McDowell’s Aristotelian near miss

John McDowell’s paper “Singular Thought and the Extent of Inner Space” made a big impression on me in graduate school, around the same time his influential book Mind and World was published.  Like a lot of philosophers, I thought there was something deep going on in McDowell’s work, though (also like a lot of philosophers, I think) I was not quite sure what to make of it.  Part of this has to do with the difficulty of McDowell’s style, but that difficulty reflects, at least in part, the difficulty of the subject matter.  The nature of thought and of experience is so close to us – like the tip of one’s nose, always in one’s field of vision, and thus rarely noticed – that it can be, precisely for that reason, harder to get hold of than the extra-mental world is. 

Hence, other than briefly alluding to his work in my doctoral dissertation, I moved on to other matters in the years immediately after first encountering it.  It wasn’t until years later, after getting hip deep into Aristotelian and Thomistic philosophy, that I realized that what was going on in McDowell was a partial rediscovery of an essentially Aristotelian-Thomistic theory of knowledge.  That is not how McDowell himself presents it, though.

The Cartesian prison

One of McDowell’s major themes is an attack on the Cartesian conception of the mind.  I’m not talking about Descartes’s substance dualism, though McDowell does reject that.  What is in view is rather what McDowell describes as follows:

In a fully Cartesian picture, the inner life takes place in an autonomous realm, transparent to the introspective awareness of its subject; the access of subjectivity to the rest of the world becomes correspondingly problematic, in a way that has familiar manifestations in the mainstream of post-Cartesian epistemology…

[It is] the idea of the inner realm as self-standing, with everything within it arranged as it is independently of external circumstances. (“Singular Thought,” pp. 146 and 152)

On this Cartesian picture, our conscious experiences could be exactly as they are, without there actually being any external world corresponding to them (as in Descartes’s scenario where these experiences are hallucinations caused by an evil spirit).  You might think that the worry here is the familiar one that this picture of the mind makes the external world unknowable.  But though that is closer to the point, McDowell is primarily concerned with an even deeper problem, which is that the Cartesian conception makes external reality unthinkable.  It’s not merely that we’re locked in a Cartesian theater, having direct access only to mental representations of the external world, and cannot be certain that there really is anything outside the theater, anything which corresponds to the representations.  It’s also that the Cartesian picture threatens to make it unintelligible how our experiences could count as true representations in the first place – how they could have the intentionality they do, how they could so much as stand for or be about external objects (whether or not those objects exist).

The problem is the contingency of the connection between mind and world posited by the Cartesian picture.  Here’s an analogy (mine, not McDowell’s).  Words like “dog” and “cat” have no inherent or necessary connection to dogs and cats.  They are, of themselves, just meaningless strings of shapes or noises (depending on whether they are written or spoken).  The connection of these symbols to the dogs and cats they represent is a matter of convention.  Now, the convention gets set up because our thoughts about dogs and cats do have some kind of necessary connection to the things they are about, and the linguistic symbols inherit this connection by standing in for the thoughts – or so it seems.  But on the Cartesian model of the mind, mental states too have only a contingent connection to external reality.  For, again, the model holds that the mental realm could be exactly as it is even if there were no external world corresponding to it.  So, how do mental states have, in that case, any more power to represent external reality than meaningless strings of shapes or sounds do?  How can they have what philosophers call any “intentional content” at all?

McDowell concludes that “it [is] quite unclear that the fully Cartesian picture is entitled to characterize its inner facts in content-involving terms – in terms of its seeming to one that things are thus and so – at all,” so that the mental realm it posits is “blank or blind” rather than having any genuine intentionality or aboutness (“Singular Thought,” p. 152).  If the Cartesian conception were correct, our own experience would have the character of what William James called, in another context, “one great blooming, buzzing confusion.”  It would not even seem to be an experience of a world of tables, chairs, dogs, cats, trees, clouds, and people. 

Note that this has nothing essentially to do with Descartes’s view that the mind is immaterial.  As McDowell emphasizes, modern materialism has inherited this broadly Cartesian conception of the mind and just relocates the mind so conceived in the brain rather than in Descartes’s res cogitans.  It typically retains the idea that there is no inherent connection between mental states and the external objects mental states are said to represent.  It posits a causal correlation between mental representations and external objects (just as Cartesian dualism does) while allowing that the representations might in principle fail to represent the world as it really is – in which case, again, it is hard to see what makes them true representations at all.  Material states no less than states of a res cogitans should, given the Cartesian picture, be “blank or blind” rather presenting the world to us in the way consciousness actually does.

