1 If you prefer, say, ‘or different contributions made to what was said in (the course of) speaking the words.’

2 Several caveats had better be entered here. First, some have thought to evade the consequences here by stating some condition for the truth of a sentence which is not per se fulfilled or not by any particular state of affairs, but which is fulfilled by some state of affairs or other relative to some particular speaking of the sentence in which that is what was said to obtain. This strategy is discussed more fully in Travis, The True and the False: The Domain of the Pragmatic (Amsterdam: Benjamins 1981). Second, there is a way of avoiding the consequences which has been suggested, or anyway presupposed by David Kaplan, in largely unpublished writings. The way, e.g., in the case of a sentence, is to assume that what a sentence means determines a function from facts of (potential) speakings of it to what is (would be) said (to be so) in those speakings. Such a function, or what fixes it, is called a character by Kaplan. The thought would be, then, that meanings, e.g., of sentences could be given by describing a function from speakings to truth conditions on a speaking, i.e., by describing when the sentence would be true as spoken on such and such a production. Of course, that such functions exist, and that describing them could be saying what expressions mean are both quite heavy assumptions about the nature of the linguistic data. Over what exactly, for a start, are such functions supposed to range? I.e., what facts exactly about speakings determine the thought expressed in them? As our story continues, I hope that these assumptions will appear as highly implausible, given the data suggested, notably, in Austin's work. But this will not be argued in detail here.

3 Or show positively that there is no way of settling them - no fact of the matter.

4 Austin, J.L. ‘A Plea For Excuses,’ Philosophical Papers, 3rd edition (Oxford: Oxford University Press 1980), 185Google Scholar (footnote)

5 Such a view was encouraged by overconcentration on one small aspect of the S-view, namely, talk about illocutionary forces. The idea that this might be treated as a technicality was apparently engendered by E.J. Lemmon (‘On Sentences Verifiable By Their Use,’ Analysis, 22 (1962) 86-9.) For a classic attempt at such a treatment cf. Lewis, David ‘General Semantics,’ in Davidson, D. and Harman, G. eds., The Semantics of Natural Language (Dordrecht, Holland: D. Reidel 1971).Google Scholar

6 The locus classicus for the ideology is Donald Davidson, ‘Truth and Meaning,’ Synthese, 17 (1967) 304-23.

7 An exception to the lack of questioning has always been Michael Dummett (cf. ‘What is a Theory of Meaning?’, parts I and II, for example). Dummett's disagreement with the dominant view, however, is within a framework in which the S-view is already left largely out of consideration. The main issue is which term of assessment for sentences - ‘true’ or some other - is most deserving of the attentions of a semantic theory.

8 This was the title of Grice's William James lectures, given at Harvard in 1967.

9 For an example of such an appeal, cf. Kripke, Saul ‘Speaker's Reference and Semantic Reference,’ in French, P.T. et al, eds., Midwest Studies in Philosophy, Vol. 2 (Minneapolis: University of Minnesota Press 1977) 255–76.Google Scholar For discussion, cf. Travis, ‘Reference, Speakers and Semantics,’ Language and Communication, Vol. 1 (1981).CrossRefGoogle Scholar

10 Grice, H.P. ‘Logic and Conversation,’ in Cole, P. and Morgan, J.L. eds., Syntax and Semantics, Vol. 3 (New York: Academic Press 1975) 43–4Google Scholar

11 Grice, ‘Logic and Conversation,’ 44Google Scholar

12 Grice, ‘Logic and Conversation’, 50Google Scholar

13 ibid.

14 Grice, ‘Logic and Conversation’, 45Google Scholar

15 Grice, ‘Logic and Conversation’, 50Google Scholar

16 Grice, ‘Logic and Conversation’, 44Google Scholar

17 Qualification: The words W must, of course, be used to mean what they do mean in the language in question, and ambiguities in W in the language – as, e.g., in ‘vice’ in British, though not American English, must play no role. Such qualification will simply be assumed in what follows.

18 Perhaps Smith's weight is like (Austin's example) the shape of a cat. ‘What is the real shape of a cat?’, asks Austin (Sense and Sensibilia [Oxford: Oxford University Press 1962]67), ‘Does its real shape change whenever it moves? If not, in what posture is its real shape on display?’ The questions are not meant as invitations to seek a unique right way of providing answers.

19 One might, of course, try maintaining that, say, ‘_weighs 80 kilos’ is ambiguous between different senses having to do with different circumstances of weighing.

20 But cf. Travis, Saying and Understanding (Oxford: Basil Blackwell 1975), chapter 3, and Travis, The True and The False.

