Mental Causation

views updated


There is mental causation whenever a mental state, event, process, or activity has a causal effect. The pursuit of our lives seems replete with mental causation. It may thus seem as obvious that it occurs as we pursue our lives. But how mental causation is possible is not obvious. And therein lies a philosophical tale. Any attempt to explain how it occurs must engage the mind-body problem.

René Descartes (15961650) maintained that there is body-to-mind causation when we perceive our surroundings, and mind-to-body causation when we act. But one of the most serious charges leveled again his substance dualism, according to which the mind is an immaterial substance that is not extended in space, is that it leaves unexplained how mental states and events (etc.) have causal effects on our bodies. Descartes held that the locus of mind-body causal interaction is in the brain (specifically, in the pineal gland). His contemporary, Princess Elisabeth of Bohemia, asked how states of, or changes in, a substance not extended in space (the mind) could causally affect states of, or changes in, a substance extended in space (the brain or pineal gland), and declared such causal interaction too incredible to believe. The absence of a satisfactory answer to her "how-question" contributed to the demise of Cartesian substance-dualism (Watson 1987).

Many contemporary philosophers hold that to have a mind is not to possess an immaterial substance, but rather to possess certain capacities, such as the capacity to think and/or to feel. Brains serve somehow as the material basis of such capacities. (Whether an artificial brain could so serve is the question of whether artificial intelligence is possible.) But because of the many apparent differences between mental and physical properties, some philosophers, while rejecting Cartesian substance dualism, nevertheless embrace Cartesian property dualism. They hold that while there are no immaterial substances, mental properties are distinct from physical properties, and are related to certain of them by irreducible laws of nature. This view faces the question of how an individual's having a mental property could exert any causal influence on the course of events. Given the absence of a reality underlying both mental and physical reality, an individual's having a mental property would have to exert a direct causal influence on its initial effects in the brain, one unmediated by any mechanism.

The year 1870 marked more than a century of increasingly detailed investigation of human physiology. In that year, Ewald Herring declared at his lecture to the Imperial Academy of Sciences in Vienna that brain physiologists should make "the unbroken causative continuity of all material processes an axiom of [their] system of investigation" (translated and quoted in Butler 1910, pp. 6465). It remains an axiom of neurophysiology. The fact that there are no "gaps" in physiological brain processes for mental events to fill led Thomas Huxley (1874) to maintain we are "conscious auotmata": conscious events accompany certain physiological brain events as dual effects of other physiological events, but are causally inert. Trained as a medical doctor, William James (1890) appropriated the term epiphenomena, a medical term for symptoms of diseases, for mental phenomena that while caused, lack causal efficacy. James Ward (1903) coined the term epiphenomenalism for the view that mental phenomena have no causal effects.

The view that mental phenomena are epiphenomena has a dense air of paradox. Epiphenomenalists maintain that we are merely under the illusion that there is mental causation. But, on their view, the illusion could not give rise to our belief in mental causation, for that would require mental causation. Moreover, on pain of inconsistency, they cannot take themselves to have been led to the doctrine by theoretical reasoning, for their being so led would involve mental causation. Indeed, reasoning itself seems to be a causal process. It should thus come as no surprise that virtually no contemporary philosophers who acknowledge the reality of the mental espouse the view that no mental states or events have causal effects. But the question of how they have effects remains.

Some philosophers combine the rejection of Cartesian substance dualism with the rejection of mental and physical event dualism, while nevertheless embracing Cartesian property dualism. C. D. Broad (1925) examined a dual-aspect theory of events, according to which physiological events in "the mind-brain" (1925, p. 439) have two independent aspects, one mental, the other physiological, the two linked by contingent fundamental laws. In discussion of the view, he formulated epiphenomenalism as a disjunctive doctrine: "mental events either (a) do not function at all as cause factors; or that (b) if they do, they do so in virtue of their physiological characteristics, and not in virtue of their mental characteristics" (p. 473). If, rather than being accompanied by mental events, certain physiological events have mental characteristics, and so are mental events, then it seems, on the evidence, that they function as cause factors in virtue of their physiological characteristics, but not their mental ones. The mental qua mental seems causally inert.

