Brief Account Of Different Dhatupatras

Author(s): G B Palsule
Publisher: G B Palsule
Catalog link: https://jainqq.org/explore/269621/1

Page #1

BRIEF ACCOUNT OF DIFFERENT DHATUPAȚHAS 37
Coming to the individual classes, classes I and X show considerable difference with those in the P.Dh. Thus in the class I there are only 21431 roots as against 1050 of P. A strange feature of the arrangement of roots in this class is the complete absence of any order in it. All sorts of roots, set and anit, vocalics and consonantals and roots conjugated in different voices have been thrown together. The Anubandhas and fi appear, indicating respectively the atmanepadin and the ubhayapadin roots ending in vowels, but there is no indication whatsoever as to the voice in which a root ending in a consonant is conjugated as, apart from the absence of Anubandhas, the usual statements (viz, such and such roots are parasmaipadins....etc.) also are conspicuous by their absence. Ten roots appear twice, and one, viz. bhe even thrice (1, 17, 148). On the other hand it is very curious that out of the 74 roots in the ghat- sub-class only one, viz. the first root ghat appears here. Then the antarganas are not given separately though a few roots forming these are read here.
Similarly in the X class, yuj and katha sub-classes consist of 9 and 26 roots respectively as against 46 and 86 respectively of P.
In classes II--IX certain points are worth noticing. Some roots which have been read by P in the bhu- class follow optionally, necessarily, or under certain circumstances, different conjugation. Such roots have been removed to their proper classes. Thus the roots éru, aks and tale (also tvaka) have been read in the fifth or the su- class; roots dhiny and lernu (also inv and invas are read in the eighth or the tan-class: the sautra roots stambh, stumbh, skambh and slumbh are read in the ninth or the lericlass; and lastly the sutra 'va bhrāsa-bhidea- etc. which enumerates the roots belonging to the div. class optionally, and which is read in the sutra-patha by other schools is read here in the Dhätupatha, and that too twice (IV, 9-15;
The so-called Sarvavarman's original dhatupatha
After Panini the next school of Sanskrit grammar is the Katantra of Sarvavarman. The dhatup. that has traditionally come down as belonging to the Kt. school is not really the work of Sarvavarman, the founder of the Kt, school. It is Durga who prepared it for that school on the basis of the Candra dhatup. We shall study it at its proper chronological place. Sarvavarman, as we shall later see, did not, in all probability, write any dhätup. LIEBICH, however, has published as Anhang II to his ed. of the Kşiratarangini a dhatup. called Kalapadhâtu-sútra, reproduced from Tibetan sources which he believes to be the original dhätup. composed by Sarvavarman himself.
This dhätup. gives roots with the meanings in Tibetan translation 30 (in Roman transcription) which are the same as those given by Bhimasena. There are the same ten classes with the same sequence.
147-154).
Another important point to note is that the vikaranas as mentioned here at the end of each class, correspond, not to those given by Sarvavarman in the sütrapàtha (yan, , an etc.), but to those which are found in the Paninian system (syan, na, ka etc.).
There are some minor mistakes committed by the Tibetan translator. He has made a confusion between 'asana' and 'asana'. Twice, he has mis
28. LURICH, Einfahrung III, 6.
29. In this respect modern scholars generally follow LADIC, BELYALKAR, Systems, P. WO W INTERNITZ, Geschichte, Vol. II, p. 338; KEITH, A History of Sanskrit Literature, p. 431; Yudhisthira MIMANSAKA, Itthdaa I, p. 407.
30. The name of the translator is glven a Mafijughop-Khadga.
31. And even this figure includes double occurrences of some roots.
32. glai (28.29), vad (35.147), path (37.144), hath (57.143), bhrath (76.121.), dhuahua (17.122), arambh (78.123), dams (79.124), safty (80.125), and srp (105.141).
33. Unlike P.

