Abstract
It has been argued that nonreductive physicalism’s problems with mental causation disappear if we abandon the intuitive but naïve production-based conception of causation in favor of one based on counterfactual dependence and difference-making. In recent years, this response has been thoroughly developed and defended by James Woodward, who contends that Kim’s causal exclusion argument, widely thought to be the most serious threat to nonreductive mental causation, cannot even be given a coherent formulation within Woodward’s preferred interventionist framework. But Woodward has, even more recently, defended a pair of necessary conditions on mental causation and higher-order causation more generally, and it is here that the interventionist framework proves less friendly to nonreductive mental causation. Functionalism is arguably the most important species of nonreductive physicalism concerning specifically mental properties, but, as I argue, functional properties fail both of Woodward’s tests of causal relevance, one of them in two different ways. The problem, moreover, seems unique to functionalism, for other types of higher-order properties appear to pass Woodward’s tests. If functionalism faces defeat even in such friendly territory, its problems with mental causation are not an artifact of naïve metaphysics. They run deep.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See also Loewer (2002, 2007) and Crane (2001) for earlier defenses of this approach in general terms. Kim (2009, p. 44) has responded that productionist mental causation is the only form of mental causation worth having, partly because difference-making relations come too cheap. They may come cheap, but they are nevertheless too dear for functionalism, as we’ll see.
The interventionist account, being nonreductive, avoids many of the issues with (Lewis’s, 1973, 1986) account, for example. See Paul and Hall (2013, Ch. 2) for a discussion of these problems. See Woodward (2003, Ch. 3) for a discussion of the ways in which interventionism improves upon Lewis’s account.
Woodward (2015, 2017) argues that Baumgartner’s argument is based upon a misinterpretation of the spirit if not the letter of interventionism and proposes a condition he calls ‘independent fixability’ in response. Much more on this below. Further contributions to this debate are made by List and Menzies (2009), Raatikainen (2010), Weslake (2011), Shapiro (2012), Polger et al. (2018), Zhong (2020), among many others.
A number of other authors have argued for a condition similar to the second of these. List and Menzies (2009, 2010) are one prominent example. One point on which Woodward’s account disagrees is whether higher-level causation excludes lower-order causation. List and Menzies believe that it often does, while Woodward does not—but this is largely due to the fact that Woodward adopts a very permissive account of causation, as we’ll see below. See Woodward (2021, fn. 23) for discussion. Wilson (2011, 2021) also defends a related condition within an ontology of causal powers and without presupposing a difference-making account of causation. See Antony and Levine (1997) for an early articulation of this condition.
Role functionalism identifies a mental (or other higher-order) property with the second-order property of having some first-order property that occupies a certain causal role. Realizer functionalism, by contrast, identifies the mental property—e.g., pain—with the first-order property—e.g., C-fiber firing in humans—that occupies the associated role. It is for this reason that realizer functionalism is not generally regarded as a species of nonreductive physicalism. See, e.g., Kim (2011, pp. 186–189) for discussion. For ease of presentation, I shall sometimes drop the qualifier and speak of functionalism in what follows, but I should be taken to mean only the role-functionalist variety thereof.
Baker (2009) notes that arguments for nonreductive physicalism in the mental domain derive either from Davidson (1980) or from Fodor’s (1974) and Putnam’s (1975) role functionalist arguments. And, with apologies to Davidson, I do not believe that the arguments for anomalous monism have many adherents in contemporary philosophy of mind. The importance of role functionalist arguments is also evidenced by the fact that those who argue for type identity theory—e.g., (Polger & Shapiro, 2016)—continue to direct their attacks almost exclusively at role functionalist versions of nonreductive physicalism. As Kim (2009, p. 46) observes, “[t]he [role] functionalist view of the mind is still the most widely accepted approach to the nature of mentality” and “is arguably the ‘official’ philosophy of cognitive science.”.
I’ve argued for claims related to these last two elsewhere (see Rellihan, 2019, 2021)), but not from within Woodward’s framework. Given the importance of Woodward’s account of causation and its associated response to the exclusion argument, its it is useful to see that this framework itself poses problems for mental causation.
