% Mahabharata: General information % Last updated: Fri Sep 25 2020 % Encoding: ASCII % Electronic text (C) Bhandarkar Oriental Research Institute, % Pune, India, 1999 % ================================== Electronic text of the Mahabharata ================================== This file contains information about the electronic text of the Mahabharata. Anyone who intends to make use of the text is asked to read it carefully. It is divided into four sections: * The status of the electronic text * The history of the electronic text * The format of the electronic text * Notes on the electronic text John D. Smith University of Cambridge ------------------------------------------------------------------------ The status of the electronic text ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ The electronic text of the Mahabharata is Copyright (C) The Bhandarkar Oriental Research Institute (BORI), Pune. This authorised and regularly updated text is available only via the web page http://bombay.indology.info/mahabharata/statement.html. Please do not provide copies of the text to others. The text is as accurate as possible, but errors may well still appear. If you believe you have found an error, please report it by email to Suvarna Deshpande (suvarnad09 [at] gmail.com) or Pranav Gokhale (pranavpg88 [at] gmail.com). Before reporting any error, please read the rest of this document! Corrections will be made to errors remaining in the text as they come to light. Check the “Last updated” date at the head of each parvan against the current dates listed on the distribution website (URL https://bombay.indology.info/mahabharata/welcome.html) to see whether you need to download newer versions. ------------------------------------------------------------------------ The history of the electronic text ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ This text has its origins in the work of Prof. Muneo Tokunaga of the University of Kyoto. During the late 1980s and early 1990s, Prof. Tokunaga typed the entire text of the Critical Edition of the Mahabharata (BORI, 1933-66) into a computer, and in 1994 he placed the electronic text he had thus created on the Internet. The enormity of the labour involved in the creation of the text is matched only by Prof. Tokunaga's generosity in making it so freely available to the scholarly world. This version of the text derives ultimately from that first version, and I am deeply grateful to Prof. Tokunaga for permitting this use of the products of his labour. In transcribing the Mahabharata, Prof. Tokunaga took a number of policy decisions which can cause difficulties for would-be users whose requirements differ from his. Most of these -- for example, his use of “m” for both m and anusvara, and “n” for the dental, palatal and velar nasals -- can be overcome with a little effort. Much the most serious problem is caused by his use of a single separator to divide word from word and compound-member from compound-member (something he did not do in his transcription of the Ramayana). The result is that, for example, the Gita begins dharma kSetre kuru kSetre, not dharmakSetre kurukSetre. In 1995 I devised a method which I believed would allow me to rejoin a significant percentage of these “split” compounds. It involved using a computer to analyse differences between substantial passages of Mahabharata text before and after full “correction” by hand. In this way the computer could learn to recognise large numbers of typical split compounds, and could therefore be used to correct them wherever they occurred in the text as a whole. Applying this method over a period of months led to the elimination of perhaps 75% of the split compounds occurring in the electronic text; I also made large numbers of corrections of other kinds. Then, like Prof. Tokunaga before me, I made the resulting text available via the Internet; I also began to make use of it in my own scholarly work. However, this kept its remaining deficiencies all too permanently before my eyes, and I determined to find a way of bringing the text to (notional) perfection. With the aid of generous funding from the Humanities Research Board of the British Academy, the Society for South Asian Studies and the Smuts Memorial Fund (University of Cambridge), I was able to set up a project at BORI to achieve this goal. A team of five assistants -- Indu Deshpande, Kirti Sharad Thakar, Pradnya Anant Rayrikar, Samita Vasant Shinde and