Phenomenal Conservatism Über Alles

Here, I defend PC against some objections.*

[ *Based on: “Phenomenal Conservatism Uber Alles,” in Seemings and Justification: New Essays on Dogmatism and Phenomenal Conservatism (Oxford University Press, 2013), pp. 328-50. ]

Background

Phenomenal Conservatism (PC) holds that, in the absence of defeaters, you have justification for believing whatever seems to you to be the case. I discussed this idea previously:

• https://fakenous.substack.com/p/compassionate-phenomenal-conservatism

• https://fakenous.substack.com/p/phenomenal-conservatism-and-self

1. What Are Seemings?

1.1. Against Analysis

Michael Tooley objected that I offered no analysis of “seems”. Is this bad? No, because

a) No philosopher has ever correctly analyzed any concept.

b) The way we learn words is not from reading definitions but from seeing examples of their usage.

c) I am not introducing a special usage of “seems”. I am, by stipulation, using “seems” in the ordinary English sense (whatever that is, even if I am mistaken about the nature of seemings).

1.2. Disposition to Believe

Some people think that “It seems to me that P” means that I have a disposition to believe that P. This is mistaken because it’s possible to be disposed to believe that P because, say, you want P to be true, and this would be quite different from being disposed to believe P because P just seems true.

1.3. One Sense of “Appear”

Chisholm thought that there were three uses of “appear” words. But I think his three uses can all be explained by a single sense of “appear”. That single sense is one in which “appear” and “seems” report a particular type of propositional attitude (one that normally causes belief but is distinct from belief). Chisolm’s 3 uses were:

1. The Phenomenal Use: In this usage, “This apple appears red” reports that the apple causes a particular quale, a red quale. I disagree; I think “This apple appears red” reports a sensory experience that represents the apple as being red. (It happens that there is a particular quale that goes along with this, but that’s not what you’re saying.)

2. The Comparative Use: In this usage, “This apple appears red” means that the apple appears the way that red things usually appear. Again, I disagree that this is a standard English meaning of that sentence. I think it means that one’s sensory experience represents that the apple is red. (Of course, when that happens, it will also generally be true that the apple appears the way that other red things appear.)

3. The Epistemic Use: In this usage, “This apple appears red” means that the speaker believes or is inclined to believe that the apple is red. Again, I don’t think it ever means that. I think that is just an implication that people commonly draw because, when your sensory experience represents something as being red, then you normally are inclined to believe that it is red.

1.4. Probability Beliefs

Berit Brogaard proposes that at least some seemings are just probability beliefs, i.e., “It seems to me that P” sometimes just means “I believe that probably P”. I think this is mistaken because

a. It could seem to you that P even if for some reason you don’t believe in probability (like you’re a probability skeptic);

b. You could believe that P is probable without its seeming to you that P. Suppose that I believe that there is probably life after death, but I believe this solely through wishful thinking, i.e., just because I want there to be life after death. That does not count as its seeming to me that there is life after death.

1.5. Awareness of Apparent Evidence

Maybe an appearance that P is just a mental state that the subject takes as evidence that P. But what is this attitude of “taking”?

a. If we say the subject has to believe that his mental state is evidence that P, that’s false. E.g., an external world skeptic might deny that appearances are evidence; this skeptic does not, however, lack appearances.

b. Maybe appearances are just states that seem to the subject to be evidence that P. But this is circular, using the notion of seemings to explain appearance. You might as well just say an appearance is a state of its seeming that P.

Conclusion: The best account is that a seeming is a particular type of experience, different from belief but normally causing belief, which represents that P.

2. The Reliability Problem

Say you believe that P because it seems to you that P. I say this is enough (absent defeaters) to have some justification for your belief. But some epistemologists disagree; they think that you also need some positive justification for thinking that your seemings on the subject of P are reliable. Otherwise, they say, it will appear to you that it must be merely an accident if you got the right answer about P. So they would say, while seemings might be necessary for justification, they aren’t sufficient (even in the absence of defeaters).

My main problem with this is that it quickly leads to global skepticism. Whatever belief-forming method you use to conclude that your seemings are reliable, we can also ask how you know that that method is reliable. (In fact, I think that method will pretty much always appeal to other seemings.) Any rationale that lets you dismiss this second question would probably also let you dismiss the original question, about how you know that your first seemings are reliable.

So the demand for independent evidence of reliability generates an infinite regress. Instead, we should merely require that the subject lack grounds for doubting that his seemings are reliable.

3. How Does Inference Justify?

3.1. Inferential Appearances

It turns out that there are two kinds of appearances: Ordinary (categorical) appearances, and inferential appearances. An ordinary appearances just represents some proposition, P, as being the case. An inferential appearance represents some proposition, P, as being the case (or being likely to be the case) given (or in the light of) some other proposition, E.

