Reductionism in the Philosophy of Mind

views updated


Reduction can be understood in a loose or in a strict sense. In the loose sense, entities (or expressions) of a given type are reduced if they refer to "nothing over and above" other entities (expressions) that we consider well established. This is consistent with the conclusion that the reduced entities are among the posits of a mistaken world view and thus have no place in our ontology, and it is also consistent with the conclusion that the reduced entities are conserved among other accepted, better established or understood entities. In the first case we have elimination, and proposing this for entities of a given kind makes us eliminativists about those entities. In the second case we have reduction in the strict sense, and proposing this for a given kind makes us reductionists (sometimes called "conservative" or "retentive" reductionists). Reductionist projects can also be semantic or theoretical. A semantic reduction attempts to show that items belonging to a certain class of expressions are semantically equivalent tothat is, definable in terms ofanother class of expressions. A theoretical reduction aims at showing that a given scientific theory can be fully subsumed under (that is, derivable from) another more basic theory.

Types of Mind-Body Reductionism

In the philosophy of mind, reductionist projects come in all formats. A reductionist effort will typically be directed against the claim that the mental has some real, independent status. But this claim has a range of versions that go from the mind being a nonphysical/biological object, to mental properties constituting a level of sui generis properties of organisms that is in some sense autonomous vis-à-vis the physical/biological properties, to mental expressions possessing meanings that cannot be accounted for in purely behavioral/physical terminology.

The substance dualist assertion influential until the twentieth centurythat the (human) mind is an immaterial object or substancehas faced widespread philosophical criticism of an eliminativist type: "Immaterial mind" or "soul," like "élan vital," "elf," or "chupacabras," are ghostly expressions that come from mistaken frameworks or conceptions and do not refer to anything. An influential formulation of this view is Gilbert Ryle's claim that the immaterial entity posited by substance dualism is the result of a category mistake in which we reify our mental activities by placing a ghost in charge of our body. Another major reason for the eliminativist consensus about nonmaterial substances is the inability of a nonphysical substance to causally interact with the physical world, because of conservation of energy considerations and because of the difficulty of making sense of bridging mechanisms between the two ontologically diverse realms. Absent causal interaction, the argument goes, postulating souls seems pointless if not absurd.

Eliminating mental substances, however, does not directly lead to a reductive view of the mental. In the twentieth century substance materialism or physicalism has been the orthodoxy in tune with modern science, but "the reducibility of mind" has remained as a philosophical issue of first importance. It is only that the focus of the debate has now shifted to the ontological or semantic autonomy of mental properties or predicates. The first systematic attempt to fully reduce the mental to the physical comes from logical behaviorism, a position championed by Rudolf Carnap, Carl Hempel, and Gilbert Ryle in the 1930s and 1940s. The view has doctrinal connections to methodological behaviorism, the dominant methodology of psychology in the first half of the twentieth century.

Based on the logical positivist's verification criterion according to which the content of an expression is just the expressions' verification conditions and on the assumption that these conditions have to be publicly observable, logical behaviorism argues that in order for sentences including mental expressions to be meaningful they have to be translatable without loss of content into sentences including just behavioral and other physical expressions. This implies that mental expressions should be defined in terms of behavioral and other physical expressions. Following the model of definitions of dispositional properties in the natural sciences, these definitions standardly include conditional sentences showing dispositions to behave under given environmental circumstances including stimuli. So logical behaviorism is a form of semantic reduction of the mental.

Logical behaviorism has been largely abandoned for several reasons, one of them being its inability to meet the positivist standards in its own reductionist strategy. Most mental terms cannot be associated with a single behavioral disposition; there is no single behavioral manifestation of, say, "believing in God" or "loving one's country." If mental terms denote behavioral dispositions, these dispositions must be "multitracked," and this would make behavioral definitions of mental terms enormously complex. This makes the behaviorist project of defining mental terms a highly dubious project.

