Śabda (testimony in Indian philosophy)

Śabda, literally “word” or “sound”, is verbal testimony considered as a source of knowledge in classical Indian epistemology. The Nyāya school counts it as one of four pramāṇas, or means of valid knowledge, alongside perception (pratyakṣa), inference (anumāna) and comparison (upamāna). Gautama’s Nyāya-sūtra (1.1.7) defines it as the statement of an āpta, a reliable speaker, and Vātsyāyana’s commentary glosses the āpta as one who knows the matter and wishes to communicate it without deception. He adds that even outsiders (mleccha) can be āptas: what qualifies a speaker is knowledge and sincerity, not birth.

Not every school agreed that testimony is a distinct source. Vaiśeṣika and Buddhist epistemologists reduced it to inference, from the premises that the speaker is trustworthy and has said that p. Nyāya held that understanding a reliable speaker’s statement normally yields knowledge directly, and that the conditions for certifying inference and testimony differ. On its view the validity of testimony derives from the speaker’s competence and honesty, not from the words alone, and chains of testimony are acceptable when they reach back to someone who knows. Mīmāṃsā, defending the authority of the Veda, held that the Vedic texts have no author and that a sentence can be meaningful without a speaker who intends to communicate; most Naiyāyikas replied that the Veda, like every sentence, has an author, and inferred that the author is Īśvara.

I find the Nyāya placement close to the one I defend: what makes words a source of knowledge is a fact about the relation between the one who utters them and the truth. Its conditions on an āpta ask for more than the Tracking Minimum, which requires neither knowledge nor an intention to communicate, but they include it. And the debate over an authorless text is the nearest classical precedent I know for assertion without a speaker, though the analogy should be held loosely.