A Problem of Puranic Text-Reconstruction

A Problem of Puranic Text-Reconstruction [puranapathanirdharanavisayaka eka samasya] / By Anand Swarup Gupta ; Purana Deptt., Fort Ramnagar / 304-321

[ srasmin nibandhe nibandhaka puranapathanirdharana visayakah katicit prasna vicaritah | nirdharita puranapathe kesucit hastalekhesu adhikyena praptanamamsanam kim sthanam bhavitavyamiti ca vivecitam | atra visaye kecid vidvamsah purananam hastalekhesu pradhikyena varttamananamam- sanam samsodhita purana samskarane'samavesamicchanti, anye ca punarvidvamsah sarvesameva adhikyena varttamananamamsanam nirdharita puranapathe samyaggrahanam samarthayanti | atra nibandhe hastalekhesu praptanam adhikanamamsanam prathamam catursu vargesu vibhajanam vidhaya tatsamavistanamadhikamsanam upabrha- ranatvam pratipaditam nirdharitapathe ca tesam grahanamucitamiti matamupasthapi tam | tatasca pancame vibhage puranapustakesu praptanam lekhaka pathakadibhih praksiptanam amsanam vimarsam krtva nirdharitapathe tesamagrahanameva samartha - tam | prathamavargacatustayantargatanam adhikamsanam puranagranthasya samsodhita- samskarane grahanaya ka paddhatiranusaraniyetyapyatra kurmapuranodaharanamukhena vicaritam, tadante ca vidvamso'smin visaye svamatapradanaya prarthitah | ] There has been a controversy as to the need and value of a critical edition of the Epic or Purana. Dr. V, S. Sukthankar was the first scholar who undertook to reconstitute the text of the Mahabharata on the modern principles of textual criticism; the critical edition of the Hari-vamsa has also been prepared by Dr. P. L. Vaidya and recently published by the B. O. R. I, Poona. The Ramayana has been edited on the same principles and pattern as adopted for the Mahabharata-edition. The first critical edition of a Purana text, i. e. of the Vamana-Purana, has been published by the All-India Kashiraj Trust on almost the same principles though modified to some extent, and now the critical edition of the Kurma Purana is under preparation. In the West, however, the purpose of textual criticism was to restore a particular text to its original form given to it by its known author whose historicity could not be questioned, a text which had

July, 1970] PURANIC TEXT-RECONSTRUCTION 305 its original autograph either preseved or inferred. The Western textual criticism has been mostly applied in the West to the editions of the classical and medieval texts composed and trasmitted in written form only. And for discussing the textual problems of these fixed (not fluid) texts works like The Editorial Problem in Shakespeare (1951), On Editing Shakespeare and the Elizabethan Dramatists (1955) etc. appeared there.¹ Textual Problems of the Puranas The textual problems of the Puranas are altogether different from those of the Western classical and medieval texts, which may be briefly enunciated as follows:- 1. Both the Epics and the Puranas, as we have them now, are not the works of single individual authors, for they have been compiled, redacted and revised several times and by several hands; and in the course of such redaction and revision their texts have rather been amplified than reduced, and have therefore undergone a vast change both in their forms and volumes. Not being the works of any single individual authors and of any particular time and region the question of restoring these texts to their original form does not arise; for, it is quite doubtful that in view of their fluid nature they ever existed in their purely original form. Though Vyasa has been generally regarded as the author of the Mahabharata and the eighteen Puranas², yet the historicity of Vyasa has been questioned by a number of modern scholars, specialy of the West, who regard him as a mythical person³. We Indians however, are not used to regard all our ancient sages and heroes as mythical figures. But supposing that Vyasa was a historical person and the real author of the Mahabharata and the Puranas, can we assert with any degree of certainty that we shall be ever able to restore the present texts of the Epic and the Puranas to their original form in which they had been composed by Vyasa ? It then should be considered as one of main problems of the Puranic text-reconstruction. 1. Vide Encyclopaedia Britannica, Vol. 22 (1961 edn.), p. 19. 2. Cf. astadasa puranani krtva satyavatisutah | bharatakhyanamakhilam cakre tadupabrmhitam | (Mt.- P. 53.69) 3. See 'Some more considerations about Textual Criticism' by M, Biardeau, Purana, X. 2 (July, 1968) pp. 115 ff,

