4. Reductive Physicalism
Descartes thought that animals do not have minds. Strictly speaking, animals cannot feel pain or be hungry or in distress, as these are mental states. What distinguishes animals from dead matter is that they are automata, self-moving, whereas dead matter can be set in motion only by an external force acting upon it.
Even the most hard-hearted Cartesian will agree that there are overwhelming similarities between animals and humans. Physiological processes in animals are very similar to physiological processes in humans. Mental states of humans are accompanied by physiological states of their bodies, many of which correspond to similar physiological states of animals. A human being, on Descartes’s view, appears to be an automaton, not too different from an animal, that is connected to a soul.
Due to these similarities, rather than denying that animals have mental states, it is more plausible to accept that animals have some mental states, at least those corresponding to physiological processes they have in common with humans. If animal bodies are not connected to thinking substances, these mental states would themselves be physiological states. Once we agree to this, what prevents us from concluding that all mental states are physiological states? It is plausible to extend Descartes’s views about animals to humans: they, too, are very complex automata, distinguished from other animals merely by the extraordinary complexity of their physiology.
This view was adopted by La Mettrie, who interestingly considered himself a Cartesian of sorts. Thinking is not something done by mental substances. It is a physiological process. As La Mettrie put it, the brain has muscles for thinking just as the leg has muscles for walking. La Mettrie observed that Descartes’s mental substances are superfluous to account for mentality once the implausible view is dropped that animals do not have mental states. The problem of mental causation can now be addressed by noticing that, because mental states are physical states, there is no problem of how the mind can cause something to happen: it is the same phenomenon as physical causation.
4.2 Three Reasons for Reductive Physicalism
In the twentieth century, the mind-brain identity theory has been put forward most prominently by J.J.C. Smart, following U.T. Place. It is also called reductive physicalism. Mental states are reduced to physical states in the sense that they are nothing but brain states. The reduction is ontological: there are only brain states, some of which are also mental states, but no mental states in addition to brain states. Three reasons speak in favour of it.
Simplicity. One reason to accept that mental states are just states of the brain is methodological. Science has made great progress in explaining the world in physical terms, positing only material objects subject to the laws of physics. Following this trend, even though we are not yet in a position to do so, we should expect that eventually we will be in a position to explain living organisms, including their consciousness and mental states, mechanistically in terms of their physiology and thus ultimately in terms of physics. Assuming that mental events are physical events is the simplest theory of the mental, given what else we know about the world.
The Principle of Parsimony is the assumption that the world behaves in the most economic fashion: for instance, there are no superfluous laws of nature, nor objects the existence or non-existence of which would make no difference to the rest of the world. Mental substances are superfluous in the explanation of the world, as everything can be explained only on the basis of physical substance. The simpler theory has exactly the same explanatory power as the more complex one and is thus to be preferred.
Ockham’s Razor is the principle that entities must not be multiplied beyond necessity. A theory that explains the phenomena on the basis of positing fewer fundamental entities is to be preferred over one that posits more. If the mind is identical to the brain, then this keeps the ontology simpler than assuming that there are two kinds of substances.
Although motivated by methodological principles of scientific theorising, the mind-brain identity thesis itself is not a scientific thesis. It cannot itself be decided by science, as there are no experiments which could rule out that there are mental phenomena over and above brain-processes, as long as the mental phenomena behave in as orderly a fashion as is required for them not to contradict scientific evidence. The strategy crucially depends on the absence of cogent reasons for dualism.
Explanatory Strength. The correlations between mental processes and brain processes stand in need of explanation. Anyone positing mental phenomena in addition to brain processes needs to formulate laws connecting the two to explain the correlations. The identity of mental states with brain states is the best explanation for the correlation. It is the closest possible one, as there is no gap between the mental and the physical at all. It is also the best explanation for why some mental states (e.g. pain) are followed so regularly by other mental states (e.g. distress). The identity thesis explains this as a case of causation: pain causes distress, because pain is a state of the brain which causes a state of the brain which is distress.
