The disappearing agent as an exclusion problem

ABSTRACT

The disappearing agent problem is an argument in the metaphysics of agency. Proponents of the agent-causal approach argue that the rival event-causal approach fails to account for the fact that an agent is active. This paper examines an analogy between this disappearing agent problem and the exclusion problem in the metaphysics of mind. I develop the analogy between these two problems and survey existing solutions. I suggest that some solutions that have received significant attention in response to the exclusion problem have seen considerably less attention in response to the disappearing agent problem. For example, one solution to the exclusion problem is to reject the exclusion assumption. Analogously, one solution to the disappearing agent problem could be to deny the claim that the agent-causal approach and the event-causal approach are mutually exclusive. Similarly, proportionality theories of causation, a solution to the exclusion problem, can be transferred to the disappearing agent problem. After establishing the plausibility of the analogy between the two problems, I examine how this latter solution in particular can be transferred from the one problem to the other.

KEYWORDS:

• Metaphilosophy
• Agency
• Philosophy of action
• Exclusion problem
• Disappearing agent problem
• Agentcausation

Acknowledgements

Ryan Cox and Daniel Stoljar helped clarify my thinking on the exclusion problem during a visit to the Research School of Social Sciences at the Australian National University in 2013. The paper also benefitted from comments and discussions with Richard Bradley, Richard Holton, Sebastian Köhler, Christian List, Wlodek Rabinowicz, David-Hillel Ruben, Jesse Saloom, and Ying Shi and discussions in Geert Keil's colloquium at HU Berlin in 2016. Finally, I was able to improve the paper's substance or exposition in response to comments from anonymous reviewers at this and three other journals.

Disclosure statement

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

Notes

1 The terminology follows Schlosser (2015). I take this binary division of the conceptual landscape for granted but register a point of dissatisfaction. Not all proponents of the event-causal approach take the relata in causal relations to be events. Some instead identify exercises of agency with interacting causal powers manifested in causal processes (e.g. Stout 1996, 2005, chap. 6; Mumford and Anjum 2011, 204; Buckareff 2012; and, on my reading, Steward 2012a, 2016).

2 For the sake of ecumenism, as I explain shortly, I henceforth omit the ‘qua substance’ qualification.

3 It must be noted that this problem is different from another problem that goes by the same name (Pereboom 2014a, chap. 2; Clarke 2017).

4 Although Stout (2007, 2013) does not self-identify as a proponent of the agent-causal approach, he can be read as such (Bishop 2007, 158).

5 I talk of ‘result of action’ to avoid the controversial claim that agents cause their actions (Alvarez and Hyman 1998). Furthermore, set aside for now whether ‘result’ refers to a fact, an event, a property, etc. I also remain neutral between so-called product views and component views about the relation between an action and its cause.

6 In expressions such as ‘you take the cup’, you count as the agent of taking the cup in the linguistic sense (a semantic concept). Some take this to support the claim that you are an agent of taking the cup in the philosophical sense. Chisholm (1976, 199) makes such an argument: ‘[S]tatements [such] as “Jones killed his uncle” and “Jones raised his arm” are sometimes true; … they imply that there is a certain event or state of affairs that an agent caused or brought about’. For similar arguments, see Lowe (2008, 122) and Steward (2012b, 200). However, it should be noted that Lowe (2008) also pursues further additional arguments.

7 The expression ‘disappearing agent’ seems to be due to Velleman (1992). Recently the disappearing agent problem has been raised by Alvarez and Hyman (1998, 227–28), O’Connor (2005), Lowe (2008, 159–61), Mayr (2011, 6–13), Steward (2012b, 197–247), Buckareff (2012), Pereboom (2014a, 2014b), Brent (2017), and Velleman (2020, 62–63). Aguilar and Buckareff (2010, 12–14) give an overview.

8 Goldman (1970, 82–83) is not himself a proponent of the disappearing agent problem because he denies the exclusion assumption and holds that the agent as well as their mental entities may cause an act.

9 The assignment is less clear for the latter two. Importantly, they each offer further arguments not discussed here.

10 Although, strictly speaking, the principle does not apply to this position because the two causes – the mental entity and the physiological event – are one and the same.

11 I use ‘agents’ and ‘individuals’ interchangeably.

12 It does not follow that whenever two individuals have the same mental entities they must be the same individual. Rather, the intentional level is associated with several other macroscopic events besides mental entities. The assumption could be that it is this level as a whole that supervenes on the intentional level.

13 I am indebted to an anonymous reviewer for this observation.

14 This answer, of course, goes hand-in-hand with a certain metaphysics of actions, namely, seeing them as processes (Steward 2012a).

15 To account for action at a time, I do not need to assume that agents are identical over time and I therefore do not sketch this part of the account here.