306 puranam - PURANA [Vol. XII, No. 2 2. Puranas had first been transmitted orally by the Sutas and other reciters before they were committed to writing. In course of this oral transmission the Puranic texts underwent several radical changes such as variations, interpolations (which may better be called additions or amplifications), omissions and transpositions. And when these texts were committed to writing in the form of manuscripts this fluid or changing nature of these texts was arrested to some extent; but similar changes still continued to take place in the transmitted written texts also due to the scribal slips and emendations and also to the readers' attempts to supplement or change the texts according to their own views; regional cultures and ideas also effected several changes in the Purana-texts. Consequently various versions and even recensions arose in different regions and times, which versions must have been far distant from the original Puranic text, if ever there existed such a text. Can we, or should we, then reduce these various versions to any single universal version by our text-reconstruction ? 3. Manuscripts of the Puranas generally date earliest from the 14th or even the 15th century A.D. onward, and the oldest Nibandhas (e.g. Vallalasena's Danasagara and Laxmidhara's KrtyaKalpataru) which quote the Puranas were not composed earlier than the 11th or the 12th century A.D. But many of the extant Puranic texts took their shape between the 3rd century B. C. and the 3rd or the 4th century A D., when they were probably committed to writing. Before that period Puranas might have had a long-standing oral tradition of transmission from the time of the composition of the original Purana-Samhita which according to the Visnu-Purana (III. 6. 15) was composed by Vyasa and taught by him to his disciple Suta Romaharsana. Now, on the basis of such late manuscript-and Nibandhamaterial is it possible to arrive at the reasonably oldest authoritative Puranic texts which may be nearer to the inferred original or to the lost archetype? 4. There have undoubtedly been later additions in a Puranic text from time to time and from region to region. Are all these later additions the results of the tampering attempts of the interpolators beyond any justification, or are these additions the results of the desire on the part of the redactors to revise the texts

July, 1970] PURANIC TEXT-RECONSTRUCTION 307 of the Puranas from time to time and keep them in line with the current religious and social ideas of their times in order to preserve the encyclopaedic nature of the Puranas and keep them upto-date. Should a text-critic or a critical editor spurn all such later additions as spurious matter from his critical text, giving them an inferior position in his critical footnotes, or relegating them to an appendix, or altogether rejecting them and thus driving them to oblivion ? These are some of the main problems which confront a critical editor in reconstructing his text of a Purana. There is yet another point: It is said that a Puranic text has different textual traditions preserved in the different regional versions which are available in their respective regional manuscripts written in their regional scripts, and that if a single critical text is prepared on the basis of the manuscripts of the various versions, a new recension or version might arise and the various valuable Puranic traditions may be lost to us. It is for this reason that some scholars, such as Prof. Sylvain Levi and his pupils, are tolally against any critical edition of the Epic or a Purana.¹ Each of the problems mentioned above requires some detailed consideration and discussion. I shall here confine myself to the last mentioned problem (No. 4); i. e. the problem of the later-on added material in a Puranic text; for, the main structure of a constituted Puranic text would depend on the proper solution of this important problem. But before considering this problem in detail a few words may here be said about the need and importance of the critical edition of the Puranas (inclnding the Epic): Need of Critical Editions of the Puranas The texts of the Puranas are preserved in manuscripts which are scattered all over the country. We find manuscripts of the Puranas from Kashmir and Nepal in the extreme north to Temilnad and Kerala in the extreme south, and from Asam in the east to Maharastra in the west. This wide area of the availability of the Purana-manuscripts has naturally given rise to the different versions of a Purana-text in different regions and scripts. But this does not mean that a Purana-text has different text-traditions 4. Ibid. 14

