What gives a sacred text authority?

What kind of authority do words have?

Picture two notices beside a garden gate. One says, “Close the gate at dusk.” The other says, “The gate is made of oak.” The first you can obey; the second you can check. But knowing what the first sentence means still leaves a question: why should its instruction bind you?

That question brings us to Mīmāṃsā, a tradition of inquiry into the Veda whose thinkers developed rules for interpreting sacred language and arguments about how words can give knowledge. Their central concern was dharma, a word with other meanings elsewhere. Here it means what is to be done, as Vedic prescriptions direct, together with the good those actions are held to bring. [1] [2]

Later in the lesson we turn to Vedānta, a related tradition whose central inquiry concerns ultimate reality. The two share a body of sacred texts, but sharing texts does not settle what those texts primarily teach.

What observation cannot tell us

Jaimini’s Mīmāṃsā Sūtra 1.1.2 identifies dharma through injunction, language that directs someone to act, and Śabara’s commentary asks how such knowledge is possible. You can observe a sacrificial fire, but can observation establish that a prescribed action brings a good beyond anything the senses disclose? In Śabara’s account, establishing that is the Veda’s distinctive work. [1]

Sūtra 1.1.4 draws the contrast: perception depends on contact with an object that is present, while dharma concerns an action still to be accomplished. The claim is about knowledge of dharma in this setting. It doesn’t make observation useless when deciding what to do, and it does not ask scripture to replace ordinary inquiry into visible things. [1] [2]

Lesson 9 introduced pramāṇa, a source of knowledge, and different sources can do different jobs. Seeing tells you that a gate is shut; someone’s words can tell you what you have undertaken to do. Mīmāṃsā gives Vedic injunctions a much stronger authority where dharma is concerned, and the difficulty is to explain why these particular words deserve it.

A defense without a human author

One answer is apauruṣeya: the Veda is held to have no human author. This is a philosophical claim about its origin and authority, and a much stronger one than saying we have forgotten who composed it. A human speaker can misunderstand, deceive or describe something beyond their knowledge; a text with no author would be free of the errors those failings produce. That is the defense. It is not historical evidence that the Veda was never composed. [2]

The defense also depends on an account of language. In the discussion accompanying 1.1.5, word and meaning are treated as connected independently of anyone’s first decision to join them. Anyone proposed as the inventor of language would already need meaningful words to announce the meanings being invented. The alternative to a first inventor is learning from earlier speakers. [1] [2]

Both points belong to a larger argument. Elisa Freschi explains the Mīmāṃsā view that what we take ourselves to know has authority unless a reason to reject it emerges. We need not verify every piece of knowledge by another before relying on it; if we did, each verification would require a further one. On this view, Vedic knowledge of dharma has no fallible speaker behind it and meets no defeating evidence within its own field. [2]

A critic can grant that endless checking is impossible and still ask why this tradition’s words should have unique authority. The absence of any known error is especially hard to assess when ordinary observation cannot test the claim. That question goes to the argument’s assumptions. Dismissing its defenders as merely credulous would miss their reasoning.

Authority still requires interpretation

Even an accepted authority has to be understood. Does a passage direct an action, state a condition, explain a procedure or praise a result? Mīmāṃsā develops general rules for answering such questions; a descriptive passage, for instance, may support an injunction without itself issuing a separate command. [2]

With our gate notices the distinction is easy. “The gate is made of oak” supplies information; it does not tell you when to close the gate. More complicated texts call for judgments about how sentences fit together. Nor does a statement that an action would achieve someone’s aim automatically create a duty to pursue that aim. Mīmāṃsā thinkers disagree about how prescription, desire and obligation are related; theirs is a tradition of debate, not a single automatic procedure for reading. [2]

These discussions also assumed a particular social setting: the Vedic obligations these authors considered applied to people with specified qualifications and inherited social positions. Studying their reasoning today carries no such conditions. [2]

Vedānta asks a different question

Vedānta means the end of the Veda, and the name commonly designates the Upanishads; philosophical Vedānta interprets those texts systematically. Brahman is ultimate reality; ātman is the self. As lesson 6 showed, asking how they are related takes more than recognizing the words. [4]

The Brahma Sūtra, traditionally attributed to Bādarāyaṇa, opens with an inquiry into brahman, and its second aphorism concerns that from which the world originates. The question is no longer what is to be done but what reality is and how it can be known. Andrew Nicholson warns that these terse aphorisms received competing interpretations, so later Vedānta should not simply be read back into the early text. [3] [4]

