In this post, I will primarily summarize the rest of section 5. I will discuss intuition in detail in connection with section 6, but Williamson does raise some worries about intuition in section 5. Again, Williamson wields the evidence neutrality (EN) thesis. This time he uses EN to claim that it cannot be satisfied in relation to the evidential force of intuitions. A theory of intuitions needs to be able to distinguish weak from strong intuitions. This is because a theory of evidence will need to make choices between conflicting intuitions according to their evidential strength. However, according to Williamson, philosophers will tend to overestimate the strength of intuitions they have a vested interest in seeing succeed (i.e., intuitions that support their favored theory of knowledge, evidence or intuitions). Such theory-driven wishful thinking will result in a lack of “uncontentious decidability” as inquirers disagree whether someone (or, they themselves) have an intuition with a certain strength. A distinction was raised by Derek Ball here between the phenomenological and evidential force of intuitions. He argued it is not always clear which interpretation is at stake. Williamson is overtly talking about the evidential force of intuitions, but when talking about aspects of human psychology and vested interests his discussion could be interpreted in terms of the phenomenological force of intuitions. This is because human psychology is to blame for over or underestimating the strength of intuitions. There will be gradations in felt subjective certainty accompanying various intuitions according to whether the intuitions align with one’s professional and psychological interests. Even trying to compensate for bias will be difficult because one can see bias in another person easier than one can see bias in one’s own self. The phenomenological force of intuitions as experienced in consciousness is not a guide to resolving biases. Such factors make it difficult to reach an uncontentious view of the objective facts about intuitions (i.e., their strength in relation to a hypothesis). In fact, Williamson’s point seems stronger when the phenomenological reading instead of the evidential reading is considered. This aligns Williamson’s comments with Jonathan Weinberg’s comments in his paper on the hopelessness of intuitions (i.e., nothing within the intuition signal, or human psychology, is able to adequately recognize and correct cases of intuitions-gone-astray).
In the last half of section 5 Williamson goes into a discussion on the dialectical standard of evidence. Again, Williamson is trying to address the judgment skeptic. This view of evidence avoids the trap of resorting to psychological facts to try and resolve disputes about the evidence. In a dialectical context evidence is that which is uncontroversial in that debate or context. Evidence does not have to be foundational or uncontroversial across all contexts. Instead, it only needs to be uncontroversial in that context; if inquirers can agree over what counts as evidence, then what they agree on counts as evidence in that context. Williamson finds this view of evidence wanting because it results in conceding too much to the skeptic. Accepting as evidence only propositions that are true if one is a BIV, in order to have meaningful debate with the skeptic, does not have to lead one to a wholesale acceptance of skepticism about the external world. However, according to Williamson, giving the skeptic his or her premises forces one into a conclusion that one does not endorse. If one does not play the skeptic’s game, then the dialectical standard of evidence ends up being irrelevant. Another possibility is to switch to a non-dialectical standard. Given this standard the fact challenged by the skeptic is not disqualified as evidence. For example, one might take the fact that a Gettier subject lacks knowledge or the fact that there are mountains in Switzerland as evidence even though these facts are contentious to the skeptic. To ignore these facts is to violate the Carnapian total evidence requirement. Williamson wonders if such a move is a legitimate response to judgment skepticism?
Williamson’s conclusion is that we need to widen our evidential base. Even if intuitions were the most reliable forms of evidence it does not follow that we should restrict our evidence to just intuitions. Facts that the judgment skeptic hold as highly probable (in contrast to contentious facts about the existence of mountains) are not to be the only facts considered. We do not need to play the skeptic’s game just because she holds that only facts that are certain (because they align with the micro-physical structure of the universe) can be admitted as evidence. According to Williamson there is nothing wrong with continuing to claim knowledge of truths in contention. The “dialectic” is not the measure of all things. As Williamson says:
No methodology is proof against misapplication by those with sufficiently poor judgment.
It is not the job of good methodology to silence all people who propose, for instance, astrological predictions as truths. Instead, good methodology must separate good from bad intellectual practices. I imagine Williamson thinks he is doing this by contributing to the literature on philosophical methodology, by uncovering things like the consequence fallacy and other ways methodology goes astray. This is a useful endeavor, but because of Williamson’s clear bias for knowledge-first epistemology, epistemic externalism and a whole host of other theoretical commitments in analyticity, assertion, and so on Williamson’s work could be accused of the very thing he is arguing against, namely, proposing an account of methodology to philosophy in general (even giving the book the sweeping title The Philosophy of Philosophy) in a way that is highly uncontentiously decidable and violates the total evidence requirement by primarily using evidence from Gettier cases and revisionary metaphysics. As I have studied chapter 7 in POP I have kept coming back to inconsistencies between the content Williamson is arguing for and the methodology he is using in arguing for it. I’d be curious to know if anyone else has noticed the same inconsistencies.