Moreover, it has been convincingly argued that even in simple cases a purely behavioral definition just is not possibleunless one uses some mentalistic term in the definition, which of course undermines the behaviorist enterprise. The fall of behaviorism as the accepted reductive view led to a different reductionist approach. In the 1950s U. T. Place, J. C. C. Smart, and Herbert Feigl proposed the mind-body identity theory, a simple and appealing view in line with the surge of neural research. According to the view, while there is no meaning equivalence between mental and neural terminology (thus no semantic reduction) mental states are just states of the brain or the nervous system. The claim is one of numerical identity between types of states or properties and as such it involves ontological reduction in the strict sense.

A main line of argument for the identity theory is based on ontological simplicity, a standard strategy for ontological reduction. Once we have observed a pervasive set of systematic correlations between mental occurrences and neural events, the argument goes, we should conclude that the mental and the neural are identical. For while mind-brain correlations are compatible with a range of views about the mind, simplicity dictates that we should not multiply entities that are not going to enhance our explanatory power. The view is also supported by considerations of theoretical reduction. The history of science offers countless cases of predicates of everyday frameworks being reduced to predicates of explanatorily richer scientific frameworks (a standard example is the reduction of temperature [of gases] to molecular kinetic energy). Given the advances in the neurosciences we have good reason to think a neural reduction of mentality is going to be one more item in a chain of successful theoretical reductions. This theoretical reduction would proceed by establishing "bridge laws" between mental and neural predicates and then reducing all generalizations involving the mental to the more encompassing laws of neuroscience.

Of the many attacks raised against the identity theory, two have aimed at the core of its reductive stance. Donald Davidson has argued against type-identification by claiming that there cannot be laws connecting the mental and the physical (this is called anomalism of the mental, an essential part of Davidson's nonreductive view discussed below). Mental states, in particular intentional states such as beliefs and desires, are governed by principles of rationality without which attribution of mentality would be impossible. Laws connecting the physical and the mental would constrain the mental by the principles of physical theory and thereby undermine its own peculiar rationality constraints.

Another highly influential argument against the identity theory is the "multiple realization" argument initially developed by Hilary Putnam. The identity theory requires a single physical property be the reduction base for each mental state. But surely the same mental state can occur in organisms with diverse neurophysiological structures. Nonhuman animals can be in pain and we can conceive of noncarbon based species and perhaps even artificial creatures being in pain. Mental states, Putnam argues, can be implemented or "realized" in widely diverse physical/chemical structures and so there is no unifying reduction base or structure for them. (This multiple realization objection is also at the core of the nonreductive functionalist approach discussed below.)

An alternative, eliminativist stance was defended in the 1960s by Richard Rorty and Paul Feyerabend and has as more recent versions the views of Patricia Churchland, Paul Churchland, and Stephen Stich. Learning from the failure of the identity theory to establish type-type identities between mental and neuro-chemical properties, eliminativism claims that the mental expressions used in our everyday psychological talk have no more reality or significance than "phlogiston" and "caloric fluids," terms of superseded and discarded scientific theories. It is highly unlikely that these concepts of vernacular psychology could be sharpened into concepts that will be useful to the sciences and do not correspond to the concepts of the sciences (neuroscience or cognitive science) that have the task of explaining human behavior. This radical view proposes to eliminate mental terminology for the purposes of scientific theorizing and can go as far as predicting that a full replacement is possible even for everyday purposes. The analogy with concepts in the history of science that were found to be fully misguided and therefore replaced plays an important role in the argumentation in favor of eliminativism. This view has been found by most philosophers to be unacceptably extreme since it means that an essential component of our conceptual framework has to be given up. Also, some have argued that the view is incoherent since the view cannot be expressed without the very (mental) concepts it rejects (since in the very act of affirming their view, the eliminativist is expressing a belief, something that, according to their view, does not exist).

Types of Mind-Body anti-Reductionisms and the Reductivists' Reactions

Starting in the late 1960s, the problems plaguing reductive views let to the establishment of nonreductive physicalism as a reigning orthodoxy in the philosophy of mind. Its two most salient versions are anomalous monism and functionalism. Functionalism in fact has been the predominant view into the twenty-first century.