Donald Davidson (1970) proposed the doctrine of anomalous monism: every particular mental event is a physical event, but there are no strict psychological or psychophysical laws, and mental characteristics are irreducible to physical characteristics. He did not, however, embrace Cartesian property dualism, which is committed to fundamental psychophysical laws. Moreover, he regarded talk of properties as pleonastic; strictly speaking, there are only predicates, not properties. He held that since mental events (i.e., events mental predicates are true of) are causes or effects, they fall under strict physical laws, and so are physical events because physical predicates that figure in the relevant strict laws are true of them. Still the causal relation, he emphasized, is extensional: if two events are causally related, they are so related however they are described. There is no quacausation.

Many philosophers hold that properties are distinct from predicates, and indeed that predicates apply to things only in virtue of the properties that things have. And they hold that although the causal relation is indeed extensional, it is nevertheless the case that events enter into causal relations in virtue of certain of their properties. The weighs-less-than relation is extensional: If a weighs less than b, then it does so however a and b are described. Still a weighs less than b in virtue of something about each of them, namely their respective weightstheir respective masses in the gravitational context in question. Anomalous monism entails the denial of token epiphenomenalism. But its proponents must answer the charge of commitment to type epiphenomenalism, the thesis that no events are causally related in virtue of falling under mental types (McLaughlin 1989, 1994; Kim 1993; Sosa 1993; see also Davidson 1993).

In the early twentieth century, the atomic view of matter was vindicated, and in the 1930s a quantum mechanical explanation of chemical bonding was provided, dispelling the idea that there are fundamental chemical forces; and later monumental advances in organic chemistry and molecular biology led to the demise of any form of vitalism (McLaughlin 1992). It is now generally held, on empirical grounds, that: for any (caused) microphysical event P there is a distinct microphysical event P* that causally determines the objective probability of P (if determinism is true, that probability will be 1).

This thesis has been called by various names in the literature, including "the closure of the physical." Given this thesis, if Cartesian property dualism is correct, then it seems that an individual's having a mental property could have microphysical effects only if it causally overdetermined those effects. Such overdetermining psychophysical causal transactions would be fundamental in that they would be unmediated by any mechanism. While that may fall within the realm of logical possibility, it is hard to see how the view that it actually occurs could be justified (Kim 1998).

Many contemporary philosophers hold that there is a stronger dependence of mental properties on microphysical properties than Cartesian property dualism allows. There is no received formulation of the dependency. But one leading view is that it is captured by the following supervenience thesis: any minimal physical duplicate of the actual world is a duplicate simpliciter of it (Jackson 1998). A physical duplicate of the actual world is any world that is exactly like the actual world in every microphysical respect, in respect to its worldwide pattern of distribution of microphysical properties and relations, its worldwide pattern of distribution of microphysical objects, its microphysical laws of nature, and so on. A minimal physical duplicate of the actual world is any physical duplicate of it that contains nothing other than what is metaphysically required to be a physical duplicate of it.

While the supervenience thesis is incompatible with Cartesian property dualism, it does not entail that every property is a microphysical property. The thesis entails that any minimal physical duplicate of the actual world will have exactly the same worldwide pattern of distribution of properties as the actual world. But, as should be made clear below, that does not require that every property be a microphysical property. Indeed, one can embrace the supervenience thesis while holding a kind of property pluralism, according to which not only mental properties, but properties that figure in the laws of the special scienceseconomics, psychology, biology, and even most of chemistryare not microphysical properties. Some proponents of the supervenience thesis are property pluralists and hold, in addition, (token) event and state pluralism, on the grounds that events and states are property exemplifications. They thus hold that mental events, and events within the domains of the special sciences, are not microphysical events. Let us label this kind of "nonreductive physicalism," which combines the supervenience thesis with property and event pluralism, "NRP."

