Self to Self

In this post, I reply to my previous post on a new motivation for positing desire-like imagination. The right response to the argument sketched there, I now think, is a combination of `Who cares?’ and `What are you talking about?’. But there’s a functionalist explanation behind the indifferent shrug and the incredulous stare.

Suppose we are all good functionalists, as we should be. Then a mental state (type) is defined by the inputs and outputs that its tokens typically have. For example, a typical belief connects to desires, affective response system, behavior-generating system, and such. To answer the question ‘what is a belief’, there is nothing over and above saying what inputs and outputs that its tokens typically have. The nature of a mental state is just its functions.

In that sense, contrary to what I claim in the previous post, Currie and Doggett and Egan have indeed explained what desire-like imagination is, in virtue of the connections that they posit. In particular, Doggett and Egan say that desire-like imagination is connected to (belief-like) imagination, affective response system, and behavior-generating system. On the other hand, Currie says that desire-like imagination is connected to imagination and affective response system, but not behavior-generating system (see Currie (2002)). The connections that they posit are their respective answers to the question `what is the nature of desire-like imagination.’ Hence, talking about desire-like imagination’s direction of fit may not add to the characterizations given, and is at any rate unnecessary. Thus the indifferent shrug.

More importantly, the functionalist framework shows that a mental state (type) does not exist in a vacuum; it exists in connection with other mental states and cognitive systems. To explain a phenomenon, such as affective response to fiction or immersion in pretense, these philosophers need to posit both desire-like imagination and the appropriate connections. Hence, even if the direction-of-fit metaphor is right and the analogies go through, as argued in my previous post, all we have done is posit a mental state. Without the appropriate connections, it exists only in a vacuum and cannot do any explanatory work. Furthermore, it is far from clear how the direction-of-fit metaphor and the analogies can help us with positing desire-like imagination’s connections in the functionalist framework of mind. So there is the feeling, as Sara commented, of “what phenomenon are you talking about” because positing a mental state in a vacuum does not help explaining the phenomena in which we are ultimately interested. Thus the incredulous stare.


6 Responses to Self to Self

  1. dtlocke says:

    Hmmm… OK, but let me play devil’s advocate for a bit and see if I can’t find some “characterizing of the phenomenon” work for the direction-of-fit stuff to do.

    Set imagination/pretense/etc aside for the moment. Let’s back way up. When people usually talk about direction-of-fit stuff, they’re usually talking about normative stuff, right? I would think the idea is roughly like this. If you have a belief about subject matter X, you ought to try to make your belief about X fit the world. However, if you have a desire about subject matter X, you ought to try to make the world fit your desire. So the direction-of-fit distinction is really this

    1. The norms for belief tell you what to believe given the facts and

    2. The norms for desire tell you what facts to “make” given your desires.

    In other words, you should make your beliefs responsive to the facts, whereas you should make the facts responsive to your beliefs.

    (Don’t for a moment think that I don’t know that the norms for belief/actions are much more complicated than the above suggest. It’s just that the complications are orthogonal to the direction-of-fit distinction I’m trying to get clear about.)

    So the question is this: can we use this normative distinction to “characterize” the distinction between belief and desire? Sure, we just ramsify. We say: belief is that state which is such that the norms for that state tell you to make it fit the facts, whereas desire is that state such as the norms for that state tell you to make the facts fit it. Is this a good definition/analysis of the belief/desire distinction? The functionalist will say “no”, since he beliefs the true definition/analysis is gotten by ramsifying the theory about the difference in function between beliefs and desires. But suppose the functionalist is wrong about this. Suppose (contra functionalism) that the right analysis/definition of the belief/desire distinction is gotten by ramsifying (a much more sophisticated version of) the above bit of normative theory. On this view, what it IS for a state to be belief is just to have such and such norms apply to it, and what it IS for a state to be desire is just to have such and such norms apply to it. I think this is what some folks have in mind when the say things like “belief is constitutively and entirely normative” or “desire is constitutively and entirely normative”. In that case, the “characterization” I just gave IS a good analysis/definition of the distinction. (I know you said that you were assuming functionalism, but hang with me for a bit; even if the functionalist is right, I still think we can get a *characterization* out of the above.)

    If the above is right (and so functionalism is wrong), then perhaps we can say something analogous in the case of i-belief and i-desire. We start with the following bit of theory and then ramsify:

    1. The norms for i-belief tell you what to i-believe given the facts-in-the-fiction and

    2. The norms for i-desire tell you what facts-in-the-fiction to make given your i-desires.

    So we’ll say that while i-belief is that state such that the norms for that state tell you to make it fit the facts-in-the-fiction, i-desire is that state such that the norms for that state tell you to make the facts-in-the-fiction fit it.

