The Nearly Global NEDP for Internalism

I have argued that the NEDP is a problem for nearly all versions of internalism.  Consider Augustine*, who in his seventy-six years of life, experiences all of Augustine’s conscious states but has no unconscious internal states.  With respect to their conscious life, there is no discernable difference.  I appealed to the intuition that the justificational status of Augustine and Augustine*’s conscious beliefs is the same.  It seems plausible to conclude that Augustine*’s unconscious, internal states are justificationally irrelevant to his conscious beliefs.  This is one version of my global NEDP for internalism.  (For earlier Certain Doubts discussion, see here.)

In his forthcoming book defending evidentialism, and in a forthcoming article in Acta Analytica (available on his website under “Research” and “Publications”, see here), Kevin McCain argues that my NEDP fails to show there is a problem for internalism.  The force of his argument is in this quote:

First, on several plausible accounts of the nature of belief it is necessary to have certain dispositional states with respect to p in order to even occurrently believe that p.  Since by stipulation Augustine* does not have any of these unaccessed internal mental states, he does not have beliefs according to these views of the nature of belief.  Second, and perhaps more fundamentally, there is a further reason to think that Augustine* lacks beliefs.  In order to have a belief that p, occurrent or otherwise, Augustine* must possess the concepts required to grasp the proposition p.  Concept possession requires unaccessed mental states.  It does not seem that one can possess the concept TREE, for example, without having some dispositional state that either represents this content or constitutes a grasping of an associated Fregean sense or at least grounds some sort of recognitional ability.  However, Augustine* lacks such dispositional states.  Without concepts Augustine* cannot grasp propositions and if he cannot grasp propositions, he cannot believe them.  Thus, it seems that there are good grounds for thinking that Augustine* cannot have beliefs at all.  But, if Augustine* cannot have beliefs, then it is impossible for him to have all of the same accessed mental states as Augustine.  Therefore, AUGUSTINE fails to depict a metaphysically possible scenario. (from pp. 3-4 of the Acta Analytica article, linked above)

(I should note that McCain means by ‘occurrent’ what I used to mean by it: conscious.) These are certainly good points put forth by McCain, points that I need to reckon with.  For my reply, I will suppose that McCain is correct.  I will argue that there is still a new evil demon problem – what I call the nearly global new evil demon problem – that applies to nearly all versions of internalism.  Some background will be necessary to present the problem.  As I understand him, McCain thinks that for any person’s conscious state at a moment of time, that conscious state entails (in the broadly logical or metaphysical sense) the presence of certain unconscious internal states. (State S entails state T if and only if, in any possible world in which S is instantiated, T is also instantiated.) To facilitate discussion, let a person’s entailed unconscious internal states at a time refer to any unconscious internal state that is entailed by the person’s conscious internal states at that time.  And let an unentailed unconscious internal state refer to any unconscious internal state that is not an entailed unconscious internal state.

I am now in a position to present my nearly global NEDP for most versions of internalism.  Suppose Augustine* is in fact impossible.  Consider, then, Augustine**, who is exactly like the original Augustine, but an evil demon has deleted all his unentailed unconscious internal states.  Unlike Augustine*, Augustine** is more clearly possible; McCain’s argument that Augustine* is impossible does not show that Augustine** is impossible.  Augustine** will have all of the same conscious internal states that Augustine has.  Now, it seems that Augustine**’s conscious beliefs will have the same justificational status as Augustine’s.  (Note this: you might be in the position of Augustine**!) Following the standard NEDP format, it seems to follow that the unentailed unconscious internal states, are justificationally irrelevant.  And what is true about Augustine is true about you, me, and anybody else.  It seems that, for any person S, we can formulate new evil demon scenarios according to which S has a twin, S**, who shares only the entailed unconscious mental states.  It will seem that the justificational status of S and S**’s conscious beliefs are the same and that any unentailed unconscious mental states are justificationally irrelevant.

This nearly global NEDP will affect virtually all internalists.  The entailed unconscious internal states are rather slim, and the sorts of unconscious internal states that moderate internalists are inclined to think are justificationally relevant are the unentailed ones.  So, I conclude that even if McCain is right that my original case fails, this refined one does not.


The Nearly Global NEDP for Internalism — 8 Comments

  1. The example seems to require further refinement. State S may not entail state T1, nor entail state T2, nor T3, etc. But S may entail their disjunction. There may be no world in which a person instatiates S without also instantiating either T1, T2, T3, etc. Each of these T-states is “unentailed” by state S, but at least one of these T-states is instantiated whenever S is.

    So Augustine** may not have all the same conscious mental states that Augustine has. If a demon deletes all of Augustine**’s unentailed mental states, and the disjunction of some of these states is necessary for some conscious state that Augustine has, then a conscious mental state has thereby also be deleted. Augustine and Augustine** would not be twins with respect to their conscious states.

  2. Matt, thanks for the helpful feedback!

    If I specified “state” to include disjunctive states, I think that would block your concern. So, if S’s conscious state C entails T1vT2vT3, then I would understand T1vT2vT3 to be an entailed internal state. Let me know if I’m misreading you or we’re missing each other.