308 puranam - PURANA [Vol. XII, No. 2 in the same way as a Vedic Samhita has different established texttraditions in the form of its Sakhas, in which case particular Vedic Sakhas are followed, studied and adhered to by particular Brahmana-families, so much so that the daily worship and the religious rites of these Brahmana-families are performed according to their own particular Sakha. The texts of these Vedic Sakhas have been preserved intact and without the least change by these Brahmanafamilies with utmost care and sanctity. It is for this reason that no single Vedic text can be reconstructed out of these various Sakhas. The study of these Vedic Sakhas had remained confined to the priestly class (Brahmanas) only and had never been allowed to be carried on by the laity. The case of the Puranas is, however quite different. The Puranas have always been the popular religious works, they were made accessible to every cast and fold of the Hindus. They are publicly recited and no such sanctity, therefore, was attached to their recitation as to that of the Veda. The whole Vedic texts had been memorised and are even today recited by memory in the particular Brahmana-families, but not so the Puranas. Moreover, no particular text-tradition of a Purana is adhered to in particular regions or in particular religious sects; for, the vulgate text of a Purana printed in Devanagari (which has been a universal script for writing Sanskrit texts) is generally used all over the country by the reciters, readers and also by Indian and foreign scholars. It is, therefore, not improper to reconstruct a single critical text of a Purana on the basis of the available manuscripts of all its versions. Such single critical text must be a conflated text by its very nature, but this defect is more than compensated by giving the readings and variants of all the available versions in the critical apparatus (in the form of the critical footnotes) of a critical edition. Again, no single manuscript of a Purana is found absolutely correct and the text found in most of the manuscripts, specially in the Devanagari manuscripts, is conflated. The printed editions of the vulgate text of the Puranas, based as they are on more than one manuscripts, are also conflated to some extent, some portions agreeing sometimes with particular manuscripts and sometimes with others. To avoid conflation the text of a single unconflated manuscript need be printed, but even then the conflation cannot be wholly avoided, for we have to take the help of other manuscripts

July, 1970] PURANIC TEXT-RECONSTRUCTION 309 also in order to corroborate and correct the readings of our manuscript and to fill up the lacuna which is sometimes found even in the best of the manuscripts. And even then we cannot have the complete or the correct pricture of the text of the Purana; for, a reading common to several versions is often (though not always) superior to a reading available in one single version or manuscript only. So the printing of the text of even a best single manuscript would not serve the end of textual criticism.5 Then again, sometimes a text in the majority of the manuscripts even of different versions gives a wrong reading or have some lacuna causing a wrong or incomplete sense; and then to our pleasant surprise the reading (which does not seem to be an emended one) found in some single or a few manuscripts gives us the clue to the correct and complete sense. In the Introduction of the critical edition of the Vamana Purana (pp. xxxiii-iv) I have given some instances to show how the Kashmirian manuscripts only provide us the text of a sloka which has been lost in all other versions, and without which the sense of some slokas would have remained incomplete or unintelligibte. Instances are also found in the manuscripts of the Kurma-Purana where majority of the versions and the manuscripts are wrong and the correct reading is found only in a few manuscripts. In one place (corresponding to I. 46.20 of the Venk. edn.) a single manuscript of the Kurma-Purana (viz. the Grantha MS. purchased from Madras) has supplied us an additional sloka which seems to have been omitted by mistake in other manuscripts, for this additional sloka of the Grantha MS. completes the sense. 5. Cf. also the following remarks under 'Textual Criticism' by its author in the Encyclopaedia Britannica, Vol. 22 (1961) "The school of textual critics stemming from Karl Lachmann favoured the acceptance of one 'best' manuscript and the rigid adherence to its readings in all possible circumstances as the only 'scientific' attitude for the editor. This school tacitly considered that textual criticism ended with the establishment: of the family tree and the consequent decision in favour of some one preserved manuscript as the 'best'......The illogic of the Lachmann position, in its abrogation of the critical second stage. in preparing a definitive text, was exploded by A.E. Housmann (Preface. The Astronomicon of Manilius, Cambridge University Press)".

