Entrepreneurship in Action, 21-24 Nov, online. Build a working product with AI

The Critical Edition of the Mahabharata | CBSE Class 12 History Notes

14 min read

On this page

Part of: Kinship, Caste and Class (CBSE Class 12 History, Themes in Indian History Part I, Theme 3). This page goes deeper into one section of that chapter note: how historians examine sources and the Critical Edition. It follows the epic from handwritten copies to print, and asks what the edition can tell the student.

What is a critical edition, and why did the Mahabharata need one?

Before printing, a long written text survived mainly through copies made by hand. Each copy could gain, lose or alter a word, a line or an episode, so over centuries the copies drifted apart.

V. S. Sukthankar, writing in 1933, observed that local versions of the Mahabharata had probably existed from very early times. He borrowed a Latin saying about manuscripts: quot codices, tot textus, or "as many manuscripts, so many texts".

A critical edition is the scholar's answer. The editor gathers as many copies, or witnesses, as possible. Then comes collation: comparing them word by word to find every point where they differ. Each difference is a variant reading.

The editor prints one main text and lists the variants in notes. A teaching page of the Pittsburgh NEH Institute (2017) calls these notes the apparatus, and notes that some editors have aimed at a text found in no single copy.

How did the project begin at Pune, and who led it?

The work was done at the Bhandarkar Oriental Research Institute (BORI) in Pune, then spelt Poona. BORI's own page dates the project from April 1919 to September 1966.

Sukthankar's Prolegomena, the introduction to the first volume (1933), describes the start. At an Institute meeting on 6 July 1918, the Chief of Aundh offered a lakh of rupees in annual grants.

In April 1919, R. G. Bhandarkar formally began the collation with an opening mantra also found at the start of some manuscripts.

NCERT credits the leadership to Sukthankar, a noted Sanskritist, at the head of a team of dozens of scholars. The method was first tested on a smaller scale: the Prolegomena records that N. B. Utgikar, whom Sukthankar calls his predecessor, prepared a trial edition of the Virataparvan (1923) that was sent to Sanskritists for comment.

The first book, the Adiparvan, appeared in parts from 1927 and as a volume in 1933. BORI names Sukthankar, S. K. Belvalkar, S. K. De and R. N. Dandekar among the edition's editors. NCERT says the project took 47 years.

How did the editors compare so many manuscripts?

BORI states that the edition consulted 1,259 manuscripts. Sukthankar says the editors sought copies in as many scripts, and so from as many provinces, as possible, preferring old ones.

For the first book alone, he knew of about 235 manuscripts in ten scripts: 107 in Devanagari, 32 in Bengali, 31 in Grantha, 28 in Telugu and 26 in Malayalam, with a handful in Nepali, Sarada, Maithili, Kannada and Nandinagari.

About 70 of these were collated, and about 60 were used to fix the text.

Sukthankar sorted the copies into two broad recensions, or families of the text: a Northern and a Southern. He split each into versions and noted that a version's script marks its region.

NCERT sums up the last step: the verses that appeared common to most versions became the main text.

StageWhat the editors didSource
CollectingGathered manuscripts in many scripts and regions; BORI counts 1,259 in allSukthankar; BORI
CollatingCompared the copies verse by verse and noted the differencesNCERT; Sukthankar
GroupingSorted copies into Northern and Southern recensions and their versionsSukthankar
Constituting the textPrinted the verses common to most versions as the main textNCERT
Recording variantsPut divergent readings in notes below the text, and longer additions in appendicesSukthankar; NCERT
PublishingIssued the books in volumes, from 1933 to 1966Sukthankar; BORI

What did the finished edition look like, and what did it reveal?

Sources measure it differently. NCERT gives over 13,000 pages, more than half of them on variations. BORI lists 19 volumes and more than 15,000 demi-quarto pages, covering the 18 parvans. BORI calls the main text the constituted text.

BORI records that the completed edition was released on 22 September 1966 by S. Radhakrishnan, then President of India.