NRP theorists acknowledge that every event is such that its objective probability is causally determined by some microphysical event occurring across some cross section of its backward light cone. But they deny that this excludes higher-level events from being causes. Some defend this denial by distinguishing causation from causal determination (Yablo 1992). They hold that to be causally related, events must be appropriately proportional, and that microphysical events are typically disproportional to the higher-level events they causally determine, and are thus disqualified as causes of those events. On this view, when the turning of a key causes a lock to open, some microphysical event will causally determine that the lock opens. But it will not be a cause of the lock's opening. The reason is that it contains too much superfluous detail to be suitably proportionate to the opening of the lock. Had the key turning occurred without that microphysical event, the lock would still have opened. The key turning thus "screens off" the microphysical event vis-à-vis the lock's opening. Of course, in the counterfactual situation that is stipulated, some other microphysical event will underlie the key turning and cause the microphysical event underlying the lock's opening. But it is claimed that is so because higher-level causal transactions are implemented by lower-level ones, and ultimately by microphysical ones.

One charge against this view is that it mistakes causal explanation for causation. Any microphysical event that causally determines the opening of the lock causes it. Nevertheless, an explanation of why the lock opened in terms of a microphysical cause would be an extremely poor one indeed in a typical context since it would contain far too many details that are superfluous to understanding why the lock opened. But whether that charge can be justified remains a matter of dispute. The dispute turns on controversial issues about the nature of causation and the individuation of events.

Many NRP theorists hold that every event is caused by some microphysical event that determines its objective probability. They maintain, nevertheless, that higher-level events are causes. One concern about this view is that if higher-level events were causes, then their effects would include microphysical events. If my decision to walk into the next room causes me to walk into the next room, a result will be that many of the physical particles making up my body at the time of the decision will end up in the next room. The decision would be a cause (though not of course a sufficient cause) of the movements of the particles. Such "downward causation" is regarded by some philosophers as untenable (Kim 1998). NRP theorists respond that while the movements of the particles are in a sense causally overdetermined, such overdetermination is not the objectionable sort to which the interactionist Cartesian property dualist is committed. For the psychophysical causal interactions are not fundamental: They are implemented by causal transactions between microphysical events. Mechanics can ignore them. Still some critics charge that the fact that the microphysical event was brought about by another microphysical event leaves no work for the decision to do in bringing it about (Kim 1998). Some NRP theorists reply that this sort of worry is based on a productive conception of causation, and that we should eschew such a conception as unrealistic (Loewer 2002). They maintain that this sort of overdetermination can be accommodated by a kind of regularity account of causation (Melynk 2003), or a kind of counterfactual account of causation (Loewer 2002). This strand of the debate also leads to issues concerning causation and event individuation.

Given the supervenience thesis, any minimal physical duplicate of the actual world would have the same worldwide pattern of distribution of mental events and special science events as the actual world. Why is that the case if mental and special science events are not microphysical events? The leading NRP answer is that all mental and special science events are realized by microphysical events and such realization guarantees this result. While there is no received view of realization, the leading notion is the functionalist notion, according to which the realization relation is the relation of role-occupancy: a realization is a role-player. This idea, however, has been implemented in two different ways (see Block 1980). Role-functionalism implements it one way; filler-functionalism implements it in another (see McLaughlin forthcoming).

According to role-functionalism, every event token of a mental type M is a higher-order event token, an event of participating in some event or other that occupies a certain role R, which includes a causal role. Events that occupy R realize M events, that is, realize events that are exemplifications of M. On this view, higher-order events are never identical with lower-order events. Thus, even if mental events are always realized by microphysical events, no mental event is a microphysical event; similarly, for special science events. This event pluralism is compatible with the supervenience thesis because the basic roles could be filled by microphysical events that fill them in virtue of microphysical laws and conditions.