    The trouble, which I’m sure you’ve picked up on, Sam, seems to be this: we need to be very careful that facts-in-the-fiction is distinguished from the facts (here in the real world) about what the facts-in-the-fiction are. Moreover, it’s not at all clear that there IS a state such that the norms for that state tell you to *make* the facts-in-the-fiction fit it. Remember, we’re not talking about “making it a fact that it is a fact-in-the-fiction that…” we’re talking about “making it a fact-in-the-fiction that…” However, there might be. Consider Walton’s favorite example concerning the game with stumps that one pretends are bears. Suppose we’re playing this game. It *does* seem like there is a norm that tells me (given that I am in a certain state–namely, one of i-desiring to avoid bears) to make it a fact-in-the-fiction that I run away from the bear.

    So if functionalism is false and the above thesis about how to analyze belief/desire is correct, then maybe we can use the direction of fit stuff to give an analysis of i-belief and i-desire. But what if functionalism is, as you say, correct? In that case, can’t we still give a *characterization* of belief and desire in terms of the direction of fit stuff just mentioned? That is to say, can’t we say something like: look, I know it’s not analytically true that belief is that state such that the norms… and desire is that state such that the norms… and it might not even be necessarily true. But I’m not trying to analyze/define these states. I’m just trying to say something that is true of them and only them (in the actual world) in an attempt to “single out” these states from other states. Thus, if you’re new to english and don’t know what I mean by “belief” and “desire”, I might be able to tell you, provided that you are capable of recognizing the states I’m singling out—not defining; singling-out is more like pointing!

    Similarly, even if functionalism about i-belief and i-desire is correct, perhaps we can use the above characterization to *single out* the states we’re interested in, so as to explain what we mean to someone who is new to Eganease in order to get them to understand (without defining or analyzing!) what we mean by i-belief and i-desire.

    Ummm… yeah, something like that. What i just said is too long to proof for typos, so I’m just gonna hope I didn’t leave out any “not”s.

  2. dtlocke says:

    Haha, that comment looks even longer now that it’s been posted! You’ve been Howarded!

  3. Shen-yi Liao says:

    Hey Dustin. Sorry I’ve been traveling.

    I think the move you made is exactly right. The directions of fit stuff may have a normative role to play distinctive from the descriptive functionalist theory of mind. At any rate, the normative stuff plays a crucial role in accounting for our ATTRIBUTIONS of mental states. Since Velleman, for example, is more concerned with giving an account of attributions of mental states, it’s probably true that he would want to give a normative story about desire-like imagination.

    In addition to some troubles you raised, I think there is an additional, more general worry with this approach, with respect to desire-like imagination. Normally, when we give a normative account, we are only interested in giving such an account for notions that we already employ. For example, we want to tell a story about belief attribution because belief is a notion that exists in folk psychology. The parallel holds for desire and imagination. But the parallel does not hold for desire-like imagination–we have no such notion in our folk psychology.

    I propose the following weakened conclusion. If we have antecedent reasons for positing desire-like imagination, then the normative characterizations may indeed be informative in the ways you have suggested. However, the direction-of-fit analogies do not provide that antecedent reason. On the other hand, if we have no reason to posit desire-like imagination, then we do not need to give a normative account in the first place. Hence, while the direction-of-fit analogies might be suggestive of what desire-like imagination is supposed to be, it cannot be part of an ARGUMENT FOR desire-like imagination.

  4. Dustin says:

    “But the parallel does not hold for desire-like imagination–we have no such notion in our folk psychology.”

    I’m not sure I agree. And the reason is that I’m not sure I agree with

    “while the direction-of-fit analogies might be suggestive of what desire-like imagination is supposed to be, it cannot be part of an ARGUMENT FOR desire-like imagination.”

    Now I don’t think it can be *all* of such an argument, but I do think it can be *part of* such an argument. Let’s just agree for now that there is such a thing as i-belief. My thought was that (A) there seem to be norms that govern our action in pretense in much the same way that there are norms that govern our normal actions but (B) those i-norms seem to prescribe certain actions in the pretense conditional on having certain i-beliefs + being in some desire-like state. Now, if we can establish that that desire-like state *cannot* be (normal) desire, then we have a reason to posit another desire-like state. We are then justified in calling that state “i-desire” because of its distinctive role with respect to to norms governing behavior in pretense.

    My thought is that folk normativity–that is, folk theory about what the norms are–not folk pyschology, seems to posit the existence of a state which plays a certain role. If we can *then* argue that this role cannot be played by ordinary desire, we have here an argument (via folk normativity) for i-desire.

  5. Shen-yi Liao says:

    I see. So the argument would be more like what Velleman and Doggett and Egan give about behavior motivation during immersive pretense, but adding in folk normativity as a crucial element. That strikes me as more plausible than the argument I was running in the previous post as well as the actual arguments V and D&E give. So it’s at least a promising move.

    The hard parts would be, of course, to (1) clarify the notion of folk normativity, and (2) demonstrate that the desire-like state in behavior motivation during immersive pretense cannot be desire, via (1). I am skeptical of the prospect of (2), in particular, but this new line of argument might be worth trying.

  6. Dustin says:

    Cool. So when should we start working on our joint paper?

%d bloggers like this: