Thursday, June 17, 2010

Stoljar on intentionality

Daniel Stoljar’s new book Physicalism is a very useful overview of its subject. However, its brief treatment of intentionality (and of some other topics too) evinces a deep misunderstanding of dualism, a misunderstanding that seems to be very common in contemporary philosophy of mind.

Stoljar begins his discussion with the following characterization of intentionality:

The intentionality of a mental state is its aboutness. When I think of Vienna or believe that the computer is on the desk or fear that the planet will get hotter, I instantiate mental states which are in a hard to define sense about Vienna, or the computer on the desk or planet Earth. The idea is that mental states (and speech acts) have a property rather like signs, sentences, and gestures; that is, they are about or represent things other than themselves. (p. 200)

So far so good; at least, that is an accurate characterization of what modern philosophers, whether physicalists, dualists, or idealists, tend to mean by “intentionality.” (Whether they are right to think of it this way is a question I will return to later.) Stoljar then suggests that the reason intentionality is philosophically problematic is that it is supposed to involve a relation that might hold between a thinker and something else, and yet lacks three key features one would expect such a relation to have. First, if I bear a relation to something else, one would expect that that something else exists; and yet I can think about something that does not exist (e.g. Valhalla). Second, if I bear a relation to something else, one would expect that there is some particular thing I bear it to; but I can think about a man without there being some man in particular I am thinking of. Third, if I bear a relation to some thing A and A = B, then one would expect that I thereby bear that relation to B; but if I am thinking about Vienna, then even though Vienna is the birthplace of Schubert, it doesn’t follow that I am thinking about the birthplace of Schubert, about whom I may know nothing. (To use the technical jargon, ascriptions of intentional mental states are often non-extensional or intensional, insofar as we cannot always substitute co-referring expressions salve veritate; that is to say, intentionality-with-a-t is often – though, it is important to note, not always – associated with intensionality-with-an-s.)

Now, Stoljar acknowledges that these features of intentionality are philosophically puzzling. But he claims that they pose no special difficulty for physicalism. In particular, they give us no reason to favor dualism over physicalism, for they are as problematic on the former view as on the latter. Says Stoljar: “[S]uppose classical dualism is true and I am some sort of complex of an ordinary physical object and soul; it is still impossible for me to stand in a relation to things that don’t exist! In sum, the paradoxes of intentionality will remain whether physicalism is true or not, hence they do not concern physicalism.” (p. 201)

There are two problems with this. First, it does not get to the heart of the problem of intentionality. Second, it rests on a misunderstanding of dualism. Let’s take them in order.

Consider the following dialogue:

Policeman: Ma’am, some bad news, I’m afraid. Your son just robbed a liquor store. Caught him red-handed with the cash tucked in the glove compartment, along with a few bottles of Tanqueray, vermouth, and tipsy olives that he tossed in the back seat.

Mom: Oh dear. I suppose he’s in trouble for being under 21. Or was he speeding in the getaway car?

Policeman: Well, there is that, I guess. But here’s the main thing: He robbed a liquor store.

Stoljar reminds me a little bit of Mom. Yes, the “paradoxes of intentionality” that he calls attention to are important. But it is intentionality itself, and not the “paradoxes,” that is of the greatest interest. Even if the objects I thought about always existed, or were always particular, or never generated non-extensional contexts – that is to say, even if intentionality exhibited none of the “paradoxical” features in question – the “aboutness” of my thoughts would remain. And it is that “aboutness” that the dualist takes to pose the chief difficulty for physicalism.

There are at least two ways to see how – a commonsense way and a more technical way. The commonsense way is this. Consider the word “cup” as you might write it in ink. Now consider a set of splotches that forms after your ink bottle leaks overnight, among which there are three right next to each other that by chance look vaguely like this: CUP. The set of splotches looks like the word, but it isn’t. The word has meaning, the splotches do not. But this has nothing to do with the physical properties of either. The ink is the same in both cases, as are the shapes. We can even imagine a case where your penmanship is bad enough and/or the splotches are distinct enough that their appearance is indistinguishable from the word “cup” that you’ve written. In general, it is not the intrinsic physical properties of letters, words, and sentences, whether written or spoken, that give them the meaning they have. Rather, their meaning derives from the conventions established by language users. It is an accident of history that the sequence of shapes cup has meaning and the sequence of shapes - ( ^ does not. Intrinsically, the first sequence is as meaningless as the second. But what is true of ink splotches and sounds seems no less true of all other physical phenomena. They all seem obviously devoid of meaning until someone decides to use them to convey meaning. As John Searle puts it, words, sentences, and the like, considered as material objects, have only “derived intentionality.” We are able to impart meaning to them by virtue of having thoughts with “original intentionality” – your thought about a cup represents or means cup without anyone having to form a convention of using it to mean that. But if neural processes are as devoid of original intentionality as ink marks, sounds, and the like, then it is hard to see how thoughts could be identified with neural processes, or claimed to supervene upon them. And the same is true of any other purported physicalistic basis for mental phenomena.

Now, there are various things a physicalist might want to say in response to this, but the point is that the problem intentionality is claimed to pose for physicalism here can obviously be stated in a way that makes no reference to the “paradoxes of intentionality.” If the commonsense point just made constitutes a difficulty for physicalism, it would do so even if the paradoxes in question did not exist.

The more technical way of making the point is to emphasize that the conception of “the physical” that physicalism typically presupposes is a mechanistic one – that is to say, one which (as I have discussed ad nauseam, e.g. here) takes matter to be devoid of any immanent or intrinsic final causality or teleology of the sort affirmed by the Aristotelian-Scholastic tradition the early moderns sought to overthrow. For the Scholastics, efficient causes in the natural order inherently “point to” or are “directed at” their effects, and in sweeping aside immanent final causality the moderns rejected the claim that any natural phenomenon inherently and irreducibly “points to” or is “directed at” anything at all. Intrinsically, the natural world is for them comprised instead of “nothing but” meaningless, purposeless particles in motion or the like. (Descartes, Locke, Boyle, Newton, and other early mechanists did of course think of ends or goals being imposed on the world by God, but precisely because they were mechanists opposed to Aristotelianism and Scholasticism, they saw the resulting purposes or meanings as extrinsic to the world rather than inherent. See the great many posts on this theme that I’ve written over the last several months, as well as the discussions in The Last Superstition and Aquinas.)

Now, if intentionality involves something “pointing to” or being “directed at” or “about” something beyond itself, and the mechanistic conception of matter underlying physicalism holds that there is no such thing in nature as something inherently and irreducibly being “directed at” or “pointing to” something else, then it seems at the very least difficult to see how intentionality could possibly be something material or physical. I had reason to make this point in my recent post on Chomsky. But though Stoljar quotes the same passage from Jerry Fodor that I cited there, he does not see (as Fodor does, though Fodor does not make explicit reference to the anti-Aristotelian mechanistic revolution) that it is the moderns’ own conception of matter, rather than the “paradoxes of intentionality,” that generates the difficulty.

Again, the point is not that the physicalist might not have a good response to points like the ones I’ve been making – I don’t think so, but that’s another issue. The point is rather that it misses the point to address the problem of intentionality as if the paradoxes Stoljar calls attention to were at the heart of it, and as if it had nothing to do with the nature of “the physical.” Both the commonsense point and the technical point (as I have called them) show that the problem has very much to do with the nature of the physical, and nothing essentially to do with the paradoxes.

But how, the physicalist might still ask, does dualism fare any better? For as Stoljar suggests, wouldn’t any objection to a physicalist account of intentionality apply mutatis mutandis to any dualist alternative? Or as Clayton Littlejohn once objected in a remark in Victor Reppert's combox: “It seems like causal pathways in an immaterial substance would have the same content fixation problems as causal pathways in a physical substance.”

As I have said, this sort of objection seems increasingly common in contemporary philosophy of mind, but it is deeply confused. What dualist ever said anything about “causal pathways in an immaterial substance”? Stoljar and Littlejohn seem to think that what the dualist means by an immaterial substance or soul is something that is just like a material substance – and in particular, something with distinct and causally interrelated parts – only not material, but instead “made out of” some other kind of “stuff” (“ectoplasm” maybe). In short, a kind of ghostly machine, but a machine all the same. But that is precisely what dualists – whether of a Platonic, Thomistic, or Cartesian stripe – do not think the soul is. For dualists have typically held that the soul is simple or non-composite, and thus not “made out of” causally interrelated parts of any sort. That its activities cannot be modeled on those of a material substance is the whole point.

How should we think of it, then? For the Cartesian, the essence of the soul is thought, and that is the entirety of its essence. Descartes does not say: “Gee, it’s hard to see how intentionality could be explained in terms of causal relations between physical parts. I therefore postulate an immaterial substance with immaterial parts whose causal relations are capable of generating thought and intentionality.” That would imply that in addition to thought, a soul has of its nature the various parts in question and their characteristic interrelations. And that is just what Descartes denies. A Cartesian immaterial substance doesn’t generate thinking. It is thinking, and that is all that it is. For that reason, and contrary to what Stoljar assumes, the Cartesian conception of intentionality cannot possibly be open to the same objections raised against physicalism. To say “Maybe a Cartesian immaterial substance – that is to say, something which just is its activity of thinking – could, like a physical substance, exist in the absence of intentional mental states” is just incoherent. A physicalist might want to raise some other objection to the Cartesian view, but Stoljar’s tu quoque is not open to him.

Now, for the Thomistic or hylemorphic dualist, the soul is to be understood, not as pure thought, but rather as the substantial form of the living human body. And qua form, it is not a complete substance in the first place, much less a material or quasi-material one. (Talk of the soul as an “immaterial substance” is thus for the Thomist at least misleading, though he does hold that the soul subsists beyond the death of the body as an incomplete substance.) Here too, though, talk of interrelated quasi-material parts, “causal pathways,” and the like is completely out of place. But for the Thomist, the Cartesian’s talk of inner “representations” is out of place too; as I have discussed elsewhere (e.g. here and here) the “representationalist” conception of the mind is an essentially modern one that the ancients and medievals generally would have rejected. As a consequence, the ancients and medievals would reject too the essentially modern way of framing the issue of intentionality that I have, for the sake of argument, been following up to now in this post. For instance, if a broadly Aristotelian-Thomistic conception of nature is correct, then natural phenomena really do have intrinsic final causes after all, so that (for example) material efficient causes inherently “point beyond” themselves to their effects. It would follow that a thought’s “pointing beyond” itself cannot be what makes it immaterial; and indeed, that is not the sort of argument the ancients and medievals gave for the mind’s immateriality. (Nor did they argue from “qualia” – that too, as I have noted many times before, is a very modern sort of argument for dualism, and presupposes a mechanistic approach to nature.) The ancients and medievals focused instead on such features of our thoughts as their universality and determinacy, which they took to be essentially incompatible with thought’s having any material organ. (See here, here, chapter 4 of Aquinas and chapter 7 of Philosophy of Mind.)

But that is a gigantic topic of its own. Suffice it for present purposes to note that with respect to Thomistic dualism no less than the Cartesian version, contemporary physicalists would do well to try better to “know their enemy” before dismissing him.

[For more on this theme, see my posts on Paul Churchland and dualism, here, here, and here.]


  1. Hey Ed,

    Great post as ever. One question I have. You say...

    The set of splotches looks like the word, but it isn’t. The word has meaning, the splotches do not.


    But what is true of ink splotches and sounds seems no less true of all other physical phenomena. They all seem obviously devoid of meaning until someone decides to use them to convey meaning.

    Fair enough. Later you say...

    For the Scholastics, efficient causes in the natural order inherently “point to” or are “directed at” their effects, and in sweeping aside immanent final causality the moderns rejected the claim that any natural phenomenon inherently and irreducibly “points to” or is “directed at” anything at all.

    Doesn't this mean that those "ink blots on the page", under the Scholastic view, actually could/would have some kind of meaning? Maybe not explicit 'intentionality' - but what about some other kind of 'aboutness' or 'proto-intentionality', so to speak?

  2. Hello Crude,

    What they'd have is certain causal powers, and causal powers entail final causality insofar as an efficient cause is "directed toward" its characteristic effect(s). But they would not intrinsically have any sort of linguistic meaning.

  3. Prof. Feser: Nor did they argue from “qualia” [...] ancients and medievals focused instead on such features of our thoughts as their universality and determinacy

    I reckon those aren't completely independent; after all, something is a quality rather than a quantity because it's not, well, quantifiable, not here or there, not this much or that much — which is of course what universals are. But there still is something interesting about qualities because they can't work the same way quantities can.

    I can see two apples, and get the two-ness in my mind, easy. (Well, not so easy, since the information is encoded in light and in signals in my brain, which have to be decoded — but the point is, the two-ness is out there, so it can end up abstracted in my mind.) Similarly with the roundness of the apples: physical stuff (molecules, etc.) has actual position, and shape is just a certain positioning of the apple's matter, and those relative positions (albeit encoded in some way) can reach my mind where that form of roundness is intelligible, and I can perceive that the apples are round.

    But the apple's redness is different. Position and number are quantities which I can perceive through my physical senses, but "redness" does not seem to be. What I perceive with my sense of sight is light of a certain frequency, and while that is the direct and natural result of the light's interacting with the nature of the apple, it's not colouredness. The transmission via light via my retina via synaptic signals, etc. isn't the point; that happens with the size or shape of the apple, but in those cases, the transmission is carrying the right kind of in-form-ation. In the case of colour, it isn't. The apple may really be red after all (i.e. participating in the form of redness), but there doesn't seem to be any way for that information to be communicated to my senses. Augustinian illumination or Leibnizian pre-harmony could explain it, but it seems more direct to conclude that the redness isn't in the apple itself; or if it is it's a coincidence. (The apple could participate in blueness, but we'd still see red because it's not really the blueness that we're "seeing".)

    This isn't to say that the colour is arbitrary or subjective (other than the obvious sense that it is a sensation in a subject); presumably God simply constituted our minds so that certain physical frequencies would appear to us as colours, and certain other kinds of physical frequencies as sounds, and so on. That would be part of human nature, and He could have given us a different nature just as much as He could have given us eight fingers each or twelve. But no matter what kind of nature God gave us, two would still be two and round would still be round.

  4. David, even if the situation in the apply under which the frequency that the apply reflects is a frequency that our brain apparatus reports as "red" needn't "BE" red in the apple in and of itself (in some fashion) the physical foundation for its reflecting some specific frequency is a real physical attribute of the apple. This real attribute of the apple can be called "red" without distortion of language or meaning.

    Let's go back a minute: when we hear a doorbell, the hearing reports first A, and then F. The brain interprets, in addition to simply the sound alone, also "two," separated by time. Two can come from senses other than sight.

    Properly speaking, the eye's (and the ocular center of the brain) primary report is only of color, not of size, nor of shape as such, nor of distance, nor even of number. If you were out in space and were in a large sphere made of semi-translucent red glass, and behind the glass were about 1 million tiny lights, all you would see is color, namely red. Without any variation. Without variation, you could not distinguish any of the OTHER sensibles, like distance, etc. Although you can tell that this part of the red is not the same part of the red that that part is (i.e. part outside of part), there are no shapes to report. And no number. If an object were inserted half the distance between you and the wall, which had exactly the same intensity of red light to your eye, you could not distinguish it as a distinct object at all, even though you would see its redness perfectly well. So you could not state it's position in the least.

    My point is that common sensibles like number and shape are not primary objects of the sense at all. The "shape" that your brain signifies about the apple is a product of, first, the color of the apple, and color of the other things that are not apple, received in such a way that the one color is not where the other color is. The apprehension that the differences in color imply a distinct object, (rather than, say, just 2 colors side by side on the same object) and further that the object has a certain recognized attribute "round" are, actually, additional operations beyond the simple reporting of the multiple colors.

  5. Tony: My point is that common sensibles like number and shape are not primary objects of the sense at all.

    Thanks for the reply. I understand what you're saying, and perhaps I expressed myself poorly — what I don't see is how to explain the connection between physical and mental qualities. Number (shape, etc.) are clearly not objects of sight in that if you put the platonic number Two itself in front of me (or the platonic Circle, etc.) I couldn't see it — they're not visible things. But when I see two patches of colour, there really is twoness there; if I saw only colour, then I could never distinguish two bits of red from one bit of red. A red circle participates in roundness as well as colouredness, and the image on my retina participates in roundness, and so on, so I really am "sensing" the roundness, or the twoness, and am able to abstract it from the redness to understand that there are two of whatever it is.

    (To be precise, it's really the form(s) as processed by my sense organs; so as you point out, if two red things are too close or too identical in intensity for my eye to distinguish, that means that in fact the image on my retina would just be one red patch, thus having the forms of oneness and redness, so that is what my mind would abstract. And of course there are further opportunities inside my brain for the signal to be processed or altered or degraded in some way, but for simplicity we can stick to examples that aren't too fuzzy or too ambiguous.)

    Anyway, my point is that when there are two (visible) objects, and I come to know in my mind that there are two of them, there's an obvious connection. (Perhaps God could have constructed me so that when I saw two things, I experienced "three(ness)" — but even if that behaviour were completely consistent in me, it would in some obvious sense be "wrong"; or at least, my thoughts would obviously disagree with reality.) So far so good.

    But when I see a visual frequency of 700nm there is no obvious "right" colour I should experience. If God had designed humans to experience "blueness" when their senses perceived 700nm light, how could you say that disagreed with reality? Or if God made us experience middle C when we experienced 700nm light? Not only which colour, but even whether we mentally experience colour or sound seems completely contingent.

    The only way it could be "wrong" is if the object participated in the form of Redness but we did not come to abstract redness in our minds. But then it's contingent that the apple has the form of Redness in the first place; if the apple participated in Squareness instead of Roundness, then it's physical construction would have to be different. The same atoms couldn't be in a round configuration if the apple is square; that's what roundness is, having parts arranged in a certain way. But the pure platonic form of Redness is not tied to the reflection and absorption of photons in that way. That's the form of Photoreflectivitiness, which is a very different thing. So if we can't abstract redness from "EMF wavelength of 700nm", then how do we get the red experience in our minds?

  6. "But that is a gigantic topic of its own. Suffice it for present purposes to note that with respect to Thomistic dualism no less than the Cartesian version, contemporary physicalists would do well to try better to “know their enemy” before dismissing him."

    Generally speaking only, admittedly, but it would seem the dismissiveness in question is the primary intention, while the rationale/argument follows. Iow, the lack of rigor and conscientiousness, the lack of probity, typically appears to be premeditated and conscious, not a mere absentminded lacuna.

    The lack of probity and conscientiousness alluded to is so often on evidence that it's difficult (imo) to draw a more charitable conclusion.