Page 64 - Encyclopedia of Philosophy of Language

Language, Metaphysics, and Ontology
Sortal Terms E. J. Lowe
'Sortal terms' are a species of general term falling different sets of identity-conditions, on pain of
within the grammatical category of common or count nouns (or, more generally, count noun phrases). The adjective 'sortal' was apparently coined by John Locke, who used it to describe names for sorts or species of individual substances, and modern usage more or less agrees with this. Typical examples of sortal terms in English are horse, tree,and clock.Some sortal terms, like horse, denote natural kinds and others, like clock, denote artefactual kinds. Yet others denote abstract kinds, for example, the logico-math- ematical sortal terms set and number.
1. The Semantics of Sortal Terms
A distinguishing semantic feature of sortal terms is that the application of any such term is typically gov- erned by criteria of individuation and identity—prin- ciples determining the individuation and identity- conditions of the particular items characterizable by that term. Thus a full grasp of the meaning of the term horse will involve an understanding of what makes for the individuality of any particular horse (how, for instance, the horse as a whole is related to its parts) and what defines the persistence of an individual horse over time. Possession of such an understanding is a necessary condition of a speaker's being able to ident- ify, distinguish, and hence count individual horses. The question of how many individual items charac- terizable by a given sortal term exist subject to some further condition (for instance, how many horses exist in the UK today) is one which always makes sense, even if it cannot always be answered. It should be noted, however, that not all count nouns (nouns which form plurals and admit numerical adjectives) satisfy this condition, and hence that not all count nouns qualify as sortal terms. It makes no determinate sense, for instance, to ask how many things exist in a given room at a certain time, because the noun thing is not governed by any distinct criterion of identity. Such nouns are sometimes called 'dummy sortals.' It is important to realize too that not all sortal terms are governed by the same criterion of identity, a fact which Locke himself clearly appreciated. (Locke argues at length that the sortal terms man and person convey different identity conditions.) However, where the sorts denoted by two sortal terms are related as species to genus (as, for example, with horse and mammal), the terms in question must indeed be governed by
the same criterion of identity, for one and the same individual may be characterizable by both such terms and such an individual cannot be subject to two
contradiction.
2. Classificationsof SortalTerms
Sortal terms may be further subclassified in various ways. So far only syntactically simple sortal terms have been mentioned, but there are also syntactically complex ones, exemplified by such count noun phrases as white horse and tree which sheds its leaves in winter. Some sortal terms which are syntactically simple appear nonetheless to be semantically complex: for instance, mathematician is arguably semantically equi- valent to something like person who studies math- ematics, and is therefore governed by the same criterion of identity as governs the unqualified sortal term person (though whether or not the latter is in turn semantically complex is a matter for debate). Sortal terms like boy and tadpole are examples of what David Wiggins has called 'phased sortals,' because they characterize the items to which they apply only during one phase of their existence. It appears that many but not all phased sortals which are syntacti- cally simple are semantically complex. For instance, boy is arguably analyzable as meaning young male human being but it is debatable whether tad- pole is synonymous with immature frog with gills and a tail, for even if the latter two terms are neces- sarily coextensive this is not obviously something of which competent speakers of English have a priori knowledge.
Sortal terms bear certain close affinities to mass terms like water and gold. Many terms in both classes are alike in designating natural kinds. But mass terms are unlike sortal terms in being dissective (gold is divisible into parts which are themselves gold, but the parts of a horse are not themselves horses) and in not supplying principles of enumeration for their instances. Thus, whereas it makes sense to ask how many horses, say, exist in a given region, it only makes sense to ask how much water or gold does. Even so, a mass term like gold clearly does have a criterion of identity governing its application, for it makes sense to ask whether the gold existing in a region at one time is the same gold as that existing there or elsewhere at another time. Where mass terms crucially differ from sortal terms, then, is in lacking criteria of indi- viduation governing their application. There are how- ever standard ways of creating complex sortal terms out of simple mass terms: for instance, drop of water and nugget of gold. It should also be noted that some general terms are ambiguous in that they admit both a sortal term and a mass term interpretation, as is
42