The Bhagavad Gita Is 700 Verses in 18 Chapters Inside the Mahabharata, and Eight Major Commentators, From Shankara to Radhakrishnan, Have Read Its Three Paths in Different Ways

The Bhagavad Gita is a dialogue of 700 verses in eighteen chapters, and it sits inside the Bhishma Parva, the sixth book of the Mahabharata. S. Radhakrishnan, in his 1948 edition, numbered those eighteen chapters as Chapters 23 to 40 of that book. Kisari Mohan Ganguli’s English translation of the epic uses a different numbering and places them in Sections 25 to 42.

Tradition ascribes the Gita, like the epic that holds it, to the sage Vyasa, and holds that Krishna spoke it to Arjuna in his chariot, between the two armies at Kurukshetra, before the fighting began. In Ganguli’s translation the Gita opens with King Dhritarashtra asking Sanjaya what his sons and the Pandavas did once they had gathered on the plain, and Sanjaya answers by reporting the conversation. The tradition classes the Mahabharata as itihasa, a Sanskrit word that splits into parts meaning thus indeed it happened.

Where the date and the count stand

Radhakrishnan put the date of the Gita around the fifth century BCE and said the text may have been altered afterward. He recorded that European scholars of his day had cut it into layers, and he added that the Indian tradition has always felt the different elements were fused in the mind of one author.

The count is steadier. Shankara, in the introduction to the oldest commentary that survives, speaks of the “seven hundred verses” in which Vyasa set down the teaching, and Radhakrishnan’s edition counts 700 as well. The Kashmiri version of the text, on which Abhinavagupta wrote around the year 1000, has 716 verses by the count in Partha Sarathi Sil’s 2020 study. Radhakrishnan, citing F. Otto Schrader’s 1930 book on that version, wrote that its variations do not affect the Gita’s general teaching.

Eighteen chapters and a division nobody wrote into the text

The eighteen chapters were originally untitled, according to Shyam Ranganathan’s entry in the Internet Encyclopedia of Philosophy, so the headings in printed editions come from editors. The text itself moves between three subjects that later readers named. Karma yoga is action done as duty without clinging to the result. In Ganguli’s translation Krishna says in the second chapter, “Thy concern is with work only, but not with the fruit (of work).” Bhakti yoga is loving devotion to God, and jnana yoga is knowledge of the self.

Commentators then sorted the chapters into three sets of six. Yamunacharya, a Sri Vaishnava teacher whom the Encyclopedia dates to 916 to 1041, did it in his Gitartha-sangraha, a summary of the Gita’s import. In Ranganathan’s account the first six chapters stress karma yoga and jnana yoga, the middle six stress bhakti yoga, and the last six sum up the earlier arguments and take up the relation of nature (prakriti) to the self (purusha). Ramanuja followed that frame. The edition of his commentary used here opens its seventh chapter by noting that the first six dealt with knowledge of the self, and that the middle six teach the nature of the Supreme Being and worship of Him called bhakti.

Madhusudana Sarasvati, an Advaita commentator of the sixteenth century, is quoted in Sil’s study with a verse that names three sections in order, action, worship and knowledge, one for each set of six chapters, and gives devotion to the middle six. Sil adds that Abhinavagupta drew up no such division himself. Tilak records yet another scheme, which gives six chapters each to the three words of the Upanishadic phrase “tat tvam asi” (“that art thou”), and he calls it a queer idea.

Shankara: knowledge first, with action as preparation

Shankara’s is the oldest Gita commentary that survives. Radhakrishnan dates Shankara to 788 to 820 and notes that his introduction mentions older commentaries that have not come down to us. Shankara says he wrote because, although others had explained the text word by word, “to the laity it appears to teach diverse and quite contradictory doctrines.”

In Alladi Mahadeva Sastri’s 1897 English translation, Shankara gives the aim of the Gita as “the Supreme Bliss, a complete cessation of samsara or transmigratory life and of its cause,” samsara being the cycle of rebirth. That comes from “a steady devotion to the knowledge of the Self, preceded by the renunciation of all works.” Works keep a role. Done in devotion and without regard to results, they purify the mind, and a purified mind is fit for knowledge. Shankara cites Krishna’s closing words, “Abandoning all dharmas, come to Me alone for shelter,” as support. Radhakrishnan sums the position up this way: action is essential for purifying the mind and falls away once wisdom arrives, and Shankara rejects the view that knowledge and works can be pursued together.

Not everyone accepts that reading of the book as a whole. Ranganathan argues that the Gita’s own theme runs the other way, with knowledge serving action and devotion, and that Shankara’s reading holds only if the reader already shares his position.

Abhinavagupta, Ramanuja and Madhva each read it through a school

Abhinavagupta worked in Kashmir between about 950 and 1015, by the dates in Sil’s study. His Gitartha-sangraha reads the Gita in the light of Kashmir Shaivism, the school of Shaiva philosophy that grew up in Kashmir, and it comments on the Kashmiri version of the text. Sil writes that Abhinavagupta joins knowledge and action, with devotion as the link between them.

Ramanuja, whom Radhakrishnan places in the eleventh century, founded the school called Vishishtadvaita, or qualified non-dualism, which holds that the world and souls are real and are God’s body. According to Radhakrishnan, his commentary refutes the doctrine that the world is unreal and the path of giving up action, and it holds that the Gita mentions knowledge, devotion and action but puts its main weight on devotion. Ranganathan reports that on the verse telling Arjuna to forsake all duties, Ramanuja offers two readings and insists that it cannot mean giving up ethical action, since doing one’s duty is itself a means of devotion.