See Woodward (2021, p. 242) for a formulation very close to this. That formulation omits the clause about holding off-path variables fixed but is otherwise identical. Without this clause, we have an analysis of what Woodward elsewhere calls a ‘total’ cause; with it, we have an analysis of what he calls a ‘contributing’ cause (2003, pp. 51–59). Contributing causation is the more general notion and thus the appropriate one for a general discussion of higher-order causation. (Note that a total cause can be thought of as the special case of a contributing cause in which the set of off-path variables that have to be held fixed at certain values is empty.) Woodward often frames his discussion of higher-order causation in terms of the simpler analysis and thus in terms of total causation. This is a harmless simplification because the issues surrounding higher-order causation run orthogonal to the distinction between total and contributing causes. I’ve nevertheless chosen to frame the discussion that follows in terms of the more general notion so as to make clear that Woodward intends his constraints on higher-order causation to apply to both types of the causal relation.
A much more rigorous definition of an intervention can be found in Woodward (2003, Ch. 3).
Consider an illustration. In the aggregate, birth control pills may not affect the probability of developing thrombosis, for the effects along the distinct pathways may cancel each other out. If so, birth control pills are not a total cause of thrombosis. (See fn. 8). But, according to M*, they are nevertheless a contributing cause of thrombosis because when we hold the value of the off-path variable corresponding to pregnancy constant, changes to the variable representing the ingestion of birth control pills are associated with changes to the variable representing thrombosis. That is, if we hold constant the fact that a woman is not pregnant, changes to the birth control variable correspond to changes to the thrombosis variable. The example is due to (Hesslow, 1976), who uses it for a different purpose. See the discussion of Hesslow’s example in Woodward (2003, pp. 49–51).
As a reviewer points out, it is not altogether clear that this is a genuine relationship between parts and wholes—the spoke is part of the wheel, but it’s less clear that the position of the former is part of the position of the latter. It is still less clear how talk of parts and wholes applies to relations between variables. My point here, however, is simply to motivate the need for IF. Woodward proposes IF as a way of precisifying the intuition that causal relata must be distinct. That it is unclear whether spokes and wheels are appropriately distinct is enough to motivate the need for IF.
See Woodward (2020, pp. 860–862) for further discussion of this and similar cases. Woodward does not, of course, believe that channel conductances and membrane potentials are actually related as parts and wholes, for then they would not be sufficiently distinct. The point, again, is that we cannot rely on our intuitions concerning what is or is not a part of what. Sometimes causal relations between apparent parts and apparent wholes are objectionable, sometimes they are not. We therefore require a clearer account of distinctness than intuition affords. This is what IF is intended to provide. I am thankful to a reviewer for pointing out the need for this clarification.
Woodward develops IF in more detail and argues that it is consistent with the interventionist approach in Woodward (2015).
This example comes originally from Spirtes and Scheines (2004). Woodward discusses it in his (2015). I follow Woodward’s presentation even though it is medically inaccurate in way that makes no difference to the underlying moral—total cholesterol also measures the level of triglycerides in the blood.
More precisely, this is what is required for LDL to be a contributing rather than total cause of A. (See fn. 8). It is the former that is the central notion. In what follows it can be assumed that I am speaking of contributing causes.
I’m assuming nonreductive physicalism here, according to which mental states supervene with metaphysical necessity on physical states. It’s worth noting, however, that Woodward’s defense of mental causation applies to dualistic accounts holding that the mental supervenes with only nomological necessity on the physical. Chalmers (1996, famously defends such a view). This is because, as Woodward makes clear in his definition, IF is meant to cover not only metaphysical but also ‘causal’ possibility, which I take to mean nomological possibility. If M supervenes with nomological necessity on P, it is not causally possible to manipulate M while holding P fixed. Woodward’s defense of nonreductive physicalism thus applies, mutatis mutandis, to at least some versions of dualism. I am thankful to reviewer for requesting clarification on this matter.
Though Woodward’s use of ‘if’ instead of ‘if and only if’ suggests that these conditions are only intended to be sufficient for conditional irrelevance, this is not the case, as his clarification in an appended footnote makes clear: “conditional irrelevance is much stronger than multiple realizability. The latter requires only that some different values of the same or different micro-variables(s) realize the same value of a macro-variable. Conditional irrelevance requires that all variations at the micro-level consistent with the value of the macro-variable make no difference to E. As this observation suggests, multiple realizability is not sufficient for autonomy understood in terms of conditional irrelevance” (2021, p. 254; italics added and omitted). See also Woodward’s (2021, pp. 258–259) discussion of the potential failure of psychological variables to meet CI with regard to neurological variables, where this is treated as a failure of autonomy. Clearly, then, in both its use and intent, CI is treated as expressing both a sufficient and a necessary condition. See also Woodward (2020, p. 866), where it is said that “for most arbitrary sets of Us, Ls, and Es, conditional independence, or even approximate conditional independence, will fail …. The interesting question is the extent to which there are cases in which conditional independence or something like it does hold.” These remarks and the surrounding discussion also clearly imply that the stated conditions are necessary.