On each page the main text stands in large bold type, with variant readings in small print beneath, as NCERT's Figure 3.2 shows. Sukthankar promised that no line found in a collated manuscript would be left out, making the edition a "thesaurus" of the tradition.

Source box: Bhandarkar Oriental Research Institute, "The Mahabharata" department page (consulted November 2026). "...prepared with painstaking efforts of scholars for about five decades consulting 1,259 manuscripts."

What it shows: the Institute stresses the length of the work and the number of copies compared. "About five decades" is a round figure; NCERT's 47 years run from 1919 to 1966.

NCERT draws two lessons. First, the Sanskrit versions had a number of elements in common: manuscripts from the far north, such as Kashmir and Nepal, shared them with those from the far south, such as Kerala and Tamil Nadu.

Second, the way the text was handed down varied enormously by region. NCERT reads these variations as traces of a long exchange between dominant traditions and local ideas and practices, marked by conflict and agreement.

How do historians use the Critical Edition?

For a historian, the apparatus matters as much as the main text. Sukthankar wanted readers to see all the significant manuscript evidence for each passage, so one can check whether a verse appears in both recensions or only in some regions' copies.

NCERT notes that much of what is known about early social history comes from Sanskrit texts written by and for Brahmanas. Historians who first explored social history in the nineteenth and twentieth centuries, NCERT says, tended to assume that what these rule-giving texts laid down was what people did.

Later scholars, in NCERT's account, turned to works in Pali, Prakrit and Tamil. These suggested that the Sanskrit norms were broadly accepted as authoritative, yet were also questioned and sometimes rejected.

NCERT presents the epic as a text that kept growing, and the regional variants are one kind of evidence for that view. This links to the syllabus objective on the "dynamic approach of Mahabharata". The chapter note covers the stages of composition and later retellings.

Why do scholars still debate the method?

The editor himself set limits. In the Prolegomena, Sukthankar called the epic a fluid text. He judged it useless to rebuild such a text in its literally original shape from an archetype and a family tree of manuscripts, or stemma.

His aim, he wrote, could only be "the oldest form of the text which it is possible to reach" from the manuscripts available. He thought nearly all versions had been mixed together, the Sarada perhaps excepted, and urged no change without manuscript support.

Some scholars argue that even this goal does not suit the epic. In 2023, Wendy Doniger, a translator of the epic, said a work put together by many people over centuries cannot be critically edited. She called cutting it down a mistake, though the left-out material survives at the foot of the page.

Source box: Wendy Doniger, interviewed by Avik Chanda, Scroll, 7 May 2023. "Since the Mahabharata is a living work ... you can't do a critical edition."

What it shows: one scholar's case against the method. Read beside the Prolegomena, it shows that both sides accept the epic changed over centuries; they differ on whether one reconstructed text is a useful goal.

Others take a different view. Vishwa Adluri and Joydeep Bagchee wrote Philology and Criticism (2018) as a guide to the edition. Anthem Press's description calls the edition a landmark; a longer listing says the book finds no justification for an earlier heroic epic, unlike NCERT's account of a text that grew in stages.

How has the edition been carried forward and translated since 1966?

BORI's page gives 1966 to 1969 for its two-volume critical edition of the Harivamsa, and lists a translation of the Critical Edition among its plans.

An English translation from the edition began in Chicago. NCERT records that J. A. B. van Buitenen started it in 1973 and left it unfinished at his death. He died in 1979, as the University of Chicago Press and a 2016 University of Chicago article record; NCERT's timeline prints 1978.

That article says he completed the first five books, and that the project passed to James L. Fitzgerald. His volume 7, with Book 11 and part of Book 12, appeared in 2004. Doniger says that in her own translation she added passages the edition left out, printing them in italics.

Timeline: when did the key steps happen, from the epic's composition to 2004?

The first two rows give NCERT's approximate periods; the rest come from the edition's records, BORI and publishers.