310 puranam - PURANA [Vol. XII, No. 2 The edition of the text of a single manuscript or of a single version of a Purana may, therefore not provide us a correct and authoritative text throughout. Hence there is the need of a critical edition of a Purana based on its available manuscripts of different versions and scripts. Evolution and Growth of a Purana-text According to the Indian tradition relating to the origin and evolution of the Puranas as recorded in the Visnu-Purana (III, 6. 15ff), Vayu-Purana (60. 2; 61.55ff.; 104. 2ff.) etc. Vyasa first composed a Purana-Samhita containining the akhyanas (tales and legends), upakhyanas (episodes), gathas (slokas handed down from the ancient times) and Kalpajokti s (lore coming down from the ages). Vyasa taught his Purana-Samhita to his disciple Suta Romaharsana, who also composed a Purana-Samhita, which was the basis of the other three Purana-Samhitas composed by three disciples of Romaharsana, viz. Kasyapa (or Akrtavrna), Savarni and Samsapayana, These four Purana-Samhitas were the purva (Vayu.-P.) or the the mula (Bhag.) Purana-Samhitas, each consisting of 4000 slokas excepting the Samsapayana-Samhita which consisted of 8600 slokas. Later on eighteen Puranas were evolved out of these four original Purana-Samhitas, which all ascribed to the authorship of Vyasa. The Vayu-Purana says that originally a Purana consisted of four pada s, and contained 12,000 slokas only (32.62-63). The Bhavisya-Purana (I. 1. 103) also corroborates the Vayu-Purana when it says that all the Puranas are said by the wise each to have been of 12,000 Slokas ( "sarvanyeva puranani sajneyani nararsabha | dvadasaiva sahasrani imate Halfaft: 11") But the Bhav.-P. further remarks that later on they increased in their extent by incorporating in them akhyanas of various kinds ("gaf"). Still later on, the extent of the Purana-literature swelled up to four lacs of slokas ("evam puranasamsthanam caturlaksamudahrtam | " Bhav. P.) Thus the Purana-literature from the few thousand slokas gradually grew up to an enormous extent of four lacs of slokas, and the authorship of all these four lacs of slokas was attributed to Vyasa ("agdame di anata-P., 53.57). Thus, all this growth in the extent of the Puranas is admitted by the Puianas

July, 1970] PURANIC TEXT-RECONSTRUCTION 311 themeselves; not only this, but all this addition or growth is also held to be valid and authentic by ascribing it to Vyasa The Puranas have always been a living literature, and have This living acquired a unique popularity and sanctity in India. literature of the Puranas may be likened to a living human organism. The spontaneous growth in the form of additions and amplifications (upabrmhana) in the body of the Purana-purusa is like the natural growth of a human organism. The grown-up form of a Purana-text, therefore, is as important and valuable as its original or the pristine form, just as a grown-up human body is not less important than its early and undeveloped baby-form. But sometimes spurious matter also makes its appearance in the Puranas, like some unnatural redundant growth of a limb in the human body. In that case the spurious matter is to be carefully distinguished and separated from the naturally grown-up (upabrmhita) body of a Purana-text. Nature of Additions in the Puranas Additions in the Puranas may be classified as follows:- 1. The natural growth or upabrih na of a Purana-text, i. e. the growth or upabrmhana effected by the redactors or the Suta s. Such growth or addition is accepted and assimilated by the Purana in its body of the text and is found in nearly all of its versions and manuscripts. The following instance from the Kurma-Purana (KP.) will make this point clear: The KP., I. 51.14 (Venkt. edn.) reads as follows:- sibirindrastathaivasicchatayajnopalaksanah | babhuva samkare bhakto mahadevarcane ratah || The first line of this sloka has its verbatim parallel in the Visnu-Purana (III 1.17ab), but the second line ( 1) is not contained in the Visnu-Purana. Now, if we regard the entire Purana-literature as one single whole, of which the various Puranas form different parts or chapters, then it can be safely said that the KP., being predominently a Saiva Purana, had added this line in the body of its text and so this line is found in all the versions and manuscripts of the KP.; it has, therefore, to be given a place in its constituted text.

