Abstract
In this chapter we shall present arguments that have been put forward against certain descriptivist approaches, such as linguistic descriptivism, the classical descriptivist theory of proper names or Reichenbach ’s account of indexicals. These arguments have cast discredit on descriptivism in general and have contributed to promote referentialism to the role of standard theory. This is so also because some of these arguments rely on data that appear to be handled straightforwardly from a referentialist standpoint. In Chapter 8, we shall consider the extent to which these arguments are ultimately telling and whether the data that they put forward can be accommodated by the descriptivist theory, CD, to be presented in due course.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
This is, for example, Kaplan ’s terminology (1989, pp. 557–558). In the terminology of Perry 1979 we could say “the essentiality of indexicals”. Both authors accept the thesis just like Castañeda . For earlier defences of it, see Burks 1948–1949 and Bar-Hillel 1954. See also Burge 1977. The references in Burge 1983, p. 81, are also worth mentioning. It should also be noted that, although Castañeda argued against linguistic descriptivism, he accepted, I would say, a descriptivism of sorts, to the extent that his guises are viewed as doing the job of descriptive contents (see note 6 in Chapter 1, above).
- 2.
More precisely we could say this: There can be a context C in which a token s(i) of a sentence S(I), containing a token i of an indexical I, such as “I”, “here”, etc., expresses a meaning M such that, for any non-indexical sentence S(I/T), which results from replacing the indexical term I with a non-indexical term T, the following holds. If a token s(i/t) of S(I/T) (involving a token t of T) is uttered in C together with s(i), s(i/t) expresses a meaning different from M, even in those cases in which i and t are co-referentially equivalent. And if a token s(i/t) of S(I/T) had been uttered [instead of s(i), or in addition to s(i)] in an otherwise identical context, it would have expressed a meaning different from M, even in those cases in which t has the same referent as i.
- 3.
Kaplan ’s main purpose in proposing the story is to show that “ignorance of the referent does not defeat the directly referential character of indexicals” (Kaplan’s (1989, p. 536) corollary 2). As I see it, Kaplan’s example is best seen as a way to support II.
- 4.
The point forcefully made by Perry with this example was in essence conveyed also by his example based on the sentence “a bear is about to attack me” in Perry 1977, or by Kaplan ’s example based on “my pants are on fire” (Kaplan 1989, p. 533).
- 5.
For the view that linguistic descriptivism is defeated by arguments due to Castañeda , Perry , and Kaplan , see, e.g., the papers mentioned in Burge 1983, p. 81; see also Burge 1977 and Bar-Hillel 1954. We shall see, however, that although these arguments, or at least some of them, are effective against linguistic descriptivism, they are ineffective against the quite different version of descriptivism that I try to support in this book.
- 6.
This example was formulated when Carlo Azeglio Ciampi was the president of Italy.
- 7.
Thus, Castañeda (1980a, 1989, Chapter 5) prefers to use “internal” rather than “ de dicto ” and “external” rather than “ de re ”. As noted above, a de re interpretation of a sentence such as (3) is taken to involve the attribution to the speaker of an existential commitment (in our example, a commitment to the existence of a president of Italy). In contrast, Castañeda uses “external” in a way that does not involve such an attribution. For example, given the common knowledge that The Bible asserts that Moses is the person who received the ten commandments on Mt. Sinai, we can accept an external interpretation of a token of “Tom believes that Moses is a Jew” (imagine that Tom is known to have said: “the person who received the ten commandments on the Mt. Sinai is a Jew”). However, that the token is so interpreted does not imply attributing to Tom a commitment to the existence of Moses. Castañeda’s internal/external distinction is thus meant to be more general and versatile than the traditional de dicto/de re distinction and accordingly I have myself appealed elsewhere to the former rather than to the latter (e.g., in Orilia 1994, 1994a). But for present purposes, we may stick to the better-known de dicto/de re distinction.
- 8.
Similar attacks can be found in Donnellan 1970.
- 9.
For the explicit recourse to an existential clause of the kind, “if X exists”, in stating Kripke ’s argument (“if Gödel exists” in the example we are discussing here) see, e.g., Donnellan 1979, p. 48 and Soames 2005.