CI is needed to establish the autonomy or distinctive causal efficacy of higher-order properties. It is not needed to establish the causal relevance of higher-order properties, full stop. Indeed, it presupposes this sort of causal relevance. The point is that if higher-order properties meet the constraints imposed by M* without meeting those imposed by CI, they make no independent causal contribution and are therefore explanatorily and ontologically otiose. Failure to meet CI would thus leave mental properties in a position similar to that envisioned by the exclusion argument. See Woodward (2018) for further discussion.
It is true that many changes to these kinetic energies will not affect E—those changes that balance each other out and thus don’t affect the average—but recall the M* requires only that some changes are associated with changes to E. This, again, illustrates just how weak a constraint M* is.
Again, the lower-level realizers—such as being maroon, for example—are unconditionally relevant by the weak standard set by M*. Some changes to this variable—the change from being maroon to being cyan, for example—are associated with changes to the effect variable. It is irrelevant that some other changes are not. It is only if one adopts a stronger criterion, such as the more restrictive M** discussed below, that higher-order causes exclude lower-level causes. Adopting such a standard would also require us to modify the definition of conditional irrelevance.
Or whatever type of entity it is that stands in first-order causal relations. There’s no need to take a stand on the nature of causal relata, as interventionism, we have seen, takes no such stand. I’ll speak of properties and states—instantiations of a property by an object at a time—for ease of presentation. Everything I say can be translated into an alternative idiom.
This manner of illustrating the distinction comes from Bennett (2003, p. 485).
Failure to attend fully to the distinction between core and total realizers has perhaps obscured this problem. C-fiber firing is sufficiently distinct from pain’s characteristic effects to be their cause, and it is true that C-fiber firing ‘realizes’ pain, but it does not follow that pain is sufficiently distinct from its characteristic effects to be their cause, for C-fiber firing is only the core and not the total realizer of pain.
A reviewer points out that (Block’s, 1989) ‘dormitive virtue’ objection to functionalism is also similar to the one being discussed here. For Block, the crux of the problem is that functional properties are defined in terms of their effects. Interestingly, Block believes that functionalists can avoid his objection by adopting a counterfactual-based theory of causation: “Here … the lesson is that if you want to avoid epiphenomenalism, go for a counterfactual theory of causal relevance, not a nomological theory” (p. 159). But this is only true if counterfactual-based accounts do not adopt IF, and, as Woodward argues, such accounts are not viable.
I argue for a similar conclusion in Rellihan (2021), but from within the powers subset framework of Shoemaker (2007) and Wilson (2011). That framework is usually interpreted as presupposing a productionist account of causation, so it is useful to see that a similar conclusion follows from Woodward’s interventionist account.
Whatever else is true of the relation of a realizer to the property it realizes, it must be the case that the instantiation of the former is sufficient for the instantiation of the latter, and this is true of total rather than of core realizers.
Kim (1992) is well known for introducing this principle—which he calls ‘Alexander’s dictum’—into the mental causation debate. Kim (1993) elsewhere considers the sort of ‘local reductionism’ discussed above and argues that it is inconsistent with the view that multiply realized properties are genuine or scientific kinds: “It must be admitted,” he says, “that pain as a kind does not survive multiple local reduction” (p. 333). Genuine kinds, he argues, are individuated by their causal powers, and local reductionism denies that multiply realized properties are causally homogenous. CI is one way of precisifying the relevant sort of causal homogeneity, so the arguments I make in the body of the paper can be used to reinforce Kim’s conclusion.
It’s also worth noting that in both the thermodynamic case and in the case of determinables like being red the relevant causal models meet the conditions imposed by IF.
The research is described in (Musallam et al., 2004).
