Phenomenal Conservatism and the Internalist Intuition

Here, I explain why Phenomenal Conservatism is the preferred form of epistemological internalism.*

[*Based on: “Phenomenal Conservatism and the Internalist Intuition,” American Philosophical Quarterly 43 (2006): 147-58.]

1. Internalism vs. Externalism

Internalists in epistemology are basically people who think that knowledge requires justification, and justification is determined by “internal” factors, i.e., things in your own mind, or things that you have introspective access to.

Externalists are people who deny internalism. I.e., they either deny that knowledge requires justification or say that justification depends partly on something external to the mind, or that you lack introspective access to. E.g., reliabilism holds that knowledge is, roughly, true belief formed by a reliable method—note that it’s the actual reliability of the method, regardless of whether the subject knows that it’s reliable, that counts. Some also say that the “justification” of a belief is determined by the actual reliability of the belief-forming method.

2. A Central Internalist Intuition

I thought of this hypothetical to illustrate a core internalist intuition. Let’s say Sue has two belief-forming methods, say, sensory perception and clairvoyance. Both of these methods seem to her relevantly alike. Neither seems any more reliable than the other, nor does she have any reason for suspecting that either of them is more reliable. But suppose that in fact one of them is more reliable. E.g., she has reliable clairvoyance, but her sensory perception is completely unreliable (she’s constantly hallucinating, maybe because she’s a brain in a vat).

One day, she seemingly sees a squirrel, which makes her inclined to believe P, “There is a squirrel here.” At the same time, she clairvoyantly perceives a unicorn on Mars, which inclines her to believe Q, “There is a unicorn on Mars.” Again, unknown to her, her perception is unreliable, whereas her clairvoyance is reliable. What should Sue believe?

According to the reliabilist theory of justification, she should believe Q but not believe P. Suppose she did that. Suppose someone then asked her why she did that. What could she say? She would have to say something like this:

“P and Q both seem to me equally likely to be true. My clairvoyance and my perception both seem equally reliable, and I have no reason for doubting either P or Q any more than the other. Nevertheless, I accept Q but I refuse to accept P, for no apparent reason.”

That seems like an absurd thing to say; this could not be a correct report of a rational state of mind. Therefore, this combination of attitudes is not rational. Therefore, the reliabilist theory of justification must be wrong.

This argument generalizes to any externalist theory of justification. On any such theory, justification depends on some external factor, which (by the meaning of “external”) could differ between two beliefs even while the beliefs seemed exactly alike to the subject in all epistemically relevant respects. So any externalist theory would license the above absurd speech, in some possible circumstances.

So that is what I take to be one of the central motivations for internalism in epistemology, perhaps the central motivation. Internalism is motivated by taking the subject’s perspective, thinking about what could be rational to say from the subject’s own point of view.

3. The Natural Form of Internalism

Epistemologists have a couple of different ways of formulating the internalist view:

a. Access Internalism: The (degree of) justification for a belief is entirely determined by factors that the subject has introspective access to.

b. Internal State Internalism: The (degree of) justification for a belief is entirely determined by what goes on in the subject’s mind.

I suggest a third formulation:

c. Appearance Internalism: The (degree of) justification for a belief is entirely determined by how things seem to the subject.

This third account is more specific than the previous two (since there could be accessible internal states other than appearances). According to (c), there could not be two cases in which everything appears to the subject exactly the same and yet in one case the subject has a justified belief and in the other an unjustified belief.

Appearance Internalism is supported by the argument of section 2 above. As long as you make justification dependent (even partly) on anything other than the appearances, you can construct examples in which a supposedly rational subject would have to say something along the lines of the absurd speech above, i.e., something like, “These two propositions seem to me to be relevantly alike in all respects, yet I just believe one and not the other, for no apparent reason.”

So I suggest that if you’re an internalist, you should be an Appearance Internalist.

4. Phenomenal Conservatism

If you’re an appearance internalist, the most simple, natural view is phenomenal conservatism. If appearances determine what you’re justified in believing, presumably it’s that appearances give you justification for believing what appears to you to be the case — rather than, say, what appears not to be the case, or some other proposition unrelated to what appears to be the case. Also, if appearances are a source of justification, presumably it’s that they provide justification when we lack grounds for doubting them — rather than, say, that they provide justification even when we have grounds for doubt, or that they provide justification when some completely different condition holds.

That’s basically the phenomenal conservative position, i.e., that if something seems correct to you, and you have no reason to doubt that appearance, then you have at least some justification for believing that thing.

In sum,

1. You should be an internalist, to avoid endorsing absurd speeches.

2. If you’re an internalist, you should be an appearance internalist.

3. If you’re an appearance internalist, you should be a phenomenal conservative.

4. So overall, you should be a phenomenal conservative.