Download pdf - Vasudev a Hindi

8/12/2019 Vasudev a Hindi

1/15

The Vasudevahindi a Specimen of Archaic Jaina-

Abbreviations used : Vh = Vasudevahindi I= MBhlrfigtri ; J M = Jaina-Mihirfigtri g = gauraseni JS = Jaina-Sauraseni AMg = ArdhamzgadhiP = Pischel's Prakrit Grammar JErz = Jacobi's Ausqewahlte Erzahlungeni n Mcihicrci~tri Mah. Nis. = Schubring, Das M8hEnisiha Sutta Pkt, Skt= Prakrit, Sanskrit.

N the Introduction to his edition of the Avaiyaka tales the lateI Professor Leumann wrote on p. 1 : Since-as may be easilyunderstood-Jacobi, when he endeavoured to utilize the languageand contents of the Jaina tales for Indology, started with theUttariidhyayana Tikii of Devendra which was in his possession, andsince this author, belonging as he does to the period of decadence andtherefore more familiar with Sanskrit, writes a rather doubtfulPrakrit-therefore the editor was chiefly concerned with theestablishing of a thoroughly reliable Prakrit text . that was to besuitable for clearing the way for a more correct judgment andutilization of medieval Jaina Prakrit. The appearance of thefirst portion of Leumann's text was welcomed by Pischel in his Pktgrammar ( 21 in the following terms : The most important textin JM is : Die Avaiyaka-Erziihlungen. Herausgegeben von ErnstLeumann. 1 Heft. Leipzig 1897. The absence of any commentaryunfortunately renders the understanding rather difficult ; somepassages remain wholly obscure. But even these few forms show thatfrom texts in JM we may yet expect much new and importantmaterial.''

In spite of all that has been done since then in the domain of Pktliterature and grammar, the hope expressed in Pischel's last sentencecan hardly be said to have been fulfilled. The work so ably begun byLeumann has never been continued : the first portion of his AvaiyakaErzahlungen has remained the last. Many other texts have becomeaccessible, but they are almost without exception open to Leumann'scriticism of Devendra : works like those of Haribhadra, Somaprabha,etc., bear witness to the skilful handling of a fixed-not to say

dead -literary language by later authors, but they are not likelyCf. the following quotation from P.


2/15


3/15


4/15

322 L. ALSDORF-91, 14. To these must be added two optatives : ptisijjam, 6, 18, andpassejjam, 125, 3. A 1st sing. opt. in -jjam is also wanting in P wherewe find only -jjE, -jja, -jjEmi. That both forms (ind. -am, opt. -ijjam)are genuine archaisms is proved beyond doubt by the fact that theyhave counterparts in Pali (e.g. gaccham, labheyyam, cf. Geiger, Ptili,5 122, 127, 128). There -am instead of -Emi is peculiar to the languageof the gathas, i.e. the oldest stratum of the language. The only otherPkt text where a 1st sing. ind. in -am occurs seems to be theMahiinisiha Sutta, from which Schubring (Mah. Nis., p. 90) quotes theSloka-plda gaccham cetJham suvam ugham dhtivam nzsam palti~ziu p , and three single forms, also from verses, caram, na ninhavap,sakkunam = Saknomi 2 . This is interesting because we shall presentlysee that another peculiarity of the Rlahiinisiha is also shared by Vh.

In the Mahlnisiha, the 1st plur. is often constructed with ahamor is otherwise used as a 1st sing. e.g. ahayam apucighimo, ndham. . . cukkimo, etc. In Vh we read, e.g. 84, 7, aham . dacchtimo ; 290,28, tao 'ham ttio luvtimo ; 291, 24, lavai ya me icchtimo . . ; 172,11, samcaramtinT . . . supimo ; 178, 22, mayti bhaniyti jti~ThEmotti. tao niggayti jtinihisi tti vottQym.According to P 3 457, the 1st person plur. of the iitmanepada iswanting in Pkt. In Vh I have noted it seven times, but in all casesit is used as a 1st sing., twice even with aha? : 147, 18, ahampadicchtimahe ; 206, 18, aham apuvatttimahe; 144, 7, ughio mi

kattha m p ne vatttimhe tti cintayanto ; 352,22, cintemi : kummipaesarnmi watttimahe 139, 24, ramcimahe (v. l., mi ya ) ; 155, 16,uvabhunjtimhe . . ., bhutta-bhoyanp ya . ; 330, 19 (in a gahl ),bhapai : . bandhtimahe . .".This form in -timahe, used as a 1st sing., probably helps to explainanother very strange and hitherto utterly unknown form. Eleventimes there occurs a 1st sing. in -ahe : acchahe, 180, 14 206, 12 ;247, 1 ; aticchahe, 319, 24 tisahe, 289, 13, 29 ; dissahe (pass.), 199, 6 ;passahe, 218, 10 ptisahe, 293, 5 ; vattahe, 247, 12. Failing any otherexplanation, I believe that -ahe is formed from -timahe after the analogyof the proportion : 1st sing. act. -timi : 1st sing. med. -e (vandtimi :vande = vandtimhe : vandahe). And, lastly, a single form may benoted which-if not a mere mistake-looks like a compromise between-ahe and -e : 156, 17, 1st sing. acchae instead of acche.According to P 5 455, the 1st plur. ind. (which is used as 1stplur. imp. as well, 5 470) ends in -mo which may be replaced by -muin verses only. Yet in the prose of Vh -mu occurs not infrequently,