Opening up the mind

Hence McDowell concludes that the Cartesian picture is not correct.  He argues that we ought “to picture the inner and outer realms as interpenetrating, not separated from one another by the characteristically Cartesian divide” (“Singular Thought,” p. 150).  What does that amount to?  McDowell proposes several ways of spelling the idea out.  One of them involves the notion of a singular proposition (also sometimes called a Russellian proposition after Bertrand Russell, who developed the idea).  A singular proposition is a proposition about some particular individual thing, where the thing itself is a constituent of the proposition.  For example, the proposition that the Wilshire Grand Center is the tallest building in Los Angeles will be a singular proposition in this sense if the Wilshire Grand Center itself really is a constituent of that proposition.  (Whether a particular individual thing really can be a constituent of a proposition, and thus whether there really are singular propositions, is a matter of controversy.)

A singular thought (using “thought” here to refer to a psychological episode of the familiar sort) would be a thought whose content is a singular proposition – and thus a thought which has, as a constituent, some particular individual thing.  For example, if I am thinking that the Wilshire Grand Center is the tallest building in Los Angeles, then the Wilshire Grand Center itself would be a constituent of my thought.  On this conception, suggests McDowell, we could take the “inner space” of the mind to extend outward to include such external objects themselves.   And in that case, “there is now no question of a gulf… between the realm of subjectivity and the world of ordinary objects” insofar as “objects themselves can figure in thoughts which are among the contents of the mind” (“Singular Thought,” p. 146).

Another way of spelling out the “interpenetration” of mind and world is developed in Mind and World, where McDowell rejects the idea that there is a sharp divide between the content of a thought, on the one hand, and, on the other hand, the facts in the world that the thought is about.  He writes:

[T]here is no ontological gap between the sort of thing one can mean, or generally the sort of thing one can think, and the sort of thing that can be the case.  When one thinks truly, what one thinks is what is the case… [T]here is no gap between thought, as such, and the world. (Mind and World, p. 27)

Commenting on passages like these, Tim Thornton attributes to McDowell an “identity theory of thoughts and facts.”  And if thoughts and facts are identical, that would (so the argument goes) rule out a conception of the mind that makes it “blank or blind,” devoid of intentionality.  We cannot say that the contents of the mind would be just as they are, independently of whether there was an external world, if those contents just are the same things as the facts comprising the external world.

Developing the argument of Donald Davidson’s classic paper “On the Very Idea of a Conceptual Scheme,” McDowell rejects the conception of experience as devoid of structure or intelligibility apart from some conceptual scheme we impose upon it from outside, as if the former could exist apart from the latter.  Only what is already conceptualized could ever serve as a rational justification for anything, so that if experience was, considered by itself, devoid of conceptual structure, it could never play any justificatory role in anything we believe.  What McDowell calls “the space of reasons” (a phrase he picks up from Wilfrid Sellars) – the order of logically interrelated concepts, beliefs, and inferences – would thus float free of empirical reality.

The right way to think about human experience, then, is as already, of its nature, saturated with conceptual content, and the right way to think about the external world that experience reveals to us is as itself having a structure that corresponds to this conceptual content.  McDowell contrasts this with the “disenchanted” view of nature we’ve inherited from early modern science, and from empiricists like Hume.  He writes:

[W]e cannot suppose that intelligible order has completely emigrated from the world we take to be mirrored by intellectual states… We have to suppose that the world has an intelligible structure, matching the structure in the space of logos possessed by accurate representations of it.  The disenchantment Hume applauds can seem to point to a conception of nature as an ineffable lump, devoid of structure or order.  But we cannot entertain such a conception.  If we did, we would lose our right to the idea that the world of nature is a world at all (something that breaks up into things that are the case), let alone the world (everything that is the case).  (“Two Sorts of Naturalism,” in Virtues and Reasons, edited by Hursthouse, Lawrence, and Quinn, at p. 160)

But it isn’t really science itself that presents us with such a picture of nature.  It is the interpretation of science put forward by scientism and reductionistic brands of naturalism that does so.  “This kind of naturalism tends to represent itself as educated common sense, but it is really only primitive metaphysics” (Mind and World, p. 82).

Kant or Aristotle?

Even those sympathetic to McDowell’s position might have at least two concerns about it.  The first is that it seems to rely too much on metaphorical ways of characterizing the position McDowell wants to put in place of Cartesianism and reductionistic naturalism.  How exactly should we cash out talk about the mind and world “interpenetrating”?  What is the nature of the “inner space” of the mind, given that it is not like the literal space that material objects outside the mind occupy?  What exactly does the Wilshire Grand Center being a constituent of my thought about it amount to?  Obviously it is not a constituent of my thought in the same sense in which it is, say, a constituent of a certain city block in Los Angeles. 