- 10.
According to Reichenbach ’s analysis, the expressed proposition is something like |the individual uttering token i is uttering nothing|.
- 11.
Suppose we take, as I am inclined to do, a possible world to be something like a maximally consistent conjunction of propositions. Accordingly, let us say that a proposition is true at a world when it is entailed by the world in question. Intuitively, certain entities, such as Clinton or Paris are contingent, in that they may fail to be constituents of some gigantic conjunction of propositions, w, that counts as a possible world, in which case we may say that the entity in question does not exist in w (vice versa, if it is a constituent of w we say that it exists in w). Consider now a proposition P(x 1, …, x n ) where x 1, …, x n are all the contingent entities that the proposition contains as constituents. We can say that P(x 1, …, x n ) is necessary 1 if and only it all the possible worlds that contain x 1, …, x n entail it. For example, any proposition of the form |Fx ∨ ¬ ␕Fx|, where x is a contingent entity, is necessary1 because any world in which x exists entails it (if we assume classical logic). But it is not necessary2 because a world without x as constituent fails to entail it.
- 12.
The clause “if N exists” is not explicitly provided by Kripke (1980, p. 14), but, we can assume, it is presupposed by him.
- 13.
At least by the speaker who introduced the name, in virtue of the very fact that she introduced the name in that manner. Strictly speaking, the speaker makes recourse to the a posteriori knowledge that she introduced the name in the way indicated, but let this pass, since this is neglected in current discussions of the matter; we can still grant that we are using “a priori ” here in a less stringent sense. Yet, according to Kripke ’s view, this sentence may very well express a contingent proposition (if F is a contingent property; cf. Kripke 1980, p. 14).
- 14.
Actually, it is not clear, given Kripke ’s account, that any proposition is expressed at all, let alone known a priori to be true, if F fails to be (uniquely) exemplified. And it can hardly be taken for granted that the speaker knows a priori that F is (uniquely) exemplified. But let this pass at this juncture.
- 15.
Some referentialists such as Salmon and Soames however do not agree that proper names can be used to provide examples of contingent a priori truths. See Soames 2005, p. 424.
- 16.
Kaplan also thinks that “Dthat ” can provide examples of contingent a priori propositions; Donnellan discusses this in his 1979, p. 56.
- 17.
When Kaplan presents his approach informally the contents of sentences and predicates are indeed characterized as PRPs (just as I have characterized as PRPs the pragmatic meanings of sentence and predicate tokens). However, in the model-theoretic framework he describes these contents as intensions, understood as functions from circumstances (pairs of possible worlds and times) to extensions (e.g., sets of objects or truth-values). Like Braun 2001, I shall follow the former option in outlining Kaplan’s formal framework, since it makes for an easier presentation and nothing crucial for our purposes is missed in proceeding in this way. It should also be noted that in the informal presentation of his theory Kaplan describes characters as rules that allow one to associate contexts to contents.
- 18.
I am simplifying a bit, since Kaplan proposes that in the case of “this” what is endowed with a character is not just the expression, but the expression with the accompanying demonstration, i.e., a pointing by the speaker at an object (Kaplan 1989, p. 527). Moreover, in Kaplan 1989a (p. 582) he suggests that the role assigned to demonstration in 1989 should be given to the “directing intention” of the speaker. Kaplan also says that when there is no demonstratum in a given context c, then “this” in that context refers to nothing (1989, p. 527). In his framework this means that it stands for a certain designated “alien object” (p. 544, # 8 (ii)), which is like the null object assumed in certain systems of free logic . For present purposes, these details are immaterial.
- 19.
This is what we might want to say in describing the formal framework (Forbes 2003), but Kaplan actually allows for a more articulated notion of circumstance (1989, p. 502).
- 20.
It is assumed of course that “if …, then ..”. corresponds to the material conditional, ⊃, and that |x is H ⊃ x is H| is a proposition involving the material conditional and the proposition |x is H| (taken twice) as constituents.
- 21.