Davidson's anomalous monism is a physicalist view that eschews reduction. From the principles that every singular causal relation needs to be backed by strict laws (nomological character of causation) and that there are no "strict" laws about mental properties (mental anomalism), together with the assumption that at least some mental events causally interact with physical events, Davidson concludes that mental events must be identical with physical events. According to Davidson, this provides causal efficacy to mental events, even though there are no strict psychological laws governing them, and it also leads to a nonreductive view of the mental because there are no laws connecting mental properties with physical properties.

Many critics have argued that Davidson's view leaves the mental with no causal role to play. Davidson is entitled to affirm that a mental event causes a physical event (by being identical to a physicalprobably neuralevent). Now, an event instantiates a lawrequired for causationin virtue of some of its properties, or, in other words, in virtue of falling under some event-type. Since anomalism entails that there are no laws involving mental properties or event-types, it is the physical (neural) properties of the cause event that are efficacious in the production of the effect. The fact that the cause event falls under a mental type, or the fact that the event has mental properties, is completely irrelevant for the event's causing the effect. Thus, critics conclude, Davidson's anomalous monism renders the mental epiphenomenal, making it an easy target for elimination.

The functionalist view of the mental defended by Putnam and Jerry Fodor, among others, starts with the anti-reductivist stance included in the multiple realization argument. Its positive view includes the claim that mental properties are functional properties, rather than physical/neural properties as claimed by the identity theory. On the functionalist view, for something to have a mental property M is for it to instantiate some physical property P that has the right causal connections with inputs, behavioral outputs and other mental states. Thus, a mental property is a second-order property of having a (first-order) property that fulfills a certain specified causal specification. A first-order property meeting the causal specification is called a "realizer" or "realizing property" of the second-order functional property. For any given mental property there will likely be indefinitely many realizing properties satisfying its causal specification.

The reductionist can challenge the functionalist by suggesting that the mental property be identified with the disjunction of realizers. Settling this challenge would require a metaphysical discussion on the nature of disjunctive properties. A more powerful challenge raised by Kim is the claim that since having the functional mental property implies having one of its realizing properties and since the casual powers of the instance of a functional property must be considered to be inherited from the causal powers of the realizing property, mental properties have no autonomous causal powers and so are epiphenomenal. To the reply that it is the mental kind and not the instance that has its own causal powers Kim answers that the sheer heterogeneity and diversity of the realizers of a functionally conceived mental property deprives the property of the kind of causal-nomological unity required for nomological and causally efficacious properties.

All versions of nonreductive physicalism (including anomalous monism and traditional functionalism) are targets of the exclusion argument initially put forth by Norman Malcolm and developed by Jaegwon Kim. Physicalists, even those in the nonreductive camp, accept the primacy of the physical not only in terms of substance monism but also in terms of physical properties being primary vis-à-vis mental properties. This commitment includes, according to Kim, accepting the causal closure of the physical and accepting a strong sense of dependence of the mental upon the physical. Thus, every physical event, including human behavior, has to have a complete physical cause. The mental event that is supposed to be the cause of behavior is preempted of its causal role by the physical state upon which it depends and which is the required physical cause of behavior. The upshot is that we cannot attribute a causal role to the mental unless it is identified with the physical, transforming nonreducible mental properties into epiphenomena. And epiphenomena, Kim thinks, should be cut from our ontology because they serve no purpose.

A common theme across several discussions so far has revolved around whether the mental, on one view or another, has autonomous causal powers. It is not obvious whether causal reduction or elimination implies full ontological reduction or elimination, that is, whether putative entities that are causally inefficacious or epiphenomenal can still be bona fide entities. To achieve full reduction we need the extra assumption that independent causal powers are necessary for the very reality of an entity. This view has been explicitly defended by Kim and Sidney Shoemaker, among others, and is largely the orthodox view. A negative answer (supported for instance by Elliott Sober and Marcelo Sabatés) makes room for epiphenomenalism as a nonreductive option about the mind.