312 puranam - PURANA [Vol. XII, No. 2 Compare also the following sloka of the KP.- svarocisascottamasca tamaso raivatastatha || priyavratanvaya hyete catvaro manavah smrtah | (I. 51. 19-20) This sloka of the KP. is the same as Vis-P. III. 1. 24, but the Vis.-P., being a Vaisnava Purana, adds the following Sloka immediately after this sloka :- - visnumaradhya tapasa sa rajarsih priyavratah | manvantaradhipane tallabdhavanatmavamsajan || 11 (Vis.- P. III. 1.25) The KP. does not contain this sloka, so it should be considered as an addition in the Visnu-P. rather than an omission in the KP.; it has become an inseparable part of the text of the Visnu-Purana. of these 2. There are found certain passages or Slokas in a Purana which are available in some of its versions and manuscripts only, while in others they are missing or omitted. Now if this omission is genuine in these versions and manuscripts and not due to haplography or other scribal slips, then the presence passages or slokas in the other versions and manuscripts should also be regarded as genuine additions there, additions which have been accepted and assimilated by the text-tradition of at least the versions and manuscripts containing them. Sometimes such addition is found in one version only and sometimes in more than one. But this additional matter should also be considered as valuable and fit to be preserved, although it is not uniformly available in all the versions and manuscripts of the Purana. Now, if we exclude such additional matter from our constituted text and give it in the critical apparatus or relegate it to an appendix, then there is every chance of its being lost for ever from our text, if the constituted text only is to be separately printed without the critical apparatus and appendices for the purpose of translations or for the purpose of general reading, recitation or study. It may be noted here that in the case of the Critical Edition of the Mahabharata the editors' efforts have been to restore the text, as far as possible, to its pristine form, and so they have tried to exclude from their constituted text all the additions which they considered might have been introduced later on, and have given

July, 1970] PURANIC TEXT-RECONSTRUCTION 313 all such additional passages in the critical footnotes, and the longer passages in the Appendix. This is one view of dealing with the inferred additional passages in the Epic and the Puranas. But strictly speaking this kind of application of the canons of textual criticism can be justified in the case of the edition of a Western classical (Greak or Latin) text only, for which purpose these canons were evolved in the West. Another extreme view is that of Sylvain Levi and his followers which would not allow any exclusion from the extant texts of the Epic and the Puranas as preserved in the various versions and manuscripts. In the case of the Critical Edition of the Vaman Purana (All India kashiraj Trust, 1967), however, we resorted to a middle path. The Saro-mahatmya text of the VamanaPurana consisting of 27 Adhyayas, is not available in the Kashmirian (only first 10 Adhyayas available), Bengali and South Indian manuscripts of the Vamana Purana, but we retained this text within the constituted text of the Vamana Purana; we, however, numbered its chapaters separately from the main text of the Vamana Purana to show that according to the editor's view this Saromahatmya text might not have been the original part of the Vamana Purana text. But if we think that a critical editor must exclude all the additional (called as spurious and interpolated) matter from his constituted text in order to arrive at the oldest possible text of his Purana, even then it is not possible; for no critical editor has got the entire and the complete manuscript-material at his disposal for scrutiny. In the Introduction (p. XVII) of his critical Edition of the Harivamsa Dr. P. L. Vaidya says that "the Critical Text in this edition represents a phase which is positively older than A.D. 1050, the age of Ksemendra...... But this does not mean that there are no interpolations in my text." And he further says (pp. XXXIV-V)-"Even my Critical Text might be an expended version of the original text of Harivamsa, but I cannot further reduce it because of my MSS. material and its abridgement preserved in the Manjari." So if the complete manuscript-material, without any break, from the time of the first archetype upto the time of the latest manuscript, had been available to us, we would then have been able to show that such and such additions were made at such and such time to the original or the oldest text of the Purana, and would then have been justified to exclude all the

