Smith's Master Argument Against Naturalistic Accounts of Perceptual Knowledge in Naturalism and Our Knowledge of Reality

Last time we discussed Smith's book, we focused on his master argument against naturalistic accounts of concept acquisition and correction. Here I'd like to focus on his master argument against naturalistic accounts of perception. As with the former argument, Smith nowhere explicitly lays out the argument in standard form. However, I think it can be fairly expressed as follows:

1. If naturalism is true, then either naturalistic direct realism (think Armstrong, Dretske et. al.) is true or some version of indirect realism is true.
2. If some version of indirect realism is true, then we don’t have adequate internally accessible grounds for believing that our perceptual experiences are caused by and accurately represent the external world.
3. If we don’t have adequate internally accessible grounds for believing that our perceptual experiences are caused by and accurately represent the external world, then we don’t have perceptual knowledge of the external world.
4. Therefore, if some version of indirect realism is true, then we don’t have perceptual knowledge of the external world.
5. If naturalistic direct realism is true, then a causal chain stands between the external world and our perceptual beliefs.
6. If a causal chain stands between the external world and our perceptual beliefs, then we don't have internally accessible grounds for believing that our perceptual beliefs are caused by and accurately represent the external world.
7. If we don't have internally accessible grounds for believing that our perceptual beliefs are caused by and accurately represent the external world, then we don’t have perceptual knowledge of the external world.
8. Therefore, if naturalistic direct realism is true, then we don't have perceptual knowledge of the external world.
9. Therefore, if naturalism is true, then we don’t have perceptual knowledge of the external world.

The argument is clearly valid. However, it's also clear that one can raise a number of worries about its premises and their justification. Here though, I'd like to raise three worries for the argument. First, perhaps the most obvious worry is that the argument relies on the truth of internalism with respect to knowledge (See premises 3 and 7). It would thus be helpful if Smith engaged with the recent literature on the internalism/externalism divide, and  provided an argument for epistemic internalism in his book. Unfortunately, one will look in vain to find such. 

Second, it seems that Smith is operating with a very narrow notion of internally accessible grounds, viz., (i) immediate, direct acquaintance with the referent (for basic beliefs), and (ii) an argument from (i)-type grounds (for non-basic, inferential beliefs). But this is to ignore an account of internally accessible grounds from one of the dominant versions of epistemic internalism today, viz., phenomenal conservatism. According to the latter sort of account, perceptual seemings constitute a basic source of evidence or justification for beliefs, even if these seemings don't involve immediate, direct acquaintance with their referents. So, for example, it seems to me that I'm in my office. In at least this sense, then, I have non-inferential, internally accessible evidence that I'm in my office. Why isn't this sufficient internally accessible (defeasible) evidence for the perceptual belief at issue? Again, one will look for an answer to this reply (and, more generally, for engagement with the literature on phenomenal conservativism) in vain if one looks for it within the pages of Smith's book.

Finally, Smith doesn't adequately develop his grounds for thinking that perceptual knowledge is impossible even on the conjunction of indirect realism and his narrow conception of epistemic internalism. This is troubling, given semi-recent arguments that begin with Smith's starting points and reach the opposite conclusion. See, for example. the case for an abductivist route to internalist accounts of knowledge and justification for perceptual beliefs from Bonjour, Vogel et. al. 

In short, Smith fails to show that naturalists who are (i) externalists about both knowledge and justification, (ii) externalists about knowledge but internalists about justification, or even (iii) internalists about both knowledge and justification -- i.e., virtually everyone -- can't have perceptual knowledge or justified perceptual beliefs.  For at least these reasons, then, Smith's master argument against naturalistic accounts of perceptual knowledge is unsuccessful.