 4. April 1996
historically and linguistically, the form paduma originated first. And from paduma the Māh. word paüma comes, because in Māh. the intervocalic -d- can be elided, but not in Sau. and hence paduma is Sau. because of the retention of intervocalic -d-, and paüma is Māh. because of the loss of intervocalic d-. In the case of põmma which is mainly found in the Amg. text the form is an assimilation of -d to m, again following the sūtra of the grammarians (cf. Hc. II. 77). That is to say, padma> 'paümma >pomma (by euphonic combination of a + u). Psychologically in the formation of this word 2 things are operating :
the form paüma has influenced the word in making it paü and secondly, the assimilation of dm occurs. These 2 processes occur simultaneously and as it is not a regular form and at the same time used in the Amg. text, the word is regarded as ārsa (something like irregular) and hence Amg. (because it is found in the Amg. text). This process should be taken into consideration when we see that some words are irregularly formed.
The main point of my emphasis is that although overlappings or lapses are possible in lg., particularly in Pkt., this must be corrected as far as possible with the help of grammar and linguistics. This is the case particularly when we think of a text written in Sau. It is the editorial discipline that makes the text a better representation than mere adhering to the blind allegiance to manuscript copying. If this distinction is not maintained, then there cannot be any features of any lg. at all. There would not be any conception of dialects in Pkt.
Arbeits India
Export House Recognised by Govt of India
Proprietor - Sanjit Bothra
8/1 Middleton Road 5th Floor Room No.4
Calcutta - 700 071 Phone : 296256/8730/1029
Telex : 021 2333 ARBI IN
Fax No. : 0091-33290 174

Page #27

A PRIMER OF ŚAURASENI
with a comparison between
Māhārāştri and Māgadhi For use in readings for lectures
compiled
by
Richard Schmidt
Translated into English from Original German
by
Pranabesh Sinha Ray
Hannover 1924 Orient-Bookseller Heinz Lefaire

Page #28

126
JAIN JOURNAL VOL-XXX No. 4, April 1996
Preface
[3] I am not alone to feel the want of a handy manual of Śauraseni by far the most important of prose dialects of the Indian drama. Who would ever base the works (surely for their period very valuable) of Hoefer or Lassen?. Or who would have the courage to guide his hearers through the primeval forest of Pischel's grammar ? Even a specialist, who is, indeed, accustomed to all sorts of fares sees himself, in all respects, in this stupendously diligent and fundamental work to which he is constrained to pay heed confess that Pischel's treatise is nearly not to be perused. The overwhelming fullness of materials in sheer discriminationless types squeezed into 400 pages and more makes the hunt for a single form agonising: the book is a typographical Hydra which is repelled by the head definitively notwithstanding the index prepared by Wickremasinghe.'
Now, it is self-evident that the advanced student will soon need to be acquainted with one or the other of the dramas and if he at all entertain any serious endeavour he must rest contented with reading the Prakrit portions merely with the help of their chāyās but will be also expected to be led into the understanding of the dialects per se as well. The present manual would serve that purpose. I have drawn up and worded it in such a cryptic, concise manner, and as its title [4] gives out the principal stress has been placed upon Śauraseni. Of the other dialects occurring in Indian dramas we have Māhārāṣtri and Magadhi considering that habitually Śakuntala will be studied, at least, by its chief divergencies.
For those who would like to make an intensive study of the Prakritspeeches I am giving here a list of the most important and pertinent works:
1. Bharatiyanāṭyaśāstra ed. Kāvyamālā No. 42. (XVII, 6-23).
2. Cowell, The Prakrita-Prakāśa: or, The Prákrit Grammar of Vararuci, with the Commentary (Manorama) of Bhámaha. Second Issue. London 1868. (1. Ausgabe Hertford 1854).
3. Hemacandra's Grammatik der Prakritsprachen, herausgegeben, übersetzt und erläutert von Richard Pischel. I Teil Halle 1877. II. Teil Halle 1880.
4. The Päiyalachchhi Namamālā, a Prakrit Kosha by Dhanapāla. Edited with critical notes, an introduction and a glossary by Georg Bühler. Göttingen, 1878.

1. With this naturally a contradiction does not arise, if I say, that Pischel with his grammar has produced a really fundamental work, on which my modest Handbook thoroughly rests.

Page #29

RICHARD SCHMIDT : A PRIMER OF ŚAURASENÍ
127
5. Hoernle, The Prākṣita-Lakshanam or Canda's Grammar of the An
cient (Arsha) Prākřit. Part I. Text with a critical introduction and
indexes, Calcutta 1880. 6. The Deśīnāmamālā of Hemacandra. Part I. Text and critical notes.
By R. Pischel. Bombay, 1880. 7. Namisādhu's Kommentar zu Rudrata's Kāvyālamkāra II, 12 (ed.
Kāvyamālā No. 2, Bombay, 1886). 8. Kramadīśvara, Prākstapāda ed. Rājendralāla Mitra, BT.; neue
Ausgabe Kalkutta, 1889 9. Trivikrama, Prākstavyākaraña. Ed. Granthapradarśanī 1-Dazu
Tukaram Laddu, Prolegomena zu Trivikrama's Prākrit-Grammatik,
(Diss.) Halle a.d. S. 1912. 10. Praksitarūpāvatāra. A Prakrit Grammar based on the Valmikisūtra
by Simharāja, Son of Samudrabandhayajvan. Edited by E.
Hultzsch. London 1909. 11. Mārkandeya Kavindra, Prākstasarvasvam, ed. Bhațţanātha
Svāmin (-Granthapradarsani Heft 34-38, Vizagapatam, (5) Februar-Juni 1912).-Dazu E. Hultzch, Neue Beiträge zur
Kenntnis der Sauraseni, ZDMG 66, p. 709-726. 12. Hoefer, Albert, De Prakrita dialecto libri duo, Berolini, 1836. 13. Lassen, Christianus, Institutiones linguae Pracriticae, Bonnae ad
Rhenum, 1837. 14. Delius, Nicolaus, Radices Pracriticae, Bonnae ad Rhenum, 1839. 15. Haag, Vergleichung des Prakrit mit den Romanischen Sprachen,
Berlin, 1869. 16. Cowell, A Short Introduction to the Ordinary Prākrit of the Sansksit
Dramas, with a list of common irregular Prākrit words, London,
1875. 17. Hoernle, A.F. Rud., A Sketch of the History of Prakrit Philology.
Calcutta Review, October, 1880. 18. Weber, Albrecht. Das Saptaçatakam des Hâla AKM VII, 4. Leipzig,
1881. 19. Rishikesh [Hrşīkeśa] Sastri, A Prakrita Grammar with English
Translation, Calcutta, 1883. 20. Jacobi, Hermann, Ausgewählte Erzählungen in Mâhârâshtri. Zur
Einführung in das Studium des Prakrit. Grammatik. Text. Wörterbuch. Leipzig, 1886. [Dazu : Hindu Tales. An English Trans
lation ... by John Jacob Meyer. London, 1909.) 21. Pischel, R., Grammatik der Prakrit-Sprachen. Straßburg, 1900.
(-Grundriß der ind.-ar. Phil. I, 8).

Page #30

128
JAIN JOURNAL: VOL-XXX No. 4, April 1996
For the examples, it is to be noted that almost always only the actually recorded ones have been admitted, above denotes that the concerned word so runs in Māhārāṣṭi also, as in Śauraseni, while *indicates agreement with Māgadhi. M. and Mg. are abbreviations for Māhārāṣṭri (the language of Mahārāṣṭra, the Marhatta country of today) resp. for Magadhi, the language of Magadha, Ś for Śauraseni. When the sign will be occasionally used for a marker of abbreviation, it will be always clear from the context, what is meant. The same is valid for *which is sometimes added to Sanskrit forms in order to inform [6] an apparent equivalent form though not the expected one; for ex. naia (*nayiya=nitvā). Forms which do not have any supplement (Ś, M or Mg.) are Śauraseni; where several examples from one dialect are given, these are separated by commas, while a semicolon separates dialects. Thus ṇāgadā, M ṇāgaā (nāgatā); necchadi etc., means that nagada the Sauraseni form, nagaa is that of Māhārāṣṭrī; necchadi is then again Sauraseni.
Somewhat more than an elementary knowledge of Sanskrit is presumed. The Indian grammarians also do that -- those who have written on Prakrit which is only reasonable.
prakṛtiḥ samskṛtam; tatra bhavam Prākṛtam ucyate.
Münster W., 24, Sept. 1914 (may it be so)
PHONOLOGY

[7] 1. Śauraseni, the language of Śurasena with its capital at Mathura, has the sound repertory of Sanskrit, lacking, however, in ṛ, ṛ, l, ai, au, simple n, and further y, ś, ș, has also the single voiceless consonants. [Likewise in M, Mg. though Mg. has y resp. yy at the beginning of words, retains s in resp. changed from s and s into ś, and instead of r there stands 1 throughout].
Richard Schmidt.
r becomes a for ex. in tana (tṛṇa), M. ghaa, kaa (ghṛta, kṛta); most frequently becomes i; *ghida, *kida, kisa, *giddha, 'diṭṭhi, siāla, °*hiaa (ghṛta kṛta, kṛśa, gṛddhra, dṛṣṭi, śṛgāla, hṛdaya); particularly after labials it becomes u: nihuda M. nihua (nibhṛta); nivvuda, M. nivvua (nirvṛta); pucchadi, M. pucchai, Mg. puścadi, (pṛcchati); °muṇāla (mṛṇāla); mudanga, M. muinga Mg. midanga (mṛdanga); vuttanta (vṛttänta). Similarly before the suffix ka in jāmādua, M. jāmāua (jāmātṛka); bhādua (bhrātṛka).
Initial becomes either ri as in 'riddhi (ṛddhi), °rikkha (ṛkṣa), risi [Mg. lisi (rşi) beside isi: mahesi, rãesi (maharṣi rājarși) or u: udu, M. uu (ṛtu); ujju (ṛju).

Page #31

RICHARD SCHMIDT : A PRIMER OF ŚAURASENI
129
If correspondingly changes into i or ū, however, illustrative example is lacking!
Į becomes ili : kilitta (kļpta).
ai is changed to e: "Erāvana (Airāvaņa), 0*tella' (taila), 'vějja (vaidya), not to aï (Pischel 8 61) as in M. daïcca (daitya).
au becomes o, very seldom aü in M. : 'Kosta (Kausika), orasa (aurasa); "osaha (ausadha), komudi, M. komui (kaumudi); Kosambi (Kaušāmbi), jõuvaņa) (yawana), 'đóbballa (daurbalya); pora, M. paura, Mg. pola (paura)
[8] n always is changed to ņ: "ņa (na),"naaņa (nayana). It remains unchanged in consonant groups like nt, nd : pěkkhantu (prekşantām), cando (candraḥ).
yfalls out in intervocalic position : Chiaa (hrdaya), 'chāā (chāyā), "pia (priya)
Ś and ș become s : 'kesa, 'pusu, osisira, 'esa, 'sesa (keśa, pasu, śiśira, eșa, sesa)
In Mg. as already stated șand s become ś, śby itself is retained: keseśu (kešeșu), salila (salila), sahassa (sahasra); see however & 8!
aḥ becomes o : 'aggado (agratah, rão (rāgah), ņioo (niyogah).
In Mg. in the Nom. sing, there is e for it : eśe puliše (eşoseşa]puruṣaḥ). -- The treatment of aḥ ih etc. as much as of medial ḥ see 89 as in accidence.
In the final position only the vowels and anusvāra are permitted (but cf. g 3 at end); therefore end-consonants fall out : "tāva, pacchā [Mg. paścā and paścādo] (tāvat, paścāt).
2. The rest of the vowels and diphthongs become, in general, retained. a ā become (indeed only apparently) è in "*ěttha= sponding to the Vedic itthā and in metta (mātrā); u appears as a in maülida, M. maülia (mukulita); u as i in opurisa’, Mg. putiśa (puruşa); uas o in pokkhara (puşkara), Mg. poskaliņi (puskariņi) omottā (muktā).
A part of or final vowels sometimes appears before in between the parts of a consonant group : 'siņeha (sneha); radaņa, M. raaņa, Mg. ladaņa (ratnā); pekkhia, Mg. pěskia (prekşya); kilammadi, M. kilammaï (klāmyati); "kilanta (klānta); Mg. kośiņa (koşņa); ocoria (caurya); •āmarisa, Mg. āmaliśa ämarsa); tuvaradi, M. tuvarai, Mg. tuvaladi (tvarate); duvāra, Mg. duvāla (dvāra); "due (dve).
An initial vowel falls out particularly in enclitics : *pi (after
1. Doubling of the consonant, as is often found also elsewhere (cf. Pischel
S 90) 2. According to Mārkandeya ix, 9, this form for Ś. is false and should run
as purusa.

Page #32

130
anusvāra], "vi [after vowels] for api; "dāņim (idānīm). At the beginning of the sentence avi resp. stands idānīm.
aya becomes e in verb forms as in nedu3 (nayatu); M. kahei, Mg. kadhedi (kathayati); ava becomes o: odaradi [9] (avatarati); M. oāra (avatāra); Mg. odala (avatara), onomāliā (navamālikā); *bhodi M. hoi (bhavati).
JAIN JOURNAL VOL-XXX No. 4, April 1996
-
A long vowel in a closed syllable becomes shortened: kavva (kāvya); "gimha (griṣma); mamsa (māmsa); also in cases like "puttāṇam (putrāṇām), mālam (mālām) sahim (sakhim). Just as e and o before double consonants – jěvva, jõvvaṇa (eva, yauvana) · -as also before kkhu (=khalu) and tti(=iti); for ex, asamaě kkhu (asamaye khalu), mahanto [Mg. mahante] kkhu (mahantaḥ khalu), bamhanotti (brāhmaṇa iti). Besides, there is khu after anusvāra and *kkhu after a short vowel: mā khu (mā khalu), kim khu, M. tam khu (tat khalu); na kkhu (na khalu). In M. after all vowels khu becomes hu: na hu, mā hu, ko hu; but Mg. sě kkhu (sa khalu), āadā khu (āgatā khalu). tti also stands after short and long vowels: esa tti (eṣā iti), duhida tti (duhitā iti), Kandappakeli tti (Kandarpakeli iti); ti stands after anusvāra : kim ti (kim iti). Like that in M.
--
eva after short vowels and e, o (which then become shortened) becomes jjeva or jjěvva, Mg. yyěvva; other long vowels remain retained; jěvva or jeva stands after ansusvāra : for ex. ajjassa jjěvva (āryasya eva); idha jjeva (iha eva), muhě jjeva (mukhe eva), idõ jjěvva (ito eva); tadhā jeva (tathā eva), devi jeva (devi eva), tam jĕvva [jeva] (tam eva). Mg. mama yyěvva (mama eva), damsaantě yyěvva (darśayanto [=darśayann] eva), disanti yěvva (dṛśyamānaiva).
―
3. Sandhi Rules. In compositions similar vowels as in Sanskrit become contracted; so a a+a a become °*ā, i, i+i i become °*i; u ū+u ū become ū, whereby before consonant group or double consonants there must, happen shortening: jammantare (janmāntare). -- a ă + a dissimilar vowel gives rise, as in Sanskrit the guna of the second vowel: 'mahesi from mahā + isi (=maharși). One takes note of cases like M. gaïnda (gajendra), narinda; Mg. nalinda (narendra); 'mahinda (mahendra), niluppale (nilotpala) 'mahusava (mahotsava)
ii u ūremain unchanged before a dissimilar vowel as a Rule[10]: santiudaa (sāntyudaka), Uvvasiakkhara (Urvasyakṣara), M. rattiandhaa (rātryandhaka).
Udvṛtta-vowels i.e. such as remain left after the fall of consonants, do not admit of any further change: jaï (yadi), not je; ogaa (gaja), not gā!

Inside of sentences the final and the initial vowels usually remain
3. But jaadu (jayatu): see § 26, 1. paragraph.

Page #33

RICHARD SCHMIDT : A PRIMER OF ŚAURASENİ
131
unchanged, only the negative na (na) is an exception:oņatthi. Mg. nasti (năsti); ņāham (nāham); ņāgadā, M. ņāgaā (nāgatā); ņěcchadi, M. ņěcchaï, Mg. něścadi (necchati). Mark mamovari (mamopari), M. sahasāgaassa logatasya).
Ending nand m become anusvāra :'bhavam (bhavān), Oʻiam (iyam); mcan, however make a samdhi with the following vowel, when metri causa short syllable is requisite. Cf. S. 27, jěttiam ihantaram téttiam imāņam
4. Initial consonants remain (exceptions, naturally are nyt ś ) unchanged; only in a compound they are mostly treated as medials with the exception of the initials of verbal roots after prefixes with final vowels which are often retained : ajjaütto (āryaputrahy, but ākaņņido (ākasitaḥ); M. bhamcraüla (bhramarakula), āirna (äkirņa). Note that adha im (atha kim), * cim una (kim punar) etc., further dāva (tāvat) and de (te) after vowels and anusvāra : cittha dāva; Mg. cistha dāva (tiştha tāvat); čuvam dāva (evam tāvat); sutthu de (şuşthu te), sāadam de (svāgatam te); Mg. tado de (tatas te), čuvam de levam te).
5. Medial consonants. k, 9. c. j. d. y fall out intervocally : O*loa, anurāa, ņārāa, "gaa, omaana, ovioa (loka, anurāga, nārāca, gaja, madana, viyoga); pļ: kadaa (kataka); tod: gadā (gatā), in M. it falls out: raaa (rajata); pand b become v: pāveņa (pāpena), M. āava (ātapa); okalevara (kalebara), "kavala (kabala), 'savara (sabara); Mg. śivila (sibira).
kh, gh, dh, ph, bh become h: 'muha (mukha), 'meha (megha), 'mahu (madhu), 'muttahala (muktāphala), (1 1Joahiņava (abhinava); th, th become resp. dh, dh : kadhiņa (kathina), Mammadha (Mammatha), but M. Vammaha (so also Mg. in the verses with initial changes).
6. Consonantal groups are separated by means of a participating vowels (see & 2) or it is assimilated (customarily) the first consonant with the second. Thereby there appears initially barring quite a handful of exceptions (for ex. mhi = asm), only a single consonant: pia (priya), and not ppia. Permitted groups are : 1) double consonants : oratta (rakta); 2) a consonant + its aspirate : 'kapphala (katphala); 3) nasal + the consonants of the corresponding class : 'accanta (atyanta); 4) the groups ņh, mh, lh : "cinha (cihna), Obamhaņa (brāhmana), 'palhatta (prahlasta). In Mg. is yet found (partly doubtful) the group yyh, śk (?), śkh [?], śc, st [?], sth [?], sth [?], sk, skh, st, st, sp, sph,ḥk (consequently jihvāmūliya+k): ņiyyhala (nirjhara), suśka beside śuska (suşka),
4. becomes j : jadhā, M. jaha (yathā); in Mg. it is retained : yadhā. 5. Mg. keeps ś, and changes ș, s into ś.

Page #34

132
JAIN JOURNAL : VOL XXX No. 4, April 1996
dhanuskhanda beside dhanuso (dhanus"), puscadi (prcchati), kasta (?) (kaşta), kosthāgāla [?] (kosthāgāra), hasta (?) (hasta), isthia [?] (itthia = strikā), maskali (maskarin), paskhaladi (praskhalati), kasta (kasta), hasti (hastin), saspa (saşpa), ņisphala (nişphala), pahka (pakşa) cp. $81 !
kt becomes tt : ratta (rakta); kth becomes tth : rittha (riktha): kp>pp : vappadi, M. vappai (vākpati); gdh>ddh: "duddha (dugdha); gbh>bbh: opabbhāra (prāgbhāra); tk>kk: chakka (şațka); șc>cc : chaccarana (şaţcaraņa); ţt>tt: chattala (şaţtala); tp>pp: 'chappaa (şațpada); tph>pph: "kapphala (katphala); dg>gg: 'chaggunaa (şadgunaka); dj>i: sajja (şadja); (dd>dd: chaddisim (şaờdiśam)]; dbh>bbh: chabbhua (şadbhuja); (du>uv : chavvisam (saqvimšati)]; tk>kk: M. ukkanthā (utkanthā); balakkāra, Mg. balakkāla (balātkāra); tkhkkh: M. ukkhaas (utkhāta); tp>pp: 'uppala (utpala); tph> pph : 'upphulla (utphulla); Mg. upphala (utphāla); dg>gg : omoggara (mudgara); dgh>ggh: ugghāda, M. ugghāa (udghāta); db>bb: ubbandhia (udbandhya); M. [12]bubbua (budbuda); dbh>bbh: 'ubbhaça (udbhata); pott : 'sutta (supta); bj>i : "khujja? (kubja); bd>dd : 'sadda (sabda); bdh>ddh :'laddha (labdha).
Double consonants mostly remain unchanged when it is the question of a final sound of the same class, for ex. datta in personal names. In Mg. tb>st: pasta (patta), bhastālikā (bhattārikā). Stenzler writes in his edition of Mşcchakațikā śť: for ex. bhastaka (bhattaka). According to Pischel (8271) st is to be written.
Nasal + consonant remains preserved : 'sanhalā (srnkhalā), Omañjari, kantha, "khanda, 'anta, jambū. But it is written dimmuha (dinmukha), 'pamti (pankti), 'Vimjha (Vindhya), osamjhā (sandhyā), because in these cases the original nasal comes out of its class.-ñj in Mg. becomes ññ: aññali (añjali), Dhanaññaa (Dhananjaya).
s nassimilate themselves with the foregoing consonants : M. rugga (rugņa); "aggi (agni), 'uvvigga (udvigna), 'savatti (sapatni). --jñ>nn (Mg. ññ: avaññā (avajñā), initially ņ:'pannā (prajña), 'anavedi (ajñapayati), km>pp : Ruppiņi (Rukmiņā); gm>gg: jugga (yugma).-ātman has N. sg. o'appā, otherwise the stem runs 'attā,-nmmm: Ojamma (janma); mnoņņ: Pajjunna (Pradyumna). 7. Groups with semivowels.
a) ky>kk: Caņakka (Cāņakya); khy>kkh: osokkha (saukhya); gy>gg: Ojõgga (yogya); cy>cc:vuccadi (ucyate); M. muccaï (mucyate);jy>ii:jujjadi, M. jujjai (yujyate); ty>țț:ņațţaa (nāțyaka), M. tuttaï (trutyati); dy>ļļ : okudda (kudya); dhy>ddh : 'addha (adhya); ny>rņ: 'purna (punya);
ya.
6. With shortening of long ā, when the accent rests on the first syllable. 7. Aspiration (for non-aspirate sounds in Sanskrit) is often found in Prakrit.

