Śabda
from 200 BCE Nyāya Sūtra; contested by Buddhists and Cārvākas throughout
Testimony treated as a genuine means of knowledge in its own right, not as a poor substitute for seeing it yourself.
- Active
- from 200 BCE
- Where
- North India
- Also
- śabda
- Status
- In use
- Tradition
- Non-Western procedures
What it is
Śabda is knowledge acquired from the words of a reliable speaker. Nyāya treats it as a pramāṇa in its own right — not reducible to inference about the speaker's reliability, but a distinct route by which knowledge arrives.
That claim is the substantive one. The Buddhist logicians denied it, holding that testimony is a species of inference; the Cārvākas denied it entirely. Where you stand determines whether almost anything you know about the world counts as knowledge.
How it works
Identify the speaker. Check the two conditions: that they are in a position to know, and that they are not disposed to mislead. Check that the statement was understood as intended — that the words were being used in the sense the speaker meant. Accept it as knowledge on that basis, without needing to verify the content independently.
The third condition matters more than it looks. Most failures of testimony in practice are failures of understanding rather than of honesty.
The form of the inference
- A speaker has asserted that P
- The speaker is in a position to know whether P
- The speaker is not disposed to mislead about P
- The words have been understood in the sense intended
- Therefore P is known through testimony
The argument
Proposition The bearer rating can be relied on and the fixing detail cannot.
- Both figures come from the same manufacturer's data sheet, and neither has been independently verified.
- For the bearer rating: the manufacturer is in a position to know, having tested the section; they are not disposed to mislead, since an overstated rating exposes them to liability and the figure is certified to a standard with an audit trail.
- For the fixing detail: the sheet specifies a screw pattern for the manufacturer's own bracket. The manufacturer is in a position to know about that bracket and is not in a position to know about the third-party bracket actually being used.
- The third condition catches the rest. The rating is given for a single span, and the drawings show a continuous span over three supports — the words are being read in a sense the speaker did not intend.
Therefore The bearer rating is known through testimony and can be relied on for a single span; the fixing detail is not testimony about this installation at all, and the continuous-span case needs its own figure.
Nobody is being accused of anything. Two of the three conditions failed for reasons that have nothing to do with the manufacturer's honesty, which is the usual case and the reason the third condition is in the list.
What it is good for
Everything you know but have not checked, which is nearly everything. It is a working framework for deciding which specifications, reports and expert statements can be relied on and for what.
Where it fails
The conditions are assessed by the hearer, who is usually the person least equipped to assess them. Whether a speaker is in a position to know about a technical matter is often exactly what the hearer lacks the expertise to judge, and the framework offers no help there.
The deeper dispute is whether it is a distinct means at all. If checking the speaker's reliability is itself an inference, then testimony collapses into anumāna and the Buddhists were right — and Nyāya's reply, that the knowledge arrives directly from the words rather than from reasoning about them, has never been decisive.
And where the speaker is a text rather than a person, the second condition cannot be checked at all, which is precisely where the doctrine was doing its heaviest work.
Common misconceptions
That treating testimony as a means of knowledge means accepting whatever you are told. The conditions are demanding, and most Indian discussions of śabda are about the ways they fail.
Where it came from
The Nyāya Sūtra lists it as the fourth means. Mīmāṃsā develops it furthest, since the authority of the Veda depends on it.
What became of it
Continuously argued about, since it is the pramāṇa on which scriptural authority rests and the schools divide sharply on it.
Lasting influence
Anticipates the twentieth-century Western recognition — Coady's Testimony in 1992 — that testimony is not reducible to inference, a position Nyāya held for two thousand years.
Why it still matters
Directly relevant to how anyone assesses expert evidence.
Put another way
Taking the reading from a gauge you did not build.
Texts
Gautama, Nyāya Sūtra
Book II defends testimony as irreducible, against the position that it is a form of inference.
Where to start
Read Coady's Testimony alongside a summary of the Nyāya position. The convergence is striking.
The surprising bit
Western epistemology treated testimony as second-rate until the 1990s, when it arrived at a position Indian philosophy had defended since the second century.
In their own words
Verbal testimony is the instruction of a reliable person
Who held it
C. A. J. Coady
Connections
- Relies on
- Pramāṇa theory
- Influenced
- Mīmāṃsā rules
The verdict
The most practically important means of knowledge, and the one whose conditions are hardest for the person relying on it to check.