The New Evil Demon Problem for Internalism

A standard NEDP case will take a twin who is internally identical to me but whose beliefs are unreliably formed because of a demon.  I share the intuition with internalists that this twin and I are justificationally identical.  So, it seems that reliable belief formation is not directly relevant to justification.  We can make variants of this case by picking any external property (e.g. sensitivity, safety) and imagining that our internally identical twin lacks that property because of a demon.  Intuitively, such believers will be justificationally identical.

I’ve argued recently (in Episteme, see here) that the new evil demon problem (NEDP) also applies to most internalist views, those which say that nonoccurrent (dispositional, background, unconscious) mental states are directly relevant to the justification of our beliefs.  Here is the case I formulate:

“Consider Augustine, who lived for seventy-six years.  During this time, he held a great many beliefs; some were justified and some were not.  Let us only consider the beliefs when they were occurrent.  Now imagine Augustine*, a creature who also lives for seventy-six years and who experiences all of the same accessed internal states that Augustine had throughout his lifetime.  The only difference is that the demon is manipulating Augustine* in such a way that he never has any unaccessed internal states.  Augustine* might think that he has dispositional beliefs at various times in his life, but in fact, he has none.  Whenever he thinks that he is recalling one of his memory beliefs, the demon is actually creating a newly formed occurrent belief.”

Augustine and Augustine*’s occurrent mental states (phenomenology, stream of consciousness, whatever) are identical throughout their lives.  It seems that the occurrent beliefs of Augustine and Augustine* have the same justificational status.  Or consider that you yourself have no way of telling whether you are in Augustine*’s situation.  Imagine your twin, You*, who is in such a situation, who has all of the same occurrent mental states as you.  It seems that You*’s occurrent beliefs are justificationally identical to yours.

Insofar as I share the internalist’s intuition in standard NEDP cases against externalism, I have them here against internalists who think that nonoccurrent mental states are relevant to justification.  Hence, I think that most internalists face the same sort of NEDP that externalists have been facing the last couple of decades.


The New Evil Demon Problem for Internalism — 9 Comments

  1. Hi, Andrew,
    I’m interested in your claim that there is an NEDP for internalism. But the post is giving me a hard time. (I haven’t yet had the time to read the paper.) Here’s what’s troubling me.

    1. Are the terms “dispositional belief” and “unconscious belief” synonyms? I don’t have a problem with the notion of a dispositional belief (or, maybe more properly, the notion of one’s having a disposition to form a given belief). But I can’t say I know that there are unconscious beliefs. Note, if there are repressed beliefs, they surely are unconscious (if I’m not much mistaken about these things), but the talk of repressed beliefs suggests that they are occurrent (in that they are supposed to affect behavior occurrently in various ways). I know some people have treated “unconscious” and “dispositional” as synonyms in the literature, but I’m asking if it’s not a mistake.

    2. I’m also surprised to hear that a memory belief is regarded as dispositional. I think most would say that it is occurrent. (Maybe Robert Audi would disagree. I’m not sure.)

    3. You claim that the NEDP “also applies to most internalist views, those which say that nonoccurrent (dispositional, background, unconscious) mental states are directly relevant to the justification of our [occurrent] beliefs”. The problem for me here is twofold: First, on my understanding of the literature, most people think of background beliefs as occurrent. (Again, maybe Audi would disagree.) Is that a mistake on my part? Second, what are your sources for the view that non-occurrent mental states “are directly relevant to the justification of our [occurrent] beliefs”? You claim that most internalists subscribe to the view. But I can’t remember claims to that effect by notable internalists. You may well be right. But I’d like to see that claim in your sources. In any case, for an internalist to make a prima facie compelling case according to which non-occurrent mental states “are directly relevant to the justification of our [occurrent] beliefs”, we need an explanation of how non-occurrent states may affect the justification of occurrent doxastic states. Can you explain how that is supposed to work? (Maybe my problem here has to do with the fact that you regard background beliefs as dispositional, and I don’t see that in the literature. So, maybe the problem for me here is not twofold after all. If background beliefs are dispositional, then everybody thinks dispositional beliefs affect one’s justification/warrant for an occurrent belief.)

    4. The issue of whether the phenomenological twin (Augustine*) is internalistically justified in his false belief about what’s non-occurrent in his mental life may depend on the kind of justification that he has for it. If his justification comes from introspection, those who think that introspective justification is infallible will demur. If I were a Fumerton, or a BonJour, or a McGrew, I’d think that the twin’s introspective justification for a belief about his mental states is infallible. So, for some influential internalists, the twin would not have that false belief and be justified in having it. Is there something I’m missing here?

    I hope the discussion is not useless to you and to CD readers.

  2. Claudio,
    Thanks for the response. I think dialogue can proceed better if we deal with one issue at a time. You said, ” But I can’t say I know that there are unconscious beliefs.”

    So, I’ll just start with one question. Do you think you still have beliefs when you go into dreamless sleep?