One later answer comes from the eighth-century thinker Śaṅkara. In his commentary on 1.1.4, he defends the Upanishads’ power to disclose brahman. Their teaching need not serve a ritual instruction to have a purpose, because knowledge can remove ignorance. On his account, liberation is not a new result manufactured by an action. [3] [4]

Śaṅkara distinguishes knowing from action, which you can choose to perform differently. Our gate, an analogy of our own, shows the difference: you can choose whether to inspect it, but what you then know about its material depends on the gate rather than on your wishes. His further claim, that the Upanishads disclose brahman, still needs a defense. [3]

So Mīmāṃsā’s concern with prescribed action and Vedānta’s inquiry into brahman lead in different directions, and it is their shared texts and interpretive methods that let them disagree within a common conversation. Neither name stands for every Hindu view of authority, self or liberation. [2] [4]

Pause and try

A reader finds an unsigned notebook giving instructions for a good life. “There is no author’s name,” she says, “so Mīmāṃsā proves that its instructions must be reliable.” What has she confused? How would her task change if a sentence in the notebook described reality rather than directing an action?

Read an explanatory answer

She has confused an unknown author with no human author at all, which is what the Mīmāṃsā claim about the Veda asserts. The defense also invokes a theory of language, knowledge and dharma, and a missing signature establishes none of that. Interpreting an injunction requires asking what is prescribed, to whom and under what conditions; interpreting a claim about reality requires asking what it says is the case and how that could be known. Even accepting a text’s authority would leave this work of interpretation to be done.

Carry this forward

Words can direct conduct, make claims about what is real and challenge the terms in which we think. The next lesson returns to Buddhist inquiry with Nāgārjuna, who asks what follows if things lack an independent nature of their own. His argument will then have to explain how words can teach if the things they describe have no such nature.

Sources and limits

[1] Jaimini’s Mīmāṃsā Sūtra with Śabara’s commentary, Shabara-Bhasya, volume I, translated by Ganganatha Jha, Oriental Institute, Baroda, 1933, for injunction, the limits of perception and the connection between word and meaning.

• Passages read: sūtras 1.1.2–5 with their opening explanations, and selected later discussion attached to 1.1.5, on printed pp. 4–9 and 24–27. The whole commentary on 1.1.5 was not read.
• In the scan: these are PDF pages 26–31 and 46–49 of the consulted 752-page scan. The title page and printed pp. 4, 8 and 24 were also checked as images.

Jha translates autpattika as “inborn” and the commentary explains it as “constant”; Freschi uses “natural” and discusses the alternatives. The lesson explains the argument without offering a new Sanskrit translation. The edition has several layers: Jha’s bracketed explanations and references to later commentators are distinct from Jaimini’s sūtras and from Śabara’s commentary, which itself transmits earlier commentary. No independent Sanskrit analysis was undertaken.

[2] Elisa Freschi, “Mīmāṃsā,” chapter 15 in History of Indian Philosophy, edited by Purushottama Bilimoria, Routledge, 2018. The opening and all six substantive sections were read, from “Historical development” through “Mīmāṃsā and God,” plus the two notes. Freschi distinguishes schools and developments, and the usual dating of Śabara to the third–fifth centuries CE remains tentative. The lesson selects arguments about dharma, language and authority; it offers no instructions for ritual performance.

[3] Bādarāyaṇa, Brahma Sūtra, with Śaṅkara’s commentary, translated by George Thibaut, The Vedânta-Sûtras, part I, Sacred Books of the East 34, 1890. The opening aphorisms and selected commentary were consulted at 1.1.1, 1.1.2, 1.1.3 and 1.1.4. In 1.1.4, the selected discussion concerns three points: whether knowledge serves injunctions, liberation as distinct from a produced result, and knowledge as depending on its object. The early aphorisms, Śaṅkara’s later interpretation and Thibaut’s English explanations are separate layers. The lesson claims neither a complete reading of the commentary nor a collation against page images.

[4] Andrew J. Nicholson, “Early Vedānta,” chapter 22 in the same 2018 collection. “Introduction” and “The Brahma-Sūtra” were read in full. Nicholson’s reconstruction stresses early diversity and the danger of projecting later views backward. The chapter as supplied appears inconsistent in its dating of the compilation, so the lesson adopts no precise compilation date. The gate and notebook situations are original teaching examples. Neither the collection nor any source volume was read in full.