Evolutionary explanation, mimics, and the doctrine of double effect

Quite a title for what will actually be a slim post! Here's the sitch—we're trying to answer the question why certain "imagings" of images or sentences hold our attention and are experienced as contentful, telling us things, when "this is an illusion", because "images are not missives sent from oneself to oneself".

Here is a possible answer. Suppose that one's mind includes many system,s attentional, computational (comprised of many task specific computational modules), memorial, affective, visual, auditory, speech-producing and consuming, and so on. One also has self-monitoring systems, built to register bodily states and needs … The information stored and processed continually by all these systems is quite enormous; but attention is limited, as are resources … In this competition for agential, person-level, resources, it gives an attitude, a need, or an emotion a decided edge if it can cause representations that have a powerful pull on the attention. The attention is commanded by representations with auditory and visual aspects, and held by information that tells a coherent story. So if a desire can cause an image or images that catch and hold the attention, relevant systems of intention-formation may more readily engage in ways favorable from the perspective of the desire. … No agency needs to design the image so that it speaks just so about the desire. It is enough that having such effects in creature [sic] like us is a way for desires to get their way with us.

All this is from p 64 of Krista Lawlor's "Knowing What One Wants" and it is, seh allows, "entirely speculative". But if this is the account she's going to go with, it seems to open up some territory that ought to be acknowledged and addressed. (Does one address territory?) The account is evolutionary in spirit, I take it, in that desires need not to be conceived of as having designs or being capable of fashioning actual messages to us, it's just that the desires that happen to cause representations that are as-if-of the things the desires themselves are concerned with fare better in actually being acted on than other desires do, so … something something evolutionary pressure, I suppose; there seems to be a missing element to the story, given that there isn't a notion of reproduction at play that would explain why desires would come more often to tend to cause imagings that catch and hold the attention and that correspond in some systematic way to what the desires are of. Given a soup of desires, one that tended to catch and hold the attention might meet with greater success than the others, in that intentions and actions would be formed and performed that led to the desire's being fulfilled. But that doesn't mean that it will transmit that same tendency to its successor desires, whatever that notion might mean. Whatever, though; let us wave our hands over that question in the manner prescribed by ritual and custom. Here's a more interesting issue.

We know from the evolution of creatures that some of them are both poisonous and brightly colored, the coloration serving to warn predators of the poison. (Handy at the individual level, one must admit.) And one knows also that some other creatures have (as it were, not by design, etc.) cottoned on to the utility of bright coloration and taken the shortcut of just being brightly colored, and omitted to actually be poisonous—so much work! The coloration gives you most of the benefit of being both brightly colored and poisonous at a fraction of the cost. Couldn't something similar happen with desires, on this sort of account? The benefit that accrues to a desire that p of catching and holding the attention and intention is the benefit of my actually bringing it about that p, fulfilling the desire; this, however, will be a benefit to any desire that would be fulfilled by my doing whatever it is I'll do to bring it about that p, i.e., not only by desires actually about p. Suppose I harbor the perverse desire to degrade the soles of my shoes; couldn't it catch and hold my attention by causing representations as if of the pleasantness of going for long walks? Going on the walks will also affect my shoes, and such a representation might actually be more reliable, from the perspective of this desire, than a representation of the pleasures of down-at-the-heels shoes, which would be apt primarily to puzzle me. The desire that q, which I find distasteful, might clothe itself in representations of p, which I find more reasonable. Of course this as-it-were strategy is not without its risks, since I might hit on a way of bringing it about that p that doesn't bring q about too, but it might on the whole be as it were worth it to the desire that q, depending. So there seems to be a question, on this account, that faces any would-be agent inferring from representations to the desires that supposedly caused them: were these representations of doing this caused by a desire with the selfsame content (is that what I desire)? Or were they caused by a more vulpine desire, which comes to me with ovine representations? How could one tell? One might think that Lawlor is actually acknowledging this possibility, albeit not drawing attention to it, when she writes that "typically causing the right kind of images gets the desire a better chance of being fulfilled." One will likely assume, on first reading, that "the right kind of images" means the kind of images that will be interpreted to have the same content the desire has. But if the point of these images is to catch and hold the attention and thereby make more likely the undertaking of actions that will get the desire fulfilled, the right kind could well be quite other.

