Dualism: An Empirical Test?

The philosophy of mind is one of the most fertile breeding grounds for the thought experiment, sending up clouds of the things like pregnant mosquitoes in a conceptual swamp. But, like mosquitoes, the thought experiment is apt to bite the hand that feeds it. If one is not scrupulously careful, an inadvertently begged question can turn an argument for one's theory into a test of one's underlying assumptions. Indeed, what was presented as a thought experiment can turn out to be a potential (if usually far-fetched) empirical experiment. A prime example is to be found in Wiggins (1967), and was later discussed in Parfit (1971). It's an extension of the familiar brain-transplant examples, and Parfit describes it succinctly:

My brain is divided, and each half is housed in a new body. Both resulting people have my character and apparent memories of my life. (Parfit [1971], p. 5)

There are three possible results: that I don't survive the operation, that I survive as only one of the two people, and that I survive as both of them.

Now, as a matter of fact it is possible to survive with only half a brain (though one is restricted to a career in politics), but for the rest we have to make a couple of assumptions. First, we have to assume that a brain can be transplanted from one human being to another, the `owner' of the brain surviving the operation. Secondly, given the possibility of survival with only one brain hemisphere, and given the first assumption, it seems reasonable to assume that a semi-encephalic patient could survive if her remaining hemisphere were transplanted. Given all this, there seems to be no obstacle to a patient's survival if her brain were extracted and only one hemisphere transplanted, the other being destroyed. But then there's surely no obstacle to such a patient (call her Renée) surviving the split-brain transplant, in which the two hemispheres are transplanted into different skulls. As Parfit asks: "How could a double success be a failure?";(loc. cit.) that is, it seems unreasonable to suppose that Renée wouldn't survive at all.

Moreover, what reason could we have for claiming that Renée survived as one of the resulting people rather than the other? We needn't even make the simplifying (and false) assumption that the two hemispheres are identical; there's no reason to choose one as being Renée (the other being... who?).n1 So that leaves only the possibility that she survive as both people (the possibility with which Parfit is concerned).

It might, of course, be objected that we need to ask various questions concerning the factual back ground of the thought experiment -- for example, we need to know more about the actual nature of the two halves of the brain. One approach to this is typified by Dan Robinson's argument against the production of two Renées:

Since brain function is not a constant over the life of the individual and since we already know that the two hemispheres are neither symmetrical in function nor identical in, to use a less than felicitous term, content, we have no reason to expect that transplanted hemispheres will constitute transplanted identities.(Robinson [1976], p.77)

Of course, Robinson is concerned here with something like identity rather than with Parfit's notion of survival, though I take it that he's denying that Renée survives as either of the new people. But in any case this isn't the sort of objection I have in mind here; on the one hand, as I've said, I'm happy to accept all sorts of simplifying assumptions, and on the other, I'm not centrally concerned with exactly who results from the transplant. My worry centres on Wiggins' and Parfit' shared assumption that each of the new body-and-half-brain combinations will be persons. Each writer sees the only question as concerning which, if either, of those persons Renée should survive as: both, one, or neither. This is more than just a simplifying assumption; it surely begs a question at the heart of the philosophy of mind.

Let us return to the thought experiment, flesh it out a little, and consider an extension to it. Imagine that, after years of successful brain transplants and of thorough research into the physiology of the brain, medical science is ready to try the split-brain transplant. A brain is divided, and each hemisphere placed in its new body; the shocking result is not two surviving persons, but one person and a vegetable on a life- support machine. The second hemisphere to be transplanted always produces a new person (the survivor), while the first always produces a vegetable. Extensive investigation reveals no significant physical difference between the two hemispheres, and no physical difference between the two parts of the operation. The experiment is tried again, and again -- and the result is always the same: one person, one vegetable. What's the explanation? How could a double (physical) success result in a such a (mental) mixed bag? There is one clear and obvious answer: each person has one indivisible mind, whose connections with that person's brain are intimate and strong. We can divide the brain, but not the mind, which therefore attaches itself to just one hemisphere.n2

Such an experiment becomes, then, (allowing for the problems indicated by the simplifying assumptions) an empirical test of dualism. It's rather one-sided, of course, for even if dualism is the only (or at least the best) explanation of the failure of split-brain transplants, their success would not count against a dualist theory. Modified success, indeed, might count as evidence for a different form of dualism; for example, both halves might give us surviving persons, but only one of them be self-conscious, or adult, or recognisably Renée -- but that sort of speculation goes beyond the very limited aims I've set myself here.n3

Returning to the original Wiggins-Parfit example, we know that, if we assume that a logically possible result of the experiment is two people, we are ruling out a certain dualist conception of the mind. If the mind is a distinct, indivisible substance, then it's logically impossible that Renée survive as two people. Even if the result of the split-brain transplant is two people, only one of them can be Renée.