Page #35

RICHARD SCHMIDT: A PRIMER OF SAURASENİ
Mg. sahilanna (sahiranya) beside abamhañña (abrāhmaṇya), puñña (punya); ty>cc:accanta (atyanta); thy>cch: 'nevaccha (nepathya); dy>jj, in Mg. yy:vějja (vaidya), but Mg. ayya (adya); dhy>jjh, Mg. yyh: uvajjhaa (upādhyāya), Mg. mayyhanna (madhyāhna); ny>nn : **anna (anya); py>[13] pp: kuppadi, M. kuppaï (kupyati), bhy>bbh: M. abbhantara (abhyantara); 'abbhuvavanna (abhyupapanna); my>mm: kilammadi, M. kilammai (klāmyati); yy>jj: osejjā (sayyā); ry>jj : 'ajja, but Mg. ayya (arya); ly>ll: "mulla (mulya); vy>vv: M. kavva (kāvya); hodavva (bhavitavya).
133
b) rkkk: takkemi (tarkayāmi), M. akka (arka); kr>kk: °cakka (cakra); rkh>kkh: 'mukkha (mürkha); rg>gg: niggama (nirgama); gr>gg: agga (agra), M. (g)gaha (graha); rgh>ggh: 'digghia (dirghikā), M. nigghina (nirghma); ghr>ggh: vaggha (vyāghra), M. agghāi (aghrāti); rc>cc: kucca (kūrca), M. accā (arca); rch>cch: muccha (mūrchā); chr>cch: samucchida (samucchrita); rj>jj; Mg. yy: dujjana (durjana), M. Ajjuna, but Mg. Ayyuna (Arjuna); jrjj: 'vajja (vajra); rjhjjh; Mg. yyh: nijjhara, Mg. niyyhala); rn>nn: 'kanna (karna), rt and tr3>tt: °*kitti (kirti), 'kalatta (kalatra); rth>tth: 'attha (artha), but in Mg. st: tista (tirtha); rd and dr>dd: 'duddiņa (durdina), 'bhadda (bhadra); rdh and dhr>ddh : **addha (ardha), *giddha (grdhra); rp and pr>pp: °dappa (darpa), 'appia (apriya); rband br>bb: nibbandha (nirbandha), M. dobballa (daurbalya); abbamhanna (abrāhmaṇya), Mg. (b)bamhaṇa (brāhmaṇa); rbh and bhr>bbh: dubbhejja (durbhedya), M. gabbha (garbha); paribbhamadi, M. °maï (paribhramati); rm and mrmm: dummanussa (durmanusya), M. dhamma (dharma); 'Dhummakkha (Dhumrākṣa); rlll: 'dullaha (durlabha); ru and ur>vu: 'puvva (pūrva) 'parivvajaa (parivrājaka).
c) lk and kbkk: 'vakkala (valkala), vikkava (viklaba); lg>gg: °phagguna (phalguna); lp and pl [14]pp: kappa (kalpa), vippava (viplava); lph>pph: guppha (gulpha); lb>bb: kibbisa (kilbiṣa); lbh>bbh: pagabbhadi (pragalbhate); Imand ml>mm: 'gumma (gulma), M. vammia (valmika); (m)měccha (mleccha); lu>ll: gallakka (galvarka), M. pallala (palvala).
d) ku>kk: pakka or pikka (pakva); jv>jj: pajjaladi, M. °laï (prajvalati); nu>nn: Kanna (Kanva), M. kinna (kinva); tu>tt: 'satta (sattva), Mg. cattāli (catvāri); dv>dd:'(d)diuna (dviguna), or vv, particularly with ud: for ex. 'uvvigga (udvigna); dhv>ddh: uddhattha (uddhvasta), but sajjhasa (sādhvasa); nu>nn: annesanā (anveṣaṇā), Mg. mannantala
8.
tr in tth is apparent only in adverbs in tra and its compounds: annattha (anyatra), atthabhavam (atrabhavaan), M. kattha (kutra); jattha (yatra), O*tattha (tatra), tattha bhavam (tatrabhavan); M. savvattha (sarvatra). Cp. Pischel § 293.- Sometimes it undergoes cerebralisation: for ex. bhaṭṭā (bhartā, "lord", but 0*bhatta "husband"; "vaṭṭadi, M. vaṭṭai (vartate); Ovicchaḍḍa (viccharda), vaḍdhadi, M. vaḍdhaï (vardhate).

Page #36

134
JAIN JOURNAL: VOL-XXX No. 4, April 1996
(manvantara). Take note evam nedam (evam nu etat), 'kim nedam (kim nv etat).
8. Groups with sibilants
a) śc>cch, unchanged in Mg. "accharia, accharia, Mg. aścalia (aścarya); if an s stands at the end of the members of a compound then śc is changed cc: niccala [but Mg. niscala] (niścala); duccarida, M. duccaria (duścarita); tavaccarana (tapascaraṇa).-sch>cch: M. nicchallia (nischallita); śn>nh:panha (prasna), M. aṇhaï (aśnāti), sm>mh: Kamhira, M. Kamhāra (Kaśmira); in Mg. unchanged; rasmi becomes rassi; initially śm>m: masāṇa, Mg. maśāņa (śmaśāna).-śy, śr, śt, śv>ss, Mg. śś: avassam (avaśyam); Mg. viņaśśadu (vinaśyatu); missa, Mg. missa, M. misa (misra); (s)sāhaniā (slāghaniyā); Vissavasu, Mg. Vissavasu (Viśvāvasu).10
b) şk and ṣkh>usually kkh : põkkhara (puṣkara), M. vikkhambha (viskambha), the aspiration can, however, also be left: Kikkindha (Kişkindha), °dukkara (duṣkara). In Mg. şk is changed to sk, skh to skh11: suska (suska), dhanuskhanda (dhamus),-ks>cch or kkh: 'acchi (akṣi) but also akkhi; riccha beside orikkha (ṛkṣa); (k)khattia (kṣatriya), sikkhida, M. sikkhia (sikṣita), dakkhina (dakṣina). [15] kṣaṇa "twinkling of eye" becomes 'khana, in the sense of a festival (c)chana. In Mg. ks sk: laskase (rākṣasah), daske (dakṣaḥ). 12-kon and kṣm>nh, mh resp.: tinha [beside °*tikkha] (tikṣṇa), °pamhala (pakṣmala). lakṣmi always becomes lacchi. șt and sth>tth: 'diṭṭhi (dṛṣṭi), suṭṭhu (susṭhu). Mg. changes st and sth into st13: kasta (kaṣṭa), susțu (susthu).-sn>nh, Mg. sn: unha (usna), Mg. Visnu' (Visnu); sp and sph>pph Mg. sp resp. sph150puppha (puspa), Mg. saspa (saspa); nipphala, Mg. nisphala (niṣphala), bāṣpa “tear” becomes obāha, in the sense of "smoke, vapour" bappha.-şm>mh: "gimha (grişma); sy and su>ss, Mg. ss: 'manussa, Mg. manussa (manuṣya); parissaadi (pariṣvajate).
c) sk and skh>kkh: (k)khambha (skambha), 'parikkhalanta
9. In M. a single s possibly with the lengthening of the foregoing vowel : ņāsaï (nasyati), vesā (veśyā); Ś. has vesajana and vessāo-vesyājana). 10. A single s with lengthening for ex. an nisasadi, M. nisasaï, Mg. nisaśadu
(niḥsvasiti, niḥsvasitu).
11. According to Namisādhu (in Rudrața, Kāvyālaṇkāra II, 12) sk, skh resp. so suske (suska), dhanuskhanda (dhanus).
12. Hemacandra and Namisādhu reduce this to preks and acaks (pěkadi, acaskadi = prekṣate, acaste and instructs besides the passage of ks (intially) to hk lahkase, pahka (rākṣasaḥ, pakṣa).
13. Namisādhu writes st and sth resp: kasta, kosṭhāgarla (kaṣṭa, koṣṭhāgāra). 14. Namisādhu teaches Visņu.
15. According to Namisādhu sp and sph.

