Pretense, cancellation, and the act theory of propositions

ABSTRACT

Several philosophers advance substantive theories of propositions to deal with several issues they raise in connection with a concern with a long pedigree in philosophy, the problem of the unity of propositions. The qualification ‘substantive’ is meant to contrast with ‘minimal’ or ‘deflationary’ – roughly, views that reject that propositions have a hidden nature, worth investigating. Substantive views appear to create spurious problems by characterizing propositions in ways that make them unfit to perform their theoretical jobs. I will present in this light some critical points against Hanks’ (2015, 2019) act-theoretic view, and Recanati’s [2019. “Force Cancellation.” Synthese 196: 1403–1424. doi:10.1007/s11229-016-1223-9.] recent elaboration of Hanks’ notion of cancellation. Both Hanks and Recanati, I’ll argue, rely on problematic conceptions of fiction and pretense.

KEYWORDS:

• Propositions
• representational content
• representational vehicles
• pretense

Acknowledgements

This work received helpful comments from audiences at the Diaphora workshop on the Nature of Representation, Stockholm, the Act-Type Theory of Propositions, Donostia, and at LOGOS and MELL, LANCOG seminars. Thanks to John Collins, Richard Gaskin, Alex Grzankowski, Peter Hanks, Bjørn Jespersen, Peter Pagin, Michele Palmira, Bryan Pickel, Indrek Reiland, François Recanati, Ricardo Santos and Elia Zardini. Thanks also to Michael Maudsley for the grammatical revision.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 See Moore (1999, 3). He convincingly argues that such pretheoretical notion is, as he puts it “individuatively vague”: different contexts involving different explanatory goals for talk of sameness of content or what is said make salient different individuative criteria. See also Grzankowski & Buchanan’s contribution to this volume. This is consistent with the deflationary view endorsed in this paper, see fn. 8 below.

2 See King’s “What Role do Propositions Play in Our Theories?”, in King, Soames, and Speaks (2014), and McGrath and Frank (2018).

3 I am just suggesting that the explanation of the justification relation would thereby be easier, not altogether unproblematic.

4 Cf. Davidson’s (2005, 76–97) presentation of the early history of the debate.

5 Cf. Green (1999), Matthews (2007, 2011), Field (2016), Ball (2018).

6 Cf. King (2009), Schnieder (2010), Eklund (2019), and Collins (2018).

7 A bit melodramatically, Davidson says that, absent a substantive account, the

philosophy of language lacks its most important chapter … ; the philosophy of mind is missing a crucial first step if it cannot describe the nature of judgment; and it is woeful if metaphysics cannot say how a substance is related to its attributes. (Davidson 2005, 77)

Deflationists dismiss the concern, like this:

Not every account is an analysis! A system that takes certain Moorean facts as primitive, as unanalysed, cannot be accused of failing to make a place for them. It neither shirks the compulsory question nor answers it by denial. It does give an account. (Lewis 1983, 352)

Along such lines, Schnieder (2010, 300) writes this on the representation problem:

the truth-value aptness of a proposition is at least a good candidate for belonging to the fundamental essence of a thing. But if this is so, the only legitimate reply … may consist not in a direct answer, but instead in a rejection of the question.

8 Cf. Speaks’ “Propositions are Properties of Everything or Nothing”, in King, Soames, and Speaks (2014); Richard (2013); Sinhababu (2015); and Pautz (2016). Of course, ‘deflationary’ is a relative term; I mean my take on propositions to be deflationary exclusively in relation to the ambitions expressed by King and Davidson, which create the problems I am about to present. I don’t take the identification of propositions with properties of worlds or situations to be substantive in that relative sense – an ascription to them after all of a ‘hidden nature’ –, because I assume an equally deflationary view of properties. Note that most current views on truth that count as deflationary agree nonetheless that truth is an ‘abundant’ property (Lewis 1983). Such properties may be understood as explained by resemblance nominalism. Propositions might thus ultimately be equivalence classes of representational vehicles, as in some traditional deflationary views – cf. Grzankowski and Buchanan (2019) and their contribution to this volume, Field (2016) and Sainsbury (2018) for related views. As such properties, propositions are more or less fine-grained for different explanatory purposes, thus being ‘individuatively vague’, as Moore (1999) argues. What is essential is that they provide correctness conditions for the relevant states, how they represent the world – which is what is needed for them to properly perform the core tasks in their job description (Sinhababu 2015).

9 Ostertag (2013, 519) reports that Stalnaker pointed this out. The properties that I will take propositions to be correspond to the states of affairs that on Matthews’ (2007, 153) presentation of the measure-theoretic account are representatives of the attitudes.