Page #2

"G. B. PALSULES: taken meanings for roots; thus, rujd which is a part of the meaning mujavisarana-gatyavasddanene assigned to the root sag is torn off and given as a separate root along with saf (Sa. Dh. 1, 173), the meaning vidare going with both. Further, as though this were not enough, Gvaddana 'dishearten', is split into avasa 'not living and adana 'eating'! Similarly vrana gitracinane is given as vrana gatra cürnane (X, 262). On the other hand the sutra basta gandha ardane (P. Dh. X, 144.145) is read as basta gandhddane (X, 122) (i.e. gandha is taken as a part of the meaning). In the X class he reads vij (215) and bicch (203) along-side wuj and biech. A peculiar addition is that of an atmanepadin root kpl (32) in the IV class, whereas kşi is only a passive or reflexive stem from the root kl.
Roots peculiar to the Veda are not read in this dhātup, except Ve (III. 14). An Examination of the so-called Sarvavarman's Dhatupdtha
Now let us examine the above dhatup, and see how far the claim that it represents the original Dhātup. of Sarvavarman is tenable. At the outset we note that Sarvavarman is nowhere specifically mentioned as the author of this work. The work is only styled as "Kalapadhatusūtra's So all that this may mean is that this dhåtup. was believed by some as belonging to the Katantra or the Kalapa system at some stage in its history. It does not follow that this is the original dhătup. of the system,
Secondly, the set of vikaranas of the ten classes as given at the end of each class strongly militates against the claim of this dhatup. having come from the pen of the originator of the Kätantra system. The vikaranas are all (with the sole exception of that of the first class, which is a doubtful case and to which we shall shortly return) the same as belonging to the Papinian system, and not those as are stated by Sarvavarman in the Katantra Satrapitha. Thus byen, énu, ko, ina and S. vikarapas of classes IV-VII and IX-are furnished with the Anubandha & But what purpose does it serve here? In the Paninian system it has a definite purpose. A verbal
BRIEF ACCOUNT OF DIFFERENT DHATUPATHAS 39 suffix which is fit-which has a $ affixed to it-gets the nomenclature sirvadhātuka through P 3, 4, 113. A sarvadhituka affix, further, when it is not pit-when it has no p affixed to it-becomes as good as a nit (weak) affix (P 1,2,4) and thus prevents the guna of the radical syllable. That is how in the Papinian system provision is made to account for the absence of
or the radical vowel in forms like artyati, minoti, tudati, rumaddhi or krinati, although a guna is expected here by the general rule sirvadhatukardhadhatukayoh (P 7, 3, 84).
Does the furnishing of the vikarana with the anubandha s serve any such purpose in the Kätantra syster? No, because the technique of Sarvavarman in this matter is quite different. His general rule nämyantayor dhātuvikaranayor gunah (3.5.1) laying down a guna when an ending follows is restricted in the case of the vikaranas to an by Sa. 3.5.3. So we have the above forms like nrtyati etc. without guna. To prevent further the guna taking place in forms like tudati where also the vikarana is an, Sa. frames a special rule tudader ani (3.5.25). With these provisions in the Satrap to prevent guna in the undesirable places, the furnishing of the anubandha 6 for the same purpose becomes superfluous and consequently could not have originated from Sarvavarman. (On the other hand, the same vikaranas are given in the traditional dhatup., which was evidently so arranged as to agree with the ancient sutrap, as yan, na, aguna an, na-fabda and nd-fabdai.e. without a & and agreeing thus with those in the sutrapatha).
The same is the case of the vikarana of the third or the hw-class, which is given as slu i.e. the same as in the Påninian system. Now in the latter school it is meant to secure elision of the general vikarana fap (P 2.4.75) and also to provide for a reduplication of the root syllable (P 6.1.10). But in the Kitantra system it serves neither of these purposes. The reduplication of the radical syllable of the fuhotyadi roots is laid down in Sa. 3.3.8 while the elision of the vikarana is secured by the rule adader luk vikaranasya (3.4.92) itself, because in the Kitantra system hue class is not a separate one, but just a sub-class of the ad-class-adadyantargano juhotyadih, as Durga says. The reading of a separate class hu and assigning the vilaran su to it, is therefore not from the pen of Sarvavarman (In the traditional dhatup, on the other hand, hu-class is read within the adclass. The concluding remark there reads: ust juhotyldih. iti lugvikarard
34. Curiously enough, the same mistake found in the Klaktana Dh. (1.87) also
35. As opposed to this the current dhatup of the Kitantra school is distinctly ascribed to Sarvavarman. Cf. the initial stanza:
dhatupdhah kpto yena katantranyarthasiddhyaye/
tammal wasty au viduse natatath karvaparmanel/ How far such statements are to be relied upon to a different thing of the 3 Mss. I have used, only the Devangart Mo. of the BORI reada the stanza. The two Bengali Mss of the India Office Library do not read it.
s
36. Cl. Durga under this rule: mina wpadhyl ndyantasya es any eva vikaTone gune bhavati roonte vartate bhavati mayat any epa vikaruse ti kim? niyati
i kerinti pusina
37. Under Ss. 3.4.92.

