From the agent’s point of view: the case against disjunctivism about rationalisation

ABSTRACT

A number of authors have recently advanced a ‘disjunctivist’ view of the rationalising explanation of action, on which rationalisations of the form ‘S A’d because p’ are explanations of a fundamentally different kind from rationalisations of the form ‘S A’d because she believed that p’. Less attempt has been made to explicitly articulate the case against this view. This paper seeks to remedy that situation. I develop a detailed version of what I take to be the basic argument against disjunctivism, drawing on a framework of explanatory proportionality. The disjunctivist cannot reject this framework, I argue, because they need it to respond to another challenge, from psychological individualism. As I explain, however, the proportionality-based challenge is not in principle insurmountable, and I outline a number of ways in which a case for disjunctivism might be developed in response to it. The paper thus clarifies the dialectic around disjunctivism about the rationalisation of action and, specifically, what advocates of the view must do in order to make a compelling case for it.

KEYWORDS:

• Motivating reasons
• reasons for action
• disjunctivism
• rationalising explanation
• rationalisation

Acknowledgements

This paper is derived in part from a section of my PhD thesis, which itself drew from work done for my MPhil Stud thesis. I would like to thank Ulrike Heuer, Fabrice Teroni and an anonymous reviewer for another journal for helpful feedback on earlier versions, and especially Mike Martin for his many incisive questions and helpful suggestions. I am very grateful to two anonymous referees for this journal whose generous comments substantially improved the paper, and I gratefully acknowledge the financial support of the AHRC and London Arts and Humanities Partnership, which funded my PhD at UCL, and the Swiss National Science Foundation, which funded my postdoctoral position on the Modes and Contents project at the University of Fribourg and in the Thumos research group in Geneva.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1 What characterises a worldly reason, on the present usage, is that it is a fact about the agent’s situation to which the agent responds or by which the agent is rationally motivated (in a broad sense of ‘rationally’). It is a further question whether the fact is a good (normative) reason for doing what the agent does.

2 For the distinction between disjunctivism and non-conjunctivism, see Williamson (2000). For simplicity I will henceforth refer simply to ‘disjunctivism’. For present purposes this should be taken to include non-conjunctivism.

3 See, respectively Williamson (2000); McDowell (1998); Martin (2004, 2006). See also Hinton (1973); Snowdon (1980).

4 In particular Cunningham (2019a, 2019b); Hornsby (2008); McDowell (2013); Roessler (2014).

5 A notable exception is Dancy (2000), who argues for a version of perspectivalism. While my thinking about these issues owes a great deal to Dancy’s work, I won’t discuss his arguments in detail. Dancy is primarily concerned not with the ‘S A’d because p’ form of worldly rationalisation but with the construction ‘S A’d for the reason that p’, and he, perhaps idiosyncratically, claims that the latter construction is not factive for ‘p’. This leads to an idiosyncratic way of framing the issue. It also means that Dancy’s arguments for perspectivalism struggle to engage with those who take more seriously the ‘worldliness’ of worldly rationalisation; see for instance the exchange between Dancy (2011) and Hyman (2011).

6 See for example Smith (1994); also Davidson (1963). Dancy (2000) argues that normative and motivating reasons cannot be fundamentally distinct in this way. See Mantel (2014, 2018) for recent sceptical discussion of Dancy’s argument.

7 For the former, see for instance Comesaña and McGrath (2014); for the latter, Alvarez (2018).

8 Cunningham uses the labels ‘the Disjunctive View’ for what I am calling disjunctivism and ‘the Common Kind View’ for what I am calling (narrow) perspectivalism.

9 An interesting question Cunningham doesn’t raise is whether there might be different kinds of efficient-causal explanation (different efficient-causation relations) and if so how those different kinds relate to one another. The possibility of subdividing kinds of explanation might merit further discussion, especially for the disjunctivist, who wants to say that there are two different kinds of rationalisation. A helpful model here might be the discussion of different species–genus relations in Ford (2011).

10 Cunningham (2020); Hornsby (2008); Hyman (1999, 2011, 2015) and McDowell (2013) argue that only facts one knows can rationalise one’s actions. Hughes (2014) and Locke (2015) argue that a weaker condition is sufficient. For present purposes, I will assume that the knowledge account is correct.

11 Cunningham (2019b, 238) argues for the stronger claim that a worldly rationalisation and the corresponding knowledge rationalisation are the same particular explanation.

12 I take this ‘qua’ formulation from Cunningham (2019b). The formulation might seem obscure if we take explanation to relate facts rather than particulars (as urged, for instance, by Strawson [1992]). Cunningham seeks to give an account of how particulars (or at least ‘entities’) and relations between particulars under explanations. Cunningham takes it that when S knows that p, S's state of knowing that p is identical to S's state of believing that p, and that this state is an ‘entity’ of a sort that can underly (rationalising) explanations by standing in a rational-motivation relation to actions. Perspectivalist and disjunctivist can agree that knowledge-states thus relate to actions, grounding worldly rationalisations. The perspectivalist denies that the rational-motivation relation here depends on the knowledge-state's actually being a state of knowledge: the explanation holds in virtue of the knowledge-state but not in virtue of its being a knowledge-state. ‘Knowledge explains but not qua knowledge’ is shorthand for this idea. Thanks to an anonymous reviewer for this journal for encouraging me to address this issue.

13 See for instance Dancy (2011) and the response in Hyman (2011).

14 Thanks to an anonymous reviewer for this journal for raising this concern.

15 I assume here a broadly Davidsonian conception of psychology as the sphere of rationalising explanation. See in particular Davidson (1980, Chap. 11–13).

16 For a very helpful overview, see Dorr (2019).

17 In particular, versions of the argument are given by Peacocke (1993); Stalnaker (1989, 1990); Williamson (2000) and Yablo (1992, 2003). Compare also Davidson (1980, Chap. 11).

18 At least, its being a state of knowledge is not causally relevant; it is not causally relevant qua knowledge. If the agent’s knowledge that p is identical with their belief that p, then it might be causally relevant qua belief. See again Cunningham (2019b).

19 For some challenges to such ‘knowledge-first’ approaches to belief see for instance McGlynn (2014, Chap. 2); Hawthorne, Rothschild, and Spectre (2016); Rothschild (2020).

20 Perhaps ‘naturalness’ is no longer the right way to think of things when we enter the normative realm. But it does seem fitting to talk about fundamentality here, and this could presumably play a similar role in the argument.

21 Cunningham (2019b) gives an argument that is significantly different from Roessler’s, but which occupies a broadly similar dialectical space. Both arguments, I think, merit further attention.

Additional information

Funding

This work was supported by the Arts and Humanities Research Council and the Swiss National Science Foundation.

Notes on contributors

Edgar Phillips

Edgar Phillips is a postdoctoral researcher in philosophy, currently based at Institut Jean Nicod in Paris, working on a project supported by the Leverhulme Trust. He works in the philosophy of mind, philosophy of action, moral psychology and sometimes epistemology and philosophy of psychology, mostly on issues relating to interpersonal understanding.