- Length
- 448 pagesOlivelle, Upanisads, Oxford World's Classics, 1996 - the twelve principal texts
- Reading
- DemandingReadable with attention; assumes no other book
- Availability
- Many editionsIn print in several competing editions
The Principal Upanishads
The texts in which Indian philosophy begins, and the point at which religion in India turns from ritual into enquiry. They are conversations rather than doctrines, they contradict one another freely, and every later Indian school - including the ones that reject them - defines itself by its relation to them.
- Date
- c. 700-300 BC
- Attributed to
- various, unknown
- Region
- The Ganges plain, northern India
- Language
- Sanskrit
- 1657 Persian, for Dara Shukoh Fifty Upanishads, at Delhi; the first translation into any language Attested
- 1801 Latin, Anquetil-Duperron Strasbourg; made from the Persian, not the Sanskrit Attested
- 1879 English, F. Max Muller Sacred Books of the East; public domain Attested
- 1996 English, Patrick Olivelle Current scholarly standard; in copyright Attested
Where it stands in the argument
- Position
- Answered — Drew a named later reply, but was not itself a reply
- Form
- Compiled
- Method
- Dialogue — Argument staged between speakers
- Transmission
- Oral then written — Transmitted by recitation before it was a text
- Standing
- Live — Still argued in current philosophical work
Transmitted orally with extraordinary fidelity for centuries before being written, and continuously studied since. The oral tradition and the manuscript tradition were parallel rather than sequential.
What the book says
Not one book but a group of texts, thirteen of them usually counted as principal, attached to the end of the Vedas and reading nothing like them. Where the Vedas are hymns and ritual instructions, these are conversations - a teacher and a student, a king and a priest, a husband and a wife, a boy and Death himself - about what is really there. The question they keep returning to is what remains when everything changeable has been set aside, and the answer they keep arriving at is that the self within a person and the reality behind the whole world are not two things. A father tells his son to dissolve salt in water and then find it: it cannot be seen but it is everywhere in the water, and so is the fineness that is the self of all this. A sage asked to describe the ultimate answers by negation, saying only what it is not. A boy who has been given to Death demands to know what happens after dying and is offered every worldly reward instead, and refuses them all. There is no system, and the texts contradict each other. What they share is the move inward: away from the sacrifice, away from the gods, toward the questioner.
What it argues
That behind the changing world there is one reality, brahman; that within a person there is something that is not the body, the breath or the passing states of mind, atman; and that these are the same. The argument, where there is argument, proceeds by elimination: the self is not the body, because the body is left behind in sleep; not the dreaming mind, because that too ceases; not the state of deep sleep, because nothing is known there. What is left is the knower, which cannot be made an object of knowledge because it is what does the knowing. The practical consequence is that liberation is not something achieved but something recognised, and that the machinery of sacrifice, which the preceding literature is entirely about, is beside the point.
What answered it
Badarayana's Brahma Sutras, which are an attempt to make a single doctrine out of texts that do not contain one, and which in turn provoked the commentaries in which Indian philosophy did most of its hardest work. The Buddhist answer was more radical: that the enquiry into what remains when everything changeable is set aside terminates in nothing at all, and that atman is precisely the assumption to be given up. That disagreement is the central argument in Indian philosophy for the next fifteen hundred years.
How it argues
By conversation, image and refusal. There is little that looks like formal argument; the characteristic moves are the analogy pressed until it yields something, and the flat denial that the ultimate can be described positively at all. The most famous formula is a double negative used to stop the student from settling on any answer. This is why the texts sustained a thousand years of commentary: they state positions strongly, in different places, without reconciling them, so that the work of deciding what they say had to be done by somebody else.
The moment
Northern India in the centuries when cities were reappearing, kingdoms consolidating and the older sacrificial religion coming under pressure from within. Several of the most important conversations are set at courts rather than in priestly households, and in more than one a king instructs a brahmin, which the texts note as a reversal. The same period produced the Buddha and Mahavira, and the questions all three traditions ask - about the self, about what survives death, about whether ritual does anything - are recognisably the same questions.
How it travelled
By memory first, and to Europe by an extraordinary route. Within India the texts were preserved orally with a fidelity that has few parallels anywhere, using recitation schemes designed to make error detectable. They reached Europe through a Mughal prince: Dara Shukoh, eldest son of Shah Jahan, had fifty Upanishads translated into Persian at Delhi in 1657, believing they were the hidden book referred to in the Quran. He was executed by his brother two years later. In 1775 a manuscript of that Persian version reached the French scholar Anquetil-Duperron, who finished a Latin translation of it in 1796 and published it at Strasbourg in 1801-02 as the Oupnek'hat - the first European translation of a major Hindu text, at two removes from the Sanskrit. In the spring of 1814 Schopenhauer read it, and called it the consolation of his life.
The strongest objection
That the identification of self and ultimate reality is asserted rather than shown. The eliminative argument establishes at most that the self is not any of the items surveyed, which does not establish that it is identical with the ground of everything; that further step is stated, illustrated and repeated but not defended. The Buddhist critique presses exactly here, and presses hard: what the analysis actually finds, when the body and the mental states are set aside, may be nothing at all, and the conviction that something remains may be the assumption the analysis was supposed to test.
Why read it
Because the good passages are as good as anything in philosophy and take minutes to read. The boy Naciketas bargaining with Death in the Katha, the salt dissolved in water in the Chandogya, the conversation between Yajnavalkya and his wife before he leaves the world in the Brhadaranyaka - none of these needs a background, and all of them stay. Read a translation with notes; the terms do not survive being rendered into English on their own.
- First printed
- 1801
- First in English
- 1832
- Digitised
- Yes
The Sanskrit is stable and was fixed by an oral tradition that preserved accent and metre. Radhakrishnan's 1953 edition prints text and translation together and is still widely used.
Max Muller's Sacred Books of the East volumes of 1879 and 1884 are public domain. Olivelle, Roebuck and Easwaran are in copyright.
Where it stands
Live in every sense. They remain the foundational texts of Vedanta, which is the dominant strand of Hindu philosophy; they are studied academically as the beginning of Indian thought; and comparative philosophy continues to work on the relation between their account of the self and the positions taken in the Western tradition.
On quoting this book
The Sanskrit is public domain and Muller's English is out of copyright, but his renderings of the key terms are now regarded as misleading. Nothing is quoted.
Contested
Which texts count is not settled: ten, twelve, thirteen and eighteen are all defensible lists, and later works continued to be composed with the name for centuries. The dating is relative only. The central interpretive question - whether the texts teach that the self is identical with brahman, or distinct from it, or both in different respects - is exactly what divides the Vedanta schools, and the reason it was never settled is that the texts support all three readings. The tradition's own account of them as revealed and beginningless and the historical account of them as composed over four centuries are both given here, and neither is adjudicated.