But NRP theorists would nevertheless face a problem in embracing role-functionalism, for there is a serious question of whether higher-order events have causal effects. While every second-order event is realized by a first order event that has causal effects, a serious question remains whether second-order events themselves have effects. The role-functionalist idea seems most plausible for abilities, but abilities themselves seem not to have causal effects, rather their bases or realizations do. The role-functionalist idea has, however, also been interestingly applied to constituted dispositional states, such as water-solubility, water-absorbency, fragility, ductability, and the like (Jackson, Pargetter, and Prior 1982; Prior 1985). For something to be water-soluble is (arguably) for it to be in some state that, under appropriate conditions, would cause it to begin to dissolve when immersed in a liquid. The state that has the causal role of producing the maninfestation of the disposition (dissolving) is the basis (realization) of the disposition. (Being composed of sodium chloride is one such basis; but the dispositional property is multiply realizable.) It is, however, the basis of water-solubility that causes the substance to dissolve when immersed in water, not the dispositionif the disposition is indeed a second-order state (other accounts of such states are possible). On this role-functionalist conception, the substance's being water-soluble seems to just be the fact that there is some state of it that would (in appropriate circumstances) result in its dissolving were it immersed in water.

The concern, then, is that if (token) mental states and events were functional states and events (i.e., higher-order states and events), they would have no causal effects (Jackson 1996, McLaughlin forthcoming). That would not exclude them from being causally explanatory. The claim that a substance dissolved in water because it is water-soluble provides some information about the causal chain leading to its dissolving (see Prior 1985). But the NRP theorist is after higher-level causation, not just causal explanation. Thus, the NRP theorist must respond to this concern with a compelling account of causation according to which functional states indeed have causal effects. Suffice it to note that the claim that functional states are inefficacious does not presuppose a productive conception of causation (see Lewis 1986).

According to filler-functionalism, an event is of mental type M if and only if it occupies or plays a certain role R, where R includes a causal role. On this view, an event token realizes role R by occupying the roleby filling it. For an event to be of type M is just for it to fill the role. Thus, if E occupies R, then E is thereby of type M. Since the role includes a causal role, filler-functionalists reject token-epiphenomenalism. Note that if, on a particular occasion, event token E is the occupant of R, then "E is the M event" will be a contingent statement of identity, like "Benjamin Franklin is the inventor of bifocals." (The description "the M event" will, like the description "the inventor of bifocals," be nonrigid: it will pick out different things in some possible worlds from those that it picks out in others.)

It may well be that tokens of various types of events can occupy role R, and thus be realizations of M ; if so, then M is multiply realizable. Moreover, events of some type N can realize M, even when N itself is multiply realizable. That will be the case when an event is of type N if and only if it fills a role R*, which includes R as a proper sub-role (Shoemaker 1994). If, on a particular occasion, an event realizes M in virtue of being an N event, and realizes N in virtue of being a C event, then, on that occasion, the C event is the N event, the N event is the M event, and so the C event is the M event.

Notice, then, that, when conjoined with the thesis that every mental event is realized by some microphysical event, filler-functionalism entails that every mental event is a microphysical event. And indeed the filler-functionalist explanation of why any minimal physical duplicate of the actual world will have the same worldwide pattern of distribution of mental (and special sciences) events as the actual world is that the only basic fillers of the roles are microphysical events, which fill them solely in virtue of microphysical laws and conditions. Events are of different orders only relative to types. (Moreover, the ordering here, it has been pointed out, is not one of scale [Kim 1998].) The filler-functionalist account of realization will not serve the NRP theorist's purposes. On the filler view, every event is a microphysical event, and it is ultimately in virtue of microphysical event types that events enter into causal relations. Mental event types are not microphysical event types, both because of actual multiple microphysical realization, and because of the logical possibility of realization without microphysical realization. Nevertheless, they are relevant to whether events of one sort cause events of another since they implicitly type events in terms of patterns of causal relations. And that may very well make them indispensable to certain causal explanations. But whether such a view is correct turns, of course, not only on the nature of causation and the individuation of events, but also on the nature of mental (and special science) properties.