Page #37

RICHARD SCHMIDT: A PRIMER OF SAURASENI
135
(pariskhalan); in Mg. it is retained16: maskali (maskarin), paskhaladi (praskhalati), and in compounds usually the aspiration remains : tirakkāra (tiraskāra), M. ṇamakkāra (namaskāra).—st and sth>tth sometimes tṭh: 'atthi (asti), °avatthā (avasthā), however 'aṭṭhi (asthi); utthehi beside utthehi, utthedu (uttiṣṭha, uttiṣṭhatu), M. thia, thia, Ś. thida, thida; utthida, M. uṭṭhia, utthia (sthita, utthita); patthida, M. patthia (prasthita); °ṭhāṇa, M. has also thāņa (sthāna), ṭhidi, thidi, M. thii, thii (sthiti). In Mg. st is preserved" and sth>st: hasti (hastin), uvastida (upasthita).-ts>cch, Mg. śc: vacchala, Mg. vascala (vatsala); or it becomes [as also originally t+ś]>ss, resp. s with lengthening of the foregoing vowel, so in compositions when t stands at the end of a syllable; ussavedi (ut-śrāpayati); ūsasida, M. ūsasia (ut-śvasita); pajjussua (paryutsuka), samūsua (samutsuka), usava (utsava), 'mahūsava (mahotsava), vasantūsava (vasantotsava); Mg. ūsasadu (utsvasatu). sn>[16] nh in case the group is not separated by a semivowel: nhada, M. nhaa (snāta)18; sp and sph>pph, in Mg. they are preserved : Bihapphadi (Bṛhaspati), papphuradi, M. °raï (prasphurati), Mg. Buhaspadi, sphulanti (Bṛhaspati, spuranti)1o. Very frequent is 'phamsa (sparsa); loss of aspiration: 'parõppara (paraspara), M. avarŏppara (aparaspara).- ps>cch: accharā (apsarāḥ), jugucchedi, M. juucchaï (jugups[yati; sm>mh: 'amhe (asme); the pronominal ending smin becomes 'ssim, Mg. śśim: edassim, M. eassim, Mg. edaśśim (estasmin).-sy, sr, sv>ss,-Mg. ss: orahassa (rahasya), Mg. tassa (tasya); 'sahassa, Mg. sahassa (sahasra); tavassi, Mg. tavassi (tapasvin), M. Sarassai (Sarasvati).
9. hk, hkh, hp, hph>kk, kkh, pp, pph: antakkaraṇa (antaḥkaraṇa), "dukkha, (duḥkha), antappada (antaḥpāta).
hs, hs, hs>ss [Mg. s] or, with extension of the foregoing vowel, s [Mg. s]: cadussāla (catuḥśāla), Dussanta, Mg. Dussanta (Duḥṣmanta), cadussamudda (catuḥsamudra), Mg. nissalida (niḥsṛta); M. ṇīsanka (nihsanka); nisaha beside 'nissaha (niḥsaha).
hn, hn, hm, hbnh, mh, lh: 'avaraṇha (aparāhna), 'genhadi (gṛhṇāti); cinha (cihna); "bamhaṇa (brāhmaṇa); 'palhatta (prahlasta).
hy jjh, Mg. yh: sajjha, Mg. sayha (sahya); hv>bbh : gabbhara (gahvara); but jihā, Mg. yihā (jihvā).
10. Transition of dentals to cerebrals: maṭṭiā (mṛttikā), °*vuḍḍha (vṛddha), °tālaveṇṭa (tālavṛnta), 'ganthi (granthi), Mg. ganthiscedaa (granthicchedaka). Cp. also the remark in § 7, b.
16. According to Namisādhu śk and skh.
17. According to Namisādhu st>st.
18. For Mg. there run manuscript forms nhāāmi (snāmi), nhāde (snātaḥ).
19. After Namisādhu śp and sph.

Page #38

136
JAIN JOURNAL : VOL-XXX No. 4, April 1996
For groups of more than two consonants the same rules are valid : 'maccha, Mg. masca (matsya), 'aggha (arghya). Only two consonants written, even when three are required there; for ex. satta not sattta (from sattva). --jyotsnā> jönhā.
(17] ca "and" after anusvāra only, besides shortened to a in all the dialects.
punar "again” becomes "puņo, in the sense of “but, however* "* una.
y in Mg. before eva is indicated before a short or a shortend vowel doubled : mama yyeva, ido yyeva (mamaiva, ita eva).
ris preserved in cases like durāgada (durāgata), onirantara (nt), °punarutta (punarukta), against antakkaraña (antahkarana), antaggada (antargata), punarnava (punarnava). It is to be noted "anteura, anteuria (antahpura, antahpurikā). vis represented by vialiva), 'vuccadi (ucyate); however utta (ukta);
MORPHOLOGY
A. Noun 11. The consonantal declension is wholly upto the rest sporadically for ex. M. viaā (vipadā), disi disi (disi disi), Mg. nisi (nisi) has been lost. The consonantal stems cast off the final consonants and now were handled vocalic ones or oftener the stem was extended through a (for masculines and neuters), through ā or i (with feminines) resp. – The dual has been lost and was replaced (with noun and verb) by the plural). The dative were paraphrased with 'attham (artham) and *ņimittam (nimittam); in M. it often happens with a stems : maraņāa, hāsāa, vaņāa and so foth (maraņāya, hāsāya, vanāya).
a) Masculine and neuters in a : putta (putra), phala : Sing. N. oputto, Mg. putte; A. "puttam; I. 'putteņa [D.M., Mg. puttaa] Abl. *puttādozo; G. °puttassa?; L."putte?; [18] V. "putta23.- Plur. N. "puttā; A. "putte24; I. "puttehim; G. "puttānam25; L. 'puttesume; V. - N27.
20. M. puttão; metri causa also the ending āu: sisāu (sirsāt), and not rarely also forms in ā : vasā (vaśāt), balā (balāt), or also in hi : mūlāhi (
mūlāt), seldom with himto : hiaāhimto (hrdayāt). 21. Mg. puttaśśa; besides the ending ha "salīlāha (sarirasya) 22. M. also with mmi : jovvanammi (yauvane); Mg. also in him: kulahim (kule),
in verses also with mmi : kuvammi (kūpe). 23. Also often with lengthened final sounds : 'puttā; Mg. also uses the N. as
V.: bhāve, cede, lautte (bhāva, ceta, rajaputra). 24. M. has also sometimes forms in ā: gunā, dosā (gunan, daşān). 25. M. much oftener without nasal. 26. In M. puttesu is prevalent; Mg. putteśum 27. For Mg. the ending ho is also permitted : puttaho.

Page #39

RICHARD SCHMIDT: A PRIMER OF SAURASENİ
The neuter is just so, only N.A. Sing. "phalam, Plur. °*phalaim [Mārkaṇḍeya for S. also permits phalāņi as well], metrically M. has also phalai, phalai.
3
b) Feminine in ā : mālā (mālā). Sing. N. °*mālā. A. °*mālam; I. **mālāe28, Abl. 'mālādo, malae29; G.L. °*mäläe, V. male30. Plur. N. A.V. **mālāo31; I. °*mālāhim32; [Abl. M. mālāhimto] G. "mālāṇam33; L. °*mālāsum).
12. Masculine and neuter in i and u: aggi (agni), dahi (dadhi) : Sing.N. **aggi34; A. "aggim; I. °*aggiņā. [Abl. aggido]35; G. "aggino; L. "aggimmi; V. "*aggi.- plur N. °aggino, aggio36 A. °*aggino; I. "aggihim3 [Abl. M. aggihimto] G. "agginam38; L. "* aggisu(m).
Just so the neuter, but N. A. Sing. "dahim, Plur. dahiim.
Example for Masculine in ɩ : vāu (vāyu), Sing. N.°* vāū; A, “vāuṇ; I. °*vāuṇā; [Abl. vāudo]; G. °*vāun39; L. vāummi; V. °*vāu. Plu. N.°"*vāuņo, vāao40; A. vāuno; I. *vāūhim; G. °*vāūnam2; L. °*vāūsu(m).
137
[19] Just so the neuter, only N.A. Sing. "mahu[m] (madhu), Plur. 0* mahūim.
13. The feminine in í and u are as good as vanished and those with i and ū have fallen together. These are declined like feminine stems in a.. Full-fledged paradigm here (as elsewhere) are not to be envisaged, consequently therefore the individual cases of various stems will be given gadī, vahū, (gati, vadhu): Sing. N. 'gadt, vahu; A. ‘gadim, vahum; I. 'gadie: Abl. Sacido (Śacyah), Mg. naalido (nagaryah); G.
28. M. has metri causa oftener mālāi.
29. M. malão.
30. hala ("O lady friend" is often in conjuction with the V. of a proper noun: hala Saüntale, hala Anusue (sakhi Sakuntale, sakhi Anusuye).
31. Rarely is N.A. *mālā, N. metri causa mālāu in M.
32. M. has also mālāhi.
33. M. mālāņa.
34. According to Märkandeya; in Hultzsch p.716. In M. vahiu (udadheḥ). 35. M. beside it also aggissa.
36. M. has also aggi.
37. M. also aggihi.
38. M. aggiņa also.
39. M. vāussa; Mg. metrically Vissavaśussa (Viśvāvasoh).
40. M. vāū also.
41. M. also vāūhi.
42. Also M. vāūna; Mg. metrically bahuna (bāhvoḥ).
43. M. gai.
44. M. gaie. Note Ś diṭṭhiā (diṣtyā). M. metrically devia (devyā).

Page #40

JAIN JOURNAL: VOL-XXX No. 4, April 1996
*gadie, vahue; L. rattimmi (rātrau)45; V. °*sahi (sakhi).-Plur. N. A.V.: *gadio, °*vahuo; I. °* sahihim" (sakhibhis); G. "kāmiņīņam; L. devisum
0*
138
14. Stems in ra) Agentive noun: bhatta "husband", bhaṭṭā "lord" (bhartṛ), Sing. N. °bhattā, bhaṭṭā; A. °bhattaram18, bhaṭṭāram; I. bhaṭṭinā; G. bhattuno, bhaṭṭino; L. bhattare; V. bhatta, bhaṭṭā.-Plur. N. bhattaro.
b) Relation words pidu (pitṛ). Sing. N. 'pidā49; A. pidaram50; I. 'piduna1; G. 'piduno52. Plur. A. pidaro.
mātṛ has in Sing N. 'mādā, A. madaram54, G. madãe, V. māde55. Of the stems in e, o, au there comes here only nau to be considered with N. ṇāvā and A. nāvam (both are Ś forms).
15. The stems in at, mat, and vat cross over to the a-declension with the employment of the strong stems, ending also in anta, manta and vanta, excepting bhagavat and bhavat [20] "thou" in Ś. Mg. which go in the follwing way: Sing. N. 'bhaavam, bhavam; A. bhaavantam, bhavantam; I. bhaavadā, bhavadā; G. 'bhaavado, bhavado; V. bhaavam.--Plur. N.V. bhaavanto.
The rest of the stems: Sing. N. Mascul mahanto karĕnto56, cittavanto, (mahān, kurvan, cittavan); A mahantam, jāṇantam57, I. calantena (calatā), bhuttavantena (bhuktavatā)58; G. cintaantassa (cintayatah), mahantassa (mahatah), Himavantassa (vatah)59; L. mahante; [V. calantal1.-Plur. N. sikkhantā (sikṣantah)62: [A. sikkhante
45. Besides the formo" is like the I.G.: M. paavie (padavyām), Mg. dhalanie (dharanyām), S. masāṇavidhie, (śmaśānavithyām).
46. M. metrically taruniu (tarunyaḥ).
47. M. without nasal as well.
48. Mg. bhattālam
49. M. pia.
50. M. piaram, Mg. pidalam
51. M. piuņā.
52. M. piuno.
53. M. māā.
54. M. maaram.
55. M. mãe.
56. M. kunanto; Mg. puścamde, mahante, colaante (pṛcchan, mahān, corayan).
57. M. Mg. ebenso; note Mg. mālantam-mārayantam.
58. M. piantena (pibata); Mg. gascantena (gacchatā).
59. M. jānantassa (jänatah); Mg. alihantassa (arhatah).
60. M. most of all forms with mmi: ruantammi (rudati); Hanumantammibeside Hanumante (Hanūmati).
61. not recorded in Ś: M. sasanta (svasan), Mg. alihanta (arhan).
62. M. jāṇantā (jānantah), Mg. sasanta (svasantah).

