Gilbert Ryle, a 20th-century ordinary language philosopher, identified a unique fallacy in his book The Concept of Mind. He calls this fallacy a category mistake.
In The Concept of Mind, Ryle argues that a belief in an immaterial mind—or a non-physical mind — is an example of this fallacy. At its most basic, the category mistake is when someone misidentifies an object or concept. They misunderstand its boundaries, or confuse the salient similarities and differences between it and other objects or concepts. He offers the following example to illuminate what he means:
A foreigner visiting Oxford or Cambridge for the first time is shown a number of colleges, libraries, playing fields, museums, scientific departments, and administrative offices. He then asks ‘But where is the University? I have seen where the members of the Colleges live, where the Registrar works, where the scientists experiment, and the rest. But I have not yet seen the University in which reside and work the members of your University.’ It has then to be explained to him that the University is not another collateral institution, some ulterior counterpart to the colleges, laboratories, and offices which he has seen. The University is just the way in which all that he has already seen is organized. When they are seen and when their co-ordination is understood, the University has been seen.
This example illustrates one of the most powerful analytical tools in contemporary philosophy. The ability to delineate.
Ryle defines the category mistake to argue a specific point; he wants to challenge the Cartesian conception of the mind. It was Descartes who once argued that the mind and the body were distinct substances—he believed thoughts were composed of something other than matter.
This might seem sensible at first blush; mental stuff is different from physical stuff. For Ryle, who was a behaviorist, the mind can be entirely defined in terms of outward behavior and probabilistic dispositions. To believe otherwise was to fall victim to a pernicious category mistake.
But shelve Ryle’s arguments against an immaterial mind for a moment. (Especially because human consciousness is still very much an open question in contemporary philosophy.) What about Ryle’s category mistake itself?
The Anatomy of a Category Mistake
The category mistake comes in two forms. It is either semantic or ontological.
Semantic category mistakes are grounded in the confusion of meaning alone. It is possible to understand the real difference between two similar things, while confusing the meanings we use to talk about them. In fact, we often have words in English that do just that; we use such words to refer to two non-identical things or properties because they’re so similar it’s almost a waste to use different words. And yet they’re still different things.
For example, it is possible to understand the word ‘organization’ as meaning both a group of people and an abstract body of rules and norms populated by people. But we could conflate these two meanings in everyday conversation; we might say, “It’s up to the organization to settle this issue.” Do we mean the people or the established systems?
This is a potential semantic category mistake. If the speaker meant organizational norms, then it would be pointless to poll every member of the organization about their opinions.
On the other hand, ontological category mistakes are grounded in the confusion of actual things. Ontology is the study of existence — of what exists. And so an ontological category mistake is when we believe one thing is another. In some sense, meaning still plays a role here. But the ontological category mistakes go beyond mere meanings.
An example of this kind is easier to imagine. It’s not merely mistaking an apple for an orange in low-light conditions. Less commonly, the ontological category mistake happens when we believe two things with similar properties are necessarily part of the same category. Perhaps the best example of this type of failure is the study of convergent evolution — the process by which unrelated species develop similar traits to deal with similar environmental stressors. For example: the dolphin is a mammal, not a fish. To believe the dolphin is a fish would be to fall victim to a category mistake.
It’s hard to tell whether the salient properties in this category are merely properties like having warm-blood and lungs. In some sense, these properties seem arbitrary — we could imagine a future species descended from dolphins without lungs — but these properties are only arbitrary because they help distinguish a real difference in evolutionary history: Dolphins share a common ancestor more closely related to mammals than to fish.
And yet, ontological category mistakes can also be far more sticky — and far more arbitrary — in practice. Return to our earlier example of semantic confusion about the word ‘organization.’ What if I argued that the delineation between ‘organizational norms’ and ‘organization’ was non-existent in an important sense? It’s feasible to argue that an organization is not just a body of people but an ontological cluster of properties.
I could argue that an organization is just people organized in a specific set of relationships. And likewise, if these relationships exist, then obviously the people do as well. While these two concepts remain non-identical, to differentiate between them in practice is itself to fall victim to a category mistake — they are ontologically necessary for each other and for the organization. And this has important consequences.
To return to our earlier example, when someone says “it’s up to the organization to handle this issue,” it only seems absurd to think of every individual’s perspective if we think of an organization as inviolable norms that exist outside of the people who bring them about.
But in practice, the norms are never inviolable; they emerge from the people themselves. If an issue is dire enough, even non-democratic organizations can revolutionize norms if the common interest brings it about. This is because norms depend on people for their existence. To think otherwise is to fall victim to an ontological category mistake.
Confused yet? The reason why category mistakes can be so sticky is human categorization is itself sticky. The way we categorize features of the world are often entirely arbitrary and nebulous.
And so, it seems these concepts bleed into each other more than we are willing to admit. But there must be some basis to categorization. How does Ryle use his category mistakes? He starts by examining concepts by analogy: How are conceptual categories similar and dissimilar? What similarities and differences are salient enough to define a category or to define certain types of categories? How do we define salience itself? And even, is property salience the category of categories?
All of these questions must be answered to have a real use for our categories.
Property Salience and Ambiguity
Ryle offers us some answers to these questions in another book. This work, titled Dilemmas, is a combined series of edited lectures given by Ryle in 1953. In it, he defines a dilemma as the following:
There often arise quarrels between theories… which are not rival solutions of the same problem, but rather solutions or would-be solutions of different problems, and which, none the less, seem to be irreconcilable with one another. A thinker who adopts one of them seems to be logically committed to rejecting the other, despite the fact that the inquiries from which the theories issued had, from the beginning, widely divergent goals.