Shilpa Mulay -- were employed to work through the entire text correcting any split compounds or other errors they encountered. The task was done throughout in duplicate, so that slips made by one assistant would show up through comparison with the work of another. I was responsible for collating the results, which form the basis of this electronic text. For their dedicated work, and the levels of accuracy that they have brought to a long and demanding task, the members of the team, like Prof. Tokunaga himself, have earned the gratitude of all those who make use of the electronic Mahabharata. My own gratitude goes in addition to Prof. Saroja Bhate, without whose help the project would never have reached such successful results. ------------------------------------------------------------------------ The format of the electronic text ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. Line-numbers --------------- The text is divided into eighteen files, one per main book (parvan) of the Mahabharata. However, the format of each line is such as to identify it uniquely, whichever file it happens to come from. At the start of every line appears a nine-character line-number specifying the book or parvan (two digits), the chapter or adhyAya (three digits), the verse or zloka (three digits), and the quarter-verse or pAda (one letter, specifying the first of the two pAdas that form each line). Thus the first line of the Gita appears as: 06023001a dharmakSetre kurukSetre samavetA yuyutsavaH indicating that it represents pAdas a and b of the first verse of book 6, chapter 23. If the line is part of a verse in triSTubh or other longer metre, the division between the pAdas is marked with a semicolon: 01001065a duryodhano manyumayo mahAdrumaH; skandhaH karNaH zakunis tasya zAkhAH If the line is part of a passage in prose, the final letter of the line-number (which normally indicates the pAda) is capitalised: 01003001A janamejayaH pArikSitaH saha bhrAtRbhiH kurukSetre dIrghasatram upAste If the line is a header such as janamejaya uvAca, the final letter of the line-number is replaced by a space: 01045003 janamejaya uvAca 2. Encodings ------------ The text is available in three encodings (character sets) commonly used for Sanskrit and other Indian languages: Unicode Devanagari, Unicode Roman (using the conventions defined in ISO 15919), and ASCII (using the Harvard/Kyoto conventions). 3. Spelling conventions ----------------------- It is impossible to achieve complete consistency in a text as large and diverse as the Mahabharata, especially when it was edited over 33 years by several different scholars from a vast range of manuscript sources. So, for example, the spellings dvaMdva and dvandva both appear in several books. No serious attempt at normalisation of spelling has been made in the electronic text, but some consistency in purely editorial matters is clearly desirable, especially given that people will wish to search the entire text for words and phrases. The following rules have been applied, which may introduce minor changes from the text as printed; the words chatra etc. raise a particular problem which is addressed in some detail below. General rules +++++++++++++ Forms such as kiM cit (/cana/api) are everywhere spelt as two words in Roman representations of the electronic text (assuming that sandhi does not prevent this, as in kApi), whatever the usage in the printed text; similarly kiM svit, etc. The word kaccit of course is unchanged, as are compounded forms such as kiMciccheSa “having little left remaining”. In the Devanagari versions of the text, forms such as kiMcit are spelt as single words to respect normal usage in Indian scripts. The phrase atha vA is always written as two words. The same applies to periphrastic perfects: kathayAM babhUva, etc. Use of the word-internal avagraha in forms such as yazo'rtha is not entirely consistent in the printed text (e.g. yazo'rthaM 7.86.20c but yazortham 8.29.25b). The Errata to Book 1 contain five such cases, in all of which Sukthankar requests the deletion of the avagraha (he missed a single further case at 1.115.11d). Arguments could be mounted against this practice, but the editorial intention is very clear, and the internal avagraha has been removed in the few cases where it does appear. An effort has been made to standardise the use of double avagraha in cases where the elided vowel is not a but A. This is the normal usage in the printed