DateEvent
c. 500 BCE onwardsComposition begins, over about 1,000 years, in NCERT's dating
c. 200 to 400 CELarge didactic sections are added, in NCERT's dating
6 July 1918The Chief of Aundh offers a lakh of rupees for an edition
April 1919Collation formally begins at BORI, Pune
1923Utgikar's trial edition of the Virataparvan is sent to scholars
1927The first part of the Adiparvan appears
1933The Adiparvan volume is completed, with the Prolegomena
22 September 1966President S. Radhakrishnan releases the completed edition
1966 to 1969BORI's dates for its critical edition of the Harivamsa
1973Van Buitenen begins his English translation, in NCERT's dating
1979Van Buitenen dies, as University of Chicago records show (NCERT prints 1978)
2004Fitzgerald's volume 7 is published

Sources: which records and studies support this page?

  • NCERT, Themes in Indian History Part I, Theme 3, reprint 2026-27 (c. 500 BCE; 1919 to 1966; 1973; 1978).
  • CBSE, History syllabus, Class 12, 2026-27: Theme 3 learning objectives.
  • V. S. Sukthankar, Prolegomena to The Adiparvan, BORI, Poona, 1933, GRETIL transcription mirrored on sanskritdocuments.org (6 July 1918; April 1919; 1923; 1927; 1933).
  • Bhandarkar Oriental Research Institute, "The Mahabharata" department page, consulted November 2026 (1919 to 1966; 22 September 1966; 1966 to 1969).
  • Pittsburgh NEH Institute, "Edition terms and concepts", July 2017.
  • Avik Chanda, interview with Wendy Doniger, Scroll, 7 May 2023.
  • Anthem Press, description of Vishwa Adluri and Joydeep Bagchee, Philology and Criticism, released 29 June 2018 (publisher's page and OverDrive listing).
  • University of Chicago Press, author page for J. A. B. van Buitenen, consulted November 2026 (1928 to 1979).
  • Jeanie Chung, "An Epic Endeavor", Division of the Arts and Humanities, University of Chicago, 1 April 2016 (1979).
  • John D. Smith, review of James L. Fitzgerald (trans.), The Mahābhārata, volume 7 (University of Chicago Press, 2004), Bulletin of SOAS 68, no. 1, 2005.

Glossary

  • Manuscript: a copy of a text written by hand; BORI consulted more than a thousand of them for the Mahabharata.
  • Collation: comparing copies of a text word by word and recording every point at which they differ.
  • Variant reading: a word, line or passage in one copy that differs from the reading printed in an edition's main text.
  • Critical apparatus: the notes, usually below the main text, listing variant readings and the manuscripts that carry them.
  • Recension: a family of manuscripts sharing one form of a text; the Mahabharata survives in a Northern and a Southern recension.
  • Constituted text: the main text an editor builds by comparing manuscripts; BORI uses the term for its edition's 18 parvans.
  • Stemma: a family tree showing which manuscripts descend from which; Sukthankar judged it unsuited to a fluid text like the epic.

Common errors and misconceptions

  • Misconception: The Critical Edition prints the Mahabharata exactly as it was first composed. Correct: Sukthankar aimed only at the oldest form reachable from the surviving manuscripts.
  • Misconception: The editors threw away every verse that most manuscripts lacked. Correct: Variants were kept in notes and appendices; NCERT says they fill more than half the pages.
  • Misconception: Sukthankar prepared the edition alone, within a few years. Correct: NCERT describes dozens of scholars working for 47 years; BORI names several editors.
  • Misconception: The main text simply copies one old manuscript. Correct: It was built by comparing many copies and keeping the verses common to most versions.
  • Misconception: All scholars accept the Critical Edition as the right text of the epic. Correct: Doniger (2023) argues a living epic cannot be critically edited, while the description of Adluri and Bagchee's guide (2018) calls the edition a landmark.

Exam-style questions with model answers

Q1. According to NCERT, roughly what share of the Critical Edition's pages records variations between manuscripts? (a) About a tenth (b) About a quarter (c) More than half (d) Nearly all [1 mark]

(c) More than half. NCERT says the variations, printed in footnotes and appendices, fill more than half of the edition's 13,000-plus pages.

