. Objects around us share features with other objects. It is in the nature of most such features that they can characterize indefinitely many objects. Because of this the features are called universals and the main problem is to describe their status. Exceptions, such as ‘being the tallest of men’, can be included for convenience. The objects are called their instances. The problem is often called, especially in Greek philosophy, the one-many or one-over-many problem.
Traditionally three kinds of answer have been given: realism, conceptualism and nominalism. Realism in this sense of primarily associated with Plato, who treated universals as objects (cf. FORM, IDEA), separate from their instances, and faced great difficulties over what they were like and how they related to these instances. Plato’s Forms, in so far as they are treated rather as particulars (see below), are often said not to be universals, though doing duty for them. Platonism is nowadays any view which treats things like universals, propositions, numbers, etc., as independent objects. Frege is a noted modern Platonist. Another form of realism, often attributed to Aristotle though the interpretation of Aristotle is very controversial, denies that universals are objects or separate from their instances, but nevertheless makes them real things which somehow exist just by being instantiated. It is unclear how this view treats things like unicornhood. The labels universalia ante rem or res: universals prior to the object(s), and universalia in re or rebus: universals in the object(s), are often applied to Plato’s and Aristotle’s views respectively. Universalia post rem or res: universals, after, or derivative from, the object(s), normally applies to nominalism, thought it could apply to conceptualism. The term substantial universals is applied, like ‘realism’, primarily to Plato’s view, though sometimes also to Aristotle’s. It could, but usually does not, denote universals corresponding to substances, e.g. tablehood, as against qualitative universals like hardness. Realists often limit universals to only some general features.
For conceptualism, universals are thoughts or ideas in and constructed by the mind. This view, summarily rejected by Plato, is largely associated with the British EMPIRICISTS. It may explain human thinking and the MEANING of many words, but it can no longer explain why the world itself is as it is (which Plato claimed his Forms explained). The view thus avoids Plato’s dilemma that the universal is either outside its instances and so irrelevant to them, or inside them and so split up. But what sort of thing is this thought or idea? Does it involve images, and if so, of what sort? Can the same idea be shared by different people, which splits the universal up again, or have they similar but distinct ideas, which leads to the difficulty associated with PRIVATE LANGUAGES? Some writers include conceptualism under nominalism, e.g. Armstrong, who talks of ‘conceptual nominalism’.
For nominalism, represented especially by Ockham in the middle ages and so by many recent writers, there are only general words like ‘dog’, and no universals in the sense of entities like doghood. Cf. MEANING, and also below on ‘types’ and ‘tokens’. (For N.Goodman (1906–), nominalism means recognizing only INDIVIDUALS (second sense), which for him may be abstract but cannot include classes.)
There are two ways of defining a class of objects. One can define it extensionally or in extension, by listing its members, or one can define it as containing all those things which have a certain property or set of properties (called defining it intensionally or in intension; see INTENSIONALITY). The former way makes it impossible for a class, once it is defined, to acquire new members, and is of little use. The latter way leaves it open how many members, if any, a class has; the class of dogs contains whatever things have the properties necessary for being a dog. Nominalism now faces a difficulty, for if there are no universals, i.e. no properties, what determines whether something belongs to the class of dogs or not? This is another version of Plato’s demand for Forms to account for the world’s being as it is. The main nominalist answer to this difficulty uses the notion of resemblance. An object is a dog if it resembles some given dog which is chosen as a standard or paradigm. Two disputed objections to this are that resemblance itself seems to be an indispensable universal, and that resemblance involves partial identity, for to resemble something is to have something, though not necessarily everything, in common with it; the common feature is then presumably a universal.
A variant on the use of resemblance is Wittgenstein’s notion of family resemblance, whereby there need be nothing common to all the members of a class, nor need any member be taken as the paradigm, but the members form ‘a complicated network of similarities overlapping and criss-crossing’ like the fibres that make up a thread. An example Wittgenstein takes is that of a game: have all games got something in common? A somewhat related notion is that of clusters (Gasking).
Particulars, which are not always the same as INDIVIDUALS, cannot be instantiated, and cannot appear as a whole at separated places simultaneously though their parts may be spatially separate. A particular can perhaps appear as a whole at different moments of time (though see GENIDENITY), but these must normally be linked into a stream—though an intermittent sound may constitute one and the same particular, and see Burke. A particular’s parts may be constantly changing, as with a flame, and it need not be ‘solid’ (shadows, rainbows, clouds, can all be particulars, and perhaps the sky). It must, however, be identifiable and distinguishable from other particulars, so clouds, etc., are not always particulars. Particulars can be abstract, provided the conditions about space and time are preserved (e.g. an action or event, like the Renaissance. Rarely non-spatiotemporal things like numbers are included.) Bare particulars are particulars considered as independent of all their properties. It is therefore hard to identify or refer to them.
Particulars are like SUBSTANCES in the first Aristotelian sense of that term, though the emphasis is on being unique in space and time rather than, as with Aristotle, on existing in their own right as the bearers of attributes and subjects of change. Therefore shadows and actions are more easily called particulars than substances, while Platonist universals are more naturally called substances than particulars, especially since particulars cannot be instantiated.
As an adjective ‘particular’ has its everyday sense, and also that given under SENTENCE.
We have seen that universals are sometimes treated rather as particulars. Idealism’s concrete universal is also a kind of particular. It is a system of instances, treated as a developing individual, e.g. man in ‘Man has evolved slowly’. Bradley treats ordinary particulars as concrete universals, since they are developing individuals, though really the universe is the sole individual. He uses ‘particular’ in a more restricted sense than the present entry.