A second concern is that attributing to the world something like the conceptual structure of thought, and making external things constituents of the mind, might seem to entail a kind of idealism that collapses the world into the mind.  This worry is only exacerbated by the fact that McDowell finds inspiration in Kant and post-Kantian idealism, albeit he does not characterize his own position as idealist.

Now, as I said at the beginning, it seems to me that there are, in McDowell’s work, clear gestures in the direction of what amounts to an Aristotelian-Thomistic conception of the mind’s relation to the world.  And resources from that conception would, I suggest, rescue McDowell from the two difficulties I’ve referred to.  To be sure, McDowell does cite Aristotle prominently in his exposition of his preferred conception of nature.  But his focus is on Aristotle’s ethics as a model of how to conceive of human beings in a way that is broadly naturalistic without being reductionist.  He does not make use of the relevant Aristotelian epistemological and metaphysical ideas.

The key theme here is the Aristotelian-Thomistic idea that when the intellect understands something, it takes on the thing’s form – the same form that, when informing a bit of matter, makes of that matter a particular instance of the kind of thing the form defines.  For example, when the intellect understands what it is to be a triangle, it takes on the form of a closed plane figure with three straight sides, which is the same form that, when taken on by a bit of ink, makes of that ink a triangle. 

The reason the intellect does not itself become a triangle by virtue of taking on this form is that it takes on the form without matter, and a triangle is a kind of material thing.  Hylemorphism – the thesis that physical substances are composites of form and matter – is thus a crucial metaphysical component of this epistemological story.  It makes it possible to say that the intellect is identical to the objects of thought formally, but not materially.  Why is this important?

Not too long ago I reviewed Raymond Tallis’s book Logos: The Mystery of How We Make Sense of the World.  Tallis notes how modern attempts to close the divide between mind and world opened up by Descartes tend either to collapse the mind into the world (as reductionistic naturalism does) or to collapse the world into the mind (as idealism does).  The trick to avoiding both extremes, he rightly argues, is to preserve the distinction between mind and world without opening up the unbridgeable gap between them that the Cartesian picture entails.  We need, as Tallis says, to preserve “connection-across-separation.”  As I noted in the review, the Aristotelian-Thomistic position does precisely this.  Because a thought and the thing thought about are formally identical, the mind has such an intimate connection with the world that there is no epistemic and semantic gap of the kind deplored by thinkers like McDowell.  But because they are nevertheless not materially identical, there is no collapse of mind and world.

McDowell is by no means unfamiliar with this account.  In his collection Having the World in View: Essays on Kant, Hegel, and Sellars, he discusses it in an essay with the intriguing and playful title “Sellars’s Thomism.”  I say “playful” because Sellars was, of course, hardly a Thomist.  But like McDowell, he was keen on making dialogue partners of great thinkers of the past, including those with very different philosophical commitments than his own, and Aquinas was no exception.  McDowell discusses the use Sellars made in his essay “Being and Being Known” of the Aristotelian-Thomistic account of knowledge I just sketched. 

Sellars’s concern was the relationship between the intentionality of thought and the meaningfulness of language, and he thinks there is an interesting connection between his own views about these matters and Aquinas’s notion of the “mental word.”  McDowell tells us that his main concern in his own essay is with understanding Sellars’s use of Aquinas rather than with Aquinas himself, but he does suggest that Sellars’s naturalistic presuppositions lead him to misread Aquinas.  And he closes the essay with the following paragraph:

Now Aquinas, writing before the rise of modern science, is immune to the attractions of that norm-free conception of nature.  And we should not be too quick to regard this as wholly a deficiency in his thinking.  (Of course in all kinds of ways it is a deficiency.)  There is a live possibility that, at least in one respect, Thomistic philosophy of mind is superior to Sellarsian philosophy of mind, just because Aquinas lacks the distinctively modern conception of nature that underlies Sellars's thinking.  Sellars allows his philosophy to be shaped by a conception that is characteristic of his own time, and so misses an opportunity to learn something from the past.  (p. 255)

All the same, McDowell himself mostly just describes the Aristotelian-Thomistic position in the course of discussing Sellars’s treatment of it, rather than either endorsing or rejecting it.  Nor (as far as I know) does he discuss the matter elsewhere.  So, McDowell’s salutary critique of Cartesianism seems, for all its strengths, insufficiently attentive to an important approach to the matter – a pre-Cartesian perspective, which is importantly different from the post-Cartesian perspective represented by the thinkers who have most influenced McDowell (Kant, Wittgenstein, Sellars, Davidson, et al.).

Related posts:

Truth as a transcendental

Fodor and Aquinas on the Extended Mind Thesis

Putnam on causation, intentionality, and Aristotle

Putnam and analytical Thomism, Part I