1 Some people use the term ‘reference’ to refer, not to a relation which holds be· tween some of what is said, or some of a speaker's words, and certain objects in the world, but rather to a relation which holds between, e.g., parts of English, or certain proper names and some objects. I do not mean here to attack such uses of the term. But the present discussion is restricted to uses fitting our initial characterization.

2 By an imperfect aspect, I mean any form of a verb in which an action is represented as in progress or incomplete- e.g., ‘is running’ or ‘was running’ vs. ‘ran'.

3 It is an odd idea that having the right intentions is all that is needed for making a given reference. The presumable locus classicus for such a view is Grice's 'meaning’ (Philosophical Review, 66 (1957)). But here things are not as clear as they might be, since Grice's ‘mean _nn’ appears to swim between ‘intend', 'mean to say', and, occasionally, ‘say’ ('mean’ isn't a very happy term in any case for describing what there is to understand in speech). However, Grice does say, ‘“A meant _nn something by X” is (roughly) equivalent to, ‘A intended the utterance of X to produce some effect…', and also, ‘“X meant something” is (roughly) equivalent to “Somebody meant _nn something by X.'” One might say, e.g., ‘“The dog” (on that utterance) meant something such that the referent, if any, is …', or simply ‘“The dog” (on that utterance) meant (referred to) Ernest.', although, again, these are not the most perspicuous ways of describing references. Using Grice's suggested rough translations yields an intentional theory. That is one interpretation of Grice, and I think one that some have adopted.

4 It might be objected that one can't have such bizarre intentions as, e.g., the intention to refer to the Rottweiler in Arnhem, or to do so by pointing at the Afghan because, say, intending to do something requires believing that the intention can be carried out. But, of course, Sam might also have such bizarre beliefs. He might Just believe that Max will be taking him to be talking about the Rottweiler, or even that pointing at the Afghan is a generally recognized way of getting oneself to be so understood. That doesn't change the case in any relevant way. Having false beliefs about how one goes about saying something doesn't prevent one from availing himself- albeit unwittingly- of the means that in fact exists. If with shaky grasp of the language, I say of someone, ‘Han er lige kommet med firtoget', taking myself to be saying that he is a witty fellow in mixed company, what I have actually said, most likely, is literally, that he's Just arrived on the four o'clock train, and, figuratively, that he's a hayseed.

It should also be remembered that it is of the nature of philosophical examples to be underdescribed. If my view of reference is correct, it is probably always possible to extend a description I give to a description of some case where intuitions are the opposite of what I say they ought to be. Sam's craziness, for example, is compatible with his referring to the Rottweiler in pointing to the Afghan, in some possible case. What one needs to try to convince himself of is that my descriptions do fit easily imagined cases where what I say about what is referred to would be correct. These ought to be cases without much more in the way of relevant detail than what I give.

5 cf. Austin, J. L., How to Do Things with Words (Harvard U.P., Cambridge, Mass; 1962) pp. 105–106.Google Scholar

6 We can also say that he intended to refer to the dog in the yard, which he believed to be Fidel.

7 One source for such ideas is Strawson's Individuals ((Methuen; London, 1964), cf. esp. pp. 181·183). Making small allowances for terminology, Strawson's idea is that X is the referent of a reference made in uttering an expression E Just in case there is an object the speaker has in mind, there is an object his audience takes him to have in mind, the object satisfying the first condition is identical with that satisfying the second, and, further, is identical with X. For the speaker to have an object in mind, ‘There must be some description he could give, though it need not be the description he does give, which applies uniquely to the one he has in mind and does not include the phrase ‘the one I have in mind' (p. 182), so, ‘There must be some true empirical proposition known, in some not too exacting sense of this word, to the speaker, to the effect that there is Just one particular which answers to a certain description.’ (p. 183) Presumably, he thinks he has something in mind if he believes such a proposition, and that which he has in mind, if anything, is that of which this belief is true. Mutatis mutandis, according to Strawson, for the hearer. So actually Strawson has a rather complex theory in terms of the combined beliefs of speaker and hearer. What the examples in the text show, if correct, is that making a reference to (i.e., talking about) something does not require Strawsonian match between speaker and hearer mental states, and where reference does take place without such match, what is referred to is sometimes not what the speaker has in mind, some· times not what the hearer takes him to have in mind. All this is consistent, of course, with the view, expressed to me by Pieter Seuren, that in communication the aim is for the hearer to reconstruct what the speaker had in mind. What is important vis-à-vis what is said, though, is what the speaker ought to be taken as having in mind, and it is when the hearer has successfully reconstructed this that understanding of what was said, if not of the speaker, has taken place.

8 Just in case this view of the matter seems unproblematic to anyone, here is one suspicious thing about it: As a matter of fact, philosophers who have been inte· rested in classical theories of reference haven't, in treating them, displayed any noticeable interest in empirical psychology - that is, in studying the strategies people in fact use in identifying references or referents- strategies which are all too likely to be fallible, or yield no result in cases where there are results to be yielded. What philosophers are interested in describing is the results that ought to be yielded by an ideal reference (or referent) identification device. If the fini· tude of our actual capacities entails the finitude of such an ideal device, it re· quires considerably more argument than I have ever seen or heard to establish that fact.

9 “Speaker's Reference and Semantic Reference”, in Midwest Studies in Philosophy, 2 (1977).

10 Op. cit., p. 264.

11 Op. cit., p. 263.

12 Kripke tries to saddle Keith Donnellan with having talked about speaker's reference, at least as of 1966 ('Reference and Definite Descriptions’). My own view is that what Donnellan was actually talking about was reference, or at least that that is what most of what he says is true of. But again, such issues needn't be fought out here.

13 In particular, they militate against a certain popular view of what semantics might be, expressed, for example, by John McDowell ('Truth Conditions, Bivalence and Verificationism', in Evans, Gareth and McDowell, John (eds.), Truth and Meaning (The Clarendon Press; Oxford: 1976)Google Scholar who says, ‘The idea is that a theory of sense and a theory of force, in combination, should enable one to move from a sufficiently full description of a speaker's utterance, uninterpreted, to a description of his performance as a propositional act of a specified kind with a specified content, that is, a description on the pattern of ‘He is asserting that p.', 'He is asking whether p.', and so on.’ (p. 144). There have also been attempts to beg off this part of the project, though it is doubtful that they succeed in pre· serving the sorts of theories they are designed to favor.

14 Of course, in some cases other things might determine such conditions- where, e.g., what we are prepared to recognize is oak trees or Bassett hounds.