10 Cf. Keller (2013) for the problems the contrasting view raises for substantivists.

11 Cf. García-Carpintero (2006), Grzankowski (2015). Camp’s (2018) argument against this assumes her decision to ‘interpret propositionality in terms of functional structure’ Camp (2018, 21), something which I am arguing should not be granted. Even if the argument is valid given her assumption, from the perspective adopted here it merely reveals vehicle-level differences. García-Carpintero and Palmira (forthcoming) argue that the notion of singular proposition should also be explained on the deflationary perspective, in terms of properties of representational vehicles, as opposed to substantive commitments on features of propositions themselves.

12 Cf. Richard (2013, 707–708, 716) and Field (2016).

13 Cf. Pickel (2015) for a clear elaboration of these concerns, the most obvious ones of a modal nature. See also Speaks’s ‘Representational Entities and Representational Facts’ in King, Soames, and Speaks (2014), Caplan et al. (2013), Caplan (2016).

14 I don’t mean ‘explanation’ in an epistemological sense, but in the objective, metaphysical sense of ‘because’ common in discussions of essence, ontological dependence, and grounding (cf. Audi 2012; Correia 2008; Rosen 2015).

15 Cf. Koslicki (2018, ch. 5) and Tahko and Jonathan Lowe (2020) for discussion and further references. Both Koslicki and Tahko & Lowe show that the existential account of the distinction I assume in the main text has problems, and they offer alternatives. The existential account is simple, and it suffices to make my points; but they could also be made with the alternative notions.

16 García-Carpintero (2019a) argues for the contingency of kinds defined by constitutive rules.

17 I understand that Hanks (2015, 3–4) endorse this here:

Propositions get their truth conditions from particular acts of judgment and assertion which are themselves the original or primary bearers of truth and falsity. The source of truth conditions is to be found in the acts of representation we perform when we make judgments and assertions, not in the propositional contents we use to classify and individuate these actions.

He also says that ‘[p]ropositions are types of actions, and propositional attitude relations are tokening relations’, Hanks (2015, 7). It is only when occurring together with acts of representation that propositions get truth-conditions: ‘The representational features of these acts are not borrowed from pre-existing propositions. They are generated in the performance of the acts themselves’ (Hanks 2019, 1386).

18 Cf. Sainsbury’s (2018, 55–59) congenial discussion of cardinality worries about deflationary views on propositions. Remember also the point made on fn. 8 on the relativity of ‘deflationary’: the generic dependence of propositions on acts envisaged here may be deemed substantive enough, by some measure; the present point however is just that granting it is consistent with rejecting as unfounded the explanatory ambitions of Davidson, Hanks and King.

19 The problem is not restricted to conditionals. It also arises with disjunction, negation (Hom and Schwartz 2013, 19), the prejacent conditions (sometimes merely contextually implicit, Dowell 2012) in modals and the open sentences required to account for generality (Collins 2018, §6).

20 There are theatrical examples entirely analogous to the one by Stock below; consider for instance Frayn’s Copenhaguen. Moreover, as Ohmann (1971, 18), Lewis (1978, 266), and Alward (2009) suggest, the author of a literary fiction can be seen as an actor impersonating the explicit or implicit teller of the story, asserting its content.

21 Cf. García-Carpintero (2019b).

22 A commitment he owns, as he acknowledged to me in personal communication. Recanati (2019, 1414) is also unhappy with this aspect of Hanks’s views.

23 I am surprised by Green’s (2018, 18) claim, ‘It has for decades been widely agreed among philosophers that there can be no intimate connection between any particular form of words (such as ‘x is good’) and a pragmatic property (such as approving)’. There is no such agreement on this if by ‘intimate’ we mean ‘conventional’, as the context of that remark suggests; see below in the main text. Glüer (2013, 344–345) briefly addresses the issue. She says that Davidson’s arguments ‘are too intricate to fully unravel and to do justice here’. She does discuss ‘two basic strands’: (i) that conventional meaning should be ascribed to utterances of declarative sentences even when they occur in situations cancelling the conventionally indicated assertoric commitment; (ii) that no convention can guarantee the sincerity of the speaker. Both points are correct, but irrelevant to reject the conventionalist claim.

24 Cf. García-Carpintero (2019b) for elaboration. Kölbel (2010, §7) also critically discusses Davidson’s arguments. He considers the reply I am assuming Kölbel (2010, 127). He ends up preferring an alternative one; I lack the space here to explain why I prefer my own.

25 In discussion at the Donostia conference Hanks suggested that the regress might not be vicious. This, however, should be worked out; prima facie we have here one more of those problematic regresses that substantivism on the unity problems tends to generate.

26 Matravers (1997, 79) calls this the Report Model: ‘in reading a novel, a reader makes-believe he is being given a report of actual events. In other words, he makes-believe the content of the novel is being reported to him as known fact by a narrator’.

27 The points in this paragraph are developed at length in García-Carpintero (forthcoming-a, forthcoming-b).

Additional information

Funding

The financial support was provided by the DGI, Spanish Government, research project FFI2016-80588-R, and through the award ‘ICREA Academia’ for excellence in research, 2013, funded by the Generalitat de Catalunya.