The original statement of PC is meant to be about ordinary appearances. In the case of inferential appearances, a subject has justification for believing P provided that (i) for some E, it seems to the subject that given E, P is (likely) true, and (ii) the subject is justified in believing E.

3.2. Fallacious Inference

An interesting implication of my theory of inferential justification is that a fallacious inference can justify a belief. E.g., if you go through an apparent proof of P, starting from well-justified premises, and after reviewing it very carefully, you can’t find any mistakes in it, then you’re justified in believing P. It might later turn out that the proof contains a subtle error, which would require you to give up believing P. However, it would still be true that before discovering the error, you were justified in believing P.

3.3. Double Counting

Michael Tooley objects that, according to PC, a subject who infers P from E will often have two justifications for P: (i) an inferential justification, based on E, and (ii) it will (often, at least) seem to the subject that P is correct, simply because of this inference. Thus, this subject will have more justification for P than we intuitively think they should have.

In response, this subject does not have more justification for P than that provided by the inference from E, because the seeming-based justification is fully redundant with the inferential justification. (In my view, the inferential justification just is the seeming-based justification.)

Analogy: Suppose action A brings about some means M, which is a means to your final end, E. Then we could say you have “two reasons” for doing A: (i) that A causes M, and (ii) that A causes E. But this does not give you overall more reason to do A than you would have if A just caused E directly, because reason (i) is fully redundant with (ii), given that causing M is just the way that A causes E.

4. Enablers

In arguing for PC (https://fakenous.substack.com/p/compassionate-phenomenal-conservatism), I used the premise,

A belief is justified only if it is based on a valid source of justification.

Peter Markie objects that this doesn’t work so well for inferential justification. When I believe P based on an inference from E, my justification for P depends upon my justification for E. (If E wasn’t justified, then P is not justified.) However, my justification for E cannot cause anything, because it is a purely normative, not descriptive, state of affairs. Since it can’t cause anything, it can’t be part of the actual basis for my belief that P. Therefore, it can’t be required that my belief that P be based upon everything that explains why P is justified.

Markie goes on to propose that there are conditions called “enablers”, which merely enable some other condition to confer justification for P. And for your belief that P to be justified, the belief has to be based on the conditions that confer justification on P, but it doesn’t have to be based (even partly) on the enabling conditions. In inferential justification, [E is justified] is supposed to be an enabler, which enables the inference from E to P to justify P.

He then proposes that one of the enabling conditions for a seeming to confer justification on P is that you have background justification for thinking that your seeming is reliable. This view conflicts with PC, but it avoids my argument for PC (https://fakenous.substack.com/p/compassionate-phenomenal-conservatism) since it doesn’t propose any additional conditions that confer justification.

My main problem with this is as stated in sec. 2 above: I think that if you require independent evidence of reliability of appearances, then you get an infinite regress, leading to skepticism.

5. Cognitive Penetration

“Cognitive penetration” is the phenomenon whereby your beliefs can alter your appearances. Normally, the causal arrow goes the other way: appearances explain beliefs. But it’s possible for influence to go the other way.

So say a person adopts an irrational belief that P. Suppose that this belief then causes them to experience an appearance that P. (Example: you adopt a belief in life after death purely through wishful thinking. But having this belief subsequently causes you to feel that life-after-death scenarios are plausible.)

A popular objection to PC is that it enables the person to be justified in believing P in such cases. But many people want to reject this; they think that the irrationality of the original belief means that the appearances induced by it should not count as sources of justification.

In reply, we can imagine two cases:

a) Suppose the person knows that their appearance that P is caused by their belief that P. In that case, they have a defeater for their prima facie justification for P, so they’re not justified in believing P.

b) Now suppose the person doesn’t know that their appearance is caused by their previous belief. In that case, I don’t see why they wouldn’t be justified in believing P. I just think PC gives the right answer in this case.

Example: Suppose I recently ingested a special kind of hallucinogen that causes turtle hallucinations. Now I’m seeing a turtle. If I know about the hallucinogen, then I have a defeater for my would-be belief that there is a turtle in front of me. But if I don’t know that I ingested that hallucinogen, then I’m straightforwardly justified in believing in the turtle. And that’s true even if I irrationally ingested the hallucinogen, and even if an irrational belief caused me to ingest that hallucinogen.

6. PC Conquers All

In one early critique, Michael Tooley complained that my defense of PC implied that “direct realism conquers all in epistemology”. I think he meant to suggest that PC makes epistemology too easy.

But I think it’s a virtue that the theory explains, via one simple principle, the justification of all sorts of beliefs, from perceptual beliefs to moral beliefs, to mathematical beliefs, to memory beliefs. It may be the only simple, natural view that does this.

If you don’t agree that appearances are the source of justification, you try saying what else might possibly account for the justification of all these beliefs.