Naturalism versus Katz’s Platonism

Naturalism holds that there is nothing more to reality than the world of concrete entities causally related to one another within space and time.  Since this is the realm studied by the natural sciences (such as physics, chemistry, and biology), naturalism can also be characterized as the view that the subject matter of natural science is all that there is.  Naturalism thus denies the existence of God, of angelic intellects, of immaterial souls (whether conceived of along the lines of Descartes’ res cogitans or in some other way), and of Platonic Forms and other abstract objects. 

When people think about philosophical criticisms of naturalism, it seems they usually call to mind arguments for God’s existence or for mind-body dualism.  But it is possible to reject naturalism for reasons independent of those particular issues.  That would be the approach of those who argue for the reality of abstract objects, whether they be universals, numbers, propositions, or what have you.  Call this sort of view Platonism.  (Here I am using the term “Platonism” more or less the way contemporary analytic philosophers tend to use it, viz. to refer to belief in the existence of such abstract entities as distinct from a commitment to either theism or soul-body dualism.  In the history of Platonism there was, of course, a closer connection between these three views than one might guess from the contemporary usage.)

Platonism in this sense has had a number of illustrious defenders in modern analytic philosophy from the very beginning of its history.  Frege’s classic essay “The Thought” is a famous defense of an essentially Platonist view of propositions.  Russell’s The Problems of Philosophy defends something like a Platonist view of universals.  Popper’s “World 3” notion is in some respects similar to Plato’s realm of Forms.  Gödel was famously committed to mathematical Platonism.  When I was in graduate school, it was arguments of this kind, and especially those concerning propositions, that gradually broke the hold over me of the naturalism to which I had long been committed – and softened me up, as it were, for a reconsideration of theism and of the immateriality of the mind.

It was during that time that I first came across Jerrold Katz’s book The Metaphysics of Meaning, a critique of naturalism very much along these broadly Platonist lines.  I’d later discover that this book was part of a larger project that included other works like Language and Other Abstract Objects.  Katz’s work is analytic philosophy at its best – the rigorous argumentation characteristic of that tradition, but in the service of defending old-fashioned metaphysical positions that the tradition is often (if wrongly) thought to have undermined. 

It is not just Katz’s Platonist approach to criticizing naturalism that makes his work stand out, but also his distinctive way of arguing for Platonism.  Katz was a linguist as well as a philosopher, and it is considerations about language and logic, specifically (as opposed, say, to mathematics or general metaphysics) that drive his arguments for Platonism.  But his focus is not (as it was for Frege) on the idea that we cannot make sense of language without positing propositions as abstract objects, lying beyond language, that we access by way of language.  Rather, for Katz, a language and the sentences that can be formulated within it must themselves be understood as abstract objects.  It is this realm of abstract objects that is the subject matter of linguistics – just as, for the mathematical Platonist, numbers constitute a realm of abstract objects that are the subject matter of mathematics, and for the Platonist metaphysician, the Forms are the primary subject matter of philosophy.

Platonic realism is traditionally contrasted with nominalism and conceptualism.  Each of these three positions has variations relating to the subject matter about which one may or may not be a realist (universals, numbers, propositions, or whatever).  For example, consider the problem of universals.  The Platonic realist takes them to be abstract objects (the Forms) in a “third realm” distinct from either the world of particular material things or the world of the mind.  The nominalist holds that only particular things are real and that universals are mere fictions, artifacts of language that correspond to nothing outside language.  The conceptualist takes the purportedly middle ground position that universals exist, but are the sheer creations of the mind rather than mind-independent realities waiting to be discovered by us.  I say “purportedly” because it is difficult to spell conceptualism out in a way that doesn’t collapse into either nominalism or realism of some sort.  (I put to one side for the moment the other variations on realism, viz. Aristotelian realism and Scholastic realism.)

Where language itself is the subject matter, Katz characterizes the nominalist rival to his own position as the view that the focus of linguistics ought to be on the study of token utterances and scribblings – this particular utterance or writing out on paper of the sentence “The cat is on the mat,” that particular utterance or typing out of the sentence “The dog is on the log,” and so on.  Katz endorses Chomsky’s critique of this conception of linguistics.  But Chomsky’s own approach is not realist either in the relevant sense.  Linguistics is, for him, fundamentally about discovering the rules of Universal Grammar that are innate to the human mind.  For Katz, this amounts to a riff on conceptualism or psychologism.  It cannot do justice to the objectivity of linguistic facts.

But it is Wittgenstein and Quine whom Katz takes to have developed the main twentieth-century defenses of naturalism of the kind he has in his sights.  Wittgenstein’s “therapeutic” approach to philosophy aimed to cure us of the temptation to suppose that language is anything more than one further part of the natural history of human beings, alongside seeing, hearing, eating, walking, and so on.  Quine’s scientism sought to fold the study of language into the more general study of human behavior, understood naturalistically.  Much of The Metaphysics of Meaning is devoted to showing that neither thinker successfully made the case. 