Woodward is explicit in making the comparison between the cases: “The preference for micro or fine-grained causation that we are considering recommends that we should regard [the neural realizer] as the real cause of [the effect] on occasion t. But this seems wrong for the same reason it seems wrong to say that it is the scarlet color of the target that causes the pigeon to peck in circumstances in which the pigeon will peck at any red target and wrong to say that it is the specific molecular configuration G1 rather than the fact that the temperature of the gas has been increased to T2 which is responsible for the new pressure P2” (2008, pp. 239).
See, for example, (Woodward 2021, p. 242) where he says that neither is clearly more correct than the other.
I give more elaborate examples of such procedures and discuss their causal (but not necessarily interventionist) consequences in Rellihan (2019).
Bennett (2003, 2008) recognizes the problem pointed out here but argues that the background context is necessary for the core realizer to have its effect. Keaton and Polger (2014) show that this is not generally the case by constructing a counterexample. I have argued elsewhere (Rellihan, 2019) that it is, in fact, very rarely the case. Simply consider the neuron diagrams for causal modeling that Lewis made famous—which, of course, are modeled on actual neurons. It is rarely the case that disrupting any one causal connection disrupts other, disparate regions of the network. The onus is at the very least on functionalists to give some reason to believe that Bennett’s happy coincidence is the norm. See Rellihan (2019) for further discussion.
Note that this variable set does not meet IF, for P cannot be set independently of T and T cannot be set independently of C and B. This is irrelevant for present purposes, however, for we can simply test each of the potential causes of A independently. In a variable set that includes P but not T, C, or B, P causes A. In one that includes T but not P, C, or B, T causes A. And so on. This procedure suffices to show that both P and its various realizers are unconditionally relevant to A. We can then apply the CI test as in the text.
References
Antony, L., & Levine, J. (1997). Reduction with autonomy. Philosophical Perspectives, 11, 83–105.
Armstrong, D. (1978). A theory of universals: Universals and scientific realism (Vol. II). Cambridge University Press.
Baker, L. R. (2009). Non-reductive materialism. In B. McLaughlin & A. Beckermann (Eds.), The Oxford handbook of philosophy of mind (pp. 109–127). Oxford University Press.
Baumgartner, M. (2009). Interventionist causal exclusion and non-reductive physicalism. International Studies in the Philosophy of Science, 23(2), 161–178.
Baumgartner, M. (2010). Interventionism and epiphenomenalism. Canadian Journal of Philosophy, 40, 359–383.
Bennett, K. (2003). Why the exclusion problem seems intractable and how, just maybe, to tract it. Nous, 37(3), 471–497.
Bennett, K. (2008). Exclusion again. In J. Hohwy & J. Kallestrup (Eds.), Being reduced (pp. 280–307). Oxford University Press.
Block, N. (1978). Troubles with functionalism. Minnesota Studies in the Philosophy of Science, 9, 261–325.
Block, N. (1989). Can the mind change the world? In G. Boolos (Ed.), Meaning and Method: Essays in Honor of Hilary Putnam (pp. 137–170). Cambridge University Press.
Chalmer, D. (1996). The conscious mind. Oxford University Press.
Crane, T. (2001). The elements of mind. Oxford University Press.
Davidson, D. (1980). Mental events. Essays on actions and events (pp. 207–224). Oxford University Press.
Fodor, J. (1974). Special sciences, or the disunity of science as a working hypothesis. Synthese, 28, 97–115.
Hesslow, G. (1976). Two notes on the probabilistic approach to causality. Philosophy of Science, 43(2), 290–292.
Keaton, D., & Polger, T. (2014). Exclusion, still not tracted. Philosophical Studies, 171(1), 135–148.
Kim, J. (1992). ‘Downward causation’ in Emergentism and Nonreductive Physicalism. In A. Beckermann, H. Flohr, & J. Kim (Eds.), Emergence or Reduction? (pp. 119–138). de Gruyter.
Kim, J. (1993). Multiple realization and the metaphysics of reduction. In J. Kim (Ed.), Supervenience and mind (pp. 309–335). Cambridge University Press.
Kim, J. (1998). Mind in a physical world. MIT.
Kim, J. (2005). Physicalism, or something near enough. Princeton University Press.
Kim, J. (2009). Mental causation. In B. McLaughlin, A. Beckermann, & S. Walter (Eds.), The Oxford handbook of philosophy of mind (pp. 29–52). Oxford University Press.
Kim, J. (2011). Philosophy of mind. Westview.
Lewis, D. (1973). Causation. Journal of Philosophy, 70(17), 556–567.