5/15

A SPECIMEN OF ARCHAIC JAINA MAHARASTRI 323e.g. ind. ainemu, 108, ; nu-yfinfimu, 117, 8 ; acchfimu, 115, 24 ;karemu, 117,17 ; imp. vasiimu, 8 2 , 3 ; harEmu, 10 0 ,2 anuvaydmu, 138,2 ; anumaggfimu, 138, 12 ; h rem u , 8 5 , 1 5 ; 109, 12 153, 15, etc.Now since, as we ha ve seen, th e 1st plur. m ay be used as a 1st sing.as well (prob ably originally as a plura lis majestatis), I have l i tt le doubtthat the 1st s ing. imp. in -mu taught by the grammarians but-according t o P , 467-not found in literature,l is nothing bu t a 1stplur. used as a 1s t sing. I n suppo rt of this explanation I can quotefrom V h a t least one 1st sing. imp . in -m_o: 12 2, 5, na me sobhai ihamacchium, avakkamfimo tti, it is no good for me t o st ay here, I willru n aw ay -Pischel believed t h a t -mu, -su, zc as terminations of theimperative corresponded to the indicative terminations -mi, -si, -i ,and mainly for this reason he disputed the usual derivation of -sufrom Sk t. -sva. If my ex planation of th e 1s t sing. imp. in -mu isaccepted, Pischel's view of -su-not ve ry convincing considering Pal i-ssu (cf. Geiger, Pfili, 126 an d note 1)-becomes wholly ~ n t e n a b l e . ~

It has till now been taken for granted that forms of the aoristhav e survived in ARlg only. F ro m Vh we learn that they occurin archaic J M as well.3 W e even find in Vh several forms which areeither quite new or hitherto known only from grammarians.

The 1st s ing. act . of th e aorist is very rare an d ends in -is sa p(P, $516 ; e.g. akarissam), with double s as in Pa li " (where, however,th e more usual form has a single s, cf. Geiger, P d i , $159 , iv). One suchform occurs Vh 225, 17 : d h a m tubbham kuppissam, I was notang ry with you. The context excludes every possibility of th e formbeing taken a s a future. Y et th e formal identity of a 1st sing. aor.in -issum with the 1st sing. fut. cannot be overlooked, and I believethat the doubling of the s (for which neither Pischel nor Geiger offer

Jl . Schubrlng has now traced it in the hlahgnisiha, where a 1st sing. imp. demuoccurs twice M a h . X i s . , p. 91 .In Vh there occurs in a number of cases a 2nd sing. imp. in -dsu (e.g. vaccdsu,93, 15 179, 18 ghaddsu 94, 24 lcardsu 96 10). The long i of these forms may,of course, be explained by the influence of the parallel form in .ah;. We m y , however,also have before us the regular compensative lengthening due t o the double s (whichis seen in Pali - m u ) having been simplified.

It might be argued that such forms are to be regarded as mereArdhamiigadhisms ", which do not prove anything for JM This argument would

perhaps hold good if we had t o do with legendary, dogmatical, or disciplinarian texts.Where th e contents are so closely akin to those of the canonical scriptiires i t is notsurprising if the language betrays a strong AMg influence (cf. Mah. X i s . , p. 86 .But, except for some inserted legendary tales, the contents of the Vh are as secularand non-canonical as possible, and there can be no doubt that its language is J Mthroughout.


6/15

324 L LSDORFan y exp lanation) is actual ly d ue to th e influence of th e s imilar futureform. This assumption receives a welcome s up po rt b y anoth er formwhich is hi therto quite unknown an d without a paral lel in ei ther P k tor Pali. Vh 289, 27, we read : dacchihum c'aham, a n d I beheld.Now it is well known th a t besides th e normal futu re in -issiimi (-issum),-issai, -issfimo, etc., there exists in Pkt a second futilre in -ihiimi(-ilzam), -ihii, -ihEmo, etc . B u t if th e h of the se forms-as m ustobviously be th e case-goes back to sy > ss > s, we should expect beforei t (with compensat ive lengthening) rather th an i. This form with along i , not recorded by th e gramm arians an d not found in P or J E rz ,actually occurs in Vh, e.g. 51, 22, giiattiham ; 22, 28, bhunjiham ;78, 22, jiinihiimo ; 89, 21, pucchThCirno ; 91, 8 , j'ivihZmo 138, 7,dacchihiimo, etc.l The 1 st sing. belonging t o th e last of these form swould be dacchihum, i.e. the very same form we have just noted as a1s t sing. aor. This use of a 1s t sing. fu t. dacchihum as a 1 st sing. aor.seems to prove th a t a connection was felt to exist between th e 1st sing.fut . an d ao r, in - issum : th e iden ti ty of these two forms seems to hav ejustified th e use of a ny 1st sing. fut. as 1st sing. aor.

