Epistemological Asymmetries Between Belief and Experience

Here, I explain how belief differs from experience in its epistemological role, and I address an objection to Phenomenal Conservatism raised by Susanna Siegel.*

[* Based on: “Epistemological Asymmetries Between Belief and Experience,” Philosophical Studies 162 (2013): 741-8. ]

1. Siegel’s Objection

My own view in epistemology, Phenomenal Conservatism (PC), holds that when it seems to you that P, and you have no specific grounds for doubting that appearance, that gives you some justification to believe that P.

One important objection to PC points to appearances that have irrational or unreliable causes, inviting us to agree that such appearances fail to supply justification, even if the subject is unaware of the irrational causes. Thus, PC is overly liberal about justification.*

[*Susanna Siegel, “The Epistemic Impact of the Etiology of Experience”, Phil Studies 162 (2013): 697–722.]

Example: S is a Democrat and, like all Democrats, is deathly afraid of guns. S is so preoccupied with the danger of guns that she tends to see guns everywhere. For instance, one day, she opens her refrigerator and looks inside. There’s a banana next to the egg carton (see image).

But S’s paranoia makes her hallucinate a gun in place of the banana (see image).

Assume that, like most Democrats, S is blissfully unaware of her own biases and thus has no idea that her fear of guns is making her hallucinate. As far as she can tell, she is just seeing a gun in the normal way. According to PC, S is justified in believing that there is a gun there.

Siegel disagrees. She cites the following principle (my paraphrase):

DP If an experience has an etiology such that a similar belief with a relevantly similar etiology would be unjustified, then that experience fails to confer justification on any beliefs based on it.

If a person had a belief that there was a gun in the refrigerator based purely on the fear of guns, we would say that belief was unjustified. Accordingly, that belief could not be legitimately used to justify any other beliefs.

Similarly, Siegel thinks, if an experience represents that there is a gun in the refrigerator, and the experience is caused purely by the fear of guns, then that experience cannot be legitimately used to justify any beliefs.

2. Reply

The foregoing doesn’t seem right to me. It seems to me that, in the scenario as described, S would be justified in believing that there was a gun in the refrigerator. This strikes me as a perfectly clear case of justified belief.

If S didn’t think there was a gun there, after having the experience described above, she would be crazy. How could she explain what happened? “Well, it looks to me exactly like there’s a gun there, but I refuse to accept that”? Why? She couldn’t say “because my sensory experience is caused by my fear”, because she doesn’t know that. More generally, she couldn’t cite any reason why she was refusing to accept the evidence of her senses, since by stipulation, she has no defeaters. (If she has defeaters, then PC agrees that she lacks justification, so there’s no counter-example to PC here.) So she’d have to say she was just randomly deciding not to trust her senses on this occasion, for no apparent reason.

This seems irrational. Therefore, I think DP is false, and S ought to believe that there is a gun in the refrigerator.

3. How Does Experience Differ from Belief?

As noted in sec. 1, if you had a belief caused directly by fear, that belief could not be used to justify other beliefs. Therefore, you might think, if you have a sensory experience caused directly by fear, that experience also should not be usable to justify beliefs. What about experiences might be relevantly different from beliefs, such that this experience could justify beliefs?

Background point: To have justification for believing P, two things must happen: (a) you must have an adequate source of justification, and (b) you must not have any defeaters for that justification.

Now, perhaps Siegel’s picture is something like this: both beliefs and experiences are potential sources of justification. E.g., in normal observation, sensory experiences justify beliefs about our immediate environment. Also, in inference, beliefs justify other beliefs. However, in inference, if the starting belief (the premise) is unjustified, that defeats its ability to provide justification for the conclusion belief. Similarly, you might think, in observation, if the starting experience is somehow irrational or otherwise defective, that defeats its ability to provide justification for the belief it would ordinarily support.

What’s wrong with that, in my view, is that only experiences (in particular, seemings) are an ultimate source of justification. Beliefs are not a justification source. So when you infer a belief, B, from another belief, A, A is not the source of justification. Rather, the source of justification is whatever justified A; the inference then merely transmits that justification from the premise to the conclusion. That is why, if there was no original justification for A, then there is none for B either. It’s not that there’s a defeater in this case; it’s that there’s no source.

But since experiences are an ultimate source of justification, an experience justifies beliefs that match its content, just by default, as long as there are no defeaters. And defeaters must be reasons that the subject has for doubting the belief in question. The unreliable source of an experience is thus not a defeater if the subject has no reason for thinking that it has an unreliable source.

4. Knowledge

Having said that, I of course agree that the subject in the fridge-gun case does not know that there’s a gun in the fridge, nor does she have warrant for that belief.

Background: “Warrant” in epistemology is defined to be the property such that, when combined with true belief, it gives you knowledge. I.e., by stipulation of the meaning of “warrant”, S knows that P iff (i) S believes that P, (ii) P is true, and (iii) P is warranted for S.

But to explain why the subject lacks knowledge of a refrigerator gun, we need not deny that the subject has justification for her belief. On almost any account of warrant, the subject lacks warrant even if she has justification. E.g.,

a. The subject doesn’t track the truth, since if there weren’t a gun there, the subject would still “see” a gun there.

b. The subject’s belief-forming mechanism (involving hallucination) isn’t reliable.

c. The subject’s belief isn’t formed using a properly-functioning, truth-directed faculty.

d. There is a genuine defeater*, viz., “this experience is a hallucination”.

[*Note that this is a different use of defeater from that above. In the defeasibility theory of knowledge, this just means that there is a true proposition such that, if the subject believed it, she’d no longer be justified in believing that there was a gun in the fridge.]

You can imagine a case of “veridical hallucination”: this would be a case where the subject hallucinates a gun (perhaps due to her obsessive fear), but coincidentally, there is actually a gun there. This subject still doesn’t know that there’s a gun. We can explain this using almost any popular account of warrant, while still holding that, per PC, the subject is justified in thinking there is a gun there.