21 Alright, so Smith wears heavy clothing.

22 cf. also footnote 19.

23 For example, instead of imagining Smith stepping into the elevator after his encounter with Melvin, we could equally well imagine him having done so instead of having an encounter with Melvin. At any rate, nothing about times of utterances plays a role here in generating the contrasts.

24 Again, meanings, time of utterance and referents fixed.

25 Grice's candidate needn't decide every such question, of course - only a goodly number of them. Where it does not decide, Grice is simply ruling against our intuitions in the situations where, e.g., we would call the kettle black even though it is only soot, or even though it has been enamelled (e.g., where we are just about to strip the enamel off.) And one wants to see an argument for the ruling. It is just as much in need of specific motivation, and just as much a dismissal of our intuitions as coming down squarely on one side or the other.

26 One must be careful here about the extent to which we are obligated (in point of language) to follow the scientist, or even to choose either to do so or not. For the scientist may well have special purposes in mind, and we may follow him when, but only when our purposes coincide with his. Note, e.g., that when a doctor hands a patient a recommended height and weight chart, he, or the chart will very likely specify various things (surely not everything!) about how the weighings are to be done for purposes of using the chart. That such specification is needed is a good sign a) that that which is specified is not part of, or determined by the meanings of the words, ‘weighs kilos', and b) it very well is part of what is being claimed in using the chart (e.g., ‘Your ideal weight is 70 kilos.’)

27 Perhaps quite a vague one, as in, ‘He weighs a lot.’

28 Some might insist here on a distinction between weight and mass – one alleged to distinguish the English and metric systems. For scientific purposes this is no doubt correct. I doubt that any such distinction shapes the ordinary notion of kilo, as used daily, say, throughout continental Europe. (If so, would it also inform e.g., the Italian ‘etto’ or the Dutch ‘pond’ and ‘ons’ [defined, respectively, as 500 and 100 grams, but used exclusively and unofficially for marketing].) Nor need it, on the S-view presented here. For some speakings of ‘kilo,’ such a distinction may be relevant to working out what something was thereby said to be, but it will not follow that it is therefore thus relevant for all speakings of the word. Neither policy need be built into the (normal) meaning of the word - despite what physicists may do.

29 Maybe you wouldn't call it a color. But it isn't always wrong to do so (e.g., ‘This shirt comes in seven colors’). And it isn't wrong here. Sometimes it might be. Again, no single policy is called for.

30 There are the inevitable and omnipresent gaps between one's intentions and what one actually says. Through mishap one might, on occasion, say something other than what he meant or thought. But if that happened, that would be through mishap indeed, and not because the words in question just weren't suitable for expressing one's thoughts tout court.

31 There is another way of slicing up what is said, and sometimes, though not always, there may be an element of free choice as to which way one chooses, One might detect, e.g., one thing which is said in all speakings of ‘The kettle is black.’, and which, furthermore, gets evaluated as true or false, but which may be true as said in some speakings, not true as said in others. There are certain attractions in this way of talking, though it must be noted that it is another alternative which Grice finds distinctly unattractive. On this approach, to be in a position to determine whether someone spoke truly or not, one would have to be told not only what was said, but also all other relevant facts about the speaking. The problem is, just what would one have to be told about a speaking? Perhaps for a given occasion, we could be told or shown what it would be reasonable to take being black to come to, in advance of or independently of the speaking. But if one wanted to defend our way of speaking of saying in the text against this alternative view, what one might try to show is that in general no fact about the occasion would be adequate for determining whether to count what was said on it as true if it stopped short of the facts about how the words as uttered were to be understood. Fix, e.g., a time, place and an audience with given mental states and there are still a variety of distinct things which might be said. Where this can be shown, it would seem to be straining at a gnat not to allow that such facts about how words were to be understood are facts about what was to be understood to have been said in them. Be such things as they may, neither way of carving up sayings is compatible with Grice's way of distinguishing saying from implicating.

32 cf., however, Travis, Saying and Understanding and The True and The False for further discussion.

33 The ‘part of’ metaphor, I realize, is annoying. The following section, I hope, will help clarify it.

34 Logic and Conversation, lecture 2.

35 Logic and Conversation, lecture 3.

36 Logic and Conversation, lecture 1.

37 On the difference between the options Grice offers, see footnote 31. The term ‘statement,’ of course, is always liable to be a trick one. But a third option open to the S-viewer - one which seems a live possibility in the case Grice discusses (involving trying to cash a cheque) - is that in one speaking of the sentence or another, no statement was made at all - i.e., nothing said of a sort which might be either true or false.