Problems remain, moreover, that are specific to the mental. Some philosophers maintain that neither a role nor a filler-functionalist view is tenable for mental states with qualitative or phenomenal characters: states such that it is like something for the subject of the state to be in the state (e.g., the state of feeling pain). And some embrace Cartesian property dualism for phenomenal mental properties ("qualia"; Chalmers 1996, Kim 2005). They thus reject the psychophysical supervenience thesis. They hold that there could be an exact physical duplicate of the actual world that, unlike the actual world, is entirely devoid of phenomenal consciousness (a "zombie world"; Chalmers 1996). But they do not deny the closure of the physical. And they acknowledge that they may thus very well have to hold that an individual's having a phenomenal property has no causal effects. Suffice it to note that even this restricted epiphenomenalism has an air of paradox. It entails, for instance, that our feeling of pains never cause our pain-behavior, or even our beliefs that we are in pain.

Moreover, even if Cartesian property dualism is rejected for all mental properties, problems remain. Intentional mental states are explanatory, in part, by virtue of their propositional contents. For example, the content that there is a snake in the room figures essentially in both the rationalizing explanation, "He decided not to enter because he believed there was a snake in the room," and the nonrationalizing explanation, "He began to quiver because he feared that there was a snake in the room." The leading theories of content, however, are externalist theories, according to which the content of a mental state fails to supervene on intrinsic states of the subject (Putnam 1975, Burge 1979). On these views, two intrinsic duplicates (e.g., an inhabitant of Earth and her doppelgänger on Twin Earth) could be in intentional states with different contents. Indeed, according to some externalist theories, content depends on historical context (Dretske 1988), and according to others, on social context (Burge 1979). Some philosophers maintain that such highly relational properties are causally irrelevant to behavior, and so must play a noncausal explanatory role. But some philosophers defend the view that intentional states cause behavior, despite being essentially extrinsic (Yablo 1999). Others claim that wide content is causally explanatory because it provides information about the causal history of the agent's behavioral dispositions (Dretske 1988). And others contend that intentional states have an externalist or wide content in virtue of having a "narrow content" in a causal environmental context, and that it is narrow content that is causally relevant to behavior (Jackson 1996). There are other views as well that are as yet less explored. Suffice it to note that these content issues too are matters of ongoing philosophical investigation.

See also Anomalous Monism; Artificial Intelligence; Broad, Charlie Dunbar; Cartesianism; Consciousness; Content, Mental; Davidson, Donald; Descartes, René; Dualism in the Philosophy of Mind; Elisabeth, Princess of Bohemia; Functionalism; Huxley, Thomas Henry; James, William; Kim, Jaegwon; Mind-Body Problem; Nonreductive Physicalism; Philosophy of Mind; Putnam, Hilary; Qualia; Supervenience.


Block, N. Readings in Philosophy of Psychology. Vol. I. Cambridge, MA: Harvard University Press, 1980.

Broad, C. D. The Mind and Its Place in Nature. New York: Harcourt Brace, 1925.

Burge, T. "Individualism and the Mental." Midwest Studies in Philosophy 4 (1979): 73121.

Butler, S. Unconscious Memory. London: A. C. Fifield, 1910.

Chalmers, D. J. The Conscious Mind. New York: Oxford University Press, 1996.

Descartes, René. Meditions on First Philosophy (1641). Edited by Stanley Tweyman. Ann Arbor: Caravan Books, 2002.

Davidson, Donald. "Mental Events." In Experience and Theory, edited by L. Foster and J. W. Swanson. University of Massachusetts Press and Duckworth, 1970. Reprinted in Donald Davidson, ed. Actions and Events. Oxford: Clarendon Press, 1980.

Davidson, Donald. "Thinking Causes." In Mental Causation, edited by John Heil and Alfred Mele. Oxford: Clarendon Press, 1993.

Dretske, F. Explaining Behavior: Reasons in a World of Causes. Cambridge, MA: MIT Press, 1998.