Page #41

RICHARD SCHMIDT : A PRIMER OF ŚAURASENÍ
139
= śikṣata]63; I. 'gacchantehim (gacchadbhiḥ)64; G. pěkkhantānam (prekşamāņānām); L. vaţtantesu (vartamāneșu).
The neuter has anomalous N. Sing. M. bhaṇantam, S. disantam (bhanat, drśyamānam); but Mg. dahante (dahat). A. as in the masculine : mahantam (mahat), M. santam (sat); likewise in rest of the cases (Plur. N.A. not recorded).
16. Stems in n:rāa, Mg. laa (rājan). Sing. N. Prāā66; A. Praanam67; 1. Orannāle; G. Oranno, orāiņo69, V. raam?'.-Plur. N.A. 'rāāno. The remaining cases are not documented.
At the end of a compound rājan is handled in the following manner : N. mahārāo, A. mahārāam, I. mahārāena, G. mahārāassa (beside Kalingaranno), Mg. mahālāaśća; V. mahārāa; Plur. N. mahārāā
ātman forms Sing. N. 'attā??; A. 'attānaam (=ātmānakam)73; I. appaņā?4; G. 'attaños; IV.M. appa).
[21] Words ending in varman have vammā or vammo: Vijaavammā, Vijaavammam (Vijayavarmā, 'varman), Cittavammo (Citravarmā), Miankavammo, 'vammassa (Mrgārkavarmā, 'varmanah).
The neuters ending in- an go usually as those in a: pěmma, nāma (preman, 'nāman) : Sing. N.A. opěmmam, ņāmam; I. opěmmeņa, oņāmeņa; (Abl. jammado=janmataḥ); G. opemmassa; L. opěmme76. - Plur. I. dāmehim (dāmabhth); L. kammesu, Mg. kammeću.
17. Stems ending in in, min, vin : Sing. N. 'hatthi, Mg. hasti (hasti); tavassi, Mg. tavassi (tapasvī); A. kañcuim(kañcukinamy, but elsewhere piaāriņam, uaāriņam (priyakāriņam, upakāriņamy; I. sāmiņā (svāminā); G. virohiņo (virodhinah)79; V. kañcui (kañcukin). Plur. N. pakkhino (pakṣiṇaḥ), kañcuino (kañcukinaho; (I. Mg. vamdihim=vandibhiḥ.-G.M. barahīņa=barhiņām; Mg. śāmīņam=svāminām L. sāmisu (svāmişu)81. V-N: adhivāsiņo, Mg. vamdiņo Pvāsinaḥ, vandinaḥ).
63. M. unnamante (unnamatah). 64. Mg. pavisantehim (pravisadbhih). 65. M. also without nasal :
cintantāna (cintayatam); Mg.
alihantānam (arhatām). 66. Mg. laa. 67. Mg. läānam. 68. Mg. laññā. 69. Mg. lanño and laino. 70. Mg. lāam. 71. Mg. läāno. 72. M. appā. 73. M. appanam 74. M. also appaņeņa.
75. M. appaņobeside attaņo; Mg. also
attamaaśća. M. yet also has the forms appano, attano (ātmā). G. appāņaassa, L. appāne; Plur. N.
appā (ātmānah). 76. M. also pěmmammi Plur. N.
pěmmāim, G. pěmmāņam. 77. M. hatthim (hastinam). 78. Mg. śāmiņā. M. sasiņā (sasinā). 79. M. sasiņo, Mg. śāmiņo. 80. M phañiņo, sankiño etc.
(phaņinaḥ, sankinah). 81. M. panaisu (pranayişu).

Page #42

140
JAIN JOURNAL: VOL-XXX No. 4, April 1996
18. Stems ending in as. N. Pururavā (vās)82; A. Pururavasam; I. Pururaveņa; G. Pururavassa; L. Pururavast. The inflection is customarily after the a-declension; N.M. vimano (vimanāḥ), neuter N.M. dummanam (durmanah); A. neuter uram, jasam, naham, siram (uraḥ, yasah, nabhaḥ, sirah4; I. M. vacchena (vakṣasā); °sirena (sirasā); M. maṇena (manasā)85; G. tamassa, rajassa (tamasaḥ, rajasah); L.M. ure beside urammi (urasi), nahe and nahammi (nabhasi), sire, mane (sirasi, manasi), sarammi (sarasi), S. sotte (srotasi). Plur. N.M. sară (sarāmsi); I.M. sarehi, sirehi and sirehim (sarobhih, sirobhih); G.M. sarāṇa (sarasām)).-Feminine forms are: N.M. vimana vva(vimanā eva), Ś. pajjussuamaṇā (paryutsukamanāḥ); AM vimanam; I. taggadamaṇāe=tadgatamanaskayā; M. vimanai (metrically for vimaṇāe); Plur. N.M. gaavaão-gatavayaskāḥ; A. sumanão-sumanasaḥ; vimaṇāhim, [22] mangalamaṇāhi; G.M. gaavaāṇa, (gatavayaskānām).—apsaras runs : N. acchara, Plur. N. accharão, I.
L.M.
accharāhim.
19. Stems ending in is, us. Apart from the rest like Auso (Ayuṣaḥ) they figure vocalical: N. dihāūs dirghāyuḥ), neuter M. havim, dhanum (havih, dhanuḥ); A. dihāum (dirghāyuṣam), dhanum (dhanuh); I. dihāuṇā (dirghāyuṣā)97.
B. Comparison
20. The suffixes are the same as in Sanskrit: mahattara, Mg. mahattala, M. tikkhaara (tikṣṇatara); piadama, M. piaama (priyatama); kaniasi (kaniyasi), kaniṭṭha (kaniṣṭha); bhūo (bhūyas) beside bahudara (tara); jěṭṭha (jyeṣṭha). Sometimes the positive stays in the sense of the comparative: tatto vi... pia tti (tvatto 'pi... priya iti), "himself dearer than thou"; M. ovaaṇāhi vi lahuam (avapatanad api laghukam) "quicker than the plunge".
82. Similarly in Mg. samassasidamanā = samāśvastamanāḥ.
83. Mg. silasi (śirasi).
84. Mg. manam (manas).
85. Abl. M. sirāhi (sirasah).
86. Besides (of the extended stems) dihauso, M. adiharauso (adirghāyuḥ; dihara stands for *diraha-dirgha, Pischel § 354).
87. L.M. dhanuhe (dhanuși). asis forms the stem āsisä-whence I. äsisae, Plur. āsīsāhim and asisā : A. äsisam, G. äsisae, Plur. āsisānam. The stem vāc is extended to vāca and therefore runs: N.M. vāā; A. M. vāam. I. O*väãe, also M. vāāt; G. Mg. vāãe; L.M. vääi; Plur. N.M. vää and vāão; L.M. vääsu.

Page #43

RICHARD SCHMIDT : A PRIMER OF SAURASENI
141
C. Pronoun 21. The pronoun of the first and second person.
Singular N. 'ahamas A. "mamo 1. Omae Abl. matto, mamādooi G. Omama, Omaha, Omega L. mai93
o* tumam O* ", te, de o'tae, ''tues
tatto96
otuha "[te], "de97 "tai, tul8
Plural
[23]
N.°*amhe99 A. no 100 I. 'amhehim101
otumhel03 tumhe tumhehim104
Abl.
G. 'amhānam102, ņo L. amhesu
*tumhāṇam, 'vol05
22. Third person pronoun
Singular Masculine
Feminine N. oso106
0'sā
neuter
oʻtam
88. Mg. hage, hagge; M. Mg. sometimes ham after vowels 89. M. also mamam, Mg. mamă. 90. Mg. also seldom mai and me. 91. According to Mārkandeya ix, 96, by Hultzsch p.717. 92. M. rarely mama; otherwise yet maham, majjhalm). 93. M. mamammi. 94. M. also tam and rarely tum 95. M. tai, tui, tumae, tumāe, tumãi, tume also. 96. M. tumāhi (mto), tumão; S. also tuvatto. 97. M. also tuham, tujjham), tumham, tumma, tu; Mg. tava instead of tuha; te
in Ś. Mg. is apparently false for de. 98. M. also tuvi, tumammi, tume. 99. In Mg. it is better to write asme, beside hage and in Ś. vaam also. 100. M. ne, also amha; Mg. asme; $. amhe also. 101. M. also amhehi; Mg. asmehi 102. Mg. asmānam; M. also amhalm and 'mha. 103. Mg. better more rightly tusme or tuyhe. 104. M. tumhehi. 105. M. oftner tumha. 106. Mg.se.

Page #44

142
A. °* tam
I. °* tena107
G. °tassa, se108 L. tassim109
Masculine N. °eso116
A. 'edam117 I. 'edena118 G. edassa119
L. edassim120
118
N. 'ede122
A.
I. *edehim125
G. edāṇam126
L. edesu(m)128
Masculine N. °*te (de)112
A. °*te (de)
I. °*tehim113
** tāhim °tāṇam
G. °tāṇam114
tāṇam
L. °tesu115
tāsu
°tesu
[24] The stem etad on the whole runs likewise :
Singular
Feminine
JAIN JOURNAL: VOL-XXX No. 4, April 1996
0*tam
*tāe110
'taell
[tie, tia]
Plural
Feminine
0* tão
0* tão
esa121
'edam
*edãe
'edãe
*edãe
Plural
edão123

107. Abl. °tado (adverbial is used) and O*tā.
108. Mg. tassa, tāha and se.
109. M. tammi, Mg. tassim.
110. M. tie, tia.
edāim124
'edehim
edāṇam127 edāṇam
23. The relative is almost entirely treated like tad; for Mg. it is to be noted that y is kept maintained. N. ojo, jā, jam. A. jam129; I. jena130,
111. M. tissā, tie, tia.
112. *de after all other pronouns :
*ede de.
°*tam
°*teņa
o'tassa
tassim
Neuter
*edam
*edam
113. M. tehi resp. tahi.
114. M. also tāņa.
115. S. also tesum.
116. Mg. ese. Ś. frequently, Mg. sel
dom esa
117. M. eam.
neuter taim
taim
** tehim
118. M. eeņa; Ś. Mg. much oftener ediņā.
119. M. eassa, Mg. edassa and edäha. 120. M. eassim, eammi; Mg. edassim. 121. Mg. eśa.
122. M. ee.
123. M. eão.
124. M. eãi.
125. M. eehi(m). 126. eāņa.
127. M. eaṇā(m).
128. M. eesu(m).
129. Mg. has yad iscase (yadicchase). 130. M. also jenam.