Note that the dualism for which this is a test is the real, full-blooded sort -- the postulation of (at least) two kinds of thing, one mental, the other physical, which interact in complex ways (and whose combination is what we call a person). One response to my argument comes from those who reject such a notion out of hand, and who consider that, "since there are good empirical arguments against [such `interactionist' dualism], there is good reason not to expect the experiment to work as [I] suggest it might."n4 Apart from the dubious assumption that it's impossible for there to be empirical evidence on two sides of a disagreement (especially a metaphysical disagreement), this sort of comment raises an important question for the philosophy of mind.

It's often stated, though generally informally and out of print, that substance, or interactionist, dualism is ruled out on empirical grounds. Now, what exactly are these grounds? Of course, if one defines causality as being a physical relation (whatever `physical' means in modern physics), one can simply rule out mind-body causal interaction by definition, but presumably that's not what's meant; apart from its question-begging nature, such a response is hardly empirical. The same goes for appeals to the principle of causal closure; whatever its status, it isn't an empirically discoverable claim.

Perhaps there's a confusion (in fact, there's certainly a confusion among some people, including many philosophy undergraduates and non-philosophers) between the claim that there's empirical evidence for a close relation between mental and neurophysiological processes, and the claim that there's empirical evidence against the distinct existence of the mental. Or perhaps anti-dualists have in mind the sort of criticism that was offered against Descartes' position by his contemporaries (and considered important and difficult to deal with by Descartes himself) -- criticism based upon their shared, crudely mechanical notion of efficient causation. Yet, even apart from the question of the acceptability of that concept as central to such criticism, we should note that, again, it's simply not empirical but philosophical.

This isn't the place to go into the notion of causation in any depth, but it's worth noting that W.D. Hart, in his fascinating and detailed defence of interactionist dualism (Hart [1988]), considers the implications for this issue of regularity theories of causation, and concludes that they trivialise the problem, because they imply that

there is no more difficulty about brute constant conjunction between mental and physical events than between some physical events and others (Hart [1988], p.59).

When we examine the work of those many philosophers who reject interactionist dualism, such as Donald Davidson, Peter van Inwagen, and David Lewis, John Searle, we find that none of them offers (or claims to offer) empirical arguments against the theory. Insofar as they appeal to empirical facts at all, these are intended to back up their arguments for the truth of their own positions.

The rejection of what I've called full-blooded dualism is in fact an assumption made by `cognitive scientists', neurophysiologists, and the like, not a conclusion drawn from their work. That this is not noticed by many philosophers is more than a little worrying. Happily, the philosophical fashion that, for example, encouraged the sneering use of `Cartesian' as an insult, often by those who have hardly taken the trouble to read or think about Descartes, shows some sign of passing. The sooner the better.

My point stands; if the sort of experiment that I described above were performed, and if the result were as I've discussed, that would constitute empirical grounds for (interactionist) dualism. There are no reasons for ruling out that (or any other) result. When philosophers do make assumptions about the outcome of such thought experiments, they risk falling into the trap that I described at the beginning of this paper, by failing to notice that an inadvertently begged question can turn an argument for one's theory into a test of one's underlying assumptions. The responses to this paper indicate how attached we often are to those assumptions, and thus how careful we should be, as philosophers, not to allow them to become unnoticed and unquestioned dogmas.n5

W.D. Hart

[1988] The Engines of the Soul (Cambridge: Cambridge University Press)

Derek Parfit

[1971] `Personal Identity' (Philosophical Review, LXXX, pp 3-27)

Roland Puccetti

[1976] `The Mute Self: A Reaction to Dewitt's Alternative Account of the Split-Brain Data' (The British Journal for the Philosophy of Science, 27, pp 65-73)

Daniel N. Robinson

[1976] `What Sort of Persons are Hemispheres? Another Look at "Split-Brain" Man' (The British Journal for the Philosophy of Science, 27, pp 73-78)

David Wiggins

[1967] Identity and Spatio-Temporal Continuity (Oxford: Basil Blackwell)

We do, however, have to make the assumption that both hemispheres are persons. Various writers have argued that this isn't the case -- that the left hemisphere (in most well-lateralised adults) lacks self-consciousness, mainly on the dubious, and in any case disputed, grounds that it lacks linguistic abilities. We can either ignore this claim, account it false (my choice), or pick our experimental subjects from among "adult females, sinistrals of either [sex,] and ambidextrals", who show "a tendency towards bilateral speech functions" (Puccetti [1976], p.66).

I don't want to suggest that there's no work being done in this area -- that philosophers have all accepted the dogma against interactionist dualism. Relevant work includes, apart from Hart [1988]:

John Bricke

`Interaction & Physiology' (Mind 84, 1975)

John Foster

The Immaterial Self: A Defence of the Cartesian Dualist Conception of the Mind (Routledge, 1991)

Stewart Goetz

`Dualism, Causation, & Supervenience' (Faith and Philosophy 11, 1994)

R.C. Richardson

`The "Scandal" of Cartesian Interactionism' (Mind 91, 1982)

Catherine Wilson

`Sensation & Explanation' (Nature and Systems 4, 1982)

It will, however, be noted that little has appeared for some years; this may not be unconnected with the fact that other writers have had the same experience as me -- papers rejected on the grounds that `there are good arguments against your position, so we shan't publish your arguments for it'.