Abstract
No contemporary compatibilist account of free will can be complete unless it engages with the consequence argument. I will argue that Wilfrid Sellars offered an ingenious version of compatibilism that can be used to refute the consequence argument. Unfortunately, owing to the opacity of Sellars’s writings on free will, his solution has been neglected. I will reconstruct his view here, demonstrating how it represents a powerful challenge to the consequence argument and tying it to some recent developments in the compatibilist literature.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Van Inwagen is, of course, aware of Harry Frankfurt’s famous essay—published six years prior to this quoted piece—and mentions it in a footnote, but he does not defend the “Principle of Alternate Possibilities” (PAP) in the quoted piece. He does, of course, address the PAP a few years later (van Inwagen 1978). However, the present piece is concerned not with moral responsibility, but instead with free will as involving the ability to do otherwise than one actually does, and so I will not discuss the PAP.
The full version of the argument is presented in van Inwagen (1989, p. 405), but this shortened version gets all the essential premises on the table. The simplified version was suggested by an anonymous referee for Philosophical Studies. Also, as noted in the text, this argument employs a redefined version of the operator N. McKay and Johnson (1996) presented a counterexample to van Inwagen’s Principle Beta; van Inwagen (2015) avoids this counterexample by simply redefining N.
See, for example, AAE.
For a detailed Sellarsian argument for the irreducibility of MI concepts (including intentional concepts) to SI concepts, see chapter 2 of Koons (2019).
In this connection, an important role that concepts play in cognition and communication is in aggregating a number of inferential proprieties under a single term, thereby allowing thought and communication to “move” larger packets of information more cheaply. Connected to this point, and to the point made in the text, it is worth quoting Sellars from “Volitions, Re-Affirmed”: “To every explanation in mentalistic terms there corresponds in principle an explanation in non-mentalistic terms. I say ‘in principle’ because, in even the simplest cases, the complexity of such an explanation would be unmanageable” (VR §19/p. 51). Of course, these two explanations will fundamentally differ, one being a rational explanation, and the other one a purely causal explanation.
Sellars makes the connection between (a) mental states being defined by normative functional roles and (b) multiple realizability in a number of places; see, e.g., SRLG, MP (esp. pp. 237ff).
A reviewer has commented that it may not be obvious that “neurons and synapses” are functional terms. I would argue, however, that biological subsystems (including neurons and synapses) are functionally defined and therefore multiply realizable: This is why (say) an artificial heart (or an artificial hip) is still a heart, or a hip. What matters is the job done; the physical constitution of what does the job is important only (!) insofar as it enables the doing of the job.
Sellars reiterates his support for a functional notion of the mental—and the corresponding commitment to multiple realizability—in several places: “Concepts pertaining to mental acts are ‘functional’ in a way which leaves open the question as to the ‘qualitative’ or, as I prefer to say, contentual character of the items that function in such a way as to be the kind of mental act they are” (“Metaphysics and the Concept of a Person,” p. 237); “the classification of thoughts, construed as classical mental episodes, permits of no such easy retreat to a non-functional level. Roughly, our classification of thoughts, construed as episodes which belong to a framework which explains the kaleidoscopic shifts of sayings and propensities to say, is almost purely functional. We have only the foggiest notion of what kinds of episodes, non-functionally described, perform the relevant functions” (AAE §34/p. 189). He does say, in the immediately preceding section (§33) of AAE that you can talk about the correlates of mental acts in non-functional terms. But to talk about such states, non-functionally described, is not to talk about conceptually contentful mental acts.
To head off an objection, I do not think that the assumption of normativity here begs the question against the incompatibilist (who after all is arguing that moral normativity presupposes free will). The question of free will is not obviously relevant to conceptual normativity, which is governed by ought-to-bes, not ought-to-dos. Sellars’s considered opinion is that inferring is not under our voluntary control and is not, properly speaking, an action. And again, the point here is intelligibility—Jones’s behavior is intelligible because he conforms to certain normative proprieties, such as inferring “Simba is a mammal” from “Simba is a lion.” Smith’s behavior is unintelligible precisely because she fails to conform to these proprieties—precisely because she is disposed, say, to assert both “It is raining” and “It is not raining” at the same time. The normativity of the MI and the sense in which it makes human behavior intelligible are not separate dimensions of this explanatory framework.