Page #45

RICHARD SCHMIDT: A PRIMER OF SAURASENİ
jāe, jena; G. jassa131, jāe132, jassa; L. °jassim. Plur. N. °je, jāim; G. jāṇam133
The interrogative: Sing. N. 'ko, 'kā, okim; A. °kam, kim; I. °kena, kāe, keṇa; G. 'kassa134, kāe135, kassa; Abl. okisa136; L. kassim137, Plur. N. 'ke, kao138, kāim.
The following forms belong to the stems ida and ima :
Singular
Singular
iam140
imam
imae143
'imādo
imāe145
imassim(!)
Masculine
N. aam139
A. 'imam
I.
*iminā142
Abl."imado G. imassa144
L. imassim146
[25]
Masculine N. 0ime
A. ime
I.
imehim148
Abl.
G. imāņam149 L. imesu(m)150
134. Mg. käha.
135. M. kie.
Plural
Feminine
imão147
°imāṇam
136. Mg. kisa.
137. M. kammi, Mg. kassim °*kahim,
where.
138. Ś. also kā.
0* kattha whereto, whither,

Neuter
°idam141
Sporadic forms are M. ena, assa (asya), assim (asmin); 'anena (anena); amum (A. neuter.), M. amuņā, ami (Plur. N. Masc.); °*enam (A. Masc. and Fem.) and "nam (A. Masc; Fem. and Neut.)
The pronominal adjectives run like tad or etad. M. parāhimto (parasmāt); annassim (anyasmin), aņņe (anyān), savvāṇam (sarvāṇām) M. annanam (anyānām).
imam
*imiņā
131. Mg. yaśśa and gäha.
132. M. jia, jie, jissä.
133. M. has also I. jehi and L. jesum. 142. M. imena.
H
143. M. imie, imia. 144. Mg. imassa. 145. M. imie, imia. 146. M. imammi, Mg. imaśśim. 147. M. also imā, imiu. 148. M. imehi. 149. M. imāna.
139. In M. only once documented; 150. M. imesu.
lacking in Mg.
imassim
Neuter
imāim
imehim
imāṇam
140. Only in Ś. 141. Mg. only imam.
143

Page #46

144
JAIN JOURNAL : VOL-XXX No. 4, April 1996
D. The Numerals
24. Cardinals : 1. '*ēkka (eka) : "ěkko, °ěkkā, °ěkkam, I. °ěkkeņa, G. Mg. ekkāha, L ekkassim (Mg. ekkassim, M. ekkammi); Plur. 'N. 'ěkke.-°aņea (areka), aneaso (anekaśaḥ).
2. N.A. Oduve151, I. duvehim, G. duvenam, L. duvesu.
3. N.A. Masc. and Fem. tao (trayaḥ) Neutr. otiņņi (trīņi); [G. 'timnam;) L.M. tisulmy, metrically tisu.
4. N. cattāro, Fem. catasso! (should be cadasso); Neutr. 'cattāri, Mg. cattāli; G. I'cadunnam, M. caünham.-- In compounds : M. caürāņaña (caturānana), caüddisam (caturdisam); $. cadussālaa (catuḥsalaka); M. caümuha (caturmukha), coddaha (caturdasa).
(26) 5. pañca, I.M. pañcehim
6. cha. In compounds : 'chaggunala) (şadguna[ka]), M. chappaa (şatpada), ochammāsia (saņmāsika), S. chaccaraña (şaţcaraṇa).
7. 'satta, Mg. satta. — [8. attha; 9. nava;) 10. dasa, Mg. daśa, M. dasa or daha; I.M. dasahi, Mg. dasehim; G. Mg. daśāņam; L.M.
sasu - 11. M. eāraha: 12. M. bāraha: 13. M. teraha: 14. M. coddaha. caüddasa; 60. saţthi; 100. sada, M. saa, Mg. sada; 1000. 'sahassa, Mg. sahassa; 100000. Mg. laskam (lakşam).
Ordinals : 1. opadhama; 2. 'dudia, M. duia; 3. tadia, M. taïa; 4. 'caduttha, M. cáüttha; 5. "pañcama; 6. O'chatthal52; 7. osattama; 8. atthama; (9. navama;] 10. M. dasama.
E. Verb
25. The verb has lost very many forms, the a-stems there preponderate; the middle is narrowed down to the 1. Sing Present, apart from the participle and from sporadic forms; the dual is totally disappeared.
a) Example for the indicative of the 1st class Vvst: vațţa - (varta) : Sing. 1. 'vattāmi153, 2. vastasi154, 3. 'vattadi155; Plur. 1. Ovattāmo, 2. 'vațţadhal56, 3. vaţtanti. Middle : 1. Sing. vatte (2. vattase157,
151. M. frequently N.A. do for all the three genders and also at the beginning
of the compound : domuha (dvimukha), certainly for it oftener du: duuna (dviguņa). There form the neuterodonmiwhich is also combined with Masc.
and Fem; I. M. dohim), G.M. donhalmy, S. (Mg) dannam; L.M. dosulm). 152. Feminine 0.pancami, 0*chatthi; the rest in ā. 153. M. has sometimes i instead of ā: jāņimi (jānāmi), bhanimo (bhaņāmah). 154. Mg. has naturally si instead of si: gascasi (gacchasi). 155. M. vattai 156. M. vattaha 157. M. lajjase, jānase (jānise).

Page #47

RICHARD SCHMIDT: A PRIMER OF SAURASENĪ
3. vaṭṭade158; 3. Plur. vaṭṭante]. For the first Sing. frequent examples are jane and 'na āne (jāne resp. na jāne) just as 'manne (manye); sporadically lahe (labhe), icchel59-vardhate naturally becomes vadḍhadi (M. vaḍdhai), vardhāmahe vaḍdhāmo, prekṣadhve pekkhadha160, nayatha nedha.
b) Optative: 1. bhaveam and bhave, pahave, laheam and lahe, jiveam (bhaveyam, prabhaveyam, labheyam, jiveyam)161; 2. M. hasějjahi, karějjāsu and kuņijjāsu galijjāsu (haseḥ, kuryah, [27] galeḥ)3. "bhave, purae, uddhare (bhavet, pūrayet, uddharet)162- 3. Plur. bhave (bhaveyuh). Notice M. hojja (bhāyāt).
c) Imperative : 2. Person 'karesu (kuru), āṇesu (anaya), 'bhuñjasu (bhungdhi), kadhesu (kathaya); Mg. lahkasu (rakṣa), āgaścesu (agaccha), desu (dehi), dhālesu (dharaya).- Middle forms like pěkkhassa (prekṣasva) are to be considered as Sanskriticisms and to be emended.
Present stems with a short final, as a rule are formed as in Sanskrit: "bhana (beside 'bhanahi), gaccha, Mg. gasca; pekkha, M. pěccha, Mg. peska (prekṣasva); "hasa, Mg. hasa (hasa).-With long finals there comes in hi: 'karehi. Mg. kalehi (kuru); mantehi (mantraya), kadhehi (kathaya), siḍhilehi (sithilaya), jālehi (jvālaya); Mg. malehi (māraya), ghosehi (ghosaya); "hohi (bhava)
The 3. Sing ends in 'du, M. u: 'bhodu, M. hou (bhavatu); pasidadu (prasidatu), kadhedu (kathayatu); Mg. muñcadu, nisidadu (muñcatu, niṣidatu); M. maraü, deu, paaṭṭaü (maratu, dayatu, pravartatu, i.e. mriyatām, dadātu, pravartatām).
The 1. Plur. has mha ending: gacchamha, uvavisamha, uvasappamha pekkhamha, karemha, nivedĕmha, 'kilěmha, homha (gacchāma, upavisāma, upasarpāma, prekṣāmahai, karavāma, nivedayāma, krīḍāma, bhavāma); M. abbhatthemha, (abhyarthayama); Mg. anneśamha, pivamha, palāamha, kalĕmha (anveṣayāma, pibāma, palāyāmahai, karavāma).
145
2. Plur. parittaadha [Mg. pali], samassasadha, avanedha, hodha (paritrayadhvam, samāśvasita, apanayata, bhavata); M. ṇamaha, rañjeha, raeha (namatha, rañjayata, racayata); Mg. ośaladha (apasarata), suṇādha (smuta), māledha (mārayata).
3. Plur pasidantu [Mg. pasi], pěkkhantu, hontu (prasidantu, prekṣantām, bhavantu); M. nandantu, vilihantu (nandantu, vilikhantu).
d) of the Imperfect only 'asi, āsī (āsīt), however, used for all persons and (numbers) has been kept.
158. M. pěcchae (prekṣate), palambae (pralambate).
159. Mg. gãe (gāyāmi).
160. Mg. pěskadha.
161. M. kuppějja (kupyeyam).
162. M. jivějjā! (jivet), dharejjä (dhriyeta), viharějja (viharet); Mg. müse (muṣet), khayye (*khādyet = khādet).

Page #48

146
Remarks on the isolated classes
26. Verbs of the first class in i, u are as in [28] Sanskrit: jaadi, M. jaaïljayati); jaadul63, (jayatu); those in i change aya into e by Samprasarana : parinedi, M. parinei (parinayati), and analoguously also the 1. Person: avanemi (apanayami), M. anemi (anayāmi); němha (nayama), nedha (nayata). However, beside these we have nedu (nayatu) parinaadu (nayatu), M. naaï (nayati), Mg. naante (nayan really nayanto).
JAIN JOURNAL VOL-XXX No. 4, April 1996
bhū forms the Optative after Sanskrit *bhaveam, *bhave, pahave (bhaveyam, bhavet, prabhavet); besides bhavidavvam164, *bhavidum, 'bhavia (bhavitavyam, bhavitum, bhūtvā) and Mg. bhavāmi, M. hava (bhava), havanti (bhavanti). In combination with prefixes the stem is bhava (M. hava) predominating: anubhavanto (anubhavan), anubhavida (anubhūta) M. anuhavaï (anubhavati). M. shows the stem huva in huvanti (bhavanti) and huvia (bhaveya), Mg. in huviadi (bhūyate). As a rule usually we say homi, hosi, "bhodi165, honti; Imper hohi, 'bhodu166, homha, hodha, hontu167; 'hodavva (bhavitavya), bhūda, M. hūa (bhūta), Mg. pahūda (prabhūta).
Roots ending in r,rform the stem of the present with ara; besides frequently in e: dharami (dhriye), pasaradi168, Mg. pasalasi (prasarati, prasarasi); anusaramha (anusarama), maradi169 (mriyate), Mg. malāmi170 (mriye); uvahara, uvaharantu, avaharadi, odaradi, odaramha, odaria11 (upahara, upaharantu, apaharati, avatarati, avatarāma, avatirya); Mg. halami, haladi (harāmi, harati), odala (avatara); M. dharemi, dharaï, dhareï, dharenti, osaraï, osaria172 (dhriye, dhriyate, dhriyante, apasarati, apasṛtya)-smṛ builds 'sumarami, sumaresi173, sumaredi, sumarehi and sumaresu, sumaredha, sumarida [Mg. sumalida] (smarāmi, smarasi, smarati, smara, smarata, smṛta).
Roots in ai: gāāmi, gāadi, gāadha (gāyāmi, gāyati, gāyatha); M. gāanti, uggaanti (gāyanti, udgayanti); Mg. gae [29] gäidam (gaye,
163. jedu is wrongly attested, even if used often; likewise Mg. yedu. 164. In addition to it bhavidavvada (tavyatā).
165. M. hoi.
166. M. hou.
167. M. havantu.
168. M. pasaraï
169. M. maraï.
170. Mg. however maledu, malĕnti (mriyatām, mriyante).
171. Mg. odalia.
172. Mg. ośalia.
173. In a lesser way attested sumarasi. Mg. sumalāmi, sumalesi, sumaledi; Imper. sumala and sumalehi.

