Page 548 - Encyclopedia of Philosophy of Language

Key Figures
role. The other thought is that the right way to explain their semantic role is to say how they contribute to the meaning (not here equated with reference) of whole sentences in which they occur. His 'On Denoting' (1905) contains a rather creaking attempt to provide standard disquotational clauses for quantifiers.
Russell allows that some denoting phrases denote. For example, a definite description 'the F' denotes if and only if something is uniquely F. However, a phrase may be a denoting one without denoting: it is not an expression whose semantic role is to denote. If 'referring expression' means an expression whose semantic role is to refer, Russell's theory of descrip- tions places definite descriptions outside the category of referring expressions.
The theory is a major step towards permitting the identification of meaning and reference, subject to two restrictions: it is to apply only to genuinely sem- antically simple expressions, and it is not to apply to the logical constants. This provides solutions to the two problems mentioned earlier. Since the semantic complexity of definite descriptions ensures that they acquire their semantic role derivatively, they are not to be thought of as expressions whose role is to refer to or indicate entities, so there is no problem about meaningful but nonreferring descriptions. Moreover, since definite descriptions are no longer classifiable as referring expressions, there cannot be 'coreferential' definite descriptions, in the way that seemed to ensure synonymy. Of course, there can be codenoting descriptions, but their different semantic structure
explains their nonsynonymy.
2.2 The Theory of Names
Both the original problems resurface in connection with apparently simple expressions. Some, like 'Vulcan,' are meaningful yet do not refer; and some pairs, like 'Hesperus' and 'Phosphorus,' are appar- ently coreferential while apparently nonsynonymous. Russell's solution is to deny that these expressions really are semantically simple. Rather, they are 'trunc- ated' or 'abbreviated' definite descriptions, and the two problems disappear.
However, questions arise about what it could mean to say that a name like 'Vulcan' is 'really' complexand about how could it be settled, given the idiosyncratic nature of the information that people possess about individuals, which description a name abbreviates. Russell gives clear answers to these questions, answers which involve a modification of one standard interpretation of his views.
In his Problems of Philosophy (1912), his discussion of names is guided by the 'principle of acquaintance,' according to which one can understand an expression whose semantic role is to refer only if one is acquainted with its referent. Since he held that the only things with which we can be acquainted are sense data and
(perhaps) ourselves, this principle places severe limi- tations, derived from a source quite different from the two problems mentioned, upon which expressions can be counted as genuine names.
Russell takes the example of the name 'Bismarck,' which Bismarck himself can use as a genuinely sem- antically simple expression, but which his friends and anyone else cannot, since he is not a sense-datum. When people 'make a judgment about him, the description in our minds will probably be some more or less vague mass of historical knowledge... for the sake of illustration... "the first Chancellor of the Ger- man Empire."' The problem is that intuitively we want a stable role for 'Bismarck,' common to speakers and hearers in successful acts of communication, yet this cannot be provided by the idiosyncratic and vari- able descriptions we associate with a name. Russell's solution is to stress that the story about associated descriptions is intended only to give a correct account of what is in the mind of an individual speaker or hearer. In order to achieve a correct account of com- munication, one must see Bismarck himself as the common object of the judgments at which speakers are aiming:
What enables us to communicate in spite of the varying descriptions we employ is that we know there is a true proposition concerning the actual Bismarck, and that however we may vary the description (so long as the description is correct) the proposition described is still the same.
(1912:31)
A major question in the philosophy of language is how to relate a notion of meaning appropriate to individuating units of communication (what is shared by a speaker and a hearer when they understand one another) and a notion of meaning appropriate to describing the states of mind of individuals when they think, speak, and understand. Russell's theory that many names are really truncated descriptions is clearly located by him as belonging to the latter enterprise and as unsuited to the former. So the many contemporary criticisms of Russell (e.g., Kripke 1980) which take him to hold that for each proper name there is a definite description that specifies the name's invariant contribution to communicative acts have not correctly identified their target. The most with which one could credit Russell is the view that on each occasion a proper name is used, there is a description that accu- rately represents what is going on in the mind of the user of the name on that occasion.
The difficulty with this weaker view is that it is unclear whether or not it speaks to the original prob- lems of apparently meaningful yet bearerless names, and of apparently coreferential yet nonsynonymous ones. Perhaps the tendency to think of Russell's view as stronger than the texts justify is to be explained by the thought that only a stronger view could hope to resolve these problems.
526