It is important that for Sellars, the space of reasons is within the space of causes; he writes that “action in terms of reasons is a special case of explanation in terms of (occurrent) causes” (RD, p. 178). But to treat something as merely in the space of causes is not to treat it as an agent at all. To treat something as an agent is to offer a specific kind of causal explanation of its behavior—namely, rational explanation.
See also “Reply to Alan Donagan” (RD, p. 182): “Only the confusions of the vulgar determinist could lend plausibility to the idea that the conceptual resources of the (first level) manifest image are rich enough to generate (even in principle) a universal derivability of events, including human actions, from antecedent events.”
This is how Sellars casts the debate, but it is not really a felicitous way of stating the problem. As we will see, Sellars’s argument ultimately amounts to the claim that there is no way to formulate the second conjunct of PD-I’s consequent—N[a(x,t′) → A(x,t)]—so that it comes out true. Thus, it is more accurate to say that the principle of determinism cannot even be stated in a way that mentions actions qua events within the MI framework. The principle of determinism is true within the SI, but as such cannot mention agents, nor actions, nor contain any intentional elements. It is not entirely clear (to me, at least) whether Sellars recognized this consequence of his argument, but it follows naturally from his other commitments, and he should welcome it.
Koons (forthcoming).
Adapted from the version introduced at FD, p. 173.
An anonymous referee for Philosophical Studies suggested something like the following objection.
Sellars is explicit about this toward the end of FD, where he writes (for example) that “our concept of ability to do … applies only to doings in the conduct sense” (FD, p. 173). He goes on in the following pages to define a broader sense of “ability” that applies not only to actions, but also to volitions (which for Sellars are acts, not actions), and he writes in the concluding paragraphs, “It is with reference to ‘real’ circumstances that abilities and hindrances are defined” (FD, p. 174). Sellars’s larger point is, I hope, clear: Modal terms (like “able”) are defined relative to a particular framework, such as the framework of actions (or acts).
Peter Olen raised this concern about Sellars’s account.
There is also some similarity between Sellars’s view and that advanced by Kenny (1978), especially chapter 2.
This way of framing Sellars’s argument was suggested by an anonymous referee for Philosophical Studies.
It is also worth noting that token identity (which Sellars endorses) will not by itself secure the transmission of modal statuses across frameworks. Let us use the following notation:
-
m = a mental state, specified in the language of the MI
-
b = a brain state, specified in the language of the SI
-
M = a proposition asserting that m obtains
-
B = a proposition asserting that b obtains
-
P = a proposition representing the state of the entire universe at time t, stated in terms of the SI
Even if m = b (i.e., the two states are token identical), it is not the case that □ (P ⊃ B) entails □ (P ⊃ M).
-
The following argument was suggested by Anjana Jacob.
Brandom (1979) offers a superb articulation of this view.
I owe this way of putting the issue to Mark Lance, who has made this point to me a number of times over the years.
Wolf and Koons (2016, pp. 92–93).
There is a sense in which SI facts can legitimately be appealed to in rational explanations, e.g.: “Why is Pierre celebrating? Because the collision in the particle accelerator released a particle with the mass of 125 GeV/c2.” This is a claim from within the SI, but it is not being used as a deterministic pseudocircumstance to explain Pierre’s action. Rather, it is being situated within the MI framework of agency—we must also take it as read that Pierre knows of these results, knows that they confirm the existence of the Higgs boson, understands the importance of this result, and so forth. What is disallowed, for Sellars, is a modal statement in which a pure SI proposition (with no background premises, enablers, etc. from the MI) entails an agency claim from the MI. That would rather be like an “is” entailing an “ought”.
I make this comparison in more detail in Koons (2019, chapter 5).
References
Birch, J. (2020). Free will and the cross-level consequence argument (unpublished manuscript). http://philsci-archive.pitt.edu/18413/. Accessed 14 April 14 2021. Cited by permission of author.
Brandom, R. B. (1979). Freedom and constraint by norms. American Philosophical Quarterly, 16(3), 187–196.
Brandom, R. B. (2015). From empiricism to expressivism: Brandom reads Sellars. Harvard University Press.
Gebharter, A. (2020). Free will as a higher-level phenomenon? Thought: A Journal of Philosophy, 9, 177–187.