  3. Good, Andrew! I misspoke. So, let’s say that I know that there are unconscious beliefs. What I don’t know is that every unconscious belief is a dispositional one. So, I can’t see that “unconscious belief” and “dispositional belief” are synonyms.

    • Okay, good we’re connecting. There might be a substantive issue here, or I could just define things so that they’re trivially true. =) In the paper, I was following Thomas Senor:

      “Thomas Senor’s (1993: 461–462) definitions: “A belief is occurrent at t iff it is conscious at t… S’s belief that P is dispositional at t iff at t, S believes that P and P is not occurrent.”” (footnote 9 of my paper)

      I analytically define ‘occurrent belief’ and ‘dispositional belief’ in terms of what is conscious or not conscious. I furthermore think that this is how at least a lot of epistemologists today use those words.

      And I think that both terms are misleading and apt to produce confusion (as, perhaps, in our current discussion!), especially ‘dispositional belief’, since it is so easy to mistake ‘dispositional belief’ with ‘disposition to believe’. If I had it my way, I would just talk about ‘occurrent beliefs’ and ‘nonoccurrent beliefs’, or perhaps ‘conscious beliefs’ and ‘nonconscious beliefs’.

      So, really what I am interested in is what is conscious versus unconscious. A mental state’s being occurrent/accessed/experienced are understood in terms of what is conscious, and a mental state’s being dispositional/nonoccurrent/unaccessed/stored-in-the-mind/background are understood in terms of what is unconscious. Maybe it is best not to use all those words. But for the substantive issues of my paper, what I’m interested in is what is conscious versus what is not conscious.

      Are we together?

  4. Are we together? Andrew, I wish I could say that we are, but I can’t — not yet anyway. I’ll explain why.

    It would be exciting to conclude, as you do, that there is an NEDP for internalism. But your case for that claim depends on a premise that I suspect is either false or unintelligible: the one according to which “most internalist views […] say that nonoccurrent (dispositional, background, unconscious) mental states are directly relevant to the justification of our [occurrent] beliefs”. If this is false or unintelligible, you don’t have a target for the Augustine* case. And the truth of the claim crucially depends on your being right about internalists’ thinking that background/memory beliefs are non-occurrent. If you are wrong about the importance of this notion to internalism, your case collapses. Is this a fair assessment of the situation? If it is, we must look at what’s supposed to count as an “occurrent” belief in the literature. And, if “occurrent” is not synonymous with “conscious”, that is relevant to whether your understanding of internalism is defensible. Fair? So, we go to the literature — because I just can’t see my understanding of the term “occurrrent belief” represented in your case. I have always thought that occurrent beliefs need not be ones that you are currently attending to, so to speak, but include memory/background beliefs, all those that have already been formed and kept in memory, as well.

    It turns out that the literature on the dispositional/occurrent, conscious/unconscious distinctions is a mess. There may be more than one concept of an occurrent belief in the literature. And, if that’s the case, you must be targeting the one that matters to internalism. Agreed?

    Consider our source for the occurrent/dispositional distinction, Fumerton’s 1976 JP paper on “Inferential justification and empiricism”. (If I’m not mistaken, this is where the distinction comes from.) Fumerton’s distinction is given as follows:

    “Each of you believes or accepts many propositions that you are not now entertaining, about which you are not now thinking. Five minutes ago, for example, you all believed that 7+5=12 even though none of you were thinking about that proposition. If an occurrent belief in a proposition can be thought of as a psychological state attending the consideration of that proposition, then the most straightforward analysis of a dispositional belief would employ a subjunctive conditional: […] S dispositionally believes P =Df S would believe (occurrently) P if he were to consider it. Now there seems to be no a priori reason why a person may not have an infinite number of dispositional beliefs, understood this way. Five minutes ago, for example, you presumably believed dispositionally that the number 2 is greater that the number 1, that the number 3 is greater than the number 1 […], etc. ad infinitum

    Notice that, according to this distinction, dispositional beliefs include both the ones that you have already formed (memory/background beliefs) and the ones that will/would become occurrent if you will/would only consider the proposition. (You have infinitely many dispositional beliefs.) And, if you can attend to only one belief at a time, then, considered synchronically, the set of your occurrent beliefs contains only one member. In many contexts, it will be misleading to speak of your “occurrent beliefS“, since, at any given time, you only have one such belief.

    Fumerton’s distinction drew fire from Gilbert Harman. In that same 1976 issue of JP, Harman writes:

    “[Fumerton] mixes up two different distinctions: that between beliefs of which one is consciously aware, and that between what one explicitly believes and what is only implicit in what one explicitly believes.”

    There, he does not elaborate on the explicit/implicit distinction. He does so in his book Change in view. I should quote at some length from the relevant passage (pp. 13-4).