38 You might also put it in other ways, One thing that will not do: ‘She said that Smith was putting on weight.’ For the purposes at hand, who is Smith? What you might do is preface your remark with a story: There's this guy who Odile and I know who,., and she said that he is putting on weight.’ But note: it wasn't said that (the story) and without that, you wouldn't have said what was said, Further, you haven't yet specified what may be an important part of what was said, namely that she called him George. That could be done too. But then, it is necessary somehow to fix how her speaking of ‘George’ is to be taken (understood).

39 Cf. Travis, ‘Reference, Speakers and Semantics’ for further discusssion.

40 For the scholarly it may be helpful to provide a full list of the examples cited by Grice in the William James lectures. Aside from the Ryle example, these include:

1. a.

   a. Norman Malcolm (Philosophical Review 1949) ‘accused Moore of having misused the word “know” when he said that he knew that this was one human hand and that this is another.’ The problem: lack of a problem about why or how it might fail to be a hand.

2. b.

   b. Benjamin (Mind 1956). ‘X remembers’ implies there is some special effort made or difficulty about it.

3. c.

   c. Wittgenstein: One does not ‘see a knife and fork as a knife and fork.’ Suggestion: it must be a somewhat abnormal way of looking at the thing in question.

4. d.

   d. A common point: When you're staring at the pig, full frontal in the barnyard, you don't normally say, ‘It looks to me as if it/there is a pig.’ or ‘There seems to be a piglike shape ahead.’, or ‘The evidence of pig is mounting.’ and so on. Grice refers to more Wittgensteinian versions of the same sort of point.

5. e.

   e. You can't say, ‘X tried to do A.’ unless there was some (presumed) difficulty for X in doing A.

6. f.

   f. ‘carefully’: Someone who is unreasonably neurotic about something (a teller recounting the banknotes 15 times before handing them to the customer) is not properly described as ‘being careful.’

7. g.

   g. Austin's maxim: ‘no modification without aberration’. ‘It is bedtime, I am alone, I yawn; but I do not yawn involuntarily (or voluntarily!) nor yet deliberately. To yawn in any such peculiar way is just not just to yawn.’ (‘A Plea for Excuses.’ Philosophical Papers, 3rd edition, 190)

8. h.

   h. ‘expressions which are candidates for being natural analogues to logical constants, and which may or may not diverge in meaning from the related constants. E.g., one doesn't say ‘He got into bed and took off his trousers.’ where this reverses the actual temporal order.

9. i.

   i. ‘true’ is sometimes claimed to serve a performative function: “to say” it is true that P” is not just to assert that p but also to confirm, concede or agree to its being the case that p.’ Note the intention here, viz., to deny the redundancy theory of truth, a general drift with which one might have a lot of sympathy without agreeing to the above details. (Such could well be a typical feeling about all of the above cases.) For my own views on truth and redundancy, cf. Travis, The True and The False.

41 cited in Grice's first lecture. The passages from Ryle are in The Concept of Mind, 67-8.

42 Not that there need always be something precise to be understood on this point. Understandings are generally tailored to fit our needs at the time, and rarely extend much further - one of the troubles with the curious idea that predicates (of a language) have extensions. But then, quite vague understandings are also specifiable.

43 Austin, J.L. Sense and Sensibilia, 3.Google Scholar

44 Austin, Sense and Sensibilia, 118.Google Scholar

45 On this point, cf. Travis, ‘Reference, Speakers and Semantics.’

46 Grice, ‘Logic and Conversation’, 56.Google Scholar

47 It has also often been remarked that should this in some instance fail to be the case, that should by no means increase our confidence that an action was done voluntarily.

48 The particular examples (after Grice) might also quite reasonably suggest that no action was performed at all.

49 Of course, I am skipping over the numerous interesting grammatical distinctions to which these philosophers also pointed - e.g., the fact that ‘involuntary’ does not mean ‘not voluntary.’ I do not mean to minimize the importance of such points.

50 Note, however, counter to what Ryle (and to a lesser extent Austin) sometimes suggest, that what we are looking for, to fix the thought expressed, are features of the expressing of that thought - i.e., features of a (context of a) speaking - and not features of the deed as such. We regard the deed as suspicious - ‘Why would old Hugo have done that when it's well known that he … ’ Thus our interest is focussed on certain aspects of things, so that when we now express a thought to the effect that he did it voluntarily/under duress/ … , we express something which clears up our problems on a given point, and thus for which there are given things which would count for or against its truth. But of course on some other occasion of wondering about Hugo's deed, given our interests and attitudes there, the whole thing may be perfectly unsuspicious - or suspicious for quite different reasons.

51 Assuming, of course, that you have not been too calloused by doing philosophy.