Page 249 - Encyclopedia of Philosophy of Language

is no contradiction in the statement New College is not new;in the former the description 'new college' is active semantically, bringing the attribute of newness into the truth-conditions, while in the latter the descriptive content in 'New College,' used as a proper name, is semantically inert. The phenomenon seems to confirm Mill's basic intuition that proper names have a different semantic role from definite descrip- tions.
2. Problems for the Pure Denotation View of Proper Names
However, Mill's view, as applied to ordinary proper names, faces a number of serious difficulties. The first, originating from Frege (1892), concerns identity state- ments. Consider the two statements:
nothing. However, what if the name 'Jonah' does denote someone, a barkeeper, say, in Cleveland, Ohio: does that make it false that Jonah did not exist! The obvious response is that it is not (hat Jonah who is being referred to. But in determining which Jonah is the relevant one some kind of connotation or con- ceptual content seems to be required for the name.
Another related problem comes from the use of proper names in fiction or mythology. Speakers seem to have no difficulty using names which have no actual denotation: Pegasus is the winged horse of Greek myth- ology, Sherlock Holmes is the famous detective, etc. These sentences are readily comprehensible. An unsat- isfactory solution is to postulate a special class of objects—fictional objects—as the referents of such names; apart from the suspect ontological impli- cations, this move is in danger of trivializing Mill's theory, in stipulating an arbitrary denotation for every problematic name. Another move might be to suggest that fictional names denote objects in the mind. But that threatens to produce counterintuitive truth assessments: the seemingly true statement Pegasus does not exist would become false if 'Pegasus' refers to something real—an idea—in the mind and Sherlock Holmes is a detective, which seems to have some truth, would also turn out false because being a detective could not be true of anything mental. These common uses of proper names do not fit the model of naming as pure denotation and demand careful logical analysis.
A fourth and final problem for Mill's theory, which like the first emanates from Frege, centers on the use of proper names in nonextensional contexts. So-called 'prepositional attitude' verbs—hope, believe, fear, think, dream, etc.—are thought to generate such con- texts. So, to use a standard example, consider the statements:
John believes that Cicero denounced Catiline. (2a)
Cicero is Tully. (2b)
If both are true, it would seem on the pure denotation view that (2c) could be inferred without further ado:
John believes that Tully denounced Catiline. (2c)
In other words if the name 'Cicero' in the first state- ment has only the semantic role of denoting Cicero then any other name with an identical role, like 'Tully,' should be substitutable without affecting the truth- value of the whole. The trouble is the inference does not seem to go through; it seems quite conceivable that the last statement might be false—where John has never heard the name Tully'—even though the first two are true. If that is so, then denotation cannot exhaust the semantic role of the name 'Cicero' in this context.
3. Frege on the Sense and Reference of Proper Names
What these problem cases suggest is that something more is needed than just denotation to account for
Dr Jekyll is Dr Jekyll.
(la)
DrJekyllisMrHyde. (Ib)
Ignoring the fictional nature of the example, it seems
clear that the statements have different 'cognitive value,' to use Frege's term; the first is trivially true, while the second conveys significant information. Yet if Mill is right that the sole function of a proper name is to denote an object, and if the names 'Dr Jekyll' and 'Mr Hyde' do denote the same object, then there could be no difference of meaning between the state- ments. Nor does it seem satisfactory to claim that such identity statements are really about names themselves, not about the objects denoted. Perhaps the statement New York is The Big Apple might be construed as about the name 'The Big Apple' (i.e., stating that New York is also called 'The Big Apple') but there are other cases, like Shakespeare is Bacon, where attention is clearly directed to the identity of a person, not to either of the names per se; furthermore, the reader is being told something other than just that Shakespeare is identical to himself.
A second problem concerns existence attributions. It is common to raise existence questions or make existence claims using proper names: Did Homer exist!, Jonah did not exist, Carlos the Terrorist did exist, and so on. But on the view that the sole function of a proper name is to denote an object, it is hard to see what sense could be attached to such attributions. All negative existence statements using proper names would be at best self-contradictory, at worst com- pletely meaningless, yet as the examples suggest they need be neither. All positive existence statements using proper names would, again contrary to the evidence, become tautological, and questions about existence answered trivially. This strongly suggests that there is more to the meaning of proper names than just denotation. Perhaps, in defense of Mill, the statement Jonah did not exist should be construed as being a metastatement, not about the man but about the name, i.e., as stating that the name 'Jonah' denotes
Names and Descriptions
227