Ginet, C. (1966). Might we have no choice? In K. Lehrer (Ed.), Freedom and determinism (pp. 87–104). Random House.
Kenny, A. (1978). Freewill and responsibility. Routledge.
Koons, J. R. (Forthcoming). ‘To show the compatibility of compatibilism and incompatibilism’: Sellars’s reinvention of Kant’s conception of free will. In L. C. Seiberth & M. Ranaee (Eds.), Reading Kant with Sellars.
Koons, J. R. (2019). The ethics of Wilfrid Sellars. Routledge.
List, C. (2019). What’s wrong with the consequence argument: A compatibilist libertarian response. Proceedings of the Aristotelian Society, 119(3), 253–274.
McKay, T. J., & Johnson, D. (1996). A reconsideration of an argument against compatibilism. Philosophical Topics, 24(2), 113–122.
O’Shea, J. R. (2007). Wilfrid Sellars: Naturalism with a normative turn. Polity Press.
Schroeder, M. (2008). Being for: Evaluating the semantic program of expressivism. Oxford University Press.
Sellars, W. (1962). Truth and ‘correspondence.’ The Journal of Philosophy, 59(2), 29–56.
Sellars, W. (1963a). Philosophy and the scientific image of man. In W. Sellars (Ed.), Science, perception and reality (pp. 1–40). Ridgeview Publishing Company.
Sellars, W. (1963b). Empiricism and the philosophy of mind. In W. Sellars (Ed.), Science, perception and reality (pp. 127–196). Ridgeview Publishing Company.
Sellars, W. (1963c). Some reflections on language games. In W. Sellars (Ed.), Science, perception and reality (pp. 321–358). Ridgeview Publishing Company.
Sellars, W. (1966). Fatalism and determinism. In K. Lehrer (Ed.), Freedom and determinism (pp. 141–174). Random House.
Sellars, W. ND (FD-revised). Fatalism and Determinism. Revised, unpublished version of Sellars 1966. Accessed 13 May 2021. https://digital.library.pitt.edu/islandora/object/pitt:31735062218957
Sellars, W. (1973). Actions and events. Noûs, 7(2), 179–202.
Sellars, W. (1974). Metaphysics and the concept of a person. In W. Sellars (Ed.), Essays in philosophy and its history (pp. 214–241). D. Reidel Publishing Company.
Sellars, W. (1975). Reply to Alan Donagan. Philosophical Studies, 27(3), 149–184.
Sellars, W. (1976). Volitions re-affirmed. In M. Brand & D. Walton (Eds.), Action theory (pp. 47–66). D. Reidel Publishing Company.
Sellars, W. (1980). Concepts as involving laws and inconceivable without them. In J. F. Sicha (Ed.), Pure pragmatics and possible worlds: the early essays of Wilfrid Sellars (pp. 95–123). Ridgeview Publishing Company.
van Inwagen, P. (1975). The incompatibility of free will and determinism. Philosophical Studies, 27(3), 185–199.
van Inwagen, P. (1978). Ability and responsibility. Philosophical Review, 87(2), 201–224.
van Inwagen, P. (1983). An essay on free will. Clarendon Press.
van Inwagen, P. (1989). When is the will free? Philosophical Perspectives Volume 3: Philosophy of Mind and Action Theory, 3, 399–422.
van Inwagen, P. (2015). Some thoughts on an essay on free will. The Harvard Review of Philosophy, XXII, 16–30.
Wiseman, R. (2016). Routledge philosophy guidebook to Anscombe’s intention. Routledge.
Wolf, M. P., & Koons, J. (2016). The normative and the natural. Palgrave Macmillan.
Wood, A. W. (Ed.). (1984). Self and nature in Kant’s philosophy. Cornell University Press.
Acknowledgements
I am grateful to Peter Olen for providing me with helpful comments on an early draft of this essay. I also had helpful conversations with Anjana Jacob and Mark Lance concerning some of the issues discussed here. Finally, two anonymous referees for Philosophical Studies provided a good deal of feedback that led to the substantial improvement of the essay.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Koons, J.R. Sellars on compatibilism and the consequence argument. Philos Stud 179, 2361–2389 (2022). https://doi.org/10.1007/s11098-021-01767-1
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-021-01767-1