Could this be the meaning of a category? Perhaps categories are more easily defined as things that, once we commit to them, deny other commitments.
There is reason to believe so. If I commit to some object being red, then I cannot also commit to that something being blue. I can later argue that the object is both red and scarlet without problem because scarlet is a subcategory of red.
But not all categories are so simple. Even colors. The problem is that most categories are not logically incompatible.
Examine the above spectrum. If I asked you to point out exactly where red begins and orange ends, would you be able to pick a non-arbitrary location?
We all know generally where orange and red are, but no one can say exactly where they begin and end. If we did draw a line, it would be an arbitrary commitment — like delineating mammals based on warm-blood alone (without grounding such a distinction further in their evolutionary history).
Is it possible to have a non-porous, non-arbitrary category? After all, if we want to be able to use the category mistake meaningfully, then we have to be able to give specific answers about which things belong to which categories. Perhaps it’s better to reject categorization of more subjective properties like colors and social constructs — the thought process being that if we can find clearly delineated objects, then we can build a sturdy framework that houses the more ambiguous cases.
Take atoms of gold for example. Each atom, that is neither an isotope nor an ion, is exactly similar. Each atom has 79 electrons and protons. Surely there is a category for gold. In philosophy, we call these natural kinds. Objects that retain certain properties irrespective of human observation. We might argue that, if anything, natural kinds are categories we can depend on.
But even if we admit the existence of the natural kinds, we still haven’t rid ourselves of the problem.
A counter-example: it is necessary that all tables have mass irrespective of human observation. Are tables a natural kind in the same vein as gold atoms? If we admit tables, then we have to wonder why we can’t admit ambiguous cases — all things are things irrespective of human observation.
One might be tempted to argue that the table example is an unfair objection to natural kind categories; we might say, “All objects have mass irrespective of human observation. Physical matter is a natural kind.” But the response to this is equally simple: Quantum forces and gravity do not have mass. If we have categories for things that do and do not have mass, then why can’t we also argue that tables are a natural kind? We need to distinguish tables from things like gravity.
We could try to redefine natural kinds to exclude these problem areas like whether or not a thing has mass. But this returns us to the original problem: How do we choose which properties are salient enough to count as delineating a distinct and new category? The new question will then be: Why is the number of protons or electrons of an atom more salient than whether or not a thing has mass, or whether or not a thing is a thing?
There is no easy answer to this question.
Categorization as Process
Make no mistake. Gold is absolutely distinct from iron. Just as bright orange is absolutely distinct from bright red — the problem isn’t one of ontological distinction in obvious cases. Even the university example properly illustrated an obvious category mistake. Instead the problem of categorization is something more.
I think Ryle would have easily acknowledged this difficulty. In Dilemmas, Ryle rejects our obsession with binary all-or-nothing logical constructs. At one point he states:
I have indicated that the quandary, though relatively simple, does depend on a smallish number of concepts, namely, in the first instance, upon those of event, before and after, truth, necessity, cause, prevention, fault, and responsibility. Now there is not just one of these concepts which is the logical trouble-maker. The trouble arises out of the interplay between all of them. The litigation between the two initial platitudes involves a whole web of conflicting interests. There is not just a single recalcitrant knot in the middle of one of the concepts involved. All the strings between all of them are implicated in the one tangle.
I mention this point because some people have got the idea… that doing philosophy consists or should consist of untying logical knots one at a time…
…the idea is totally false that this examination is a sort of garage inspection of one conceptual vehicle at a time. On the contrary, to put it dogmatically, it is always a traffic inspector’s examination of a conceptual traffic-block, involving at least two streams of vehicles hailing from the theories, or points of view or platitudes which are at cross-purposes with one another.
Here, Ryle highlights the basis of a category mistake. It is not some non-relational examination of a thing in stasis — it is not the “untying of logical knots one at a time”; it is the examination of a thing in relation to dozens of other things. Each thing is itself in a kind of flux that affects all of the others.
Perhaps the category shouldn’t be understood as something non-porous and non-arbitrary. Yes, there are paradigmatic examples of categories — like what we would commonly call natural kinds—that are certainly not ambiguous. Yet these individual cases don’t prove the perfection of categories, and they still remain open to amendment as our understanding of categories change.
Instead, we should understand categories as processural things. Their delineation is an ongoing process — a constant unveiling, if you will.
If this is the case, then how should we answer the question “ Which properties are salient enough to define a category?” The answer is simple: Whatever properties we find useful or interesting or capable of making (what we perceive to be) a real distinction for our purposes. This means the answer to this question changes as we change.
To return to the dolphin example: We acknowledge warm blood and lungs as salient properties because they tell us something about evolutionary history. We are curious about evolutionary history because it fits into our over-arching biological paradigm, the theory of evolution. Evolution is the paradigm of contemporary biology because we believe it offers the most explanatory power. We believe it offers the best explanations because it’s satisfying in just the right ways: it’s simple, conceptually coherent, explicable, and “best fits” the evidence at hand.
That’s all there is to it — but note that any change to any one of the above levels would invite not just a restructuring of dolphins and mammals but dozens of categories. And no level, either alone or in combination, is sufficient to give us a perfect road-map for delineation of non-porous, non-arbitrary categories in every case.
If all of this is true, then the category mistake is not merely a mistaken premise made in a bad argument. The recognition of a category mistake also gives us the opportunity to reconceptualize the very structure of our world.