Category Mistakes

Ryle's primary concern was to refute the standard Cartesian (derived from Descartes) picture (Ryle calls it a myth) of "the ghost in the machine." By "the myth of the ghost in the machine," Ryle means the view that there are (at least) two fundamentally different kinds of things, mental and physical. Mental things each belong to a single mind, and they are private, that is, each is only part of a single mind and only in any way available to that mind. Consciousness is a fundamental property of minds, and a lot of similar stuff that probably most of you believe. Physical objects are, unlike minds, public, that is, they are, in principle, accessible to most of us and many actually are accessible to most of us. Physical objects are independent of consciousness.

Mental and physical things are of fundamentally different categories. It is perfectly conceivable that any mind could exist independently of anything physical. (Idealists make that clear.) It is also perfectly conceivable that there could be a physical world without any minds. Many have thought that the mind can survive the death of the body, and the body does survive the death of the mind.

Descartes in part adopted, in part conceived, and certainly established the kinds of things I have been saying. They are now so well established that they seem obvious to most of us. Descartes along with Newton established a radically new physics according to which everything in the physical world is subject to universal laws. Physical systems are, in the metaphor Descartes and his contemporaries emphasized, like clockwork.

If all physical events are determined by physical laws, then there is a problem inherent in the Cartesian picture: there is no room for mental events to influence physical events. So, how is it that we manage to do anything? It ought to be impossible. If, in addition, as Descartes and modern psychologists hope, the mind is also governed by internal laws, we have the same problem in the other direction.

It is completely mysterious how mind and body can interact.

What does any of this have to do with Ayer? He is interested in the foundations in terms of sense data (mental) of empirical knowledge of material objects (physical). The argument from illusion begins with Descartes. It is a product of the picture just sketched: the project of seeing how we (mental) come to know about physical things is an attempt to answer part of the Cartesian problem, seeing how the mind can learn about the physical.

If, in the Cartesian sense, sense data (mental) and physical objects (physical) are of fundamentally different ontological kinds, it isn't clear how they could be related at all. Ayer has a way around that: according to his view, sense data language and ordinary language are just two different ways of talking about the same phenomena. It isn't clear how to relate the two because they are parts of different linguistic systems. The problem isn't to see how mind and body (sense data and material things) interact (an account of that would be a causal theory of knowledge, which Ayer argues against), but rather to see how to translate. If that worked, it would be a partial solution to the Cartesian predicament. One might hope to extend it to a complete solution.

Ryle's project is very much like the one I've just suggested: according to Ryle, "I believe that grass is green," which describes the existence of a mental object, a belief, is a different way of expressing something like, "If anyone asks me "Is grass green?," that is, if the appropriate sounds are produced in my vicinity, I will produce the sound, "yes." More carefully, Ryle thinks mental language is just a special vocabulary for reporting certain kinds of dispositions to behave, and like phenomena, of our bodies.

In brief, Ryle gives a theory a lot like the language of sense data and language of material things theory for the whole mind-body dichotomy.

That can't work, as I have phrased it. Ayer thinks all the phenomena can be described either in terms of sense data or in terms of material things. Ryle doesn't think that everything we can say about physical objects can be expressed using mental vocabulary. He doesn't think that there are two languages, but that there are two parts of a single language that function differently, that can express related phenomena in rather different ways, and that the mind-body picture arises from our failure to see that those parts of the language are used in different ways.

Ayer says that it is a mistake to ask about the (for example, causal) relations between sense data and material things that arises from a misunderstanding about languages. Ryle says it a mistake to ask about causal relations between mental and physical things for an analogous reason: he thinks that they are category mistakes. The idea of a category mistake is a good candidate for rescuing Ayer from some of Austin's criticisms. Those criticisms show (and so do Quine's arguments) that there can't be two distinct languages for sense data and material things. If some of the same things that Ayer described as problems in not realizing terms are parts of different languages can be reinterpreted as category mistakes, then we revive Ayer's explanation of what has gone wrong in causal theories of the relation between the mental and the physical, of sense data and material things.

Even Austin would presumably have agreed that something along those lines would be worthwhile, since he agreed with Ayer that such causal theories and questions about idealism vs materialism, and a whole lot of the Cartesian picture are fundamentally mistaken.

So much for background. Now,

Category Mistakes

What is a category mistake? For that matter, what is a category? I have no idea. Ryle does give us a variety of examples about which he says true things that would be useful if they also applied when mind and body were the two relevant categories. Lots of people agree, I think, that there is something profoundly right about the idea, but there is no generally accepted detailed version of it.

Someone visiting Oxford might be introduced to various professors, shown buildings, museums, departments, and so forth. If, after seeing all that, the visitor asked to see the University, that would betray a fundamental misunderstanding: The University is composed of the ….

After seeing a parade of battalions, batteries, and squadrons, if someone asked when the division will parade by, that would betray a similar misunderstanding.

One can see a baseball team, stadium, players, manager, equipment without locating the team spirit. That example isn't quite like the others: the team spirit isn't the composite of all the parts, but it, like the composites in the examples above, is in some sense not something additional but a feature of how the other parts are organized.

I suspect Ryle dragged in team spirit because it lets him suggest, without saying, that the same is true of a person's spirit.

So, what mistake is made in each of the examples? I don't have a good characterization or even a characterization that I think is true to what Ryle had in mind, but, roughly, a composite or feature of a composite of things or their organization is taken to be just another thing of the sort that made up the composite.

That is also true of Ayer's analysis of how we get from sense data to material things: A material thing isn't a new object in addition to sense data, one that could, for example, be related to them causally. It is an organizational feature of a bunch of sense data.

Ryle thinks that the mind-body problem is just as silly as thinking that the team spirit is some sort of ghostly apparition and wondering how it is that the apparition gets weaker when team members don't get along. What does one have to do with the other? Ayer's rejection of a causal theory of perception rejects it along related lines. His explanation has the defect of relying on a notion of a language that is unworkable. Ryle tries to work out such an explanation for particular cases. It seems that that is a worthwhile project.

Two more cases: If beliefs are inside the mind, then how could what a person says (in the material sense of the sounds made) or does (in the material sense of bodily motions) provide any evidence relevant to what they believe? It seems that there could no point in asking someone what they believe in order to find out what they believe. The one couldn't possibly have anything to do with the other. But that is absurd: asking someone what they believe is the very type of the natural way to find out what someone believes.

If sense data are mental, then how could what a person sees provide any evidence relevant to what material objects there are? It seems that looking is pointless. Sense data and material objects couldn't possibly have anything to do with each other. But that is absure: looking is the very type of the natural way to find out what is there.

-- ShaughanLavine - 26 Apr 2007