There are, however, indications to show th a t th e relations betweenth e future and th e aoris t were not l imited to t he 1st person of th e s ing.The gram marian s know tw o ao rists formed with h instead of s which-to conclude f rom P , 516-have no t yet been found to occur in tex ts,viz. kchi besides kiisi (from k?.) and thZhi besides (hiisi (from sthii).Here th e h m ay actually go back to t he s of kiis5 an d thiisi. B u t it isvery significant th a t th ese h-aorists, too, a re identical with forms of thefuture : kiihi and thiihi are perfectly normal 3rd persons sing. of thefutures kiiham (P , 533, where kiihi is actual ly quoted) an d (Z h a mP, 524, thiihii ). Two other aorist forms (vocchiya and gacchiya)which can only be explained b y referring to t he futures vocchav an dgaccham will be discussed below. And , lastly, the Mahanisiha furnisheswhat might be called a counter-proof. It uses (Mah. Nis., p. 91)a 3rd plu r. bhavisum, once even spelt bhuwimsum, as if t h e fo rm s wereaorists ; b u t t h e co ntext, where also bhavihenti precedes the m , provesbeyond doubt th at they are futures. Here forms of th e aoris t havepenet ra ted into the future : th e exact co ntra ry of th e relations betweenaorist and future which we have noticed so far.

One of th e m ost comm on aorist form s in AMg is vayiisi he spoke ,wh ich is-like iisi-used for oth er persons, pa rtic ula rly th e 3rd plur.,

The missing link between - h a m and - i h a q is supplied by th e Mahiinisiha,where futures in - i sam occur, e g c imucc isam su j j h i sam f i l ah .Nis. p. 91) .


7/15

A SPECIMEK OF ARCHAIC JAINA-BIAHARASTRI 325as well. In Vh it occurs as 3rd sing. 351, 28 ; as 3rd plur. 33, 17. Butbesides it there occurs three times a hitherto unknown by-form withsamprascrana of the root : 284, 9 and 13, 3rd sing. inam udesi(284, 9, one MS. reads idam vadcsi ) ; 324, 3, 3rd plur. ima?cayanam udcsiTwo, as it were, normal forms are the 3rd sing. vehvbi (291,20 ; from velavai, to upbraid ") and the 3rd plur. vinnavimszc (216,2 ; from vinnavai vijfiapayati).

A more doubtful case is found, 29, 11, where a husband, referringto his previous warnings, says to his wife : kim idcnim rodasi?mamam tadc na sunesi bhannamlni Why do you weep nowAt that time you did not listen to me when I spoke to you It is,of course, quite possible to take sunesi as a 2nd sing. of the presentindicative. But a 2nd sing. aor. (which does not differ from the 3rd)would have the same form (cf. P, 516, kahesi, etc. ; the final S mayalso be short, cf. e.g. akcsi, ahesi, etc.), and the context seems to demandmost categorically a form of a past tense.

466, end, Pischel deals with some rather mysterious forms in -;awhich are mentioned by the grammarians as being used in the senseof the imperfect, aorist, and perfect tenses ", and he thinks that,

inexplicable as it seems, acchia, genhza, daliddaza, maria, hasia,huvia, dehia are really optatives of the present and kchza, thchia, hohiaare really optatives of the aorist. Whatever the correct explanationof these forms may be, the Vh furnishes the first instances of theiractual use. We read there : 289, 17, gacchiya, I went ; 289, 28,dine gamesiya, I spent the days ; 278, 32, vocchiya, he spoke ;111, 22, kesiya, he made .I According to Pischel, we should have toregard the first of these forms as an optative of the present and theother three as optatives of the aorist. But an aorist *vocchi can hardlybe explained except as having been derived from the future voccham(P, 529), of which the 3rd sing. is vocchii, vocchi ; and this makes ita t least very probable that gacchTya does not belong to the presentgacchai, but that an aorist *gacchT had been derived from the futuregaccham P, 523). Here, again, the very close relations becomevisible that must have been established in Pkt between the future

As in the te xt of Vh ca ya t t i t are frequently inserted where they are superfluousand even manifestly impossible, there is just the possibility of declaring the finalya of the forms in question or some of them) to be such a spurious ya. Considering,however, th e unanimous testimony of th e grammarians, feel confident that we haveactually to read gacchiya etc.