Lewis, D. (1983). Mad pain and Martian pain. Philosophical papers (Vol. I, pp. 122–130). Oxford University Press.
Lewis, D. (1986). Postscript to ‘Causation.’ Philosophical papers (Vol. II, pp. 172–213). Oxford University Press.
List, C., & Menzies, P. (2009). Nonreductive physicalism and the limits of the exclusion principle. Journal of Philosophy, 1006, 475–502.
List, C., & Menzies, P. (2010). The causal autonomy of the special sciences. In Ed. McDonald & Ed. McDonald (Eds.), Emergence in mind (pp. 108–129). Oxford University Press.
Loewer, B. (2002). Review of mind in a physical world. Philosophy and Phenomenological Research, 65(3), 655–662.
Loewer, B. (2007). Mental causation or something near enough. In B. P. McLaughlin & J. Cohen (Eds.), Contemporary debates in philosophy of mind (pp. 243–264). Blackwell.
Mill, J. S. (1843). A system of logic, ratiocinative and inductive. University of Toronto Press.
Musallam, S., Corneil, B., Greger, B., Scherberger, H., & Andersen, R. (2004). Cognitive control signals for neural prostethics. Science, 305, 258–262.
Paul, L. A., & Hall, N. (2013). Causation: A user’s guide. Oxford University Press.
Polger, T., & Shapiro, L. (2016). The multiple realization book. Oxford University Press.
Polger, T., Shapiro, L., & Stern, R. (2018). In defense of interventionist solutions to exclusion. Studies in History and Philosophy of Science Part A, 68, 51–57.
Putnam, H. (1975). The nature of mental states. Mind, language, and reality (pp. 429–440). Cambridge University Press.
Raatikainen, P. (2010). Causation, exclusion, and the special sciences. Erkenntnis, 73(3), 349–363.
Rellihan, M. (2019). Strengthening the exclusion argument. Synthese, 198(7), 6631–6659.
Rellihan, M. (2021). A familiar dilemma for the subset theory of realization. Analytic Philosophy, 64(1), 68–90.
Rupert, R. (2006). Functionalism, mental causation, and the problem of metaphysically necessary effects. Noȗs, 40(2), 256–283.
Shapiro, L. (2012). Mental manipulations and the problem of causal exclusion. Australasian Journal of Philosophy, 90(3), 507–524.
Shoemaker, S. (1981). Some varieties of functionalism. Philosophical Topics, 12(1), 93–119.
Shoemaker, S. (2007). Physical realization. Oxford University Press.
Sprites, P., & Scheines, R. (2004). Causal inference of ambiguous manipulations. Philosophy of Science, 71(5), 833–845.
Wilson, J. (2011). Non-reductive physicalism and the powers-based subset strategy. The Monist, 94(1), 121–154.
Wilson, J. (2021). Metaphysical emergence. Oxford University Press.
Woodward, J. (2003). Making things happen. Oxford University Press.
Woodward, J. (2008). Mental causation and neural mechanisms. In J. Hohwy & J. Kallestrup (Eds.), Being reduced: new essays on reduction, explanation, and causation (pp. 218–262). Oxford University Press.
Woodward, J. (2015). Interventionism and causal exclusion. Philosophy and Phenomenological Research, 91(2), 303–347.
Woodward, J. (2017). Intervening in the exclusion argument. In H. Beebee, C. Hitchcock, & H. Price (Eds.), Making a difference (pp. 251–268). Oxford University Press.
Woodward, J. (2018). Explanatory autonomy: The role of proportionality, stability, and conditional irrelevance. Synthese, 198(1), 1–29.
Woodward, J. (2020). Causal complexity, conditional independence, and downward causation. Philosophy of Science, 87(5), 857–867.
Woodward, J. (2021). Explanatory autonomy: The role of proportionality, stability, and conditional irrelevance. Synthese, 198, 237–265.
Yablo, S. (1992). Mental causation. Philosophical Review, 101(2), 245–280.
Zhong, L. (2020). Intervention, fixation, and supervenient causation. Journal of Philosophy, 117(6), 293–314.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The author certifies that he/she has no conflict of interest regarding the submission of this manuscript. The manuscript is the author’s original work. It has not been submitted or published in whole or in part elsewhere.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Rellihan, M. Functionalism, interventionism, and higher-order causation. Synthese 203, 91 (2024). https://doi.org/10.1007/s11229-024-04500-7
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11229-024-04500-7