In the first decade of the twenty-first century reductionism has gained some momentum. Kim has developed an influential functionalist view of reduction with ties to the version of functionalism defended by David Lewis in the 1970s. Kim's position, in agreement with his criticism of traditional functionalism á la Putnam, claims that "functionalizing" a property provides a form of theoretical reduction that does not require bridge laws and fully explains its reductive relationship on the base property. The view implies that on account of its multiple diverse realizability, no mental property has sufficient causal/nomological homogeneity to count as a genuine, projectible property useful in science. Instead, it proposes that we eschew talk of mental properties in favor of mental predicates or concepts that at most we get a pragmatically useful mental predicate. In making this move, functional reductionism appears to turn itself into a form of eliminativism with regard to mental properties.

See also Alexander, Samuel; Anomalous Monism; Broad, Charlie Dunbar; Davidson, Donald; Eliminative Materialism, Eliminativism; Emergence; Frege, Gottlob; Knowledge Argument; Logic, History of; Metaphysics, History of; Mind-Body Problem; Moral Realism; Morgan, C. Lloyd; Multiple Realizability; Phenomenalism; Philosophy of Mind; Philosophy of Science, History of; Philosophy of Science, Problems of; Physicalism; Properties; Qualia; Reduction; Russell, Bertrand Arthur William; Set Theory; Supervenience.


Block, Ned. "Antireductionism Slaps Back." Philosophical Perspectives 11 (1997): 107132.

Carnap, Rudolf. "Psychology in Physical Language" (1931). Translated in Logical Positivism, edited by A. J. Ayer. Glencoe, IL: Free Press, 1959.

Churchland, Patricia. Neurophilosophy. Cambridge, MA: MIT Press, 1986.

Churchland, Paul. "Eliminative Materialism and the Propositional Attitudes." Journal of Philosophy 78 (1981).

Davidson, Donald. "Mental Events." In Essays on Actions and Events. Oxford: Oxford University Press, 1980.

Fodor, Jerry. "Special Sciences, or the Disunity of Science as a Working Hypothesis." Synthese 28 (1974): 97115.

Fodor, Jerry. "Special Sciences: Still Autonomous after All These Years." Philosophical Perspectives 11 (1997): 149163.

Hempel, Carl G. "The Logical Analysis of Psychology" (1935). In Readings in Philosophy of Psychology, vol. 1, edited by Ned Block, 1423. Cambridge, MA: Harvard University Press, 1980.

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

Kim, Jaegwon. Mind in a Physical World. Cambridge, MA: MIT Press, 1998.

Kim, Jaegwon. Supervenience and Mind. Cambridge, U.K. and New York: Cambridge University Press, 1993.

Lewis, David. "Psychophysical and Theoretical Identifications." Australasian Journal of Philosophy 50 (1972): 249258.

Place, U. T. "Is Consciousness a Brain Process?" British Journal of Psychology 47 (1956): 4450.

Putnam, Hilary. "The Nature of Mental States." In Mind, Language and Reality. Cambridge, U.K.: Cambridge University Press, 1978.

Rorty, Richard. "Mind-Body Identity, Privacy and Categories." Review of Metaphysics 19 (1965).

Ryle, Gilbert. The Concept of Mind. New York: Harper and Row, 1949.

Sabatés, Marcelo. "Being without Doing." Topoi 22 (2003).

Shoemaker, Sidney. "Causality and Properties." In Identity, Cause and Mind, edited by S. Shoemaker. Cambridge, U.K.: Cambridge University Press, 1984.

Smart, J.C.C. "Sensations and Brain Processes." Philosophical Review 68 (1959): 141156.

Sober, Elliot. "A Plea for Pseudo-Processes." Pacific Philosophical Quarterly 66 (1985).

Sosa, Ernest. "Mind-Body Interaction and Supervenient Causation." Midwest Studies in Philosophy 9 (1984).

Stich, Stephen. From Folk Psychology to Cognitive Science. Cambridge, MA: MIT Press, 1983.

Marcelo H. Sabatés (2005)