It must be noted that, as regards (7) and (8), Kaplan is trading on the intuition that utterers are contingent beings: they exist, but might have failed to exist. From his perspective an individual x exists insofar as it is spatiotemporally located in our actual world, but it may fail to exist in that it may fail to be spatiotemporally located in some other world. I am inclined to view possible worlds as maximally consistent conjunctions of propositions, as mentioned above. From this point of view, a contingent being x is a being that may fail to occur as constituent of a gigantic proposition w that counts as a possible world. Such a proposition w fails to entail the proposition |x exists|.
- 22.
As already noted, I would take a possible world to be something like a maximally consistent conjunction of propositions. From this perspective, that P is not true in possible world w means that w fails to entail P. Moreover, that x exists in w means that x is a constituent of w.
- 23.
Another example of self-supporting sentence is “If I exist, then I am an utterer”. The following sentences are also contingently self-supporting: “either a token has existed now, or it has existed in the past, or it will exist in the future” and “something either exists now, has existed or will exist”. They have been provided by Predelli 2006 in an effort to show that a type-oriented approach is to be preferred to a token-oriented approach, in that the former can correctly classify them as expressing contingent propositions in any given context, whereas the latter is forced to say that any token of them expresses a necessary proposition.
- 24.
A similar example involving a person who believes he is David Hume is provided in Perry 1977, p. 13.
- 25.
Someone might say that a content involving Castor as constituent is narrow as long as the subject of the content is Castor himself (and similarly for Pollux and any other subject). This would be so if we define “narrow content ” as a content that does not involve the existence of any worldly entity beside the subject of the content (rather than simply the mind or mind-brain of the subject, as we saw in § 1.8). But we can develop an argument analogous to Kaplan ’s, involving “this” rather than “I”. Say, Castor and Pollux utter “this is a spider” when both are facing a spider-like shape except that Castor is experiencing a real spider and Castor a fake one. The referentialist can react to this by dropping MN and MSDN and taking Castor’s and Pollux’ tokens of “this” to have as meanings the real spider and the fake spider, respectively.
- 26.
Kaplan (1989, pp. 512ff.) also proposes the following problem for the Fregean account of indexicals. He asks us to consider
-
1
(a) he now lives in Princeton, New Jersey,
on the assumption that (a) has been uttered while pointing at Paul, who indeed lives in Princeton, NJ. Clearly in this case the expressed proposition is true, because the relevant token of “he” refers to Paul. Kaplan then proposes for our attention the counterfactual situation that is exactly as before except that the demonstrated person is now Charles, who does not live in Princeton, has disguised himself as Paul and looks exactly like him. In this case, Kaplan claims, the expressed proposition is false, because the token of “he” refers to Charles. Yet, Kaplan goes on, the Fregean should still say that even in the counterfactual situation the expressed proposition is true. The reason is that, in Frege ’s view, the token of “he” should still be taken to refer to Paul, for it must be taken to have the same sense as in the real situation and thus it must have the same referent. The argument is difficult to assess, because Frege is not explicit about the nature of the senses of indexicals such as those involved in this example. Yet, it must be admitted that, given the abstract nature of Fregean senses, and given the fact that things appear the same in the two situations, it is hard to imagine how the sense of the “he” token in the latter case could be different from the sense of the “he” token in the former.
-
1
- 27.
There were other authors who, shortly before Donnellan ’s paper, raised issues and provided answers very similar to those discussed therein. See Neale 1990, p. 63, for references.
- 28.
This terminology is from Neale 1990, p. 201. Neale refers to Donnellan 1978 and Strawson 1952 in discussing this issue.
- 29.
Kripke ’s disquotational principle, here labelled “KDP”, is of course rather similar to the disquotation principle DP appealed to in § 3.2.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2010 Springer Science+Business Media B.V.
About this chapter
Cite this chapter
Orilia, F. (2010). Why Referentialism Is So Successful. In: Singular Reference: A Descriptivist Perspective. Philosophical Studies Series, vol 113. Springer, Dordrecht. https://doi.org/10.1007/978-90-481-3312-3_4
Download citation
DOI: https://doi.org/10.1007/978-90-481-3312-3_4
Published:
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-3311-6
Online ISBN: 978-90-481-3312-3
eBook Packages: Humanities, Social Sciences and LawPhilosophy and Religion (R0)