Mental Causation. The biggest problem for Descartes’s dualism is to explain how mental and physical substances can interact. Descartes claims that there is causal interaction between the two, but we have no plausible model of causal relations between material and immaterial substances. Thus mental causation is inexplicable. If the mind is identical to the brain, there is no problem: mental causation is a case of physical causation. Furthermore, it is plausible that the realm of the physical is causally closed: every physical event has a physical cause. So, if there is mental causation, either mental events are physical events, or a physical event which is caused by a mental event is causally overdetermined and has both a mental and a physical cause. The latter option is unattractive. The mental event seems to be superfluous in bringing about the physical event, so it would not actually be a case of genuine mental causation. Thus, as there is mental causation, mental phenomena must be physical phenomena.
We can put the point even more forcefully, using an argument due to David Lewis. What makes a mental event what it is are its typical causes and effects. For instance, what makes a mental state pain are its typical causes and effects such as tissue damage, distress and avoidance behaviour. An event that typically fails to have these causes and effects is not pain. It suffices that these causes and effects are typically linked by pain. There may be nothing in common to all causes of pain or to all that pain causes. Occasionally, an organism may be in pain without showing avoidance behaviour, because there are other reasons for not avoiding the cause of the pain. You might be stoic and ignore your pain, or the pain might lead to pleasure that outweighs the pain. There is no requirement that the typical causes and effects of a mental event are always present or entirely uniform in every case in which the event occurs. Typical causes and effects may be fairly vague and general. Maybe all there is to pleasure is that it causes a desire to seek out its cause again. Desires coupled with beliefs of how to satisfy them cause actions. Typical causes and effects of mental events are also other mental events, but the whole network of typical causes and effects among mental states is tied to typical causes and effects in the merely physical world. A mental event is what it is because of the place it occupies in this causal network. Mental states are individuated by their causal roles.
So far, this account is metaphysically neutral, apart from the assumption that mental events have typical causes and effects. We haven’t said anything yet about the nature of mental states or the nature of causation. But if we assume the closure of physics, it follows that mental events must be physical events, as then only physical events can have physical causes and effects. The conclusion is even stronger: anything non-physical cannot be mental either.
Empirical investigation suggests that we will discover that the causal roles we attribute to mental events are occupied by neurological events. These then are the mental events. This case for the identity thesis is not conclusive, but depends on the outcome of future empirical research. It may be that we won’t find any physical events that occupy the causal roles we think our mental events have. Even where we succeed to make a good case, further enquiry may show that the initial identification was mistaken. If we have to reject the evidence as inconclusive or even defective, the only option would appear to be to reject the closure of physics.
4.3 A Dualist Rejoinder
The identity theory makes a very strong claim about the nature of the correlation between mental states and brain states: there is only one thing to start with. Being in pain just is to have a certain brain state—philosophers often use the dummy term ‘c-fibre excitation’ or ‘c-fibres firing’ to refer to whatever that brain state might be. As shown in the section on Descartes, if a and b are identical, then they are necessarily identical, and if a and b are different, they are necessarily different. For the identity theorist, if pain is identical to c-fibres firing, that is necessarily so. There could be no pain without c-fibres firing, and no c-fibres firing without pain.
On the face of it, there is no contradiction in assuming that pain might be correlated with a different brain state. The identity theorist, however, is committed to exactly that claim. If pain is identical to c-fibres firing, then assuming that it might be correlated with a different brain state is to assume that pain is and is not identical to c-fibres firing. The identity theorist cannot, as it were, lift the mental state off the brain state and correlate it with another one. There is no distinction between them that could allow such a change in correlation.
The Cartesian intuition is that the correlation between mental states and brain states is not necessary but contingent. If the correlation could be different, then pain cannot be identical to a brain state. There are no contingent identities and no contingent differences. If it is possible that pain is not identical to c-fibres firing, then it isn’t, as possible differences are actual differences.