Page #49

RICHARD SCHMIDT: A PRIMER OF SAURASENI
*gayitam-gitam).-parittaadi, parittāasu, parittaadu [Mg. palittaadul, parittaadha [Mg. pali] (paritrāyate, 'trāyasva, °trāyatām, 'trāyadhvam).-nijjhaadi, M. nijjhaaï (nirdhyāyati); parimilāadi
(parimlāyati).
kram contains a short vowel: nikkamāmi, nikkamadi [M. nikkamaï and nikkhamai, Mg. niskamadi, nikkama [Mg. niskama], nikkamamha (niṣkrāmāmi, krāmati, 'krāma, "krāmāma).
147
ruh becomes a compound verb after the inflected 6th class: āruhadi [M. āruhai, āruha [Mg. aluha], aruhadu [Mg. aluhadu] (arohati, aroha, ārohatu).
sthā as the stem for the pres ent has ciṭṭha [Mg. cistha] and forms from it: ciṭṭhāmi, ciṭṭhadi [M. citthaï, Mg. cisthadi, ciṭṭha, (tiṣṭhāmi, tiṣṭhati, tistha); M. ciṭṭhaü (tiṣṭhatu). With the ud prefix we get the forms 'uṭṭhehi beside utthehi (uttiṣṭha), utthedu, Mg. uṭṭhedu (uttiṣṭhatu).
muc has in Ś. Mg. a nasal, it can be lacking in M. muñcadi, ‘munca, muñcadu, M. muai and muncai, mua, muasu (muñcati, muñca). By the side also the e- stem 'muñcesi, muñcedi, muñcedha, muñcehi (muñcasi, muñcati, muñcatha, munca).
The verbs of the tenth class, which mostly counter acts aya in e the causatives and the denominatives seldom has aa for that and indeed so usually before consonant groups: M. kahei, Mg. kadhedi (kathayati), Ś. kadhehi (kathaya), kadhedu (kathayatu); cintemi, cintemo (cintayāmi, mas), but cintaanto (cintayan); pavesehi [Mg. pavesehi] (pravesaya) beside pavesaami (praveśayāmi); ciraadi (cirayati), viraaāmi (viracayāmi).
27. Second Conjugation: a) The roots of the 2nd class infringes in the 1st. conjugation: M. ruāmi, ruasi, ruaï, rua (rodimi, rodişi roditi, roda); Ś. rodasi, roadi, roanti, roda [Mg. loda], rodidum (rodiși, rodati, rudanti, roda, roditum); Mg. metrically luadi (roditi).-śvas forms nisasadi [M. ņīsasail, visasāmi, samassasa, samassasadu (nihsvasiti, visvasimi, samāśvasatu), Mg. sasadi, samassasadu etc. (svasiti, samāśvasatu).-cakṣ forms acakkha (acakṣva), Mg. ācaskadi (acaṣṭe); M. niacchaï (**nicakṣati=nicaṣṭe); han has paḍihanāmi (pratihanmi), vihananti (vighnati); M. hanai, nihananti (hanti, nighnanti), Mg. āhanedha (ahata).-as has 1. Omhi, Mg. smi [30] 2. °si, Mg. si, 3. °atthi, Mg. asti; Plur. 1. omha [M. also mho], Mg. sma; 2. M. ttha, 3. M. santi, Mg. santi174
174. The 2. and 3. Plur. is rare, atthi is used for all persons of the Sing, and Plur.

Page #50

JAIN JOURNAL: VOL-XXX No. 4, April 1996
b) The third class forms are retained only sparingly. dā builds *demi, desi, dedi [M. dei, "dehi [M. also desu, Mg. desu), dedu [M. deul, Mg. dedha [M. deha] (dadāmi, si, °ti, dehi, dadātu, datta).Forms from dha: M. saddahimo (sraddadhmah); S. anusamdhemi, 'dhedha (anusamdadhāmi, °dhattha, M. samdhei (samdadhāti), samdhenti (samdadhati).-bhi forms (from the stem bhai bhāāmi, (Mg. bhāāsi), bhaadi (bibhemi, bibheşi, bibheti).
148
c) Forms from the fifth class like avacinomi, avacinumo ('cinomi, °cinumaḥ) are unsure against avacinamha (°cinuma) and uccinedi (°noti). M. has, for ex. samuccinaï, vicinanti (cinoti, °cinvanti), but also uccei, uccenti, ucceu (uccinoti, 'cinvanti, 'cinotu); Mg. samcehi (samcinu).-Likewise from śru there are unsure sunu, Mg. suņu (srnu) and Mg. suņudha (smuta). We should read suna175, as in S. Mg. roots follow the 9. class: sunami176, sunādi177, sunāmo, sunahi, suṇādu178, sunamha, sunadha179, sunantu180 (smomi, ti, smumah, smu, smotu, smuma, smutha, śṛvantu). Beside it there occur forms from e-stems, as sunemi, sunedi, suněmha (sṛnomi, 'ti, sṛnuma). M. builds from the a-stems: sunai, sunimo, sunanti, sunasu, sunaha (smoti, smumaḥ, śṛvanti, smu, smuta).
prap goes with the 1st class: M. pāvasi, pāvaï [Ś. pāvedi]; °pävanti M. pāva [Ś. pāvehi], pāvaü (prāpnoși, °ti, prāpnuvanti, präpnuhi, prāpnotu).-sak has sakkanomi or sakkunomi (saknomi); Mārkeṇḍeya IX, 131 teaches sakkunadi and sakkadi.
d) The seventh class has the nasal also in the weak forms M. chindaï, bhanjaï, bhindaï (chinatti, bhanakti, bhinatti); bhuñjasu (bhunkṣva); paüñjadha (prayunktha), rundhedi (runatsi); "bhindia (bhittvā), bhañjia (bhaktvā).
[31] (e) The root kr belonging to the eighth class forms "karemi181, 'karesi182, karedi183, karemo, karěnti184, karehi185, okaresu186, karedu187, karĕmh188, karedha189 (karomi, si, °ti, kurmaḥ kurvanti, kuru, karotu, karavāma, kuruta).
175. In Mg. according to Pischel § 503 sunu is a Sanskritism for suna. 176. Mg. suṇāmi.
177. Mg. sunadi.
178. Mg. sunadu.
179. Mg. suṇādha and sunedha. 180. Mg. Sunantu.
181. Mg. kalemi.
182. Mg. kalesi, M. also kunasi.

183. Mg. kaledi, M karei and also kunaï.
184. M also kunanti.
185. Mg. kalehi.
186. M also kuna, kunasu.
187. M kunaü.
188. Mg. kalĕmha 189. kaledha.

Page #51

RICHARD SCHMIDT: A PRIMER OF SAURASENĪ
f) Ninth class: jñā loses the initial after prefixes and na(na) and forms jāņāsi190, jāṇādi191, jāņāmo192, jāņa193 and jāṇāhi (jānāsi, oti, janimaḥ, jānihi); 'aṇādi, viaṇādi, ānavedi, °du, na anadha194 (ājānāti, vijānāti, ājñapayati, tu, na ājānītha).
kri also has the present stem kina and forms nikkinasi, kinadha, M. vikkinai (niskrīnāsi, kriņīta, vikriņāti).-gṛhforms genhasi, 'genhadi [M. genhail, gěnhanti, 'genha, 'genhadu [M. genhaü], genhadha, gěnhantu (grhnāsi, 'ti, grhnanti, gṛhana, grhnitu, grhnita, gṛhnantu).-bandh forms bandhāmi, bandhasi and bandhesi, bandhedi [M. bandhail, bandhanti [M. also bandhěnti], M. bandhasu (bandhnāmi, si, avabadhnāti, badhnanti, badhāna)-bhan forms bhanasiand bhanesi [Mg. bhanesi, *bhaṇādi, *bhanadha, *bhana and *bhanahi, bhanehi; bhaṇādu (bhanasi, oti, 'tha, bhana, bhanatu).
149
g) Forms of the Imperfect (apart from °āsi, āsī = äsit), the Aorist and the Perfect are lacking.
Future
28. The endings are issam, issasi195, issadi196, issāmo, issadha, issanti, i.e. with the exception of the 1. Sing. as in Sanskrit, the sya is added with the linking vowel i; Mg. has always ss in place of ss.bhavissam [Mg. bhavissam197], bhavissasi198, [Mg. bhavissasi]; bhavissadi,[Mg. bhavissadi199; gamissāmo, bhanissadha, bhavissantí200 (bhaviṣyāmi, syasi, ti, gamiṣyamah, bhanisyatha, bhaviṣyanti).
[32] Examples from isolated classes: 1st Class : visumarissam (vismariṣyami); pariharissadi, Mg. palihalissadi (parihariṣyati); marissasi, Mg. malihisi, M. anumarihii (mariṣyase, te); gaissam, Mg. gäissam (gāsyāmi); gamissam, Mg. gamissam (isyami); M. samagamissaï (iṣyati), pahinti (pasyanti); ciṭṭhissam, Mg. cisthissam (sthāsyāmi); uṭṭhissāmo (utsthasyamah); pěkkhissam, Mg. pěkkhisśam, M. pěcchissam (drakṣyāmi i.e. prekṣiṣye); Mg. khāissam, metrically khāhisi (khāsyāmi, khāsyasi). 6th Class: pucchissam Mg. puściśśam (prakṣyāmi);
190. Mg. yānāsi; M. jāņimi (jānāmi), jāṇasi, jāņase.
191. Mg. yāṇādi, M. jāņaï.
192. M. jāņimo.
193. M. also jāṇasu; Mg. yāṇāhi.
194. M. na anasi, anai, animo āṇaha (na ājānāsi, oti, ājānīmaḥ, °tha). 195. M. ihisi.
196. M. ihii.
197. Also huvissam.
198. Also huvissasi.
199. Also huvissadi, Mg. huvissadi.
200. Also huvissanti. M (from the stem ho) hossam, hohii, hohāmo, hohissāmo, hohittha, hohinti.