314 puranam - PURANA [Vol. XII, No. 2 later-on-added (or spurious) matter from our constituted text, which we might perhaps, to some degree of certainty, claim to call as the oldest text. It is not, however, possible now, for the manuscripts written earlier than the oldest manuscripts now available to us are perhaps, irretrievably lost. In such a state of affairs why then should we exclude from our constituted text only those additions which may be known to us as such on the basis of the meagre and insufficient manuscript-evidence, and allow all the passages which might also have been earlier additions introduced at some unknown stage of the growth of the Purana an honourable place in our constituted text ? These remarks may be labelled as uncritical, but they may perhaps conform more to our Purana-tradition. In my humble opinion all such additions should also be included in the constituted text. We should, however, use some device to indicate that as they are not uniformly available in all the versions they might be treated as belonging to the textual tradition of those versions and manuscripts only in which they are available; but care should be taken to ascertain that they had not been inserted in some common exemplar of those manuscripts from a different source by the scribe or reader of that exemplar. 3. Sometimes there are found two parallel versions of some portion or portions of the text of a Purana in its manuscripts; the one version being shorter which is available in some of its manusecripts, while the other version is longer which is also available in the other set of manuscripts of the same Purana. In the case of the Kurma-Purana this kind of of double version is found in its manuscripts in several places of the text. On account of the different wordings and construction of these two versions they can not be amalgamated with each other. Both these versions-shorter and the longer-should be accepted as authentic, i. e. as acknowledged and assimilated by the two different manuscript-traditions of the Purana, althouh the longer version may be a later one (which, however, is not always the case). 4. Sometimes a text in almost all the manuscripts is wrong or incomplete and does not, in some cases, also tally with the preceding and the following text. I have already referred in these pages to such texts in the Vamana and the Kurma Puranas. In such cases the additional line or sloka available in some single version or

July, 1970] PURANIC TEXT-RECONSTRUCTION 315 even in a single manuscript, which completes or corrects the sense of the text of the remaining manuscripts, should be included in the critical text, assuming that this additional text might have been rightly preserved in that particular version or manuscript only, and some-how-or-other lost in others. 5. Besides these authentic or justifiable additions, we also find in some manuscripts of a Purana such additional matter as is really spurious, introduced by a scribe, reader or reciter according to his particular leanings or predilections either from some other source or composed by himself. Such spurious additions or interpolations are made with a desire on the part of the interpolator to amplify some description, deva-stuti, mahatmya, phala-sruti; etc.6 Some additions are tinged with sectarian zeal and motive; e. g. a Nandi-nagari MS. of the Kurma-Purana, procured from the Sringeri-Matha, Mysore, adds a sloka of Saivite nature after KP. I. 2. 16cd-17ab (Venkt. edn.), and this Sloka of the Nandi-nagari MS. is not found in the other collated MSS. of the Kurma-Purana. The Saivite spirit of this MS. is borne out when it alone subtitutes the reading "sarvam sivamayam jagat " for the reading "sarvam brahmamayam jagat " contained in all the other MSS. in I. 4.65d. Such spurious additions or interpolations are found sometimes in one manuscript only and sometimes in more than one MSS., 6. Dr. P. L. Vaidya, in the Introduction (p. XXXV) of his Critical Edition of the Harivamsa has given the following sloka which mentions the five groups of interpolations introduced in the Epic and Purana :-- samgramasca vivahasva stutirdevivaradikam | praksepakaranabhyahuh pancamam ca phalasrutih || 7. KP., I. 2. 16cd-18ab (Venkt-edn.), reads as follows:-- 15 ye yajanti japairhomairdevadevam mahesvaram | svadhyayenejyaya durattan prayatnena varjaya || bhaktiyogasamayuktanisvararpitamanasan | pranayamadisu ratan durat pariharamalan || Between these two Slokas the Nandi-nagari MS. inserts the following Sloka :-- namaskaradisamyukta siva ityaksaradvayam | jihvagre vartate yesam tan prayatnena varjayet || But no other MS. of the KP. contains this Sloka.