    “I assume one believes something explicitly if one’s belief in that thing involves an explicit mental representation whose content is the content of that belief. On the other hand something is believed only implicitly if it is not explicitly believed but, for example, is easily inferable from one’s explicit beliefs. Given that one explicitly believes the earth has exactly one sun, one can easily infer that the earth does not have two suns, that the earth does not have three suns, and so on. So all these propositions are things one believes implicitly. […] [A] belief can be explicitly represented in one’s mind, written down in Mentalese as it were, without necessarily being available to consciousness. For example, one might explicitly believe that one’s mother does not love one, even though this belief may not be consciously retrievable without extensive psychoanalysis. So the distinction between implicit and explicit beliefs is not the same as that between unconscious beliefs and those available to consciousness. […] Turning now to the distinction between occurrent and dispositional beliefs, we can say a belief is occurrent if it is either currently before one’s consciousness or in some other way currently operative in guiding what one is thinking or doing. A belief is merely dispositional if it is only potentially occurrent in this sense. Any merely implicit belief is merely dispositional, but explicit beliefs are not always occurrent, since only some explicit beliefs are currently operative at any given time. So the distinction between implicit and explicit beliefs is not the same as that between occurrent and dispositional beliefs.”

    (Note, by the way, that he avoids my conclusion that it is at best misleading to speak of one’s having more than one occurrent belief at any given time with the cagey wording “or in some other way currently operative in guiding what one is thinking or doing”. Did he say “guiding what one is thinking”?! Without being that very thought being thought itself?!)

    So, we should now be asking ourselves: Which is the distinction that matters to your case against internalism: occurrent/dispositional, conscious/unconscious, or implicit/explicit? Before you answer, I’ll give you another option.

    Here’s Paul Moser in The Cambridge dictionary of philosophy:

    “Beliefs are either occurrent or non-occurrent. Occurrent belief, unlike non-occurrent belief, requires current assent to the proposition believed. If the assent is self-conscious, the belief is an explicit occurrent belief; if the assent is not self-conscious, the belief is an implicit occurrent belief. Non-occurrent beliefs permit that we do not cease to believe that 2+2=4, for instance, merely because we now happen to be thinking of something else or nothing at all.”

    Suppose I’m now thinking that this subject is giving me a headache. Is it a self-conscious belief? Or just a conscious one? Go figure! More to the point: Notice that, for Moser, an occurrent belief can be implicit, whereas, for Harman, that’s impossible: every occurrent belief must be explicit, and an explicit belief may be unconscious.

    This is far from being a complete survey of the literature, of course, but the mess is plain to see, I suppose.

    In any case, I must go now, but I’d like to leave you with a suggestion. I suspect that much of epistemology tacitly operates with Harman’s distinction between implicit and explicit beliefs — even when we call them “occurrent” and “dispositional”, as I sadly conclude that I have been doing myself. That’s because we don’t need to confuse memory/background beliefs with the really interesting among dispositional beliefs in Fumerton’s sense, Harman’s implicit beliefs, the ones that haven’t yet been formed but will if the proposition is considered. When an internalist considers what is in one’s mental landscape, it’s not just that lonely occurrent belief that matters. She’s thinking about every explicit belief one has. But, if I’m right, can you, too, be right?

  5. Hi Claudio,
    Thanks for the continued dialogue and for taking the time to write out all those interesting quotes. I liked Harman’s especially.

    I actually don’t think things are that complicated w/r/t my argument.

    Let’s put aside my remark about what will affect MOST internalists. Suppose I just said that my nedp affected those internalists who deny that only currently conscious internal states are directly relevant to justification (those who think that some internal states that are not currently conscious internal states are also directly relevant to justification). Some internalists would then remain protected from my NEDP, those who say that justification supervenes only on currently conscious internal states (call these fine folk ‘strong internalists’, as I do in my paper and following Alvin Goldman). I can say that my NEDP only applies to those who are not strong internalists.

    I could tell my story so that the difference between Augustine and Augustine* is that the demon deletes all of Augustine*’s internal states that are not conscious. Intuitively, the justificatory status of their conscious beliefs are identical. Indeed, you could have a twin who is exactly like you w/r/t conscious internal states but does not share the internal states that are not conscious. The two of you would be phenomenologically identical. And it seems that you would be equally justified in your conscious beliefs.

    There’s still that question regarding how many people actually deny strong internalism; I’d say more (I address this in the paper), but I like to try to stick to one thing at a time, and I’m sure you’ll have lots to say about this bit I’ve written.

  6. Andrew, just a quick erratum for now. Please, ignore this remark in my previous reply: “More to the point: Notice that, for Moser, an occurrent belief can be implicit, whereas, for Harman, that’s impossible: every occurrent belief must be explicit, and an explicit belief may be unconscious.”

    There are so many warts there! But I’m afraid posting at breakneck speed is the only way to post at all. So, we get back and try to polish these things, right?

    I’ll try to give you a proper reply soon. This is just the erratum. Thanks for the patience.

  7. 🙂 That’s a noble cause right there, and I’m happy to oblige. Let us not forget: hasty grading is bad grading! If turns out to be fair, it’s just by accident; it’s not rooted in knowledge!

Leave a Reply

Your email address will not be published. Required fields are marked *