Page #3

BRIEF ACCOUNT OF DIFFERENT DHATUPATHAS 41 In his Virttikas, of which, to give only one instance, there are no less than sixty-six under one rule (3.2.42) alone. In the present case also it is just possible that Sa. gave only general rules about the vikaranas, and later on somebody filled in the details by making the necessary changes in the dhatup. much the same as Durga did it in his own way by framing Värttikas supplementary to the sutras.
G. B. PALSULES adadayo dhatavah samdptáh. adddau bhadirudddi-jakad juhotyldaya ete catvaro' ntargandh).
At the end of the bhi- class in the Tibetan dhätup. the vikarana is seen to be read as on- vikarama(h) bhūvddayah sampdh-the only case where a vikarana peculiar to Sa is apparently found in the dhätup. But even this single agreement is more apparent than real, for an here is evidently an emended text of the editor, since the sources here read only nyd. and nya- which I think are corrupt readings for nydyye. which, as is well known, means the general vikarana.
There is one more point. In his rule bruvah & ca (3.2.35) Sarvavarman lays down the vikarana nu for Viru and also enjoins the substitution of br in its place when this nu is to follow. From this it is clear that Sa. also like P had retained Varu in the I class and so along with the substitution of $r, he had also to teach the vikarana m u bhayari vidheyath bhauvidikatvat" as Durga says. In the Tibetan dhatup. however, bruis read just in the V class. Had Sa. been the author of this Dh.P. he would have taught the substitution of or only. So this fact also goes against Sarvavarman's authorship of this dhatup. (In the traditional dhatup. we find the root read in the I class in conformity with the Sütrapatha).
There is only one point which seems to be in favour of the view that the dhatup. in question comes from Sa. himself. In the chapter (III, 2) on vikaranas Sa. simply gives the vikaranas of the different classes without stating roots which belong either optionally or under certain circumstances to different classes. Durga has made good this defficiency by way of his own Värttikas under the different rules. Now, as said above, we find most of such roots actually read in the different classes in the present dhātup. and it may be argued from this that Sa. did not treat this subject in the Satrapatha simply because it was unnecessary, as he had already provided for this in the dhatup. itself by arranging his roots to that effect in the various classes and by reading the rule od bhrase-bhlasa.... etc. in the dhatup, itself. But this point can be explained in another way also. Sa is not out to give a thorough grammar of Sanskrit. He does not mind omitting whole chapters-kydanta, for instance and even when he handles a subject he thinks it sufficient to give only the most broad and general rules, dispensing with all the minor details. He disposes the whole subject of voice, for instance, in barely eight sútras, all the details being later supplied by Durga
Considering all these things, I am of opinion that the evidence does not warrant the claim that the dhatup. in question is the original dhätup. of Sarvavarman on the contrary it definitely goes against it. How in Tanjur it came to be associated with the Kalapa school, is difficult to say. Candra Dhātupátha."
A major innovation introduced by Candra is that for every root ho gives, as a rule, only one meaning. Where a number of meanings had been assigned to a root before, usually the first of these is retained by Candra. In a very small number of cases we find two meanings given to a root either expressly or by a ca added after a meaning, probably through inadvertance. .
Another important change is that Candra gives up the use of accents, both on the root and on the Anubandha syllable. The voice is indicated by the descriptive statements at the bottom of the sections, each section conaining only roots belonging to a particular pada. His terms tunin, atafuna and vibhasita correspond to Panini's ātmanepadin, perasmnipadin and ubhayapadin respectively. Similarly he frames special rules in the Satrap. (V, 4) to indicate the roots which take and those which do not take the union-vowel before an ārdhadhatuka affix.
In the arrangement of roots in the first class, again, he differs from Pånini. I. 1-305 are all parasmaipadin roots in the same order as Panini, viz. dentals, gutturals etc.), 306-522 all åtmanepadins (again in the same order; with 496 begin the sub-classes).
39. Published by Lebich in his 'Cindra-Vyakarana (pp. 1-34), Leipzig. 1912. 40. e.g. C.Dh. I, 589, 597 41. e.g. C.Dh, I, 318; 350.
12 The terms this and ateins were evidently suggested to Candra by Panini's rule tahand atmonepadam (P 1,4,100), Technique I, pp. 102-103.
43. The term appears in the P.Dh, also in the sotra faks vibhipito marrone (IV, 78), where it is interpreted variously by the commentators, one of the Inter pretations, being that the parasmalpada is made optional here.
3
bhris-bh Mha-bhramu-kamu-traal-trup-lapl-yol-aayaibhyad on under 3.2.38,
ete.

Page #4

G. B. PALSULE In the other classes also C follows the principle of grouping together the roots taking the same pada. He also brings together in a sūtra roots having the same meaning, thus effecting economy (though this is not carried to the logical extent).
Candra's difference from P is most conspicuous in the X class. Firstly, as against about 395 roots of Panini in this class, Candra has only about 114. He does not read what are apparently denominatives. He also does away with the sub-sections ákusmiya, svadiya, adhrsiya (jadi) and the kathddi (with the sub-section agarviya). The big group of bhartha (v. 1. bhairtha) is also naturally missing.
On the whole Candra's Dhātupatha does not differ much from Pinini's. He retains the sutras like 'ghatādayo mita' etc., 'yarluk' and those at the end of the X class, which others (except Kk, Kt.) properly shift to the Sutrapatha.
There is only one anubandha of P which is dropped by C viz. fil. Jainendra Dhätupitha
This dhatup, has a still different order in the first class. It reads all the atmanepadin ('haidir in J's terminology) roots first, then parasmalpadin (called mavanta), and lastly the ubhayapadins (filt)--an order which Sakatāyana entirely follows.(By the way, J retains the root bhi at its wonted top place, though he reads the section of the atmanepadin roots first whereas SA, who follows the same order, goes one better and removes bha to its proper place viz. Parasmaipadin roots in ). .
In the order of classes, the hu- class preceds the ad-class. In the case of others, the old order remains.
As regards the Anubandhas, J employs two Anubandhas not used by his predecessors, viz. ai (to denote the atmanepada) and au (to denote an anit root, this ou being later used for exactly the same purpose by Vopadeva also). A peculiarity as regards anubandhas in J is that he actually
BRIEF ACCOUNT OF DIFFERENT DHATUPATHAS reads the anubandhus und o after all those 100 which have them while P partly reads them and partly only declares, without actually adding them, certain roots as having those anubandhas. (Here sa follows him in the case of the anubandah, but not in the case of o).
With regard to meaning, J sometimes modifies the traditional meanings so as to effect economy (using mukhaikadese for vadanaika (Vgand); guptau for rakyane (Vgup) etc.), or to give them in a proper form (dasane for dandaküke (Vkhard); sarikane for sankyam (Vrag) etc.) or to make them more clear (wecaih sabde for sabde tare, etc.). He also seems fond of action nouns in-ti (thus he prefers dratau to dariane, arhitew to idane, vilikhitan to vilekhane etc.). Lastly in a few cases, the meanings are altogether different (e.g. daithilye against daurbalye (vkype, Sratha etc.), sarkpobhe against sancalane (Vrubh) etc.).
In the tenth class, J discontinues the practice of reading roots like sutra, mitra etc. (i.e. roots ending in a conjunct) with a final a, since it serves no practical purpose. Here also sa follows him.
The tenth class is broadly divided into two sections, (1) roots which belong to the tenth class alone (1-312), and (2) roots which belong to this class only optionally (313-351). Each of these two groups is further divided into the usual 3 groups, parasmaipadins (1-263, and 313-342), atmanepadins (264-311 and 343-348) and ubhayapadins (312 and 349-351).
J does not read roots restricted to the Vedic use.
The Jainendra grammar has come down in two recensions, a shorter one which is commented on by Abhayanandin, and a longer one with the commentary Sabdarnavacandrika of Somadeva. The shorter version is quite insufficient, the deficiency being filled by Abhayanandin with his Värtti kas. The longer recension is a complete one and it is to this recension that the extant J. Dh. belongs. The dhātup. of the shorter recension, if there was one, has not apparently come down. The roots cited in Abhayanandin's commentary agree with those in the P.Dh. even including the same use of the accents to denote voice and the presence or absence of the union vowel i. This would show that the shorter version had no dhatup, of its own. The Sabdarnavacandrilu dhatup., however, abandons the use of accents, and uses throughout to denote ubhayapada, and A or ai to denote Atmanepada while the absence of any of these signs denotes parasmaipada, roots belonging to this last eategory being called mabantah in the dhatup. Whereas, on the one hand, it is impossible that a new system should not have its own dhatup.. or that a Jain author belonging to a comparatively later age should resort to the use of accents for a technical purpose, on the other hand it is very
44. e.g. II, 1; 12 etc.
45. Like jal (Mila), paky (pap), rag (doa) ete. He however renda denominatives like sohordma (X, 71), mirge (X, 73) ete. Ksirasimin also has noticed this obsence of number of roots in the tenth class. Vide his remarks under P.Dh. x 24.45.49.61.65, etc.
46 Edited by Pandit Ari Lala Jain Sastar as an appendix to the Sabdarnavacandria in the Santana Jain Grantha MAI, Benaras, 1915.
47. In J, the anubandha al is added to the roots in consonants to denote that they are amanepadin.

Page #5

G. B. PALSULES likely that the present dhatup. has undergone a subsequent revision and has received a considerable addition. This is particularly so in view of the fact that it contains a number of roots which are a creation of a later age and are missing even in the Sakatiyana dhatup, (e.g. Ichurd, manth, etc.) which usually follows J.
Kasakstana Dhatupatha.
This dhātuplitha is a recent discovery. The students of Sanskrit grammar already know one Kasakstana, references to whom or to whose work are found occasionally in the Sanskrit' grammatical literature. A few quotations from his work are found scattered in different Sanskrit works. 91 This is all that was so far known about Kasakrsna. This is the first time that a whole work going by his name has been discovered.
BRIEF ACCOUNT OF DIFFERENT DHATUPATHAS
It must, however, be stated at the outset that the present work cannot have come from this Kasakytsna known so far the disabdika Kačakrtsna, to use Vopadeva's term. He is usually supposed to be anterior to Panini, evidently owing to the association of his name with that of Apišali in the grammatical literature. But unlike Apisali he is not mentioned by Pinini in the Astadhyāyi and so some have disputed that claim. Anyway the work which we have now before us must be certainly a post-Paninian work since it uses all the twenty-one Anubandhas of Panini's dhătup, exactly in the same sense. It does not reject any of Pinini's Anubandhas nor does it add any. This single fact is enough to show that the author of the present ahitup, han drawn on Panini and consequently he must have come after Pannini. Tt cannot be ar pued that it is Pinini who borrowed from K akrsna since such a wholesale borrowing on the part of P is unthinkable. Besides, Patanjali (under P 7.1.18) expressly tells us that P does not make use of the Anubandas of his predecessors."
Further, there are other circumstances which would show that our work belongs to a still later date. Thus the set of meanings assigned to the roots here is almost the same as given by Bhimasena and so cannot be earlier than the begining of the Christian era, which is approximately the date of Bhimasena.
Coming to the arrangement of roots, It has striking resemblance with the traditional Kitantra dhatup. Lastly, many of the roots in this dhatup. like dhundh (1, 191) dast (X, 50), fist (X, 48) etc. seem to be of a very late origin. In fact this is the most inflated of the Dhätupathas. All these things would tend to show that the present dhatupōtha--at least in the form in which we have it now-is the product of a very late age and cannot have come from the pen of the ancient Kasakrtsna who in any ense was anterior to Patañjali. For all practical purposes the present dhatup. may be assigned to a period not earlier than 600 A.D.
It will be advisable to give here some aspects of the work in detail because, firstly, the work, which is a new discovery, has been printed for the first time and, secondly, the Kannada script is unfortunately retained in the printed edition.
48. Kalkreme-Sabda-Kalapa Dhidhah of Cannavirskavi, ed. by A. N. NARAKIA, pub. in the Sources of Indo-Aryan Lexicography Series by Deccan College Post-Graduate and Research Institute, Poona, 1952 Cannavirskavi is only the name of the authour of the Kannad commentary which is also printed along with the text. The edition is in the Kannad characters in which both the Mss. on which the ed. is based, were written
49. MIMAMSAKA Mtihdas I, p. 79, has raised a subtle point by saying that the word Kasaktan as the name of a system of grammar presupposes (by P 42,112) K artani as the name of the author, and not Kaakana which, by P 42,114, would give the form Kalksniya as the name of the grammar. But since Kalakana also actually appears as the name of the grammarian, he concludes that the same author was sometimes referred by the name Kasaklana, and sometimes by Kaiaktsni. It seems, however, that it is not quite imperative to bring in Kasakstani. The form Kisakrlana (grammar) can be derived from the base. Kakakana (author) itself by resorting to abhedopacira, just Bhattoji does in the case of the word 'drirako (h m ) in the Kaumudi (under P IV, ill, 87). It is also worth considering whether on the basis of the Varttika
madheyasys wrddhafa vaktavy, the word Kilaksana (author) could not optiotionally cease to be technically called upddhe, in which case Kasakstana becomes as legitimate as Kisakaniya as denoting the work of Kasaktana.
50. The Kasakstana system of grammar is referred in the Bhagya (Vol. I, p. 12). From remarks like "trik Kaskady (Candravrtti 31,42) and 'trikash Kasakstanem (Kaid to P 5.1.23) it appears that Kiaksana's work consisted of three Adhylyasi also seems that he was the first to introduce the principle of brevity (laghava) in the construction of the grammatical rules, d. the oft-quoted statement "Kadakanan gurulaghavan (Kasikh to P 43.115; Comm. to SKBh. 4,3,246; Comm. to Sh. 3,1,182). In his comm. to Bhiaya under P 21.50 and 5,1.50 Kaiyata cites two rules of Kaiaketana. Kpirasimin (quoted by Sysna in the M. Dh.) in his comm. to P. Dh. II, says that the followers of Kisaktana favour the form duasta. Vopadeva (Kkd. 2) mentions Kinkana as one of the digabdikas'.
51. For a good collection of these se MINAMSAKA, Itha I p. 84.
52. CHATTERJI, Technique I, p. 2
53. This statement of Patanjali is, of course, true only to a certain extent. There are many anubandhas in the Panininn system which were used by hie predecessors for the identical purpose, see particularly in this respect Mangala Deva SHASTRE, The relation of Pesini's technical devices to his predecessors, Proceedings of the 4th Oriental Conference, Vol. II, pp. 69-472

Page #6

G. B. PALSULE 2002 The work is of the nature of a dhatuvytus that of Kpirasyämin or of Sayana; only it is on a very small scale. Out of the wealth of forms belonging to the verbal system in Sanskrit, here a single form, viz. 3 sing. Pres. ind. has been cited. The commentary also gives a few nominal derivatives for every root. Every form, verbal or nominal, as also the original Sanskrit meaning, is translated in Kannada. The sequence in each case is: root, its traditional Sanskrit meaning, the verbal and finally the nominal forms (all these elements except, of course, the root being immediately followed by their Kannada translation).
The beginning and the end are as follows: The beginning:
Sri Yagantisarabhaya namah. KätakstmaSabdakaldpa-dhatupathah. bhvadih. 1. parasmaipadabhāpai praktipratyayangiya parasmiy dtmanepade/ namo Yagantikarabhalisgesiya carantinel/ dhatupáthaḥ leto yena bilavyutpattiaiddhaye/
namas tasmai laidakstanagurave Sivardpłyel/ These two introductory stanzas of the commentator are followed by an explanation in Kannada of the second stanza, after which come the remaining two stanzas of the introduction, each followed by its explanation in Kannada. They are :
nipdtas copasargas ca dhätavas ca trayo'py"amt/ anekärtha empoh sadbhih pdfhas tendit nidardanam// sattayari mangale vyddhau nivdswydpticaspadoh/
abhipraye ca saktau ca pradurbhave gatau ca bhat//
After this the main work begins, where there is a repetition of the remarks bhvadih. 1. parc.emaipadabhast and of the introductory stanzas No. two and four. A curious feature of the initial portion is that the portion of the dhatupatha up to rutha, lutha upaghdte (comprising two hundred odd roots) is repeated; the dhatupatha alone appears in the first occurrence, while in the second it is followed by the commentary.
BRIEF ACCOUNT OF DIFFERENT DHATUPATHAS 47
At the end of every Present class there is a colophon which gives minute details about the commentator. The final colophon runs thus:
iti driYaganfisarabhalingaprasidinas TittiraYaju sikhädhyayanasya Vamadevamukhodbhutasya. Gajakarnaputrasya Atrigotrasya Viramihesvaratantrasitrasya Sivelankamaricanapanditdradhyapravarasya Kokilakundapurusya mahamathasya Sarganagurulisiganandyambakumarasya Tumukůramahamathakanthaedryapitsuya Nambyannagurukarajätasya SahyadrikatakaKeladinagarasamipasya paddesasya Kunfidpurasya Kasikinda- Cannavirakavikytaus Kasakrisna-labdakalápasya dvddašasahasradhatupashasya Karnatafikāyim in-vikaranas curadih. samāptam.
As said above, this dhātup. has striking affinities with the Daurga dhätup. of the Katantra grammar. Thus, both treat hu- class as a sub-class of ad-class. In both, the order of roots in all classes is the same. In both the wording of three dhatu-sutras in the ghatādi sub-class which differs from P and C is identical. Both agree in the matter of omission and inclusion of certain roots (e.g. both read Vi after Vkat; both read roots like huel, dhruva, urt (for ert), which are not read by any other schools), in assigning meanings different from others (thus both assign only one meaning viz. pilana to Vav against the nineteen meanings of all other schools; saukhye against sukhe to Vbhand; asamudaye against sariaye to Vcar; fighdhadyam
inst dakandane (or him.dydm, etc.). Sometimes this relationship extends even to minor things and indeed even to discrepancies (thus both assign the meaning kathans against katthans of others to Váalbh; both read Vmov twice (Kk, I 227 and 2626, Kt. I, 1636 and 1966), both read erroneously saka maa kşamdyth ca (Kk. IV, 115; Kt. 118) instead of saka kşamayath, musa ca, orsaka mia ca kepamdydm).
The only difference which Kk. shows from the Kätantra dhatup. consists in reading about 500 roots of a rather odd character which have no
55. This must be amended Into kapeh kytau; otherwise there arises the absurd position of all the previous adjectives in genitive, which are meant to qualify Cannavirakovi, going with dhatupdharya
56. Juala-lala-mala-namonupasargd i and the following two. Whereas the compounds end In genitive in P and C, they end in baninative In K. and Kt.
54. This stane, which contains a salutation to Kikana, evidently belongs to the commentary, and not to the original text, Le, the dhatup.

Page #7

G. B. PALSULE parallel anywhere. Some of these roots are added by extending the present sútras (Le new roots are added against an old meaning) while in the caso of many others new sutras (i.e. both roots and meanings) are framed. A good number of these are evidently meant to supply a verbal basis for certain derivatives which are supposed to be derived from them whilo a few are cases of error, pure and simple. The anubandhas, e.g. are erroneously taken as roots. Thus the root mim-r is separated into two roots mi and my, the comm. giving the two forms mayati and marti(?). The root o-vai is read as u vai with the forms quati and viyati. Even the preverbs are mistaken for roots. Thus ati- is given as a root with the meaning kramaNithaayon whereas really speaking atikramo-him.sayoh is the meaning of the root ap (which is here given as adi); sutra X, 175 reads as bhu klp ava Quakalpane where the first ava is really an erroneous repetition of the preverb GUG- (forming a part of the meaning) turned into a root.
If one ignores this mass of additional roots, then what remains is nothing but the Kt. Dh. There is a highly significant fact to be noted in this connection. The name of this dhatup. as it appears in the final colophon of the commentary is Kdiakstana-sabda-Kalapa. Now it will at once be recalled that Kalipa is also another name of Katantra. The commentator might have deliberately used this word to suggest that the dhatup, which he has commented upon is just a version of the Kalapa or the Katantra dhatup.
There is another aspect of this question. The commentary has quoted a number of sutras, evidently from the Kasakrana grammar, in the course of its explanations. On an examination of these sutras it is found that they betray a strong affinity with the Katantra grammar. Indeed a number of the sutras and technical terms are identical with those in Katantra." This shows that the whole system of the Kasakriana grammar is just another version of the Kätantra grammar.
BRIEP ACCOUNT OF DIFFERENT DHATUPATHAS
There are one or two peculiarities of this dhātup, which may best be mentioned here. Unlike other systems, five roots, viz, củ, tús, pus, mis and si are read with a short vowel, i.e. cus etc, and a special provision is made in the Satrapatha for their lengthening (cupider dirghah'). Sumi larly in the VI class roots pr my, dr and dhr are read as pri, mri, dri and dhri (the comm, gives only such forms as show-ris, priyate, priyah etc., forms with -- are not given). The Kātantra dhatupasha
Modern scholars headed by LIEBICH believe that what passes as the dhatup. of the Katantra school is really Candra dhatup. remodelled by Durga for the Katantra school. As we have already seen, LLEICH has published as an appendix to his ed. of the Kiralarahgini a dhatup.. reproduced from Tibetan sources, which he believes is the original dhatup. composed by Sarvavarman himself. However, even accepting that the current dhatup. is not the original dhātup. of Sa., the one published by LIEBICH cannot, on account of its numerous and important differences from Sarvavarman's grammar proper (i.e. the sutrap.), be that original one. This much I have already shown.
But did Sa, really write a dhātup.? We do not know whether Sarvavarman actually wrote a dhätup. for his grammar. Considering that his grammar was originally meant to be only an elementary one, and that consequently be ignored whole topics like the Samisa, as also the krt and the taddhita suffixes, it is not impossible that either he did not write any, or that if he did, it was incomplete. The various references to the roots or their anubandhas in the sutrap. do not necessarily imply Sarvavarman's own dhatup., for they might have some other then current dhatup. in view, preferably Papini's. Indeed there are some rules touching some aspects of the dhatup, which strengthen the suspicion that when Sarvavarman refers to the dhatup. it is to the Papinian one. While discussing the topic of the pada, Sa. (3,2,42.45) says regarding the roots in consonants that the rueddi roots take the atmanepada endings, and the yaydi roots take
57. See in this connection the present write's paper entitled 'A Glimpee into the Kasakrima School of Sanskrit Grammar read before the 17th session of the All-India Oriental Conference at Ahmedabad, 1953. As this paper is not printed so far, I quote from it a few instances here. Some technical terms used by Kk, which do not appear in Panini and otherwise first seen in the Katantra are: aghon, bandha, ap a, calradhituka, kavarga etc., caturtha, dhum i, vikaran, oh dhyakpara, samana, muara. The common vikaranas are: , ano luk, yan, mud and it. The common sútras are: pah pibah (Kk. p. 71 = Kt, 3.6.70), dhmo dhamah (Kk. 71 = Kt. 3.6.72), wo manal' and dito pacchak (Kk. 71 = Kt. 3.6.74-75), dvayam abhyastam (Kk. 175 = Kt, 3.3.5), uddhir adau sane (Kk 300 = Kt. 2.6.49) etc. - of course there are also a number of technical terms and sutras which are not found in any other known school of Skt. grammar, including even the Katantra
58. I have no knowledge of any publication of this dhatup. so far. I have here used chiefly BORI'S devanagari Ms. No, 252 of 1884-86. Two Bengali Mss. of the India Office Library, viz., No. 773 and 774 were also constantly consulted.
50. L ICH, NGGW, 1895, 316, also Kşiratarangini, p. 213, footnote: BELWALKAR, Systems, pp. 88, 90, WINTERNITZ, Geschichte, Vol. III, p. 388, footnote 3.
60. Though all these three sections appear in the current Kl. grammar, and are commented on by Durgasimha, they did not come from the pen of Sarvavarman, but are later additions, see LIECH, Einary I, 7; Ks. p. 233

Page #8

G. B. PALSULE both the sets of endings. Now what exactly do the expressions rueddi and yajadi mean? Obviously they do not reler to the ganas or classes in the usual sense of the term. While the roots rue and yaj belong to the first class. Itmanepadin and ubhayapadin roots are scattered over all the ten classes, and so it is impossible to form ganas of such roots. Taking, therefore, the word adi in the sense of prakara 'similarity', Durga interprets the word rucid: to mean 'roots with an anudatta anubandha-vowel'; the term yajadi is sunilarly explained as meaning "rools with a samahara (i.e. svarita) anubandha- vowel. Now are we to understand from this that Sa. had his own dhatup, and that in this dhatup, accents on the anubandha-vowels were actually used to indicate the pada of the roots? I hardly think so. Sa. has nowhere taught accents in the grammar and so it is inconceivable that he would employ these for technical purposes in the dhätup. Secondly, if he had actually used the accents in the dhátup, then in the sutrap. he would have straightway said e.g, something like anudattanubandhebhyal (3,2,42) instead of using the round-about expression rucadio. I think that Sa, had no dhätup of his own at all and that when occasion arose he referred to the P.Dh.
It is true that such an outright denial of any dhatup, originating from the founder of the Katantra school might not be compatible with some features of the Kätantra grammar where it differs from the Paninian school. For instance, there is no rule in the Kt. sutrap. which teaches directly the dropping of the vikarana (an) in the case of roots belonging to the third 1.e. hu class, Sa. 3, 4, 92 covering only the ad class. But in his commentary to this rule Durga cites, along with atti and hanti, also forms like juhoti and bibheti, remarking that hu- class is a sub-class of the ad-class (adadyanlargapo juhotyddir iti). From this one would naturally be led to believe that Sa had his own ditup, and that therein he had read ha class as a sub-class of the ad-class, thus dispensing with a reference to it in the sutra 3, 4, 92. I feel, however, that the position can be explained otherwise. The omission of a statement in the sutrap. declaring the hu- class to be a subclass of the ad-class might have been inadvertent and subsequently rectified in the oral vytti until finally the hu-class came to be actually read within the ad-class in the traditional Kt. dhatup. There are many other similar cases of inconsistency or mistakes of omission in the Kt. Thus Sa. 3,6,1 speaks of roots with the anubandha i, but actually it is nowhere taught in the
BRIEF ACCOUNT OF DIFFERENT DHATUPATHAS 51 sütrap;& similarly Sa. 3,2,28 teaches that the dyutadi roots, along with some others, form the Aorist stem in the Parasmaipada with an (corresponding to an of P), but there is no rule in the Kt, sutrap. (corresponding to P 1,3,91) which teaches that the dyitadi roots, which are otherwise atmanepadins, become also optionally parasmaipadins in the Aorist. I think the present case also can be similarly explained as due to inadvertence and that it need not necessarily point to a new dhatup. by Sa.
All things considered, considering that the Kt. grammar was originally highly incomplete and was meant to be only a primary one dispensing more or less with the accessory works, which came to grow around it only in the course of time; that, with a few exception, there are hardly any material deviations from the P. Dh, of such a nature as would have required Sa. to put together an altogether new dhatup: that the solitary pointers to the possibility of Sa, having composed a new dhatup. can be explained otherwise; and, lastly, that a reference to the use of accents for showing the voice of a root definitely points to the P. D.-I am of the opinion that $a, most probably did not write any new dhatup, but depended on the Paninian one.
If Sarvavarman is not the author of the traditional Kt, dhatup, then who is? In this respect the name of Durga at once comes to the mind. Besides being the earliest exponent of the Kt, grammar who wrote down the vrtti on the strap., Duren is also often mentioned in the grammatical literature as an authority on the dhatup. Comparing the views attributed to Durga in the Ksiratarangint and the Purusakira with the current Kt. dhātup., I have found that in the majority of cases they tally with it. It is also interesting to note that though Ks., whose work is replete with citations from older works in the field, obviously quotes from the Kt dhatup scores of times, he never once mentions the name of Kt, but invariably wives the quotations under the name of Durga. All these things point to the conclusion that the authorship of the traditional Kt. dhätup. is to be assigned to Durga.
Citations from Durga's explanations of the traditional meanings of roots are sometimes met with. As only a commentary is a proper place
3. On this Dures remarks: ata era redd Idenband Mole notit-The core responding rule in Pis idite um dhidroh 7.1.58
64. Ky slone, e mentions Durgs 3 times. Deur 3 times and De n trem ance. Pk. mentions Durgs 20 times and Devrih 6 times
65. C. eg. Ks. under I el blu am bhe. sambhah Ieriylnirodhoh, obho dosuddhytkyrtha i Durgh (p. 34); under 16: took free free traco grahan , rencm ifi Durgah.
arsbandha yete
61. runity audibandhopelakonan, under 3, 2, 42; m yejddayah, under 3, 2, 45.
62. The reduplication is taught in Sa. 3.2.8.