Universals, like particulars, are of many kinds. Some universals (relations) can only be instantiated in ordered pairs or triplets, etc., of objects. Others, like ‘round square’, cannot be instantiated at all, even in thought. Some can be instantiated together with their opposites: an object can be both beautiful and ugly, in different respects; or the object may instantiate the universal only if described in a certain way: something may be large if described as a mouse, but not if described as an animal (see ATTRIBUTIVE); and the instances may themselves be universals, for a universal may have universals as its instance: red may have the property of being beautiful. Moreover, stuffs, like water, are not particulars but presumably instantiate universals (though wateriness rather characterizes other things resembling water). Logically, then, it is the notion of an instance that is correlative to that of a universal, though instances are no doubt usually particulars.
A distinction closely related to that between universals and particulars, and revealing some of the complications in this field, is that between types and tokens, introduced by Peirce. The word ‘in’ appears twice in the present sentence, yet it is only one word. Peirce would call these two appearances in any one copy of the present book, two tokens of a single type. A word as found in the dictionary is therefore a type with indefinitely many tokens (written, spoken, etc.) Only types can be derived from Latin. Only tokens can be illegible. A token may be ambiguous, and then so must its type. A type may be polysyllabic, and then so must all its tokens. The distinction is significant for nominalists, for when they say there are only words and no universals, do they mean types or tokens? Also the distinction is not sufficient by itself, for the words in a speech cannot be types, for types are not limited to a single speech, nor yet tokens, since the same speech, and therefore the same words, can be recorded many times (Cohen). It is disputed how closely this distinction resembles that between universals and particulars. Word as a universal has instances (several hundred on this page); as a type it has tokens (each of just four letters). Also to what spheres, apart from words, is it relevant? Is the Union Jack, or the lion in ‘The lion is carnivorous’, a type or a universal or what? Is the lion a concrete universal? Spheres where the distinction has been used include aesthetics, in the analysis of works of art, and in the IDENTITY THEORY OF MIND. See also REALISM, CONCEPT, IDEA, SENTENCES, TROPE, THIRD MAN ARGUMENT.
R.I.Aaron, The Theory of Universals, Clarendon, 1952, revised 1967. (Universals as ‘natural recurrences’ and ‘principles of grouping’. Some history.)
E.B.Allaire, ‘Bare particulars’, Philosophical Studies, 1963, reprinted with discussion in Loux (above).
Aristotle, Metaphysics, book 7 (or Z), chapters 13–16, Posterior Analytics, book 2, chapter 19. (Cf. also Aristotle references under SUBSTANCE.)
D.M.Armstrong, Universals and Scientific Realism, vol. 1: Nominalism and Realism, vol. 2: A Theory of Universals, Cambridge UP, 1978. (Important modern work. See also his Universals: an Opinionated Introduction, Westview Press, 1989, which is shorter than his earlier work but contains important revisions.)
F.H.Bradley, The Principle of Logic, 1993, book 1, chapter 2, § 4, chapter 6, §§ 30–6. (Concrete universals. Cf. R.M.Eaton, General Logic, Scribner’s Sons 1931, pp. 269–72.)
M.B.Burke, ‘Cohabitation, stuff and intermittent existence’, Mind, 1980. (Material objects can exist intermittently.)
W.Charlton, Aesthetics, Hutchinson, 1970, pp. 27–9. (Types and universals. Relevance to aesthetics. Cf. also R.A.Dipert, ‘Types and tokens: a reply to Sharpe’, Mind, 1980; Sharpe replies in Mind, 1982.)
L.J.Cohen, The Diversity of Meaning, Methuen, 1962, pp. 4–5. (Brief discussion of types and tokens.)
D.Gasking, ‘Clusters’, Australasian Journal of Philosophy, 1960.
N.Goodman, ‘A world of individuals’, in I.M.Bochenski et al.,The Problem of Universals, 1956, reprinted in P.Benacerraf and H.Putnam (eds), Philosophy of Mathematics, Blackwell, 1964 (cf. also ibid., pp. 21–3), and in C.Landesman (ed.), The Problem of Universals, Basic Books, 1971. (Goodman’s nominalism.)
D.K.Lewis, ‘New work for a theory of universals’, Australasian Journal of Philosophy, 1983. (Discuss Armstrong’s earlier work, claiming universals are indeed needed, but for reasons different from Armstrong’s. Difficult.)
M.J.Loux (ed.), Universals and Particulars, Anchor Books, 1970. (Selected readings.)
A.Oliver, ‘The metaphysics of properties’, Mind, 1996. (Extended survey article, with references, on recent views on the nature and role of properties, concentrating especially on Armstrong’s treatment of properties as immanent universals.)
Plato, Phaedo, Republic § 596, Parmenides, esp. down to § 135c. (These are among the important passages. The Parmenides includes what seems to be strong self-criticism, including the ‘third man argument’.)
H.H.Price, Thinking and Experience, Hutchinson, 1953, chapter 1, reprinted in Landesman (above). (Moderate defence of resemblance theory, reconciling it with ‘universalia in rebus’ theory.)
W.V.O.Quine, ‘On what there is’ (see bibliography to BEING). (Offers a criterion for deciding whether there are universals or not.)
A.B.Schoedinger (ed.) The Problem of Universals, Humanities Press, 1992. (General anthology.)
M.A.Simon, ‘When is a resemblance a family resemblance?’, Mind, 1969. (Critical discussion of family resemblance view.)
*H.Staniland, Universals, Doubleday Anchor, 1972. (Elementary introduction, if inevitably a bit dated.)