  3. Hey Andrew,

    Great to see you discussing this here! I’ll just mention a couple small points for now. 1) I think that the intuitions concerning Augustine and Augustine** may not clearly favor the point that you want. Once you grant that a large number of the unconscious mental states are kept it becomes less clear (to me anyway) whether or not the internalist should think that Augustine** is/isn’t justified. I guess I simply need to know more about the case. 2) The way things are presented here makes it seem like my entire response hinges on whether Augustine* is possible or not. I think it is worth mentioning that in the paper I claim that even if I am incorrect and the sort of global case that you describe with Augustine* is possible, the points I raise about your other, localized, case apply here as well. So, the moderate internalist has other moves available to her. 🙂

  4. Kevin,
    Yep, good to hear you chiming in. I’m looking forward to your book coming out.

    On your second point, yes, there is my local version of the argument and your reply to that (which is ignored in my post). I am working on my reply to that, and that’s probably for another time.

    On your first point, it is difficult to say precisely which internal states will belong in which of the two categories of entailed unconscious internal state and unentailed unconscious internal state, since that will depend on which theory of belief or concept possession is true. Still, regardless of which theory is true, we can look at individual cases and be confident that some internal states are clearly unentailed unconscious internal states. First, unconscious beliefs that are normally thought to be positive reasons for our conscious beliefs will often be unentailed, unconscious internal states. Suppose I believe that p and if p then q, and on the basis of those beliefs, I believe q. In the moment afterward, while I reflect on and admire my new, conscious belief that q, my beliefs that p and if p then q become unconscious. Though my beliefs that p and if p then q played a causal role in bringing about my belief that q, they will, at this later moment, still be unentailed unconscious internal states because it is possible that I believe q and be in my present conscious state without having had those unconscious beliefs. It seems possible, for example, that I could have arrived at the belief that q by some other process: wishful thinking, indoctrination, testimony, or even an evil demon. At the end of any of these processes, I could both consciously believe that q and also be in the same conscious state as I was in in my initial scenario: I could have the same conscious feeling of confidence that q, the feeling of endorsement that q, the same sensory experiences, the feeling of admiring my belief that q, etc.

    The point is that the very sorts of unconscious internal states that many think are suitable candidates for positive reasons are unentailed states. And that’s enough for there to be a problem for nearly any moderate internalist view. (Notice that what I say here about potential positive reasons could also be said about potential defeaters. I spell that out in the paper, but not here, for sake of space.)

  5. Hey Andrew,

    Regarding your reply to Kevin’s first point: it’s plausible that one basis of your belief that q is (something like) your felt confidence that q. A moderate internalist could correctly count your belief that q as justified without appealing to unconscious states.

  6. (In case there’s any confusion for those who haven’t read my paper, I define: Moderate Internalism – internalism is true, and some unconscious internal properties are directly relevant to justificational properties. Matt and Kevin know my terminology, since I think they’ve both read my paper.)

    Hey Matt,
    You are correct. But the point was that any unconscious state that is a prospective unconscious positive reason will also be an unentailed unconscious state. My modus ponens example was just an example of the sort of unconscious internal state that many moderate internalists are likely to think are justificationally relevant. But I believe it’s an example that generalizes.

  7. Hi Matt (and anyone else, feel free to chime in),

    It looks like one way to resist Andrew’s point is to find something to serve as a justifying ground that the demon can’t take away from you. It seems to matter for the dialectic whether such a ground would be adequate in the absence of defeaters. (Let’s bracket the issue of defeaters and whether they also need to be demon-proofed.)

    I’m curious as to whether such felt confidence is a plausible candidate for what would justify practical responses and affective responses. We sometimes do talk about justified anger (and also potentially related notions like fitting anger or appropriate anger) and I’m curious to know what people have to say in response to four questions:

    Q1: Are the conditions that determine whether a particular affective response is justified all demon-proof?
    Q2: Is the felt-confidence that accompanies an affective response something that could constitute an adequate ground for that affective response?
    Q3: Q1: Are the conditions that determine whether a particular practical response is justified all demon-proof?
    Q4: Is the felt-confidence that accompanies a practical response something that could constitute an adequate ground for that practical response?

  8. Clayton,
    Interesting questions.

    Regarding your Q1, I imagine the following cases:

    Case 1: Suppose I have many justified unconscious beliefs about Fred’s character. I unconsciously believe that he is the sort of guy that would steal, cheat, lie, etc. I see him grab a candy bar at a store, and it looks like he is holding it in a way that looks like he is concealing it. I come to believe that he is going to steal it, and I get angry.

    The belief seems justified, and so does the anger.

    Case 2: I am consciously like I am in Case 1, but I do not have any of the unconscious beliefs. I come to believe that he is going to steal it, and I get angry.

    The belief does not seem justified, and neither does the anger. (Perhaps I ought to have the attitude of suspecting that he will steal it, but belief seems unjustified.)

    This points toward unconscious beliefs mattering

    Case 3: I am like in Case 1, but for a 3-second interval, during the period at which I feel angry, a demon deletes those unconscious beliefs. He then recreates the unconscious beliefs at the end of the 3-second.

    (These cases mirror my Fred, Sally, and Melissa case in my Episteme paper.)

    As I reflect on these cases, I’m not sure what I think about them. I’ll have to think more about them.

Leave a Reply

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