316 puranam - PURANA [Vol. XII, No. 2 generally in those which might be the direct or indirect copies from the same exemplar containing that spurious reading. These interpolations are to be excluded from the constituted text and should be given a place in the critical footnotes or appendix. text. ADDITIONS AND TEXT-RECONSTRUCTION Out of these five categories of additions found in the manuscripts of a Purana, there should be no controversy about the first category (i.e. the additions accepted by and assimilated in the text of all the versions) to be included in the constituted text, and also about the fifth or the last category (Additions which are clearly spurious and interpolated) to be excluded from the constituted But the second, third and the fourth categories of additions, as detailed above, should, in my opinion, also be included in the constituted text; in such cases of inclusion, however, we should use some device to indicate that these additions are not uniformly available in the text-tradition of all the versions as I have already mentioned above; for example, we may mark these portions in our constituted text by a vertical line and (or) number these passages separately. I shall now illustrate my point by giving a few illustrations from the MSS. of the Kurma-Purana. The following symbols have been used by us for the 21 collated MSS, of the Kurma-Purana :- = Kashmirian Version MS. No. 3563, Raghunath Sanskrit Library, Jammu. Bengali and Uriya Version 1=MS, No. 2885, Dacca Unieersity Library, Dacca, Pakistan. 2=MS, No. 3390, Dacca University Library. 3=MS, No. 398, Asiatic Society, Calcutta. 31 = Uriya MS., No., 75139, Adyar Library, Madras. (Containing only the Uttarardha of the text). Devanagari MSS. (North and West) 1 MS., E. 3346, India Office Library, London. = 2 Ms. Tod 39, Royal Asiatic Society, London; A.D. 1598. = 3 MS., No 5589, V.V.I., Hoshiarpur, Punjab. = 4 MS., No 999, Banaras Hindu University Library. 5 MS., M.C, 371, Oriental Research Institute, Mysore. ==

July, 1970] PURANIC TEXT-RECONSTRUCTION 6 MS., No. 1039, Harvard University (U.S.A.) 317 7 MS. E. 3345, India Office Library, London (collated = Bhuvan-kosa only). 8 MS., 41 of 1881-82, B.O.R.I., Poona; A. D. 1558 (the oldest MS. collated). 9= MS., 16 of 1873-73, B.O.R.I., Poona; A. D. 1565 (the next oldest MS). 10 MS., PM, 2418, Adyar Library, Madras. South Indian MSS. 1 = Nandi-nagari MS, Srngeri Matha, Mysore, Palm-leaf. a 1 = Telugu MS., D 2107, Govt. Oriental MSS. Library Madras. * 1 = Devanagari MS., No. 1588, Sarasvati Mahal, Tanjore. 1 = Grantha MS., Purchased from Madras, Palm-leaf. 2=Grantha MS., No. 5036, V.V.I., Hoshiarpur, Palm-leaf. # 1 = Malayalam MS., No. 110/19129, Kerala University Trivandrum, Palm-leaf. Now, about the place of additions of the three categories (2-4) in the constituted text we shall first take the case of the text which is found in some of the MSS. only but omitted in others. (References from Venkt. ed.). After KP., I. 53.9 we have the following three lines which are found with some variants in ka 1 ba 1-3 de 1-3.5.6 te 1 gra 1, 2 ma. 1, but are omitted in 4. 8-101 and 1. So we can mark these lines with a vertical line and number them separately in Lines :- mahayamo munih suli dindimundisvarah svayam | sahisnuh somasarma ca nakulisvara eva ca || 9 vaivaravate 'ntare sambhoravatara strisulinah | 1 astavimsatirakhyata hyante kaliyuge prabhoh | 2 tirthe kayavatare syad deveso nakulisvarah | 3 tatra devadhidevasya catvarah sutapodhanah | sisya babhuvuscanyesam pratyekam munipungavah || 10 The critical apparatus of these three additional lines may be given Line-wise ofter the critical apparatus of sl. 9.

318 puranam - PURANA [Vol. XII, No. 2 These three lines are omitted in the oldest MSS. 8.9 also; but in certain cases a text is found contained in both these oldest MSS. and also in 10 (which three form a separate Group) but omitted in all other MSS., e.g. KP. I. 26.32. In this case also these two lines (Sl. 32) may be included in the constituted text but marked with a vertical line. (2) In the Kurma-Purana MSS. there are some cases of two parallel versions, one shorter and the other longer. These should be considered as the two independent text-traditions, as as the readings and the construction of these two versions indicate: they run parallel to some extent, the longer version then gives some additional text which is absent in the shorter version, and then the two versions culminate in a similar text which has variants separately conforming to the two texts of the two versins. Both of these should be adopted in the constituted text. But the question is how to adjust these two parallel versions in the same place. If the constituted text is printed in single column, then these two texts can be given side by side in two columns in padas where they differ from each other in variants in their culminating portions. For example, compare the following text :Kp. I. 24. 11 has the following two versions :-- kathasyabhavat kuntirvrsnistasyabhavat sutah | 11 A vrsnernivrttirutpanno dasarhastasya tu dvinah | 1 dasarhaputro'pyaroho jimutastatsuto'bhavat | 2 jaimutirabhavad viro vikrtih paravirahah | 3 tasmannavaratho nama babhuva sumahabalah || 11 B tasya mimarathah putras tasmannavaratho'bhavat | danadharmarano nityam satyasilaparayanah | 5 The Here the two versions of the genealogy of Kings after Kratha, the son of Videha and the grandson of Jyamagha, are recorded. shorter version is given in 1.8 9 10 1, and the longer is given

