Last week I discussed a bit about internalism and externalism of various sorts. One of the big issues in the externalism debate is the issue of luminosity. Luminosity is the property that when a condition obtains we are always in a position to know that it obtains. Thus a philosopher might argue that pain is a luminous condition. Because the only way to be in pain is to be aware of pain. If you are asleep or unconscious you might be having the same stimulus but aren't in pain. Now it is important to realize that the externalist doesn't doubt that there are some mental states that are luminous. As Timothy Williams argues though it seems that only trivial conditions are luminous. His approach is to start discussing complex situations which are clearly mental, but which seem difficult to distinguish. If we can't distinguish two different mental states then clearly luminosity can't be true for all mental states.
Williamson's argument in Knowledge and Its Limits is a slowly changing mental state. If we can't tell when it stops, then clearly there is a problem with luminosity. It's not surprising that Williamson would make an argument like this, given his work on vagueness. It really is a very similar argument. If we have a vague knowledge of boundaries between mental states, then can we really argue there is luminosity? Williamson's argument though really is one involving time. His argument, as I understand it, is that it isn't enough to know about a mental state, we have to know (or be able to know) when it obtains. Thus it is not enough to know that we are in pain. We must for any mental state be able to tell when it ends.
I'm not sure we can do that for many mental states such as belief. Yes for many examples we can tell that we believe. But can we tell when we ceased to believe something? Now the objection might be that not knowing isn't at issue. It is that we could have known. Yet, just from my own admitted arm chair philosophizing, even analyzing pain it seems that there is a blurriness where I can't tell when it ends. That vagueness seem to apply for the mental.
This seems especially tricky if Dennett is correct in that there can be discrepancies between objective and experienced time. If we represent (and thus see) B as occuring before A, when in fact A happened (and was mentally processed) first, is that a problem for luminosity?
Good question. I don't know.
My first inclination would be to think it does. Yet at the same time I halfway think that this may also be a weakness in Williamson's argument. Why does the when matter this much? Yet this may just be a recognition that our knowledge of mental states is, in fact, vague.
I read most of the objections to Williamson quite some time ago but plan on rereading them this week to see if they address this is a fashion I'd agree with.
I should add that Williams has a more recent paper for a forthcoming book collection on epistemology that expands his argument from the form it has in Knowledge and its Limits. "Probabilistic Anti-Luminiosity".
There he argues that luminosity entails that the probability of condition C when one believes that C is 1. More or less from the same line of reasoning he used in his book he varies the state. At some point we may believe we are in pain but our belief doesn't have 100% probability.
This also deals with the time issue, Richard. Rather than saying that we know C at time t (which is problematic at small time scales) we say we have a probability that C given belief that C between some interval [to - tn]. That avoids problems due to unreliability being due to the size of the interval.
The details of Williamson's argument get a tad complex so I'd check out the above paper to get them. His new argument "requires an example in which the condition at issue fails to obtain throughout a small open interval" and it is thus "inapplicable to a slightly wider range of conditions than the original anti-luminosity argument was." Thus while the range of examples is limited somewhat, the overall argument isn't affected by there being a smallest reliable time scale. So I think Dennett's case doesn't pose problems.
Here's that paper I mentioned by Williamson. "Why Epistemology Can't be Operationalized" It looks like it was a significant rewrite of the original which was deleted. It was just added a couple of months ago so presumably replies to more recent critiques. I've not had time to read it yet.
I've closed comments in order to avoid spam since I don't check this older blog as much anymore.
and Theology
Number of unique visitors:
Blogged by Clark Goble