Q2. Explain three steps by which the editors of the Critical Edition turned many manuscripts into one printed text. [3 marks]
  1. Collecting: the team gathered manuscripts in many scripts from across the subcontinent; BORI counts 1,259.
  2. Comparing: the editors compared copies verse by verse and grouped them into Northern and Southern recensions.
  3. Selecting and recording: verses common to most versions formed the main text, while variant readings went into notes and appendices.
Q3. Read the source box quoting Wendy Doniger (Scroll, 2023). (i) What reason does she give against a critical edition of the Mahabharata? (ii) What did Sukthankar accept about recovering the original? (iii) Where are divergent readings printed? (iv) Where are longer passages left out of the main text placed? [4 marks]
  1. (i) She argues that the epic is a living work, put together by many people over centuries, so it has no single fixed original to edit.
  2. (ii) Sukthankar accepted that a fluid text could not be rebuilt in its literally original shape; he aimed only at the oldest reachable form.
  3. (iii) Divergent readings are printed in notes below the main text.
  4. (iv) Longer passages missing from the main text are placed in appendices.
Q4. Examine how the Critical Edition of the Mahabharata was prepared, what it revealed, and why scholars debate its method. [8 marks]
  1. Origins: after the Chief of Aundh pledged funds in 1918, collation began at BORI in April 1919.
  2. Leadership: NCERT credits V. S. Sukthankar with leading a team of dozens of scholars; the whole project took 47 years.
  3. Manuscripts: BORI says 1,259 manuscripts were consulted, in many scripts and from many regions.
  4. Method: copies were compared verse by verse and sorted into two recensions; verses common to most versions formed the main text.
  5. Size: NCERT counts over 13,000 pages, more than half on variants; BORI lists 19 volumes.
  6. Findings: NCERT notes several common elements in manuscripts from north and south, alongside enormous regional variation.
  7. Use for historians: NCERT warns that early historians took normative Sanskrit texts as a record of practice; Pali, Prakrit and Tamil works show the norms were questioned.
  8. Debate: Sukthankar aimed only at the oldest reachable form; Doniger (2023) rejects a single critical text; the description of Adluri and Bagchee's guide (2018) calls the edition a landmark.

Key takeaways

  • A critical edition compares many hand-copied versions of a text, prints one main text and records the differences in an apparatus of notes.
  • BORI in Pune carried out the Mahabharata project from April 1919 to September 1966; NCERT credits V. S. Sukthankar with leading it.
  • BORI says 1,259 manuscripts were consulted; Sukthankar sorted them into Northern and Southern recensions, with versions marked by script.
  • NCERT finds several common elements across the subcontinent and enormous regional variation, with more than half of the 13,000-plus pages given to variants.
  • NCERT warns against reading Sanskrit normative texts as a record of practice, and presents the epic as a text that kept changing over time.
  • Sukthankar aimed only at the oldest reachable form; Doniger questions any single reconstructed text, while the description of Adluri and Bagchee's guide calls the edition a landmark.

Test yourself

What does collation mean?

Comparing copies of a text word by word to record every point where they differ.

Where was the Critical Edition prepared, and over what period?

At the Bhandarkar Oriental Research Institute in Pune, from April 1919 to September 1966, according to BORI.

Name the two recensions into which the Mahabharata manuscripts fall.

A Northern and a Southern recension, each divided into versions linked to scripts.

How many manuscripts of the Adiparvan did Sukthankar know of, and how many were used?

About 235 were known to him; about 70 were collated and about 60 used to fix the text.

Who began the English translation of the edition, and when?

J. A. B. van Buitenen, in 1973, according to NCERT; he completed the first five books.

When was the completed edition released, and by whom?

On 22 September 1966, by S. Radhakrishnan, then President of India, according to BORI.

Write about it

Turn a question from this chapter into a paper.

The Writing Studio has a structure guide, a rubric and citation tools for 15 formats, from essays to research papers. A free account keeps up to five drafts. Learnacy Essential adds feedback from an AI writing coach and submission to The OYI Review.

Publication depends on review and is not guaranteed.