Huxely, T. H. "On the Hypothesis That Animals are Automata, and Its History." Fortnightly Review 16 (1874): 555580. Reprinted in Huxley, Collected Essays, Vol. 1. New York: J. A. Hill, 1904.

Jackson, F. "Epiphenomenal Qualia." Philosophical Quarterly 32 (1982): 127136.

Jackson, F. From Metaphysics to Ethics. Oxford: Oxford University Press, 1998.

Jackson, F. "Mental Causation: The State of the Art." Mind 105 (1996): 377413.

Jackson, F., R. Pargetter, and E. Prior. "Functionalism and Type-Type Identities." Philosophical Studies 42 (1982): 209225.

James, W. The Principles of Psychology (1890). Cambridge, MA: Harvard University Press, 1983.

Kim, J. Mind in a Physical World: An Essay on the Mind-Body Problem and Mental Causation. Cambridge, MA: MIT Press, 1998.

Kim, J. Physicalism, or Something Near Enough. Princeton, NJ: Princeton University Press, 2005.

Kim, J. Supervenience and Mind: Selected Philosophical Essays. New York: Cambridge University Press, 1993.

Lepore, E., and B. Loewer. "Mind Matters." Journal of Philosophy 84 (1987): 630642.

Lewis, D. "An Argument for the Identity Theory." Journal of Philosophy 63 (1966): 1725.

Lewis, D. "Events." In Philosophical Papers II, edited by D. Lewis, 241269. Oxford: Oxford University Press, 1986.

Loewer, B. "Comments on Jaegwon Kim's Mind and the Physical World." Philosophy and Phenomenological Research LXV (2002): 655662.

Macdonald, C., and G. Macdonald. "How to be Psychologically Relevant." Debates on Psychological Explanation. Vol. I, 6077. Oxford: Blackwell, 1995.

McLaughlin, B. P. "Type Epiphenomenalism, Type Dualism, and the Causal Priority of the Physical." Philosophical Perspectives 3 (1989): 209235.

McLaughlin, B. P. "The Rise and Fall of British Emergentism." In Emergence or Reduction?, edited by A. Beckermann, H. Flohr, and J. Kim, 4993. Berlin and New York: Walter de Gruyter, 1992.

McLaughlin, B. P. "On Davidson's Response to the Charge of Epiphenomenalism." In Mental Causation, edited by John Heil and Alfred Mele, 2740. Oxford: Clarendon Press, 1993.

McLaughlin, B. P. "Epiphenomenalism." In A Companion to the Philosophy of Mind, edited by S. Guttenplan, 277289. Oxford: Blackwell, 1994.

McLaughlin, B. P. "Is Role-Functionalism Committed to Epiphenomenalism?" Journal of Consciousness Studies. Forthcoming.

Melnyk, A. A Physicalist Manifesto. Cambridge, U.K.: Cambridge University Press, 2003.

Prior, E. Dispositions. Scots Philosophical Mongraphs. Aberdeen: Aberdeen University, 1985.

Putnam, H. "The Meaning of 'Meaning,'" In Mind, Language and Reality, edited by H. Putnam, 215271. Cambridge, U.K.: Cambridge University Press, 1975.

Shoemaker, S. Identity, Cause, and Mind: Philosophical Essays. Cambridge, U.K.: Cambridge University Press, 1994.

Sosa, E. "Davidson's Thinking Causes." In Mental Causation, edited by edited by J. Heil and A. Mele. Oxford: Clarendon Press, 1993.

Ward, S. L. "The Conscious Automism Theory." Lecture XIII of Naturalism or Agnosticism, vol. 2, 3464. Adam and Charles Black, 1895/1903.

Watson, R. A. The Breakdown of Cartesian Metaphysics. Atlantic Highlands, NJ: Humanities Press, 1987.

Yablo, S. "Mental Causation." Philosophical Review 101 (1992): 245280.

Yablo, S. "Wide Causation." Philosophical Perspectives 11 (1999): 251281.

Brian P. McLaughlin (1996, 2005)