- Beyond rigidity: The unfinished semantic agenda of naming and necessity by Scott Soames
Soames 1985 opens with the following sentence: ‘Recent years have seen a significant convergence of interest among philosophers and linguists in constructing semantic theories of natural languages’. The present work made me wonder what had happened in the interim. Beyond rigidity (BR) may be excellent philosophy of language, but it contains little that is relevant to the concerns of linguists.
As the subtitle implies, BR’s primary goal is to defend and elaborate Kripke’s (1980) thesis that proper names and natural kind terms are ‘rigid designators’. Rigid designators are referring expressions of which the semantic content is not equivalent to any description; rather, they designate their referents without the mediation of any descriptive conditions.
Moreover, statements of identity between the referents of two rigid designators, if true, are claimed to be necessarily true, even though in many cases they are true a posteriori. The classic example of such an identity statement is 1.
(1) Hesperus is Phosphorus.
Although there is no doubt that it was an empirical astronomical discovery that the heavenly body named Hesperus (also known as the evening star) was the same as the one called Phosphorus (the morning star), Kripke argued that 1 is a necessary truth. Similar arguments are made with respect to statements of identity between natural kinds, for example, Water is H2O.
Ch. 1 of BR summarizes the central theses of Kripke 1980 and identifies two unresolved problems that S seeks to solve. The first is that two sentences that, according to Kripke, express the same proposition may be used to convey different information and express different beliefs. For example, uttering Hesperus is Hesperus is uninformative, but an utterance of the purportedly semantically equivalent 1 is informative.1 The second unresolved issue BR addresses is the claim [End Page 786] that natural kind terms like water, cats, and pain are also rigid designators. S argues that although Kripke noted striking parallels between names and natural kind terms, he did not demonstrate that they shared all of the properties of rigid designators.
S’s solution to the first problem is nicely summarized in the following passage (86):
When speaking of the information carried by an assertive utterance of a sentence in a context, one must distinguish (i) the semantic content of the sentence uttered in the context; (ii) what the speaker says (asserts) by uttering the sentence; (iii) what the speaker implies, implicates, or suggests; (iv) what the speaker hopes or intends his utterance to impart to his audience; (v) what the audience does acquire from the utterance . . . I have argued that (i) is standardly included in (ii), but that in the case of many utterances, (ii) is not exhausted by (i).
Claims about rigid designation are claims only about (i), not about (ii)–(v).
In the course of arguing for this position, S responds at length to some of Kripke’s critics. Among the issues he addresses is the status of partially descriptive names such as New York City and World War I. He treats them as hybrids ‘associated with both a descriptive property Pd and a reference o [which] is determined in part by having the property Pd and in part by the same nondescriptive mechanisms that determine the reference of ordinary nondescriptive names’ (110). Later he asserts: ‘The semantic contents of the vast majority of linguistically simple proper names are their referents alone (or their referents plus the properties expressed by some very general sortals associated with the names)’ (129).
The parenthetical hedge in the preceding quote is presumably meant to cover at least number and gender. But these are by no means the only aspects of proper names that convey information. Smarr and Manning (2002) have shown that it is possible to classify the referents of unfamiliar names with a surprisingly high degree of accuracy on the basis of orthography alone. S might well regard this discovery as irrelevant since it can easily be...