‘the search for a purely operational standard of evidence is …vain. It can be argued on quite general grounds that no standard of evidence is such that we are always in a position to know what our evidence is (Williamson 2000, 93-113; 147-83).’One central argument for the claim proceeds as follows:
- Facts about S’s present mental states aren’t always cognitively accessible to S. In particular, how is S to judge whether S’s present conscious inclination to judge that P is strong enough to count as an intuition that P?
- Either all inclinations are strong enough to count as intuitions or not all are.
- If all such inclinations count, then very weak and very strong intuitions will be lumped together. (i.e. they’ll have the same evidential impact.)
- But a theory of evidence must distinguish the evidential impact of very weak and very strong intuitions.
- If the evidential force of an intuition varies with its strength, then we need a way to measure the strength of the intuition.
- To do so, we need a common scale of strength.
- But then, there will be scope for misjudging the strength of one’s intuitions, because, given human nature, there will be a temptation to overestimate the strength of one’s intuitions, even unconsciously.
- But then, if one looks to correct for this, one might underestimate the strength of one’s intuitions.
- S cannot always know the strength of S’s inclination to consciously judge that p.
- So, S doesn't always know what S's intuitions are.
That argument is perhaps not irresistible. The description a theory of evidence in (4) is ambiguous. It could mean: ‘A theory of the nature of evidence’, or ‘a theory of how evidence works’ (how it justifies). If
The internalist could say that any conscious inclination of S’s to judge that P is an intuition that P. Now, if S has a conscious inclination to judge that P, typically S is aware that he has a conscious inclination to judge that P. If I intuit that P, P seems true to me, for, surely, judging that P means representing to myself that P is the case. If P seems true to me, then I’m defeasibly justified in believing that P. If P’s seeming true to me defeasibly justifies me in believing that P, then P’s seeming true to me is (defeasible) evidence for P (for me) . So, if S judges that P, S typically knows that he has evidence that P. One can know that e is evidence for P without knowing exactly how good e is as evidence for p. Therefore, if S judges that p, S has evidence for P, even if S doesn’t know exactly how much credence he ought to attach to it.
 It’s unlikely that a healthy alert observer would be mistaken about whether he was judging that P.
 If I judge that P, then P seems true to me. This will usually mean that my personal Pr (P) has increased.