Madhva wrote two works on the Gita, the Gita Bhashya and the Gita Tatparya, and drew dualist teaching from it: God and the soul are eternally different. Radhakrishnan writes that Madhva read the Upanishadic phrase “that art thou” as an instruction to give up the distinction between mine and thine and to hold that everything is under God’s control, and that Madhva took devotion to be the method the Gita stresses. The Internet Encyclopedia of Philosophy dates Madhva to 1238 to 1317.

Jnaneshwar put it into Marathi in 1290

The Jnaneshwari was completed in Saka 1212, which is 1290 CE, at Nevasa, in the district once called Ahmednagar, in the temple of Mhalsa on the bank of the Pravara river. The life sketch printed with Ramchandra Keshav Bhagwat’s English translation says its author, Jnaneshwar, was then 15. Its original title is Bhavartha Dipika, it has eighteen chapters to match the Gita’s, and it honours his guru and elder brother Nivrittinath.

The UCLA guide to the text says it presents the Gita as the essence of the Mahabharata, which is in turn the essence of the Vedas, and that it opened the Gita to ordinary people outside learned or high-caste circles. Scholars read its leaning differently. The UCLA guide names Bahirat as seeing it as devotional and Dandekar as reading it as Vedantic philosophy. Radhakrishnan wrote that Jnaneshwar and Tukaram were great devotees who accepted Shankara’s position in metaphysics.

Charles Wilkins and the first English Gita, 1785

The first English translation appeared in London in 1785 as The Bhagvat-Geeta, or Dialogues of Kreeshna and Arjoon, in eighteen lectures, with notes. Its title page names the translator as Charles Wilkins, a senior merchant in the East India Company’s Bengal service.

Eric Sharpe, in his survey of Western readings of the Gita, writes that Wilkins reached Bengal in 1770, began learning Sanskrit in 1778 and apparently meant to translate the whole Mahabharata. In the early 1780s the governor of Bengal, Warren Hastings, urged him to print the Gita separately. By early 1785 the translation was before the Company’s directors, who ordered it published provided the total cost stayed within 200 pounds. The book carries a letter from Hastings dated Benares, 4 October 1784.

Wilkins wrote in his preface that the Gita set up “the doctrine of the unity of the Godhead” against what he called idolatrous sacrifices and the worship of images. Sharpe notes that this reading suited the age of Deism, and that Wilkins’s version remained the English-speaking world’s main source on the Gita for decades.

Tilak wrote his commentary in a Burmese jail and put action first

Bal Gangadhar Tilak wrote the Gita Rahasya, also called the Karma-Yoga-Shastra, while imprisoned in Mandalay in Burma. D. Mackenzie Brown, writing in the Journal of Asian Studies in 1958, gave the writing time as four months in the winter of 1910 to 1911. He reported that the British authorities seized the manuscript and returned it only after a government Marathi expert had declared it free of seditious passages. It was published in Marathi in 1915, after Tilak returned to Poona, and B. S. Sukthankar’s English translation followed in 1935.

Tilak argued that the Gita teaches an “Energistic” path, his term for a life of action, joined to spiritual knowledge. He wrote that Shankara’s commentary was “the first attempt to deprive the Gītā of its Energistic form” and give it a renunciatory one, and he held that Shankara’s own introduction shows earlier commentators had read the Gita as combining action with knowledge. He did not doubt Shankara’s stature: he wrote that the world has not produced a philosopher of his calibre. On the range of readings he was firm. “The Gītā is not such a pot of jugglery, that any one can extract any meaning he likes out of it,” he wrote.

Gandhi on desireless action, and an edition dedicated to him

Gandhi wrote that his first acquaintance with the Gita came in 1888 to 1889, through Sir Edwin Arnold’s verse translation, The Song Celestial. He finished the introduction to his own Gujarati translation at Kosani in Almora on 24 June 1929, after two years of waiting, and the translation was published on 12 March 1930, the day he left Sabarmati on the march to Dandi.

His reading of the whole is stated in the introduction. “That matchless remedy is renunciation of fruits of action,” he wrote, and he called this renunciation “the central sun, round which devotion, knowledge and the rest revolve like planets.” He allowed that, to the letter of the text, warfare can be called consistent with giving up the fruit of action. After forty years of trying to live by the Gita, he wrote, he had concluded that perfect renunciation is impossible without perfect observance of ahimsa, or non-violence.

Gandhi read the battlefield differently from Shankara and Tilak, both of whom write of Krishna instructing Arjuna in person. He took the Gita to describe the duel that goes on in the human heart. That is his reading, set beside the tradition’s own word for the epic, itihasa.

Radhakrishnan dedicated his 1948 edition to Gandhi. In its introductory essay he sets out the schools of Advaita, Vishishtadvaita, Dvaita and Shuddhadvaita, and says each commentator found his own system in the Gita because its author suggests that the one truth cannot be shut up in a single formula. He follows the text on which Shankara commented. Of the differing commentaries he wrote that the Hindu tradition believes the different views are complementary, and he cited a verse from the Bhagavata Purana in which the sages describe the essential truths in various ways.