Testimony
Also called: testimonial knowledge, knowledge from words, śabda
Coming to know something because someone told you; the source of most of what any of us knows.
Testimony is coming to know something because someone told you. Almost everything any of us knows about history, geography, medicine or our own early childhood we know this way, which is why Coady and others have treated the epistemology of testimony as central rather than specialised. The classical dispute is between reductionists, for whom a hearer’s warrant must rest on her own evidence that the speaker is reliable, and anti-reductionists, for whom she is entitled to accept what she is told unless she has reason to doubt it.
The modern accounts disagree about much else. Burge grounds testimonial warrant in an a priori entitlement; Williamson in the knowledge norm of assertion; Goldberg in the reliability of an extended process that includes the speaker’s production; Faulkner in the assurance a speaker gives; Lackey in the reliability of the statement rather than the speaker’s belief. I argue that they agree on one thing, which I call the Tracking Minimum: a source transmits testimonial knowledge that p only if the truth of p, or the evidence for it, figures in the explanation of its asserting p, and does so through its own production process rather than through a proxy on which that process depends.
The condition can be met originally, by a speaker whose assertion is governed by her grounds, or derivatively, by a faithful link that preserves a particular assertion from such a source; that is how testimonial chains carry warrant. On the best current evidence, large language models meet it in neither way. Their reliability is a high rate of accuracy, not a sensitivity of production to the truth of what is said, and users who learn from them stand to them as readers to an instrument rather than as hearers to a speaker (see AI testimony).
The Nyāya school, which counted testimony (śabda) as a distinct means of knowledge when it is the statement of a reliable speaker (āpta), likewise puts the question on the speaker’s side: what makes words a source of knowledge is the relation of the one who utters them to the truth.
My essays and papers on it
Related terms
Questions that turn on it
In the library
Referred to by
Further reading
- Coady, C. A. J. (1992). Testimony: A Philosophical Study. Oxford: Clarendon Press.doi.org/10.1093/0198235518.001.0001
- Burge, T. (1993). Content preservation. The Philosophical Review 102(4): 457–488.doi.org/10.2307/2185680
- Lackey, J. (2008). Learning from Words: Testimony as a Source of Knowledge. Oxford: Oxford University Press.doi.org/10.1093/acprof:oso/9780199219162.001.0001
- Saxena, A. (2026). Assertion without a speaker: Testimony, tracking, and large language models. Episteme, 1–22.doi.org/10.1017/epi.2026.10160