eternal. And when something like the soul is the example, then the reason, which
is taken to be “because of being intangible” in accordance with the su ̄tra “a reason
is that which proves what is to be proved by being like an example (1.1.34),” devi-
ates from eternality. For a thought is intangible as well as non-eternal. So, as there
is deviation in both sorts of case, there is no prover–proven relationship as the
mark of a (proper) reason is absent. One end is eternality, one end non-eternality.
So we understand “being at one end.” Opposite to that which is at one end is that
which does not remain at only one end, because it is a pervader (vya ̄paka) of both
ends.Four inferences are compared. (1) A pot is non-eternal. Sound is unlike a pot –
one is tangible, the other intangible. So sound is eternal. However, we also have
this. (2) An atom is eternal. Sound is unlike an atom – one is tangible, the other
intangible. So sound is non-eternal. Again, we have: (3) The soul is eternal.
Sound is like the soul – both are intangible. So sound is eternal. But also: (4) A
thought is non-eternal. Sound is like a thought – both are intangible. So sound
is non-eternal. The implication is that what undermines the inference is the exis-
tence of examples which do not fit, i.e. counter-examples. Another maxim of
extrapolation is in play: do not extrapolate if you know of any counter-examples
(there is no implication that the extrapolator is obliged to look for counter-
examples, however). A prover–proven relationship is one for which no counter-
examples exist. What Va ̄tsya ̄yana lacks, however, is a clear grasp of what makes
something a counter-example. He does not see that only a thing which is intan-
gible and non-eternal ought to be thought of as a counter-example to the infer-
ence from intangibility to eternality. Something which is tangible and eternal
(an atom) is nota counter-example. For the existence of tangible eternal things
is not inconsistent with the rule underpinning the inference, that whatever
is intangible is eternal It is the gaining of a clearer grasp of the notion of a
counter-example that leads one to an understanding of the proper form of
the prover–proven relationship. Again, Din.na ̄ga is extremely precise on the
nature of counter-examples, and can take a lot of the credit, even if the
essential point had been appreciated before him.
My point has been that there are many ways to arrive rationally at belief,
other than that of formal deduction. Informal argument schemes, such as the
inference from sampling, are just as much ways of reaching beliefs that it is
rational for someone to hold, and it is with this wider concept of rational belief
that we make better sense of the early Nya ̄ya philosophical enterprise.
1.8 Reason, Scripture, and Testimony
Is it rational to believe the testimony of others or the statements of the
scriptures? Does the assumption that it is rational to believe what we hear or
read require us to think of rationality in a new way? We have so far encountered
two epistemic models of rationality: the perceptual model, according to which
hinduism and the proper work of reason 437