. Ultimate or fundamental divisions or kinds. For much of its history the search for categories has wavered between seeking distinctions among things in the world and distinctions among our ways of thinking or talking about the world. Much of the difficulty in each case has lain in knowing what distinctions to count as sufficiently fundamental. It is mainly by being ultimate or fundamental that categories differ from mere classes.
This wavering appears in Aristotle, who first explicitly introduced categories. His ‘official’ list contains ten categories, but the most important are SUBSTANCE, quantity, relative and quality, and his main interest seems to lie in distinguishing substance from the others. The list is clearly derived from different kinds of question that can be asked about a person, like ‘How big is he?’, ‘What is he doing?’
Sometimes Aristotle seems to take a ‘metaphysical’ view, treating categories as the highest genera into which things in general can be divided, so that the world contains substances, qualities, etc., and anything one picks out such as a horse, or red, can be classed under one of these headings. He sometimes lets categories overlap, so that the same item appears in more than one (end of chapter 8 in his Categories). At other times he seems to take a ‘logical’ view and to be classifying the things one can say about something, and in particular about a substance, such as what it is, what qualities it has, how it is related to other things. Here he might be described as classifying predicates, but he often seems to regard predicates themselves as things in the world and not as linguistic expressions, so that the ‘metaphysical’ and ‘logical’ approaches are not clearly separate. Aristotle’s classification is not very exhaustive. The terms ‘one’, ‘good’, ‘being’, he said, did not belong to any one category. These were later called transcendentals (see BEING). There are many others, some of which he discussed, which have no obvious place, including ‘surface’, ‘sound’, ‘chance’, ‘proposition’, ‘necessity’, and complex terms like ‘multiple of three’, ‘knowledge of French’. Aristotle also argues that there cannot be a single all-embracing genus like being or unit.
Many writers have followed Aristotle in elaborating sets of categories, usually more systematic than his. The Stoics had a set of four and they apparently wanted to classify at least some of the world’s contents by examining the questions one can ask about a thing.
Among modern writers the most important contribution is that of Kant, who had a system of four groups of three. He intended these as a classification, whose correctness and exhaustiveness he claimed to prove, of the ways in which any mind recognizably like the human mind necessarily had to perceive and think about the appearances it was presented with. The categories were not a classification of things in themselves (NOUMENA), for Kant thought we could never know these and so never apply categories to them. The categories could only be applied to material given by experience, but they could not themselves be derived from experience, since all use of experience presupposes them. Kant’s general idea is that we can only make sense of the world by imposing some structure originating from the mind upon it, e.g. to choose two of his categories, by seeing it as a set of substances in causal relationships. Many who accept this general idea reject the particular list he gave, and deny that there is some one list valid for all people and times. There is still much dispute about such related questions as whether there are certain features every language must share (cf. philosophy of LANGUAGE, INNATE).
In this century two converging streams of thought have aroused interest in categories. First there are the logical PARADOXES, which led Russell to construct his theory of TYPES. This theory divides the world up by insisting that things of different ‘types’ cannot be put together into a single class. It leads to corresponding divisions in language, e.g. two sentences of which one refers to the other are on different ‘levels’, and cannot be joined into a single sentence by ‘and’. In fact ‘type’ and ‘category’ are sometimes used synonymously. Secondly, thinkers like Husserl and Ryle, among others, have tried to construct a doctrine of categories in order to systematize the ways in which a sentence can go wrong, and in particular the distinction between the false and the meaningless. Roughly speaking, the ideal of this approach would be to divide things into non-overlapping groups so that what could be said truly or falsely, but not nonsensically, of the members of one group differed radically from what could be said of the members of another, rather as most of the things that can be said of a cat differ from what can be said of a wish, or of a day of the week. Sentences which say about a subject in one category something that can only be sensibly said about a subject in another category, are called category mistakes or type confusions, e.g. ‘Saturday is in bed’. Such a doctrine cannot tell us when a sentence makes sense if we must already know this before constructing the doctrine. But the doctrine could systematize the situation, and throw useful light on individual cases through comparisons.
Many difficulties arise concerning categories. It sometimes seems to be thought that, if they exist at all, they must belong to the world and not language, because they must be found out and not created by us. But even if we create a language, we can still discover things about it. We may choose what our sentences shall mean, but once we have chosen we are committed to the implications of our choice, and we do not choose these implications. The main difficulties seem to be of two kinds. First, to think of subjects and predicates that will not go together is perhaps too easy for we may reach so many categories that the doctrine becomes rather trivial, and ‘category’ becomes a pompous name for ‘class’ as often happens in ordinary speech. Are spoons and forks in different categories because ‘This fork has lost one of its prongs’ becomes nonsensical when ‘fork’ is replaced by ‘spoon’? A distinction between absolute and relative categories has been found necessary in facing this problem (Strawson).