Katz’s positive arguments appeal to features of language and logic that the methods to which naturalists confine themselves cannot account for.  For example, there is the necessity possessed by logical truths and analytic statements.  A sound logic and linguistics cannot plausibly deny such necessity.  But the psychological properties posited by a conceptualist theory like Chomsky’s can only ever be contingent in nature.  Hence there is a deep ontological mismatch between the logical and linguistic facts on the one hand and the facts to which such a naturalistic theory can appeal on the other.

Or consider that a language’s grammatical rules allow for the construction of an infinite number of possible sentence types, and that the recursiveness of the logical connectives (and, or, if-then, etc.) allow for an infinite number of possible compound propositions.  By contrast, actual human linguistic and logical performance has yielded only a finite (albeit very large) number of concrete sentence tokens. 

Now, a naturalistic theory that identified linguistic and logical facts with psychological-cum-neurological facts would have to extrapolate, from actual performance, what our psychological-cum-neurological capacities are.  And while actual performance could justify attributing to us capacities for linguistic and logical performance well beyond what has been observed, it could not justify the attribution to us of an infinite capacity.  Again, there is a mismatch between what the logician and linguist know to be true of logic and language, and what is true of psychological capacities construed naturalistically. 

The basic problem, as Katz sums it up, is that “linguistics and logic… trade in the abstract while naturalism insists that everything be concrete” (p. 280).  And again:

Sciences like linguistics and logic are about structures which are maximally abstract… [T]he best scientific theories in these disciplines cannot be brought under constraints interpreting them as theories of concrete objects like minds/brains.  In requiring that a theory of English or a theory of implication be a theory of a concrete psychological reality, conceptualism presents us with theories that do not describe the structure of English sentences or implication relations themselves, but describe, as it were, the shadows they cast on the walls of our mental/neural cave. (p. 281)

Linguistics and logic are in this respect like mathematics.  As is well known, it is quite hopeless to try to interpret mathematical truths as nothing more than descriptions of finite, contingent, concrete entities (collections of physical objects, or of symbols, or whatever).  Katz’s point is that the truths of logic and linguistics are no less infinite, necessary, and abstract, so that it is no less hopeless to try to reduce them to descriptions of finite, contingent, and concrete phenomena of some kind (such as psychological phenomena).

Katz’s reference to “shadows they cast on the walls of our mental/neural cave” is most apt.  The modern naturalist labors under the delusion that he is more rational and scientifically informed than the traditional metaphysician, but the reverse is true.  His position is in fact radically out of harmony with the deliverances of sciences like logic and linguistics (not to mention mathematics), and it is only ideology that prevents him from seeing this.  He is like a denizen of Plato’s cave, albeit it is the brain rather than shadows that his attention is fixated on, leaving him unable to see the sunlight of the higher truths of logic, linguistics, mathematics and the like.

There is much else in Katz’s discussion that merits attention – his differences from Frege, his critique and reformulation of Moore’s notion of the “naturalistic fallacy,” and so on.  But the primary value of his work lies in the salutary reminder it affords us of a fundamental and historically enormously influential (but in recent times strangely neglected) style of challenge to naturalism.

It may seem odd for an Aristotelian like me to commend the work of a Platonist like Katz.  But I don’t see Platonism as a hostile rival to Aristotelianism, as naturalism is.  Rather, I see the dispute between Platonism and Aristotelianism as a family squabble, a disagreement over details between thinkers united on the key moves in the war against the naturalist menace.  (Aristotelianism is in this way best seen as a member of a broader “Ur-Platonist” alliance, to borrow Lloyd Gerson’s phrase.)

Moreover, though my own settled version of realism is Aristotelian (or, more precisely, Aristotelian-Thomist) rather than Platonist, I have long thought that the best route to seeing the truth of realism is through Platonism.  Given the way our minds work, I suspect, it is initially easier to see the falsity of naturalism by way of contrast with the exaggerated anti-naturalism of Platonism (and then to bring Plato down to earth with Aristotle and St. Thomas).  That was my own trajectory, in any case, since it was the Platonism of Frege and other contemporary analytic philosophers that first broke the hold over me of naturalism, and opened the way to reaching, eventually, the sober Aristotelian-Thomistic middle ground.

Related reading:

Join the Ur-Platonist alliance!

Frege on objectivity

Frege on what mathematics isn’t

The metaphysical presuppositions of formal logic

Rucker’s Mindscape

What is mathematics about?

The access problem for mathematical Platonism

David Foster Wallace on abstraction

Review of Craig’s God Over All: Divine Aseity and the Challenge of Platonism

Five Proofs of the Existence of God, Chapter 3