Page #9
52 BRIEF ACCOUNT OF DIFFERENT DHATUPATHAS . G. B. PALSULE MOSK, . for explanations, it might be inferred from this that Durga had also written a commentary on his dhatup. LIEBICH, followed by almost all the modern scholars, says that Durga's dhatup. is really just the revised and remodelled Candra dhatup. itself. It is not clear what direct evidence LIEBICH had on the strength of which his statement is based;66 but there are certainly some points of resemblance between the two dhatupathas which suggest that Durga had Candra's dhatup. before him and that he adopted some of its features which had impressed him. The striking resemblance which the Kt. Dh. everywhere (except in the class X) bears with the C.Dh. (as against the P. Dh.) in the matter of arrangement of roots will be patent even to the most casual reader. Occasionally, though rarely, Durga prefers meanings in the C.Dh. to those in the P.Dh.67 Also, Candra's peculiar term for the ubhayapadin, viz. vibhasita is frequently met with in the Kt. dhatup. also. There is one point in the tenth class which bears the most eloquent testimony to the close affinity between these two dhatupathas. Along with many other roots the yujadi class of roots i.e, those belonging to the cur- class optionally, are not read by Candra. In his dhatup. Durga also first gives all the regular roots of this class at the end of which he remarks: kaiscic curadisu wujadayah pathyante.68 and then reads the yujadi roots. This shows that he generally regarded C.Dh. as the model, and therefore was not very enthusiastic about reading the yujadi roots in the tenth class. The order of roots (excepting the tenth class) is, as said above, the same as in Candra, though occasionally Durga changes the arrangement so as to effect economy or regularity. (Thus of the anit roots, tap, yam and nam are shifted to their proper place in the general block [the others however are retained]; similarly 7 roots ri, pi, etc. (C.D. VI, 101-106) have been removed and given their proper place in other parasmaipadin roots). Durga also differs from Candra in omitting certain roots (like tang, tvang, dhvanj, tej, gaj, sphat, vraa, bhran etc.) and adding certain new ones (like dhraj, jim, cucy, jiny, taris, nis etc). He also reads some roots in different classes (e.g. Vdhinv not in the I class but in V). The wording of three dhatusutras in the ghatadi group is also slightly changed by him. In the tenth class, Durga reads, as said above, all the roots known to him and also follows a different principle in their presentation. As mentioned above, he first reads all the roots which belong necessarily to this class (including the sub-classes), then gives certain rules governing the formations of certain denominatives, and only after this he introduces the roots belonging to this class optionally. This process of segregating the optionally curadi roots and reading them at the end is also followed in the Jainendra dhatup, and the arrangement followed here is also the same as in that dhatup. Durga has taken proper care to see that the dhatup. which he wrote conformed to the Kt. sutrap, and that there was no discrepancy. The agreement of vikaranas, inclusion of the juhotyadi class in the adadi class etc. fully demonstrate this. The Kt. dhatup. however differs from the Candra Dh, in two important respects: firstly, the meanings given by Bhimasena are, as a rule, retained, and not curtailed as in C:69 secondly, the tenth class is not curtailed but given fully. As regards anubandhas, ni, the only anubandha which was rejected by Candra, is restored by Durga. 29 A peculiarity of this dhatup is that it avoids artificial technical terms and in their stead uses simple ones. Thus, panubandah (for sitah), manubandhah (for mitah) etc. Vedic roots are, as in C, retained. A fact to note about this dhatup is that it has greatly influenced the later development of the P. Dh. through Maitreyaraksita. There are many roots (like khurd, manth, sphurch, met, mut, ruth, luth etc.) which must have crept from the Kt. Dh. into the P. Dh, through M, seeing that though they have been regularly read by M, Ks. did not find them in his version of the P. Dh. but added them on the express authority of Durga. And this is not to be wondered at, since M hails from Bengal which had been a stronghold of the Kt. The close affinity between the Kk. and the Kt. dhatupathas has been already noticed above. However, in the absence of the exact knowledge of the relative chronology of Kk. and Durga, it is difficult to say who is indebted to whom; is is, therefore, equally probable, that Kk, came after Durga. Under these circumstances, the priority given to Kk. here is to be regarded as purely provisional, and subject to revision in the light of future researches. 68. The NGGW 1895, where LTEBICH has made the statement, was not available to mo. 67. Cf. e.g., klesa vyaktiyar vaci P, but klesa badhane C Kt, hatha pluti-sankubandhanayoh P, but hatha balatkare Ckt. 68. This remark is found in the BORI Ms. but not in the 10 Mss. Since, however, it also occurs in the closely allied Kk dhatup., there is no doubt regarding its genuineness. 69. The case of the Vav to which only one meaning viz. palana is assigned as against the nineteen meanings in the P.Dh. is an exception.