Page #52

150
JAIN JOURNAL : VOL-XXX No. 4, April 1996
muncissadi, M. mócchihii (moksyatii). - 4th Class : kuvissam, M. kuppissam (kopsyāmi);ovivajjissam(vipatsyāmi). 10th Class : cintaissadi (cintayisyati); Mg. gañažśćam (ganayişyāmi); kadhissam beside kadhaïssam (Mg. kadhaïssam M. kahissam(kathayisyāmi).- 2nd Class : rodissam, M. roissam (rotsyāmi); suvissam, Mg. suvissam (svapsyāmi); daïssam, Mg. daïssam (dāsyāmi, properly speaking dayisyāmi from the stem de).-5th Class : avacinissam(avacesyämi, sunissam, Mg. sunissam (srosyāmi); pāvissasi, M. pāvihisi, Mg. pāvihisi (prāpsyasi). The 7th Class is without documentation; from the 8th Class very frequent "karissam, Mg. kalissam (karisyāmi); karissasi, M. karihisi (karisyasi); karissadi, Mg. kaliśśadi (karisyati); M. karissāma201 (karişyāmahy; karissanti (karisyanti). We have beside it karaïssam, Mg. kalažšśam; Mg. kalaïssasi; karaïssadi, Mg. kalaïssadi, M. karehii; karaissanti.-M. yet also builds kāham, kāhisi, kāhii.-9th Class : jānissam; M. jāņihisi; jāņissadi; jāņissāmo, Mg. yāņiśćamo ljñāsyāmi, jñāsyasi, oti, syāmah.)-Mg. kīņiśćam (kresyāmi); kişissadi (kresyati).-gěnhissam, gěnhissadi (grahisyāmi, "syati). Passive
29. To the root resp. in the stem of the present there comes ya (whereby y202) stuck out behind vowels, (33) assimilated with consonants behind, or ia.
In M. ya becomes ija, ia is changed to ijja. - 1st Class : *ņiadi, M. nijjaï (niyate) *bhaviadi, Mg. also huviadi (bhūyate); sumariadi, M. sumarijaï (smaryate); icchiadi, Mg. iściadi (isyate); M. gammaï (gamyate), but gamijjanti (gamyante); S. gacchiadi, but gamiadu (gamyate, otām); piviadi, M. pijjaï (piyate); anuciţthiadi, odu (anuşthiyate, otām); disadi, Mg. disadi, M. disaï (drsyate).-6th Class : pucchiadi, M. pucchijjaï (prcchyate); munciadu (mucyatām), M. muccaï (mucyate). 4th and 10th Classes, Causative : M. paờibujjhijjaï (pratibudhyate); chedianti, M. cheijjanti (chedyante); pobodhiāmi (prabodhye); vinnaviadi (vijñāpyate); M. kahijjai, Mg. kadhiadu (kathyate, "tām).--2nd Class : suviadi, M. suppaü (supyate, otām); *vuccadi, M. vuccaï (ucyate); *diadi, M. dijjai
te). 5th Class : M. cinijjai (ciyate); viciadu (viciyatam); suniadi, Mg. suņiadi, M. sunijjaï (śruyate); sakkiadi, Mg. sakkiadi (sakyate).-7th Class: ochijjanti (chidyante); jujjadi, M. jujjaï (yujyate) in the sense of "to be suited to itself"; on the other hand païñjiadi, odu (prayujyate, otām).-8th Class : kariadi, odu, Mg. kuliadi, odu, M. kiraï, kiraü (kriyate, kriyatām).-9th Class : jāņiadi (najjai), odu (jñāyate, otām); anuggahiadu (anugrhyatām); M. forms ghéppaï, ghéppanti (gshyate, onte) from a stem *ghrp.
To the passive there belongs still also another future with forms
201. Metrical for 'mo. 202. Pischel says on p. 369 below ya is cast out: then the form would not be
diadi from dā (out of di-ya-di) but must sound didi.

Page #53

RICHARD SCHMIDT : A PRIMER OF ŚAURASENÍ
151
like muccissadi, chijjissadi (*mucyisyate, *chiềyisyate), M. bhajjihisi (*bhajyisyase). Causative
30. As in Sanskrit the root is upgraded, the aya suffix turns to e, paya becomes ve. kāredi, M. kārei (kārayati); *āņavedi (ājñapayati); vinnavissam (vijñapisyāmi).-ve = payais also used in roots with other
as ā,ii, consonants and diphthongs : jīvāvedu (*jīvāpayatu), Mg. palivattāvehi (*parivartāpaya); M. ramāvěnti (*ramāpayanti); moāvemi, "hi (mocāpayāmi, *mocāpaya); Mg. lihāvemi (*likhāpayāmi); M. māresi (mārayasi), ś. māredha (34) (mārayatha), Mg. mālemi, Chi, odu, odha (mārayāmi, māraya, "tu "ta).-With drś M. forms dāvemi, dāvei, dāvěnti (darśayāmi, oyati, 'yanti), Ś. damsemi, damsaïssam (darśayāmi, darśayisyāmi), Mg. damsaante (*darśayanto = darśayan.
Desiderative
31. The formation is like that in Sanskrit:jugucchedi (M. juucchai], jugucchanti (jugupsati, Onti); adijuucchida, Mg.adiyuuścida (atijugupsita); cikicchidavva (cikitsitavya); sussūsaïssam (śuśrūşayişyāmi); Mg. śuśśāśida (suśrūşita).
Denominative
32. One way of the formation is directly from the stem : M. dukkhāmi (duhkhāmi), dhavalaï (dhavalati), kahāmi, kahasi, kahaï (kathāmi, si, oti). --The usual formation is that in a (=aya) : M. umhāi (uşmāyati); S. kuravaäadi (kurabakāyate); Mg. cilāadi (cirāyati); suhāadi, M. suhāaï (sukhāyati). There are onomatopoeic forms frequently like ghumaghumaadi, tharatharedi; finally following the way of causatively formed : saddāvemi (sabdāpayāmi), suhāvedi (sukhāpayāmi); Mg. śuskāvaïśćam (suşkāpayişyāmi); M. mailei (*maïlayati), biunei (dviguņayati).
Verbal Noun
33(a) Participles. The active present participle is built from the present stems ending in anto (Mg. ante), Feminine anti: jāņanto; Mg. puscante; M. calanto (jānan, prcchan, calan); pěkkhanti, M. apāvanti (*prekşanti, aprāpnuvanti). – The present middle and passive participles has the termination māņa, māņā : ņivattamāņā, anuņiamāņā (nivartamānā, anuniyamānā). - Preterite passive participle has, as in Sanskrit the ta suffix and the na suffix (those in *da resp. become otņa, whilst M. naturally for (35) ta has a only) : *gahida, M. gahia (gshita); sumarida, Mg. sumalida (smsta); pucchida, M. pucchia (prsta); **diņņa (datta), o*mukka (mukta from mukna).

Page #54

JAIN JOURNAL VOL-XXX No. 4, April 1996
b) Gerund: *hodavva, Ś. also bhavidavva, Mg. also huvidavva (bhavitavya); sodavva [also sunidavva], M. soavva (srotavya); *kādavva, M. kāavva (kartavya).- The aniya suffix becomes *ania, M. anijja: for ex. karania, M. karanijja (karaniya); Mg. palihalania (pariharaniya).— kajja, Mg. kayya (kārya);°gějjha, Mg. dugghěyha (=*grhya, *durgṛhya).
c) Infinitive: The S. Mg. suffix dum, M. um often occurs with the pure stem or the stem of the present with the linking vowel i for example, gacchidum beside gamidum and o*gantum; pucchidum, Mg. puścidum, M. pucchium (prastum); kāredum (kārayitum); M. nivvāheum (nirvāhayitum); Mg. sodhāvedum (sodhāpayitum); more rarely being contraried: tāḍaïdum (tāḍayitum); Mg. mālaïdum (mārayitum); dharidum, M. dhārium (dhārayitum); măridum, Mg. mālidum (mārayitum). From the second conjugation: sunidum, M. soum (srotum); bhunjidum, M. bhottum (bhoktum); genhidum, M. gahium (grahitum); *kādum, karidum, M. kaum (kartum). Formations throughout corresponding to Sanskrit like jividukāma (jivitu°), M. tāḍiumaṇā (tāḍayitumanāḥ).
152
d) Absolutive. Apart from *kadua and *gadua (kṛtvā, gatvā), there is in Ś. Mg. ya (resp. a) which is the only governing suffix, while that for M. una is found. A distinction between the simple and the compound verbs is not made.-naïa (*nayiya nītvā), but āņia, avania (aniya, apaniya), *bhavia (bhūtvā); odaria, Mg. odalia (avatirya); pariharia, Mg. palihalia (parihṛtya); pěkkhia, Mg. pěskia (prekṣya); pavisia, Mg. pavisia (pravisya); gĕnhia (gṛhitvā).-Examples for M. : jeūņa (jitva), houņa (bhūtvā), hasiūna (hasitva), vihasiūna (vihasya), gaḥiūna, ghettuna, the latter metrically also in Mg. (gṛhitvā).
=
N.B. The grammatical discussions are made till the end of page 35. From the pages 36 to 43 the specimens from some texts are given. For the Specimens of Śauraseni and Māhārāṣṭri (pp. 36-38) the passages are taken from the Karpuramanjari edited by Sten konow 1.4.15-12.3. And the Specimen for Magadhi (pp. 38-43) is given from the Sakuntalā edited by Richard Pischel, S 113-116. All the Prakrit texts are accompanied with Sanskrit rendering. [Editor]

Page #55
Summer 1996 Registered with the Registrar of Newspapers for India Under No. R. N. 12121/66. BHILAI ENGINEERING CORPORATION LTD. Industrial Area, Post Box No. 31 Bhilai 490 001, Madhya Pradesh Telephones: 355417/356447/355358 Fax Numbers : 0788 355527/355427 Telex Numbers : 0771-214/245/204 Manufacturers Erectors of Equipments for Coal Mining, Steel Plant, Power Plant, Cement Plant and Manufacturers of Sulpheric and Super Phosphate Fertilizers Regional Offices 13 Masjid Moth, Local Shopping Centre New Delhi 110 048 Telephones : 6445815 6445819 6434987 6414390 Fax Number: 011 6445819 4B Little Russel Street, 2nd Floor Calcutta-700 071 Telephones : 242 7606 / 3372 Fax Number : 033-2427061 31 Makers Chambers III 3rd Floor, Nariman Point, Bombay 400 023 Telephones : 231724 2043647 244208 2873789 Telex Number : 0118 4054 Fax Number : 022 287 3611 By the courtesy of Antarctica Graphic Limited Jain Educatior1 A, Vidyasagar Street, Calcutta-700 009al Us Phone : 350-2173/5424/7658 elibrary.org