The second kind of difficulty, connected with the first, centres round the notion of meaninglessness. There are many ways in which something might be meaningless, nonsensical or absurd, as the following examples illustrate: ‘Horse whether the’, ‘Twice two is five’, ‘My wish has whiskers’, ‘I have found’, ‘I have any apples’, ‘He sleeps like milk’. Some of these may be given senses in special cases, but which of them serve to distinguish categories? One can ask how clearly in fact is the meaningless distinct from the false? ‘Absurd’ can cover both. And is every kind of predicate relevant? Are two things in the same category merely because the predicate ‘being thought about by me now’ can apply to both of them? One controversy arising out of all this is whether categories can be ultimately founded on grammatical distinctions (called syntactical in logic), or whether considerations of meaning independent of mere grammar must be used (in which case categories will have a semantic basis; cf. SEMIOTIC).
Categorial means ‘having to do with categories’. Categorical, though often misused for ‘categorial’, has something like its ordinary meaning of ‘definite’ or ‘downright’, but refers to a certain form of proposition, one which says something is the case without reference to conditions or alternatives. ‘That’s a cat’ is categorical, ‘If that’s a cat, it’s an animal’ is hypothetical. ‘That’s either a cat or a dog’, is disjunctive. Mongrel categorical is Ryle’s name for a statement overtly categorical but covertly including a hypothetical statement; e.g. ‘He drove carefully’ says, for Ryle, not only that he did something but that he would have done certain things if certain events had occurred. See also categorical IMPERATIVE.
Categorematic, rarely used, refers to words or phrases naming things in categories, especially the Aristotelian categories, substance, quality, etc. Syncategorematic applies to words or phrases that somehow link or go ‘together with’ the categories. Basically, syncategorematic words are all those which are not categorematic, and they have often been regarded as not having meaning independently. The commonest examples are words like ‘all’, ‘if’, ‘the’, which do not name substances, qualities, etc., and are often known as logical words although many other examples are often given, including most adverbs. Also some words which behave grammatically like ordinary categorematic ones, yet which are somehow incomplete in meaning, are sometimes included. Some adjectives, especially, only apply to things when the things are described in certain ways. A big mouse is not a big animal. A thing cannot be big (unqualified) as it can be red (cf. ATTRIBUTIVE). A philosophically important case is GOOD. Bigness and goodness, etc., are therefore sometimes called syncategorematic properties, because they only apply to an object when taken together with categorematic terms (‘mouse’ or ‘animal’ in the above example). Cf. transcendentals, under BEING.
Aristotle, Categories, transl. with comentary by J.L.Ackrill in Clarendon Aristotle series, 1963, reprinted (without the commentary) in the 2nd edition of the Oxford Translation of Aristotle, ed. by J.Barnes, Princeton UP, 1984. (See also his Topics, Bk I, chap. 9, and for argument that being is not a genus see Metaphysics, 998b22–7.)
K.Campbell, Abstract Particulars, Blackwell, 1990. (Uses tropes as basis of a one-category metaphysical system, in contrast to Aristotle’s.)
A.D.Carstairs, ‘Ryle, Hillman and Harrison on categories’, Mind, 1971. (Discusses some of their recent work.)
R.Grossmann, The Categorial Structure of the World, Indiana UP, 1983. (Discusses categorial status of individuals, properties, relations, classes, numbers, and facts. Full table of contents.)
Kant, Critique of Pure Reason, 1781, trans. by N.Kemp Smith in 1929. (See its index.)
S.Körner, Categorial Frameworks, Blackwell, 1970. (Discusses basic frameworks of our thinking, allowing that these may legitimately change.)
J.M.Rist, ‘Categories and their uses’, in A.A.Long (ed.). Problems in Stoicism, Athlone, 1971. (Stoic categories.)
*G.Ryle, ‘Categories’, Proceedings of the Aristotelian Society, 1938–9, reprinted in A.Flew (ed.), Logic and Language, 2nd series, Blackwell, 1953. (Attempts to construct theory of categories on semantic basis, resulting in so many category differences that the notion seems in danger of becoming trivial. See also his The Concept of Mind, Hutchinson, 1949, esp. p. 16 (p. 17 in Peregrine edn), for category mistakes, and esp. p. 141 (p. 135 in Peregrine edn) for mongrel categoricals.)
F.Sommers, ‘Types and ontology’, Philosophical Review, 1963. (One of several articles by Sommers elaborating a semantic theory of categories more rigorous than Ryle’s.)
T.L.S.Sprigge, Facts, Wordsand Beliefs, RKP, 1970, pp. 70–2. (Syncategorematic properties. Cf. N.Griffin, Relative Identity, Oxford UP, 1977, pp. 10–11 on ‘polymorphous predicates’.)
P.F.Strawson, ‘Categories’, in O.P.Wood and G.Pitcher (eds), Ryle, Doubleday, 1970. (General discussion of possibility and usefulness of categories, starting from Ryle. See p. 199 (Papermac edn, 1971) for absolute and relative categories.)
This is the complete article, containing 1,835 words
(approx. 6 pages at 300 words per page).
