We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 01 January 2020
Context plays a crucial role in our propositional attitude reporting practices. A belief-reporting sentence (containing no obvious indexical elements) which seems true in one context may seem false in another, as Kripke showed us in ‘A Puzzle About Belief.’ To put it a bit sloppily, (1) may seem true when we are discussing Peter's beliefs regarding Paderewski-the-pianist and false when we are discussing his beliefs regarding Paderewski-the-statesman.
(1) Peter believes that Paderewski is a fine musician.
A number of recent theorists have taken this contextual variation very seriously, and attempted to provide accounts which accommodate it in one way or another. Salmon and Soames accommodate the variation in pragmatics, arguing that our intuitions about attitude reporting often result from mistaking pragmatic implicatures for semantic content. Other theorists have offered semantic accounts which yield contextually varying truth conditions for belief reporting sentences. I have a great deal of sympathy for the project: I think the contextually varying intuitions do exist, and that they need to be accommodated, in one way or another. I am not so sure, however, that this can be done.
1 Kripke, S. ‘A Puzzle About Belief,’ in Propositions and Attitudes, Salmon, N. and Soames, S. eds. (Oxford: Oxford University Press 1988)Google Scholar
2 Salmon, N. Frege's Puzzle (Cambridge, MA: The MIT Press 1986)Google Scholar; S. Soames, ‘Direct Reference, Propositional Attitudes, and Semantic Content,’ in Propositional Attitudes, Salmon, N. and Soames, S. eds. and ‘Beyond Singular Propositions?’ Canadian Journal of Philosophy 25 (1995) 515–50Google Scholar
3 Such as Richard, Mark Propositional Attitudes: An Essay on Thoughts and How We Ascribe Them (Cambridge: Cambridge University Press 1990)CrossRefGoogle Scholar and ‘Defective Contexts, Accommodation, and Normalization,’ Canadian Journal of Philosophy 25 (1995) 551-70; Crimmins, Mark Talk About Beliefs (Cambridge, MA: The MIT Press 1992)Google Scholar and Crimmins, Mark and Perry, John ‘The Prince and the Phone Booth,’ The Journal of Philosophy 86 (1989) 685–711CrossRefGoogle Scholar; Forbes, Graeme ‘The Indispensability of Sinn’ The Philosophical Review 99 (1990) 535–63CrossRefGoogle Scholar and ‘Reply to Marks,’ Philosophical Studies 69 (1993) 281-95; and Schiffer, Stephen ‘Belief Ascription,’ Journal of Philosophy 89 (1992) 499–521CrossRefGoogle Scholar.
4 See my ‘The Pragmatics of Attitude Ascription,’ Philosophical Studies 92 (1998) 363-89.
5 In my ‘The Road to Hell: Intentions and Propositional Attitude Reporting,’ Mind and Language, forthcoming, my focus is Mark Crimmins's and John Perry's account (see Crimmins and Crimmins and Perry).
6 The problems for all of the accounts are parallel, in that they arise from the same defect – reliance on audience states of mind for fixing the contextually varying factor. But the details of the various accounts require that they be treated separately. Other accounts, which I don't discuss in detail, would also appear to be subject to the same problems. Schiffer and Forbes both offer accounts which invoke contextual variation in semantics in a manner similar to Crimmins and Richard. However, neither author specifies the source for the contextually varying factors (modes of presentation and linguistic counterpart relations, respectively). If it is meant to be speaker intentions or audience interests, the accounts will be subject to the same problems. If not, then we are owed an explanation of what it is.
7 (3) is obviously correct because they are all fascinated by the Most Wanted List, and Katherine Ann Power has been on it longer than anyone else: They all know that she is on it.
8 Richard's account as originally formulated allows restrictions only on individuals and so cannot handle general reports. This has been pointed out by Soames and Richard has responded by allowing the first item in a restriction to be a property rather than an individual, with the restriction applying to all those who have the property. I use the above formulation for simplicity. See Soames and Richard.
9 There is a small technical problem which arises with a case like this. According to Richard, Lydia's intentions will determine one set of restrictions for all those who have the property of being ancient astronomers. But, to secure the truth conditions we would like, we need at least to have a different set of restrictions for each language spoken by an ancient astronomer, and we need to attach these restrictions to the right astronomers. This can be done if general intentions can be taken to fix a collection of restrictions on individuals, different sets of them for different individuals. This is, in fact, the proposal Soames makes in response to the need for a way of handling general belief reports. I take cases like this to show that Soames’s solution must be adopted in preference to Richard's.
10 I take a representation associated with appearing in the morning to be a representation involved in beliefs about appearing in the morning.
11 Mark Crimmins and Mark Richard have suggested an approach like this, in correspondence.
12 ‘Adequately informed’ would require some work to cash out properly. I'm going to leave it un-cashed out for two reasons. First, dispositional accounts in any area contain handwavey terms of this sort, so this account is at least no worse off than others of its kind and it's worth seeing how well it fares if handwaving is permitted. Second, the account still turns out not to fare particularly well, so it turns out not to be worth cashing out properly. The loose formulation given above is enough to allow us to appreciate the problems faced by the view.
13 If we like, we can think of the speaker's dispositions as determining a set of restrictions on mappings. These restrictions will have the same form as they did on Richard's original account: an individual, property, or set of individuals, an annotation, and a set of appropriate annotations.
14 In order to avoid circularity worries, certain attitude claims – those involved in determining the content of ‘appropriatec’ have a context-invariant content. For such claims, there are no restrictions on translation.
15 Various abnormal speakers have been suggested to me by David Braun, Harry Frankfurt, Chris Hookway, and Scott Soames.
16 This case was suggested to me by an anonymous referee.
17 A response like this was suggested to me by Hugh Mellor and Eric Olson.
18 This case was suggested to me by Scott Soames.
19 A response like this was suggested to me by Gisela Striker.
20 He also mentions audience interests, but as I noted earlier this will only make the problem worse.
21 I am very grateful for the helpful comments I have received on versions of this paper from Kent Bach, Paul Benacerraf, Mark Crimmins, Gilbert Harman, Mark Richard, Teresa Robertson, anonymous referees, and especially Scott Soames and David Braun. In addition I am grateful for the very useful discussions I have received from audiences at Bradford, Cambridge, Cornell, Leeds, Maribor, Princeton, Rochester, Sheffield, Sussex, York, and UCL.