8/15

326 L. ALSDORFan d the aorist . Unfortunately our material is as yet too scanty toallow of a more definite description of them.

Th e verbum su bsta ntivu m calls for a few remarks. The firstpoint to at tract our at tention is the quite unusual frequency of i tsuse. E .g . th e 2nd plur. ttha is qualified by Pischel 5 498) as

very rare . He gives one single reference (to Setubandha 3, 3)an d in his parad igm ascribes ttha t o M only. I n Vh it occurs more th anfo rty tim es (e.g. 86, 6, 8, 16 92, 21, 23 ; 93, 6 ; 96, 2, 4 102,17 103, 10 107, 21 110, 5, 7, 24, etc.) . The 1st and 2nd sing. andth e 1st plur. are even more common ly used. As we hav e doubtlessto do with old forms that are gradually becoming obsolete in Pkt,their strong vitality in Vh m ust be looked upon a s a sign of antiquity.Ano ther very striking peculiarity points in the same direction. Weshould expect enclitic forms like mi, si, mo, ttha t o be no rmally placedafter the word (mostly a past part .) to which they belong,l and thisis actu ally often enough t h e case, e.g. gao mi, 182, 20 patto si, 146, 16pattG mo, 148, 15 jCya ttha, 86, 16 etc. But perhaps even morefrequ ently th e forms in question are placed before th e word t he y belongto, and they may even be separated from it by one or several otherwords, e.g. 281, 16, tao m i niggao 196, 2, tattha ya mi gao 283, 16,tao m i junn 'anteu ram gayG 229, 25, j a i si saho puraccaran e 80, 18,sumarasi, jam si bGlabhGve do vi 2 ) Aimuttaepa nabhacCrinG bhaniyii

214, 19, tattha ya mo iMandara-sam ive vutthGo; 86, 8,kao ttha-m-GgayG?; 96, 2, ja m ttha devie Gp ttG 238, 4, j a i t t h ajamm a-m arapa -bahu lam samsGram chindiu-kiimii 125, 2, tumhe t t h amayC s a y G m - o pad iniyattc akkhaya-sarZrG di tt ha , etc. cf. alsoseveral of th e passages q uoted below, 329, note , as instances ofthe conditional.All these sentences would n ot only reta in exa ctly the same meaningbu t th ey would even look more normal if mi, si, mo, ttha were replacedb y th e perso nal p ronou ns aha?, tumalyz, amhe, tumhe respectively.It is thus easy to imagine that these forms of tii could have cometo be regarded as equivalents of the nominatives of the personalpronouns. Th at this has, a t least to some extent , actually been thecase is proved by four forms which the grammarians enumerate asequivalents of aha?. Pischel has pointed out ( 41 7) h a t amhi, ammi,mmi are Sk t. asmi, while ahammi is aham mi aham asmi,and t h at there is no reason to do ubt th e strt tement of th e gramm arians

Cf. t h e quotations in P 145


9/15

A SPECIMEN O F ARCHAIC JAINA MAHARASTRI 327that these forms were used in the sense of aha? l--even thoug h noneof them was known t o occur in a P k t tex t. I n Vh we read : 217, 19,am hi pesiyE sumaramEnTe sc. devze) t u b b h a ~ Eya-samivam ; 146, 2,tao tena amhi bhanio ; 182, 22, ten'a mhi ma hu ram EbhaGho ; 210, 24,tena y'amhi bhanio ; 212, 7 , t iy ' am m i rnahuram Ghattho ; 218, 5,p a v i ~ h o y 'a n ~ m i 230, 16, tehi y 'ammi tutthehim Eruhio ; 279, 3,tass' a m m i phalam patto. It will be seen a t once t h a t these sentencescannot be separated from those just quoted above the use of amhiand avnmi corresponds exactly t o t h a t of mi , s i , mo, t tha . I n all casesi t ~ ~ o u l de possible, but is by no means necessary, to interpret theforms of -\/as as personal pronouns. Two less equivocal passagesseern to be 165, 18 keriso s i kesu bhavesu 6si ? = kTdydas t v a m k e ~ ubhavesv EsTh ? and 217, 29 : tao m i uttipno visamiimi = tato' ham uttirno~ i d r a m - m i . ~B u t 24, 3, we read t u m a y s i me bhGy6 kanit tho ESTand here it seems almost unavoidable to regard s i as a n interpolation.Yet even this passage m a y be correct : tumamsi = t ram would bean exact counterpart of ahammi t aught by the grammar ians asnom. sing. = ahum.

From the 1st sing. bemi = bravTmi, AMg and J M have deriveda 3rd plur. benti P, 4 9 4 ; Vh, 118, 11 ; 223 , 15) . A 3rd sing. bei,not recorded by Pischel, occurs Vh 35, 18.

Hem acandra adm its i , 46) datta besides d inna , but according toP, 566, it is found only in a P allava Grant (6 , 21, datE) and in propernames. I n Vh datta and dinna are about equally frequ ent. I n one-halfof the text I have counted the former about forty t imes e .g. 106,24, 26, 27 ; 181, 3 5 , 6 ; 241, 5, 8 etc.).

Even in Skt manye sometimes has almost got th e chara cter ofa particle Spe ijer, Sanskr it Sy nta x , 500, note) . I n Vh maFne isregularly used as such after interrogativ es, where i t mig ht be translatedby German zcohl . Examples could be quoted b y the dozen, bu t afew must suffice k i m nzanne, 13, 20 ; ko m. 18 , 27 ; kE m., 101, 8 ;ke?a m . kcrunenu, 133, 15, 28 ; kassa m . rZi?zo, 83 , 28 ; kisa m ., 14, 27 ;kattha m . , 20, 10 ; kiha m ., 310, 15 ; kayE m., 176, 26 ; kayaro m . esa

The peculiar use of atthi illustrated in 417 is also familiar t o Vh 57, 10, atthdkoi . paricasai ; 10, 29, n'atthi koi ciihi-doso disai ; 125 5 atthi me puno rajja-sir;h o j ja ?

There can hardly be any doubt that we have to write teg' amh i, y 'amhi, y 'am mi,etc., and not t ena 'm hi , ya m hi , ya mm i .t iya = tie, cf. below, p. 328 f.

t is grammatically possible, but otherwise very unlikely that we have to dox-ith two sentences : fato 's m y utt irna h. z idramCmi.


10/15

328 L. ALSDORF-devo, 78, 13. This use of manne throws new light on the adverb vanetaught by Hemacandra (ii, 206 ; cf. P, 457). I t seems now certainthat vane must be explained as mane =manye.Turning to the declension of nouns, we shall first of all make theimportant statement that the nom sing. masc, of A-stems ends in -0without a single exception. The nominative in -e occurs only in twoshort AMg quotations-a Vaitlliya stanza, p. 30, 2-3, and half aSloka (not recognized as such by the editors), p. 234, 24--the only onesthat I have been able to discover in the whole text. The other importantcharacteristic of AMg, the locative in -msi, does occur in Vh, but of t,hevery few forms the majority are obvious Ardhamiigadhisms, so thatthe same may safely be assumed of the rest.1

In normal AMg and JM the dative sing. m. n. and the inst. gen.loc. fem. of the A-stems end in -iie, the inst, gen. loc. of feminine I-and U-stems end in -ie, -tie. According to Pischel ($ 361) a dat. sing.in -6ya is limited to AMg verses and Mlgadhi verses. As to the inst.gen. loc. sing. of the feminines, he admits -cia only for M ($ 374 f.),-ia and -aa only in verses where the metre demands a short vowel5 385). He evidently rejects, deeming unnecessary even to mention it,the opinion of Leumann who maintains (Avziyaka-Er~hlungen,

p. 3 f.) that the feminine forms in -6ya, -i;ya, -z?ya, which are notinfrequently found in older texts such as the ciimis and even thebhiisyas, are genuine, and that -ya is the older form (as comparedwith -e) which Pkt has in common with Pali ".

In Vh the following state of things prevails. Besides the dativein -iie (e.g. vahcie, 326, 7 ; aghiie anatthfie ya, 124, 14 ; puttattiie, 76,13 ; 91, 21, etc.), that in -Eya is also found : uahiiya (= vadhZya),169, 19 ; 245, 2, ; 313, 5, 7 ; hiyiiya, 268, 5 ; uvagfirfiya 163, 4 ;vinGsiiya, 313, 8. Feminine forms in -iiya, -iya, though very much rarerthan the normal ones in -iie, -ze, are also not infrequently met with(e.g. instr. padihfiriiya panayiiya, 213, 13 ; gen. asuhfiya, 230, 6 ;kannciya, 311, 1 ; inst. tutthzya, 121, 31 ; buddhza, 10, 24 ; loc. vekiya,150,20,(parinayzya santzya, 173,1, etc.). But in addition to them Vh haspreserved in a few cases a form which seems to me to prove conclusivelythat Leumann was right in comparing a Phi. kanniiya to Pali kaEGfiya.According to Geiger s grammar ($$ 81, 86), Pali forms the inst. gen.

I have only noted the following forms. In an inserted labha-carita : kucchimsi,159 16 ; ulloyamsi, 161 15 ; paramsi, 167 17. An AMg-phrase kucchimsi puttattcie. 76 13 ; 91 21. Besides 246 27 mawamsi (in a doubtful passage), 150 20 tamsizlelciya (grammatically wrong ), 36 25 sohanamsi, and 147 16 etamsi (v. I. etammi .


11/15

A SPECIMEN OF ARCHAIC JAINA MAHARA STRI 329sing. of feminine A-stems in -8ya, but the locative in -tiya and -tiyam(= Skt. -Eytim ) ; similarly we have of feminine I-stems an inst. gen.sing. in -iyti, b u t th e loc. m ay end in -iy8 or -iyam. Of these loc ativesin -8yam and -iyam, the Pkt counterparts are preserved in thefollowing passages of Vh : 280, 20, pavitthe ya dinayare, virattayams a m j h 8 y a ~; 28 0,2 7, uttartiyam disEyam 32 3,1 8, uttartiyam sedhTyam ;310 , 22 , u t t a r8ya~sedhTya; 312, 8, datttiyam (sc. k a n n f i y a ~ )napabhavai sayano rtiyti v?i ; 210, 1, r8iyam (" at night ") ; 310, 24,Pubhankarcyam nayarTyam.

Feminine locatives in -8yam and -Syam have hitherto not beenknown to occur in any literary Pkt dialect. Their genuineness is,however, corroborated not only by the corresponding Pali forms, butalso b y one solitary form which Pischel 3 388) quotes from a PallavaGrant : Loc. sing. Apitj,iyam (6, 37), i.e. Apit(iyam = Api~t~rn,i.e. th e Pali form. Now if -Zyam an d -Syam are genuine arch aism s, i t isobvious that -tiya and -Tya must also be recognized as such-whichin the case of the dative in -tiya has, of course, never been doubted.We have t hu s throughout an archaic form in -ya, a norm al one in -e,and we may therefore ask ourselves whether we have not here to dowith a uniform and purely phonetic development of final -ya after along vowel into -e. This would remove t h e difficulties with which th eexplanation of th e da t. sing. in -tie is beset (cf. P , 3 364), and wouldmake it unnecessary to have recourse to the Briihmana form mZlCyaifor the explanation of Pkt rnClCe (P, 3 375). B ut whether thisbe accepted or not, i t is a t least obvious th at th e sub stitution of a dativein -?ie for th a t in -tiya cannot be se parated from th e s ub stitu tion of fem.-tie, -7e for -Cya, -iya ; if both changes are not the result of the samephonetic law the former must be due to analogy with the latter.

One of th e most in teresting and remark able archaic forms preservedin T h is an abl. sing. m. n. in -am. I have noted it only thrice : 6, 13,. tao nissaranto duklihn-marunam = tato nirasarisyad duhkha-mara?z?it, th en he would hav c escaped a miserable d ea th ; 146, 29,

Th e use of th e present part iciple as condi tional, ta ug ht by H emacand ra ( i ii ,180 cf. Jacobi, Bhavisatta Kaha 35. I n J E r z no i n st a nc e o c cu rs , c f. J E r z , 1 1 2 ) ,is remarkab ly frequen t in Vh. I hav e note d no less th a n nineteen cases , of which a fewmay he quoted here : 149 11 jai jicnanto nu ento H a d I known ( th i s ) , I shouldno t have come (w i th you) 126 6 jai eua vaggho honto to pad iyam m am am langhento-na esa vaqgho " If th i s were a t iger he would at ta ck me who has fallen down-this isno t iger 110 26, . k i m puna turnam si annesim sichintao " I should love to te l l~ O I I ) ,u t y o u w o ul d te l l o t h er s 120 14 aham jai pamaena niggao honto to nrbandham pacento H a d I s tepped ou t by an ove rsight I should hav e been arres ted.228 25 jai si nu i nt i ttse v6 r?lra?nnu damsenti to m i vivanno honto " H a d y o u n o t c o m e

VOI . WTI. PARTS 2 AS] 3 . 22


12/15

330 L. ALSDORF-pavvaya-kandarap winigga2yci = partlala-kundard winirgatau ; 227, 24,cukkti si ciyciram = bhra$&sy Zccircit. There can be no doubt that theforms in -am are ablatives : none of the verbs nihsarati, vinirgacchati,cukkai = bhraiyati could be constructed with any other case. Nowan abl, sing. m. n, in -am may a t first sight appear rather strange anddoubtful. Yet the development of final -cit into -am is in perfectaccordance with the phonetic laws of Pkt, cf. P, 339 ; 75 ; 114 ; 181.Pischel says in 114 : kMg sakkham = scikscit (Hc i, 24 ; Uttar.116, 370 ; Ovav.), besides sakkhci (Mallikam., 190, 19). AMg h f 3 t h a pbesides AMg JM h6 thci ($107) is an accusative besides an ablative ;the same may be the case with sakkham. The three forms preservedin Vh enable us to decide with absolute certainty that sakkham andhettham are regular ablatives going directly back to scZksEt andadhahsthcit. Their preservation is doubtless due to the fact that theywere used as adverbs and therefore no longer felt to be ablatives.

An additional proof is once again furnished by the Pali, where theabl. sing. in -am also exists, though it is unknown to the officialgrammar.1 Cf. e.g. the well-known stanza Dhammapada 314 (= Samy.Nik., vol. i, p. 49) : akatam duklcatam seyyo, pacchc tapati duklcatam /katam ca sukatam seyyo, yam katvci &nutappati, the first pads of whichis quite correctly rendered in the Udcinavarga (xxix, v. 53, B 41in Pischel's edition) by : akptam kukrtlc chreyah I t is clear that inthe 3rd piida, too, we have to translate krtiic ca sukrtam Breyah. .Or cf. Jiitaka 458, 13d : lcattha-tihito para-lolcam na bhciye ? Beingin which condition need he not fear the other world But it is wellknown that the verb bhT is never constructed with the accusative,but regularly with the ablative, exceptionally with the genitive.or had you not shown me her figure, should have died. 36, 11, ja i te piyiL jiva nto,tu m am vd is attha kusalo honto, to nu esa erisa-sirie b?u?yaaam honto evam s i ~ h d d a g a -tiya caukka caccara racc u muhesuvalalanto viharejja If your father were alive,or if you were skilled in the science of archery, this fellow would not enjoy suchprosperity nor would he thus roam about sportively in . Further instances willbe found in Vh 1 3 , 2 6 ; 1 6 , 2 5 ; 5 1 , 2 1 ; 7 1 , 2 2 ; 1 3 5 , 1 6 ; 1 3 7 , 2 , 2 1 ; 106, 18 (readd e v a v s i . .) 168, 14 ; 1 6 9 , 2 ; 2 2 8 , l ; 248,19.-The last of the sentences just quotedhas already shown that for the expression of the madus irrealis the optative may beused as well. Other instances of this occur, e.g. 17, 4 ; 109, 2 ; 135, 30 ; 322, 16.The two conditional periods 16, 25 f., and 17, 4, are both clear examples of the modusirrealis. Yet in the first case th e present pa rt. is used, in the second the optative. t isvery interesting to notice that Hemacandra in his rendering of the story in question(PariBistaparcan, i, 46 ff. translates the present participles by conditionals (i, 60 ,but the optatives by optatives (i, 72 .

owe the knowledge of the Pali ablative in - a q and of the passages quoted belowto Professor Luders of Berlin.


13/15

SPECIMEN OF ARCHAIC JAINA MAHARASTRI 331The verse Dhammapada 201 : jayam veram pasavati is usuallyinterpreted : H e who conquers ( jayan ) creates ( pa sa va ti) hatred."B u t a much more natural interpretat ion, corresponding much bet terto th e last pad a hi tvi jaya-parci jayap , is : jaylfd vwlram prasravati,

from victory arises hatred." There can also be no do ub t th a tDhammapada 49 : yathdpi bhamaro puppham vanna-gandha?ahethayam lpaleti rasa m ?idiiya, evam gcime mun care m ust be t ra ns la ted :

And a s th e bee, having tak en th e juice, flies awa y from th e flower(puspcit ) without damaging its colour and smell, even so . . . Itcan be shown th at in th e dialect from which t he Pa li scriptures weretranslated t he abl . in -am was more frequent . B ut as in AMg and JM ,only those forms are preserved which had the good fortune of beingeither overlooked or misunderstood.

A few pronominal forms also deserve to be mentioned. I n Vhmay6 = Skt. mayci is the most usual form of the inst. sing. of aham ;i t occurs several hun dreds of times. This for m is nowhere evenmentioned in P, though it had been duly recorded by Jacobi (JErz ,$ 43). It seems, however, that in JErz it occurs only once (10, I),so tha t Pischel m ay hav e regarded this sol i tary form a s a Sanskri tism.I n th e acc. sing. Pischel restric ts t h e use of me to AMg, of te t o AMg, $,Mg, bu t we find th e acc. me Vh 43, 26 ; 44, 8 ; 105, 18 ; 140 ,7 ; etc.,acc. te 65, ; 81, 29 ; 84, 7 ; 87, 30 ; 221, 20. I n th e loc. sing.Pischel knows m ai in $ only : i t occurs Vh 150, 25 ; 241, 13. Besides,we read Vh 282, 21, th e form mamamhi, which is wanting in P .

According to P , $$ 419 ; 422, th e p lural forms ne a nd bhe are used asfollows : ne, acc. M, AMg, inst. gen. only AMg ; bhe, nom. taughtb y one gram marian (Canda), ace. inst. o nly AMg ; gen. AMg and JM.I n V h we find : ne, acc. 121, 13 ; 134, 27 ; inst. 233, 27 ; gen. 43, 3 ;70, 24 ; 73, 23 ; 94, 16 et passim, besides acc. pl. nam (read ne ? ,70, 23 ; 230, 23. bhe, nom. 99, 26 ; 125, 13 ; acc. 118, 4, 8 ; 153, 23 ;367, 17 ; inst. 101, 4 ; 108, 3 ; 112, 10 ; 115, 26 etc . ; gen. incommon use. According to P , $ 422, the gen. pl. vo = vah is used inM, 6 and the Pallava Grants , but not t raceable in other dialects :it occurs Vh 211, 27 ; 224, 3 ; 351, 2. Besides, a n unm istakablenominative vo is found Vh 88, 21.