The argument generalises. If a mental state could fail to be correlated with whatever brain state it is in fact correlated with, then it is not identical to it. Descartes’s argument for substance dualism is based on the plausible assumption that it is not contradictory to assume that our mental lives might be phenomenologically just as they are now, while our physiology is different or, indeed, lacking entirely, if perhaps we are deceived by an evil demon. Cartesian Dualism, even if it may not be actually true, nonetheless appears to present a genuine possibility. If this is correct, then the correlation between mind and brain is weaker than identity, like a suitable notion of supervenience.
Saul Kripke has given this Cartesian line of argument some prominence. A rigid designator is a way of picking out the same individual in every possible circumstance. Proper names are in general rigid designators. If I say that Newcastle could have been in Scotland, then I’m talking about a possible situation in which the place we name when we use ‘Newcastle’ in sentences like ‘Newcastle is in England’ is in Scotland. Suppose we name a pain ‘p’ and the brain state it is correlated with ‘b’. It seems to be logically possible that p and b are not correlated with each other, be it either that p exists without b existing or that b exists without p existing, i.e. the pain might be there without the brain state or the brain state without the pain. If ‘p’ and ‘b’ are rigid designators, then they pick out the same thing in every possible circumstance, so ‘p’ names the same thing in a situation in which it is correlated with b as well as in those in which it isn’t. So p and b are not identical.
Kripke adds the following observation. Pain feels a certain way. It is an essential property of pains that they hurt. If something doesn’t feel like pain, it’s not pain, and whatever feels like pain is pain. We determine the referent of ‘pain’ by that essential property, the way it feels. This is why ‘p’ is a rigid designator. Brain states are not picked out by the way they feel, but we can refer to them rigidly by indicating the kind of physical thing they are, which is why ‘b’, too, is a rigid designator. This accounts for the intuition that we can be in a situation that is phenomenologically just like a situation in which we are in pain, but there are no c-fibres firing. As a situation that is like one of being in pain just is a situation of being in pain, pain and c-fibres firing can come apart and therefore can’t be identical.
The identity theorist would have to deny that anything could feel like pain unless it is c-fibres firing and conversely must hold that it is impossible that there are c-fibres firing without this being felt as pain. Maybe the apparent possibility that pain is correlated with another brain state is an illusion or a conceptual confusion. We’ll come back to this line of argument in a later section.
Kripke, S. ‘Selections from Naming and Necessity’ in Rosenthal, D. M. (ed.) The Nature of Mind (Oxford: Oxford University Press, 1991)
Lewis, D. ‘An Argument for the Identity Theory’ The Journal of Philosophy 63 (1966): 17–25
Papineau, D. Thinking about Consciousness (Oxford: Oxford University Press, 2002), chapter 1
Place, U.T. ‘Is Consciousness a Brain Process?’ British Journal of Psychology, 47 (1956): 44–50
Smart, J.J.C. ‘Sensations and Brain Processes’ The Philosophical Review 68 (1959): 141–156
de La Mettrie, J.O. ‘Machine Man’ in Thomson, A. (ed.) Machine Man and other Writings (Cambridge: Cambridge University Press, 1996)
Lewis, D. ’Psychophysical and Theoretical Identifications’ Australasian Journal of Philosophy 50 (1972): 249–258
Lewis, D. ‘Reduction of Mind’ in Guttenplan, S. (ed.) A Companion to the Philosophy of Mind (Oxford: Blackwell, 1994)
Place, U.T. ’E.G. Boring and the Mind-Brain Identity Theory’ British Psychological Society, History and Philosophy of Science Newsletter 11 (1990): 20–31
Rosenthal, D.M. ‘Identity Theories’ in Guttenplan, S. (ed.) A Companion to the Philosophy of Mind (Oxford: Blackwell, 1994)