Page 250 - Encyclopedia of Philosophy of Language

Reference
the functioning of proper names. Frege's (1892) solu- tion is to introduce the idea of 'sense' to supplement that of reference. The sense of a proper name, he argued, contains the 'mode of presentation' of the referent; the reference of a name is the object it denotes. Thus two different names might refer to the same object but possess different senses, i.e., present the object in different ways. That, Frege claims, is precisely what accounts for the 'cognitive' difference between identity statements of the kind Dr Jekyll is Dr Jekyll and Dr Jekyll is Mr Hyde.
The introduction of sense as well as reference pro- vides a potential solution also to the problem of fic- tional names: these are names with sense but no reference. Although Frege appears to endorse that solution in Frege (1892), it is a matter of controversy whether it is consistent with his wider semantic theory (Evans 1982: ch. 1).
Perhaps the most elegant application of the distinc- tion comes in Frege's treatment of reported speech and the contexts of prepositional attitude verbs. Frege's simple but important observation is that speakers do not always use names (or signs in general) to speak of their 'customary reference,' i.e., the objects they standardly refer to. Sometimes indeed they use them to talk about the names themselves: in the sentence 'Frege' contains five letters the name 'Frege' is not being used to refer to Frege the man. Sometimes names are used to refer to their 'senses,' not to their customary objects of reference. In reported speech the aim is to capture the sense of another person's remarks. Thus when someone reports John said that Frege was clever the name 'Frege' is being used not with its customary reference (i.e., Frege the man) but with an 'indirect reference' (i.e., the customary sense of the name 'Frege'). Likewise when someone reports what John believed or thought, the names within the context of the attitude verbs lose their customary ref- erence and acquire their indirect reference. This account neatly explains why names with the same denotation cannot always be substituted—'Tully' and 'Cicero,' for example—in all contexts and truth-value preserved. The names 'Tully' and 'Cicero' have differ-
ent senses and in those contexts where the senses are being referred to, substitutions would be expected to make some difference.
One difficulty with Frege's theory is stating more precisely what the sense of a proper name is supposed to be. Frege offers only general remarks about the notion. An important feature is that sense is objec- tive—it is public, graspable by different speakers, and does not reside merely in an individual's mind. In this it contrasts with subjective 'ideas' associated with names. How can the sense of a name be specified? One way—though it is far from clear that Frege thinks this is the only way—is to provide a definite description which indicates the means by which the reference is identified. It might be, he suggests, that different
people attach different senses—thus different descrip- tions—to the names they use. His own example is Aristotle which might have the sense of the pupil of Plato and teacher of Alexander the Great or the teacher of Alexander the Great who was born in Stagira or some other descriptive content. Frege thought it was an imperfection of natural languages that single names should have such variations of sense. Nevertheless, his account of the sense of ordinary proper names (like Aristotle) has been the basis for a tradition of 'descrip- tivist' theories of names, which is returned to in Sect. 7.
4. Russell on'Logically Proper Names'
Curiously, although Bertrand Russell (1905) rejected Frege's account of sense and reference, he ended up with a theory of ordinary proper names which also belongs in the 'descriptivist' tradition. In manyways Russell is closer to Mill than to Frege. He took over several of Mill's central ideas about names but dropped Mill's terminology—'connotative terms,' 'implying attributes,' etc.—and revolutionized the subject matter by bringing to bear the apparatus of logical analysis. Like Mill, and unlike Frege, Russell postulated a fundamental distinction between proper names and definite descriptions and held that proper names have a purely denotative function. Unlike Mill, though, he argued that ordinary proper names—Ari- stotle, Dartmouth, New College—are not genuine pro- per names at all but 'disguised descriptions'; and he provided a logical account of definite descriptions, which became a paradigm of logical analysis, precisely to show how they differed from genuine names.
According to Russell, a genuine or 'logically' proper name has the most direct possible relation with the object that it names. He called such names 'simple symbols,' in the sense that they could not be defined, and their sole semantic role was to denote or stand for some unique, and simple, object. The very meaning of a logically proper name, on this conception, is the object it denotes, such that should it turn out there is no such object then the name would be meaningless. Russell's austere conception of a logically proper name was partly motivated by the constraints of an ideal or 'logically perfect' language and partly by an epistemological distinction between knowledge by 'acquaintance' and knowledge by 'description.' (A similar view of names is found in Wittgenstein 1961, though withoutthe epistemological element.) Humans are acquainted, Russell believed, only with the immediate objects of awareness: individual sense-data principally, but perhaps also certain universals, and the self. These alone can be the referents of logically proper names. Knowledge of other things—all com- plex objects—like portions of matter, persons, other minds, is knowledge by description. Clearly, then, such objects as Aristotle or Dartmouth, being complex, can be known only by description not by acquaintance and cannot be directly named. The
228