if we have first to establish that the speaker is competent and sincere, then
it seems that our grounds for believing are inferential, and the rationality
implicated in testimony is nothing but a variety of extrapolative rationality.
We reason, in effect, that the present utterance is relevantly similar to past
utterances of the same speaker, which have been seen to be true. Someone
who accepts this (David Hume is a notable case) is led to conclude that it is
indeed rational to believe the utterances of a sincere and competent speaker, but
that no new model of rationality is involved.
What saves the Nya ̄ya theory is the idea that one can “monitor” the
competence and sincerity of the speaker without forming any beliefsabout her
competence or sincerity. One might simply have an internal “lie-detector”
subconsciously monitoring for signs of blushing, fidgeting, and so on. The exis-
tence of such a mechanism makes the following counterfactual conditional true:
if the speaker were lying, one would come to believe it. In the presence of a sub-
doxastic faculty of this sort, one need not attempt to acquire knowledge about
the speaker’s credentials. For one’s readiness to assent to what is being said will
be overridden if she were to lie. Assent is made rational in a negative way, by the
absence of evidence that the speaker is deceitful, rather than by positive evidence
that she is sincere. it is rational in the same way that it is rational for one to
believe that one has not just trodden on a nail. One need have no positive reason
for so believing (a visual inspection of the foot, for example) for one knows that,
if one had just trodden on a nail, one would have come to know about it. The
“reasoning” is ab ignorantiam and not inductive.^46
While the worry about the reduction of testimony to inference is raised in the
Nya ̄yasu ̄tra (NS 2.1.49–51), this defence is not to be found there. It emerges in
the later idea that a precondition for testimony is the “absence of knowledge of
unfitness” and not the “knowledge of fitness,” Va ̄tsya ̄yana says only that testi-
mony depends on the speaker’s credibility and that “inference is not like this.”^47
1.9 Reason’s Checks and Balances
I began this examination of rationality in the early Nya ̄ya with a description of
the disreputable “reasoners” mentioned and criticized in the epics. The ill-repute
was on the grounds that their use of reason was unmotivated, groundless,
unconstrained. The Naiya ̄yika is very careful to avoid this charge. The very term
nya ̄ya, which Va ̄tsya ̄yana identified with reasoned inquiry itself, is contrasted
with another, nya ̄ya ̄bha ̄sa– a pseudo-inquiry. A pseudo-inquiry is one which,
although-otherwise in accord with the rules governing proper reasoning (setting
out the demonstration in a five-limbed format and with a proper reason prop-
erty), contradicts observation and authority (see section 1.2). The same point is
made time and again. A properly conducted debate, one which is friendly and
truth-directed, is one which proceeds with the help of the methods of know-
ledge-acquisition, employs the five-limbed format, and is not in contradiction
hinduism and the proper work of reason 439