Page 320 - Encyclopedia of Philosophy of Language

Language and Logic
about him (though it is directly about the USA). If Clinton were never to have existed, the proposition that the President of the USA is not tall would still exist and could be stated in the same way; it is not internally related to Clinton; it is not identified by reference to Clinton, but only to the property of being the President of the USA; it might have been true even if Clinton were tall (so long as the President of the USA was not). The proposition that Bill Clinton is not tall, on the other hand, is internally related to Clinton. It is directly about him\ it could not be stated (in the same way) if he never had existed; it is identified by reference to him; it could not be true unless Clinton himself was not tall. Singular propositions are those of this latter kind, which are directly about particular
individuals. General propositions are about no par- ticular individuals, but only about properties. Some paradigm general propositions are that all snow is white, and that the shortest spy is human. Paradigm singular propositions include that the earth is round, and that Nixon was President of the USA. In theories of structured propositions, singular propositions are held to contain the individuals they are about as con- stituents, whereas general propositions contain only properties.
See also: Proposition; Reference: Philosophical Issues. Bibliography
EvansG,McDowellJ(eds.)1982TheVarietiesof Reference. Clarendon Press, Oxford
The type/token distinction, attributed to philosopher C. S. Peirce, is usually explained by example. Thus, there are two possible correct answers to these ques- tions: how many words are uttered in saying 'Home, sweet home'? how many numerals are written in writ- ing '422'? In each case the answer is two if 'types' of symbol are counted, and three if'tokens' are counted. The general terms 'word,' and 'numeral,' then, are ambiguous; they might stand for categories either of types or of tokens.
1. Putative Definitions
There are no generally received definitions of type and token; the closest candidate would be: tokens are simply particulars, types are universals (properties), and a given token is of a given type if it falls under that universal (has that property). In practice, not all properties are such that it is natural to call them types and to call the things that have them tokens (are you a token of the property of being far from the North Pole?); it seems natural only where the property itself (the type) can be classified with a common noun ('home' is a word, '2' a numeral, toothache a sensation), and only where having that property (being of the type) is crucial for the relevance to some inquiry of the things (tokens) having the property.
2. The Distinction Applied to Language
In linguistics, the most frequently discussed types are expressions such as words, phrases, and sentences, and theoretical entities like parse trees. The tokens of these types are (said to be) utterances and inscriptions.
Every utterance or inscription of the word (type) 'now' is a token of that word-type. An utterance is relevant to the study of a language—it is in the language— because it is a token of an expression-type of that language, and perhaps also because it is intended so to be.
It is usual to say that the question of truth arises not for sentence types but for token utterances and inscriptions, since tokens, not types, are associated with specific truth-evaluable claims, thoughts, or propositions. The sentence type 'I am hungry,' is not itself true or false, but various tokens of it are true or false. Types, on the other hand, are associated with lexical meanings and other meaning rules and con- ventions of language. The sentence type 'I am hungry' has a specific meaning in English independent of any particular occasion of its use (there are no actual tok- ens of many or even most meaningful sentence types). Lexical ambiguity involves an expression type (like 'bank') having multiple meanings.
3. Complications for the Distinction
It should be recognized that the type/token distinction is not as clear as it may sometimes seem. Com- plications arise in the case of reused tokens, such as a recording on an answering machine or an 'out to lunch' sign on the door of a business. Here it may be useful to distinguish between an occasion of use of an expression and the physical 'vehicle' carrying the expression on that occasion (the inscription, or the recording); which of these should count as the token is not clear. A token is to be distinguished also from
298
Type/Token Distinction M. Crimmins