bhe is explained by Pischel as the result of a weakening of tubbhe,due to absence of accent ($ 422). This exp lana tion is confirmed bya hi therto unknown form found in Vh ; a genitive mhe bears exactlythe same proportion to tumhe as bhe to tubbhe. I have therefore nodoubt as to its genuinenesss, even though it occurs only four times


14/15

332 L. ALSDORFwithin th irt y consecutive lines of th e te xt : 213, 8, ahavfi to [tubbhe]bhtiena kenai mh e um hf i hojjE ? Or should you have got feverthrough some demon ? 213, 15 : n u mhe um hf i sarTrassa, You haveno fever 213, 18 : ker isa m mh e sarTrassa ? How is your health(lit . the state of your body) ? 214, 2 : ma ma pasfiena mhe kajja-siddhT. The readings of th e MSS. plainly show th a t mhe was unfamiliaror even unintelligible to the scribes so that they tried to eliminateit-which for us is a n add itio na l proof of its correctness andauthenticity.It would be easy to add a long list of other forms, words, etc.,occurring in Vh which ar e either completely w antin g in P or a t t r ibutedb y Pischel to AMg only. A brief enu mera tion of a t least some of themmay be given in conclusion.

EyE = fitman ( 88, also JS d E , 130, 16 ; -tra after long vowelbecomes -ya ( $ 8 7 ) : gEya = gZtra, 73, 12 ; 196, 10 ; 320, 14 ; 328, 21 ;goya = gotra, 159, 14 ; vidiyZ = vidya (not in P) , 88 , 14 ; ni t i yam= nityarn (not in P ) , 178, 27 ; ahe = adhas ($ 345), 152, 3 ; 155, 18 ;ahe-loga, 159, 20 ; u y a h u ($ 85 ), 59, 9 ; 151, 13 ; udiihu, 12, 6 ; 37, 20 ;47, 27 ; 126, 23 ; 135, 28 ; 142, 30 ; voc. rciyam ($399, no t in JEr z) ,128, 3 0 ; 131, 9 ; 234, 4, 9 ; 244, 13, et passim ; nom. pl. bahave($ 380, in J M probably wrong for bahavo ) 52, 19 ; 234, 27 ; 310,22 ; 323, 26 ; nom. pl. gave = gavah (not in P , $ 393, but cf. Geiger,Pfil i , 88), 181, 28 ; gtio = gavah ($393), 182, 3 ; kZludhammunZsarnjutta ($404, p. 284 bot tom), 75, 25 ; 284, 20, 25 ; 286, 20 ; 287, 4 ;298, 15 ; 299, 9 ; 304, 4, 5, 7, 9. Absolutives : in -ttEnam ($ 583) :bharettanay, 53, 27 ; paklzhalettanam, 247, 20 ; uddissa = uddikya(not P , $ 59 0) , 1 4 , 8 ; paducca ( $5 9 0) , 5 , 1 7 ; 1 0 , 2 3 ; 280 ,31 ; 3 1 1 , 1 3 ;342, 16 ; 360, 26 ; pappa = prEpya ($ 591 AMg and JS), 235, 5 ;Eyiiya = GdcFya ($ 591 AMg and JS) 163, 20.Vh h as presented us a picture of J M m aterially different from th a twhich is familiar to us from P and JErz . I t s main features a re : first,an even much closer affinity to AMg than that prevailing betweenordinary ill and AMg (cf. P $ 20)-we are markedly nearer t he t imewhen JM came first into being as an individual dialect, distinct fromAMg. Secondly a gre at num ber of archaisms and other interesting

Inserted because the following mhe had become unintelligible 1Only one MS. reads thus all others have seThe editors print uggiL which gives no sense ; cf. the next passage.* One MS. reads thus the others have he.All RISS. but one read mhi.


15/15

A SPECIMEN OF ARCHAIC JAINA MAHARASTRI 333peculiarities which not only in themselves are valuable additions toour knowledge bu t also help to elucidate man y a dar k point of P k tgram ma r a nd linguistic history. Ult im ately th e gre at problems ofthe origin and true character of Ahlg and JJI of their developmentand their relat ions to each other as well as to th e other P k t dialects th ePali etc. are raised anew an d dem and new answers.l A discussionof th em is however quite beyond th e compass of th e present paperthe aim of which has merely been to make known the new materialsupplied by Vh and to direct the attention of fellow Indologists to afield of research which in my opinion badly needs new tilling.

The conclusions arrived at by Jacobi in his paper Uber das Prakrit in derErz zhlungsliteratur d rr Jainas Ricista egli Studi Orientnli 1908-9 pp. 23 f f .will have t o be modified.