The trouble is that ‘connotation’ means much the same as ‘meaning’. The distinction between ‘denotation’ and ‘connotation’ is, though, a matter of some importance to our understanding of meaning. By the process of labelling we denote a dog. ‘Dog’ denotes a certain four-legged canine, etc. But the meaning or connotation may be different whereby we think of the various sensible ways in which the word ‘dog’ can be used. A different 132
example will help. “The present King of France” denotes nothing, but has connotation, because we could easily imagine what the present King of France could be, did he exist. He would be a Frenchman of royal birth, etc.If we talk of ‘meaning’ we talk of many meanings, and in the narrow “logical” sense (Sellars, 1963), we usually mean naming, denoting and connoting. It is of interest in passing to note that Sellars regards some efforts, especially Carnap’s (1942) to formalise meaning, both premature and disastrous. A view which I myself would not accept.Sellars distinguishes five modes of meaning which he calls:
1) Meaning as translation2) Meaning as sense3) Meaning as naming4) Meaning as connotation5) Meaning as denotation One example will suffice to clarify most of the distinctions he wishes to make: 678910
6) ‘Cheval’ means horse 1)7) ‘Cheval’ expresses the concept horsekind 2)8) ‘Cheval’ names Man o’ War, Zev, etc. 3)9) ‘Cheval’ connotes the property of having four legs 4) and we could add:
10) ‘Cheval’ denotes horselike things (i.e. horses) 5) and we must note that it does not name them. If we say ‘horse’ expresses the concept horse, then it could be argued that is the same as translation meaning and is simply as 6) above.Dummett (1975) in talking of a Theory of Meaning says that a Theory of Meaning for an entire language would entail a detailed specification of all the words and sentence-forming operations of the language. This in turn yields the meaning of everything said (or written) in that language. A “detailed
specification” means presumably a formalisation in a metalanguage, or some such equivalent process. He agrees though that this is a mammoth undertaking which is not yet feasible.Then there is the notion of Quine (1960) which is mentioned by Dummett where we translate a language whose meaning is clear into a known language. More practical though this may be, it does not help if we have not already got a language whose meanings are clear.Dummett (op. cit. p 89) comes nearer to a more pragmatic view of meaning when he talks of a theory of meaning as a theory of understanding. Here he seems to be nearer to Quine (1975) when the latter says that until he can aspire to linguistic activity in physiological terms the level of work is that of “disposition to overt behaviour”. As he puts it in opposition to the mentalistic level: The easy familiarity of mentalistic talk is not to be trusted. We shall be returning frequently to these matters, but notice now that there are indeed, as Sellars suggests, various approaches to meaning, although the Quinean notion of a pragmatic view is not obviously on his list.We should also be clear (and this is perhaps symptomatic of Sellars’ omission) that the various people concerned with meaning may have different motives. We shall find, for example, that Davidson in his theories is denouncing a pragmatic view, is seeking a formal foundation for logic independent of behaviour (although he favours a theory of-translation) whereas Quine is seeking meaning in the behavioural context. It may seem strange as a result that Follesdal (1975) says: Davidson’s theory is so closely associated with Quine’s that people fail to see the difference between them. The resemblance, as Follesdal (op. cit. p 37) points out, depends upon them both holding a view of translation in which one should try to maximise agreement (whether assent or dissent in both languages). It is, of course, Quine’s search for agreement in behavioural roots that represents the difference, and one that
Follesdal himself recognises.Both Moore (1953) and Wittgenstein (1969) make the point, which we must constantly bear in mind, and which is to the effect that words like ‘real’ or ‘certainty’ have an ordinary meaning in everyday language which is understood as one might expect “more or less certain” or “beyond doubt” (and this is usually sufficient for the purpose in hand). Such words may also have a special philosophical (or scientific) meaning which will require further explanation. Such explanation usually involves more precision - a narrowing of the surplus meaning permitted by the approximation of ordinary language. We have already said something of this with respect to ‘certainty’ and ‘truth’ (Chapter 5) where Wittgenstein by his usage makes clear the difference.Another way of looking at the distinction over kinds of meaning in terms is to consider that all meaning depends on denotation (this conflates a possible distinction) and we argue that all words (or terms or indeed statements) denote the concepts concerned. It now ceases to matter whether there is in fact a present King of France or not, since we are denoting the concept.On the other hand, we may wish to make further distinctions rather than close them and C. I. Lewis (1946) has suggested a four-fold terminology as applied to terms: 1) The denotation of a term is the class of all actual things to which the term applies. 2) The comprehension of a term is the classification of all possible or consistently thinkable things to which the term could be correctly applicable. 3) The signification of a term is that property in things the presence of which indicates that the term correctly applies, and the absence of which indicates that it does not apply. 4) Formally considered, the intension of a term is to be identified with the conjunction of all other terms each of which must be applicable to anything to which the given term would be correctly applicable. Lewis uses the term ‘extension’ as synonymous with ‘denotation’.One important point to get to grips with immediately is that
terms (words or phrases which may refer to classes of objects or individuals) are not the same as statements (sentences) which express propositions and may appear as subject-predicate sentences in ordinary English, say.Pragmaticists (see Chapter 8) usually have argued that the appropriate unit of meaning is the statement and not the term. But again, before we pursue this thought for ourselves, let us look briefly at other people’s statements on the subject.Meaning takes many forms for C. I. Lewis (1946) and a whole host of types of meaning are cited. His notion of analytic meaning, depending as it does upon zero intension and on synonymity, would certainly fall foul of Quine’s (1953) well-known attack on the two dogmas of empiricism.Pap makes clear that any theory of meaning, like a theory of truth, is a matter of logical analysis and not empirical. So we need a definition or several definitions. As Pap again puts it: . . . the phrase “the concept of meaning” is again misleading, since it suggests that there is only one concept of meaning and that it is just a question of finding it.