July, 1970] PURANIC TEXT-RECONSTRUCTION 319 in the remaining MSS. The two versions representing the two separate traditions may be shown as follows:Shorter Version (de 1.8-10, na 1) katha Longer Version. (ka 1 ba 1-3 de 2-6 te 1 da 1 gra 1.2 ma 1 ) kratha kunti vrsni U 1-1-1-1-1-1-1-1-1 navaratha } = navaratha } W. Kirfel in his two works, Das Purana Pancalaksana and Das Purana vom Weltgebaude, has adopted this method of giving the two parallel texts of the Purana side by side. But if the constituted text is printed in double column, then the above method may not be convenient. In that case the longer version may be given in the constituted text and the common culminating portion of the shorter version (e.g. here the two padas of 11B) is then to be given as variant in the critical apparatus ; but that would not be a quite correct method; we have, however, to adopt it for convenience's sake.

320 puranam - PURANA [Vol. XII, No. 2 According to this latter method the above text may be given as follows : krathasyabhavat kuntirvrsnitasyabhavat sutah | vrsnernivrttirutpano dasarhastasya tu dvijah || 11 dasarhaputro'pyaroho jimutastatsuto'bhavat | jaimutirabhavad viro vikrtih paraviraha | 12 tasya bhimarathah putras ] tasmannavaratho'bhavat | danadharmarato nityam satyasilaparayanah || 13 [The vertical straight line indicates that the text marked by it is available in some of the versions only, and the horizontal wavy line denotes that the text marked by it has its alternative or variant in some versions or manuscripts. ] (3) If a version or a manuscript contains some additional line or lines which complete the sense of the text given in the other manuscripts and thus fills up their lacuna, then this additional text ought to be included in the constituted text, but marked with a vertical line to indicate that this additional text is available in some MS or MSS. only and not in all the MSS. of the critical appartus. For example, KP. I. 46.20 (Vaikt. edn.) reads as follows :- tatrapsaroganaih siddhaih sevyamano 'maradhipaih | aste sa varuno rana tatra gacchanti ye'mbudah || Here in 20d 'ye'mbudah ' is the reading in the Vefkt. edn and in de 1.8.9. only, de 10 reads 'ye budhah ', but the majority of the remaining MSS. read here 'ye mudra ' or 'ye narah ' (South Indian MSS. ; te 1 - 'ye sada '). All these readings ('ye muda ' 'ye narah ' 'ye sada ' etc.) require a clause to complete the sense, this clause or line is available in gra 1 follows :- tirthayatrapara nityam ye ca loke'ghamarsinah | This additional line in 1 completes the sense and also fits well with the Vedic and Puranic conception of Varuna and his devotees. The whole text of I. 46.19-20 may be given in the constituted text of the Purana as follows:-

July, 1970] PURANIC TEXT-RECONSTRUCTION pascime parvatavare varunasya mahapuri | namna suddhavati punya sarvakamaddhisamyuta || 19 tatrapsaroganaih siddhaih sevyamano'maradhipaih | aste pta varuno raja tatra gacchanti ye muda ( or narah ) | | tirthayatrapara nityam ye ca loke'ghamarsinah || 20 321 We should thus adopt the text as it has been accepted by the manuscript-tradition of the various versions of our Purana. The task of the critical editor and the text-critic should be to indicate in his adopted text the various stages of development or growth of the text, for which he has to use some suitable device as suggested above. I have thus tried here to place before the scholars a few suggestions for dealing with the additional texts of the manuscripts in the constituted text of a Purana. Comments and suggestions from scholars interested in the problem are welcome.