And for Lawlor, I think, the question is, suppose that the representations as if of doing p were actually caused by a desire that q, but the agent doesn't know this, and self-ascribes a desire that p. The agent might even then (mightn't she? if not, why not?) experience the "characteristic changes" (p 62) in her imagings that show the question to be settled; she experiences a "sense of ease" (p 59), the attribution "sticks" (p 57). What should we conclude about her? That despite the actual cause of the representations, her inference that she desires that p is correct? I'd be inclined in that direction, but I'm unmoved by Lawlor's position and paper overall anyway. If we do say that, then how do we justify it? If, on the other hand, we prefer to say that she's just mistaken, and really doesn't desire that p, only that q, then what are we to make of her ease, the attribution's stickiness, etc.? (Maybe we're to deny that in this case it could really happen that way.) The (limited, piecemeal) discussion of the attribution's sticking, once made, and the "characteristic changes" and whatnot after the reflection and inferring, seems to me to be a weak point in the account, one that serves as a bactrio-nasal inlet. This question doesn't actually, I suppose, require the possibility of the kind of mimicry described above (though I do think that's worth addressing)—the agent could just misinterpret the representations the desire that q caused along p-lines and find herself at ease with that interpretation. (Right? Or, again, if not, why not?) Maybe "misinterpretation" isn't really the right term, since by hypothesis these representations (and even "representations" isn't the right term, for the same reason) aren't messages or any kind of interpretandum in the first place—but the imaging characteristically caused by a q-desire can't guarantee its being taken qishly.

A bit later Lawlor writes that "whether or not one's desire actually causes one's imagings and other internal promptings is a separate question. The point I note here is that our inference is structured in such a way as to suppose that desires cause imagings and other internal promptings" (p 65). But how could this be a separate question? It seems rather central to me; if we're engaging in some kind of causal inference from an internal prompting to the existence of a desire that caused it, then the soundness of the inference seems to depend quite a bit on the desire's actually having caused the prompting. If the inferential pattern merely supposes that there are such causal relations, then, on the one hand, something that I take it is necessary to separate Lawlor's position from one like Taylor's, namely that the desire in question definitely already exists and is therefore decidedly not constituted by the interpretive activity of the agent is free to go by the wayside. (Lawlor wants to distinguish the inferential account from a constitutive account, and also describes the cognitive element of the inferential account. But someone who says, as Lawlor summarizes Taylor, that "knowing what one wants owes to the fact that self-interpretation … puts in place the very facts known" (p 51), will of course allow that there's cognitive activity associated with knowing what one wants. Hermeneusis doesn't come for free. The claim to deny is that the thing known preëxists the knowledge thereof.) One is free, in particular, to claim that the causal supposition is a dispensable manner of speaking, that what one essentially has is, say, a sandy mental irritant around which reflection and interpretation build up a pearly desire, which was not the cause of the irritation but the product of the interpretation. ("What's this grain of sand then" is of course a good question. But I think on the whole there are lots of advantages to thinking this way, among them that it can capture the path-dependence of reflection and satisfaction with its results in a way that taking the desire as independently constituted can't obviously do.) And even if we don't go that far, then, on the other hand, this pattern of inference does not seem like a very reliable way to go about getting self-knowledge—if our desires don't cause the internal promptings, then the inference is just structured wrong.

Comments

[Guy Lionel Slingsby]

I really like this -- especially the parts about path-dependency and (also) the pearl metaphor. I assume that at some point she talks about Freud? Because that is what she's talking about, isn't it? Does she mention Fingarette?

From your quotation it sounds like the evolutionary idea is meme plus pheno-type. At some point her way of thinking, or your account of animal mimicry, would have to hit the rdf of costly signaling. That is: if all frogs are blue, blue no longer works to protect them. So being blue has to be something that most frogs can't afford to be: there is a pretty low limit on how many non-poisonous frogs the blues can save. The camel may get through the -- nose? -- of the needle, but without much headroom for a free rider. (That tends to mean that blue us a hard color to produce without poison, so that if you're blue you're likely to be poisonous. Certainly true of... pink flamingoes.)

Same I imagine with her view of the meme's relation to its purpose: the purpose has to be pretty helpful and the meme pretty consonant with it for the purposer's way of purposing to survive down the generations. And then there's the question of the interpretation that others put on her expressions of desire, etc.

Just maundering out-loud here. I always hate feeling Taylor is right about something. But that's just me.

The doctrine of double effect as put to mention here wouldn't be hers, right? But more yours or Taylor's?

[ben w]

Oh hey I was pleased by the pearl metaphor as well; thanks.

No Freud in the paper—she actually specifically disavows unconscious desires as part of her topic, which doesn't really make much sense to me—nor Fingarette, though the self-deception book sounds interesting.

I'm not sure that there needs to be any sort of costliness here; I mean, not all desires can cause pleasant representations of going for a walk, but why would they all? And if the perverse shoe-abrading desire does, and maybe a few other perverse desires do, why must that be costly? Because otherwise I'll go on too many walks? (I guess because: otherwise I'll become inured to the desire. But I dunno, the whole mechanism here seems underspecified. NB I'm the one to talk about evolutionary dynamics here; they're absent from the paper. I just want to understand why desires would (only) cause images/thoughts that align with their content.)

No double-effect in the paper (or Taylor, or that part of Taylor, anyway); just being gestured at here as possibly of interest.