text, but there are a few cases where single avagraha appears instead; the electronic text corrects these. The words kArttika, kArttikeya, etc., are for the most part spelt with double “t” in the Critical Edition, but there are some occurrences of “kArtika” and “kArtikeya”. These have been silently normalised in the electronic text. Similarly, the absolutive chittvA is occasionally spelt with a single “t”: such occurrences have been normalised to the correct form. The words chatra, satra and patra +++++++++++++++++++++++++++++++++ The words chatra, satra and patra are spelt inconsistently in the Critical Edition: some editors favoured the spelling with single “t”, others that with double “tt”. It appeared undesirable to leave such inconsistency in place, and all three words are uniformly spelt with single “t” in the electronic text. This applies also to derivative forms in -in, etc. The choice of single over double “t” reflects the overall weight of opinion amongst the editors: this appeared a more important criterion than any questions of etymological “correctness”. Note, however, that the word tottra is spelt with double “t”, since the Critical Edition is entirely consistent in spelling it thus at every occurrence. ------------------------------------------------------------------------ Notes on the electronic text ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Occasionally the electronic text corrects an error in the printed edition. The following is a complete list of such corrections: each line is quoted as it is actually printed, followed by a comment explaining the correction. (In a few cases a possible correction is noted that has not in fact been implemented.) Note that changes made purely to adhere to the spelling conventions listed above are not included. Note also that all corrections included in the printed edition's various lists of errata, etc., have been applied to the electronic text: they too are not referred to here. In these notes, “PAE” refers to Phillip Ernest, who has spent many hours checking problematical readings in manuscripts used in the preparation of the Critical Edition and still available for consultation at BORI. Thanks to his efforts it has been possible to establish beyond doubt that certain puzzling readings found in the printed edition are simple typographic errors. 01001046a bhUtasthAnAni sarvANi rahasyaM vividhaM ca yat The first printing of Book 1 (in separate fascicules) has trividhaM here, but in the second printing (in bound form) this has been replaced by vividhaM. Since vividhaM is also quoted as a variant reading, it is clear that it is the latter form which is a typographic error. I am grateful to James L. Fitzgerald for drawing my attention to this problem. 01001174c ajeyaH parazuH puNDraH zambhur devAvRdho 'naghaH This is the only occurrence of the word zambhu- spelt with nasal “m” rather than anusvAra. The spelling with anusvAra occurs seventeen times, including an occurrence later in Book 1 (01058043c). It is clear that this was the spelling preferred by Sukthankar and the other editors, and I have restored it in this one discrepant instance. I am grateful to James L. Fitzgerald for drawing my attention to this problem. 01058043a tam uvAca mahArAja bhUmiM bhUmipatir vibhuH It is certain that tam is a typographic error for tAm. The editor records no variant readings for the word, and yet the Bombay Edition and manuscripts K0, Da1 and D1 all have tAm (PAE). 01179009a kejid Ahur yuvA zrImAn nAgarAjakaropamaH 02024018c uragAvAsinaM caiva rocamAnaM raNe 'jayat See the Supplementary Addenda, p. 515. 05032014c virocate 'hAryavRttena vIro; yudhiSThiras tvayi pApaM visRjya See the note on p. 726 of the edition. 05070042c prazAntAH samabhUtAz ca zriyaM tAn aznuvImahi It is certain that tAn is a typographic error for tAm. The editor records no variant readings for the word, and yet the Bombay Edition and manuscripts K1, Ds1 and D5 all have tAm (PAE). 07005034c traiyyambakam atheSvastram astrANi vividhAni ca traiyyaMbakam is also quoted as a variant reading. 07035011a tam udIkSya tathA yAntaM sarve droNapurogamAH “Coming” rather than “going” is necessary for the sense here. 07053014c utsahante 'nyathAkartuM pratijJAM savyasAcinaH Monier Williams allows anyathA-kR- as a compound verb, but all the other ten occurrences of anyathA + verbal form of kR- in the Mahabharata (01003133B, 01091013c, 03131015c, 03205010a, 05081053a, 07052011c, 07172048a, 13043025a, 13047036a, 16009027c) are printed as non-compounds. Of these, seven consist of anyathA kartum, including two further occurrences of the pAda-phrase utsahante 'nyathA kartum, one in the previous chapter (07052011c), the other later in the same parvan (07172048a). I am therefore preferring the non-compounded form. 07054005c pratisrotaHpravRttAz ca tathA gantuM samudragAH Clearly a typographic error (cf. the almost identical 07167002c: pratisrotaH pravRttAz ca gantuM tatra samudragAH). 07054010a snuSA zvazvrAnaghAyaste vizoke kuru mAdhava 07070037c sahasenaH sahAmAtyo drApadeyAn avArayat 07076029a iti kRSNA maheSvAsau yazasA lokavizrutau 07079025c bhUrizravAs tribhir bANair hemapuGkhaH zilAzitaiH 07085021c abhyavarSaJ zarais tIkSNaiH kaGkabarhiNavAjitaH 07120069e adRzyau ca zaroghais tau nighnatAm itaretaram 07164037a taM sAtyakiH pratyaviddhat tathaiva dazabhiH zaraiH It is certain that pratyaviddhat is a typographic error for pratyavidhyat. The editor records no variant readings for the word, and yet the Bombay Edition and manuscripts Dc1 and D6 both have pratyavidhyat (PAE). 07165044c mahimAnaM mahArAja yogamuktasya gacchataH It is certain that yogamuktasya is a typographic error for yogayuktasya. The editor records no variant readings for the word, and yet the Bombay Edition, P. S. S. Sastri's edition of the southern recension and manuscripts Dc1, D5 and D7 all have yogayuktasya (PAE). 07165055c palAyanakRtotsAhA dudruvuH sarvato dizam 08031004c yudhiSThiraM cAbhibhavann asapavyaM cakAra ha It is certain that asapavyaM is a typographic error for apasavyaM. The editor records no significant variant readings for the word, and yet the Bombay Edition and manuscripts K1, D4 and D8 all have apasavyaM (PAE). 08062047c sa vAjisUteSvasanas tathApatad; yathA mahAvAtahato mahAdrumaH The line is unconstruable unless the initial sa is taken as part of the following compound. 08065014c athAbravIt pANinA pANim Aghnan; saMdaSTASTho nRtyati vAdayann iva 08069017a yaH sa dyUtajitAM kRSNAM prAha satpuruSAdhamaH I believe this is a rather celebrated case, and I am grateful to Saroja Bhate for directing my attention to it. 09001052c samAzvAsayata kSattA vacasA madhureNa ha See Critical Notes. 09006023a etac chrutvA yathAbhUtaM kuru mAdhava yat kSamam See Critical Notes. 09016058e saMnyastakavacA dehair vipatrAyudhajIvitAH This line has been reproduced unchanged as it appears in the printed edition; however, it seems likely that vipatrAyudhajIvitAH is a typographic error for vipannAyudhajIvitAH, the reading found in the Bombay Edition and the manuscript D10. This reading is not recorded by the editor (though he does record other variants beginning with vipannA-). However, manuscript K2 does read vipatrAyudhajIvitAH, so I do not feel justified in modifying the printed text. Manuscript information from PAE. 09018062c punar evAnvartanta pANDavAn AtatAyinaH 10001050c zlaukau nyAyam avekSadbhis tattvArthaM tattvadarzibhiH 11008039a bhavAn karmaparo yatra buddhizreSThaz ca bhArata It is certain that karmaparo is a typographic error for dharmaparo, as suggested by James L. Fitzgerald (The Mahabharata translated, vol. 7, Chicago and London, 2004, pp. 666-7). The editor records no variant readings for the word, and yet the Bombay Edition and manuscripts K1, K2 and D3 all have dharmaparo (PAE). 12047006a etaiz cAnyair munigaNar mahAbhAgair mahAtmabhiH 12089018a sthAnAny etAni saMgamya prasaGge bhUtinAzanaH This line has been reproduced unchanged as it appears in the printed edition; however, James L. Fitzgerald (The Mahabharata translated, vol. 7, Chicago and London, 2004, p. 740) suggests that prasaGge is an error for prasaGgo. noting that this is the reading of the CitrazAlA edition, but that no variants are recorded by the editor. It is also the reading of the Bombay Edition and manuscripts Da1 and Da2; however, manuscript K1 does read prasaGge, so I do not feel justified in modifying the printed text. Manuscript information from PAE. 12155004c tapasaiva hi sidhyanti tapo mUlaM hi sAdhanam I am grateful to James L. Fitzgerald for pointing out this error (tapomUlaM must be a bahuvrIhi compound meaning “having asceticism for its basis”). 12244011a vyavasAyAtmikA buddhir manovyAkaraNAtmakam I am grateful to Kirti Sharad Thakar for pointing out this error. 12291043c yo 'haM soha 'm iti hy uktvA guNAn anu nivartate 12292027c divasAnte guNAn etAn abhyetyaikovatiSThati 12309024c RtvAsyaH samabalazuklakRSNanetro; mAMsAGgo dravati vayohayo narANAm It is certain that mAMsAGgo is a typographic error for mAsAGgo. The editor records no variant readings for the word, and yet the Bombay and Kumbakonam editions and manuscripts K1 and D4 all have mAsAGgo (PAE). I am grateful to James L. Fitzgerald for pointing out this error. 12319002c pAdAt prabhRtigAtreSu krameNa kramayogavit I am grateful to James L. Fitzgerald for pointing out this error (prabhRti when governed by the ablative pAdAt must be an independent word). 12323057e parAM gatim anuprApta iti naiSThikam ajjasA 12324024c na kSutpipAse rAjendra bhUmez cchidre bhaviSyataH 12331030a zvetadIpe mayA dRSTAs tAdRzAv RSisattamau 12331035c zvetadIpe tvayA dRSTa AvayoH prakRtiH parA I am grateful to James L. Fitzgerald for pointing out this error. 12334014c ekAntinAM zaraNado 'bhayado gatido 'stu vaH; sa makhabhAgaharas triguNAtigaH 12334015a catuSpaJcadharaH pUrte;STayoz ca phalabhAgaharaH I am grateful to Kirti Sharad Thakar and Samita Vasant Shinde for correcting what must be an editorial slip in the division of the text into pAdas. The critical note says, “Very probably ekAntinAM is a marginalia for vaH and should not be regarded as part of the line”; however, the version adopted in the electronic text is clearly preferable. 12338005a idaM puruSasuktaM hi sarvavedeSu pArthiva 12342008c provAca vacanaM zlaSNaM prAjJo madhurayA girA 13016013a sa dRSTvavAn mahAdevam astauSIc ca stavair vibhum 13023006a na brAhmaNaH sAdhayate havyaM daivAt prasiddhyati 13027098c bhajed vAcA manasA karmaNA ca; bhaktyA yuktaH parayA zraddhadhAnaH 13028022a tataH saMtApayAm Asa bibudhAMs tapasAnvitaH 13059005a mriyate yAcamAno vai tam anu mriyate dadat This line has been reproduced unchanged as it appears in the printed text; however, it seems likely that the final word should appear as 'dadat. 13061025a pitRMz ca pitRlokasthAn devaloke ca devatAH 13076007c klaizair vipraM yo 'phalaiH saMyunakti; tasyAvIryAz cAphalAz caiva lokAH 13091001c bhRgvaGgarasake kAle muninA katareNa vA 13101024c pitqNAM mAnuSANAM ca kAntAyAs tv anupUrvazaH 13110062a kalahaMsavinirghoSair nUpUrANAM ca nisvanaiH 13125022c na bhAti kAle 'bhihitaM tenAsi hariNa kRzaH 13134027c pravaktqn pRcchate yo 'nyAn sa vai nA padam arcchati This line has been reproduced unchanged as it appears in the printed text; however, as the editor's critical note observes, “nApadam, that is na Apadam (for nA padam) is certainly more natural”. 14011010a vyAptAsv athAsu vRtreNa rase ca viSaye hRte 14017035e tac chrutvA naiSThikIM buddhiM buddhyethAH karmanizcayAt 15005006a yac cAhaM pANDuputreNa guNavatsu mahAtmasu It is certain that pANDuputreNa is a typographic error for pANDuputreSu. The editor records no variant readings for the word, and yet the Bombay Edition, P. S. S. Sastri's edition of the southern recension and manuscripts K3, K4 and D9 all have pANDuputreSu (PAE). 15033012a iyaM ca mAtA jyeSThA me vItavAtAdhvakarzitA It is certain that vItavAtAdhvakarzitA is a typographic error for zItavAtAdhvakarzitA. The editor records no variant readings for vIta-, and yet the Bombay Edition, P. S. S. Sastri's edition of the southern recension and manuscripts K3, K4 and D9 all have zItavAtAdhvakarzitA (PAE). 15044008a mA sma zoke manaH kArSIr diSTena vyathate budhaH The sense demands diSTe na. 15044039c gamyatAM putra maiva tvaM vocaH kuru vaco mama It is certain that maiva is a typographic error for maivaM. The editor records no significant variant readings for the word, and yet the Bombay Edition and manuscripts K3, K4 and D9 all have maivaM (PAE). 16004004c te sAgarasyopariSThAd avartan; manojavAz caturo vAjimukhyAH 16008039a tat sAgarasamaprakhyaM vRSNicakraM mahardhimat 16009012a punaH punar na mRzyAmi vinAzam amitaujasAm It is certain that mRzyAmi is a typographic error for mRSyAmi. The editor records no significant variant readings for the word, and yet the Bombay Edition and manuscripts K3, K4 and D9 all have mRSyAmi (PAE). 17001011a mAtRbhiH saha dharmAtmA kRtvodakam atandritaH It is certain that mAtRbhiH is a typographic error for bhrAtRbhiH. The editor records no variant readings for the word, and yet the Bombay Edition and manuscripts K3, K4 and D9 all have bhrAtRbhiH (PAE). 18001016c draupadyAz ca pariklezaM na cintayatum arhasi It is certain that cintayatum is a typographic error for cintayitum. The editor records no variant readings for the word, and yet the Bombay Edition and manuscripts K3, K4 and D8 all have cintayitum (PAE).