Reading Kim’s “Physicalism…” Chapter One

I have begun reading Jaegwon Kim’s Physicalism, or Something Near Enough (Princeton University Press 2005) as part of my self-directed research program on the hard problem of consciousness.  The book seems promising, both for its clarity and readability (a somewhat uncommon virtue among works of analytic philosophy) and for its organization.  It is split into six chapters that are intended to be read as stand-alone essays (indeed, most of these began life as lecture notes) but which together form a cohesive whole.  The book is intended to provide an argument to the effect that a thoroughgoing physicalism is not an appropriate theory to be applied to ‘the mind’, but that a near-total physicalism that leaves room for qualia is necessary and, importantly, good enough.  On with it!

Chapter 1 – “Mental Causation and Consciousness: Our Two Mind-Body Problems.”

Kim argues that there are two serious obstacles in the philosophy of mind that any modern physicalist theory will have to tackle.  The first of these is the matter of mental causation, of how it is possible that “the mind [can] exert its causal powers in a world that is fundamentally physical” (pp. 7).  Mental causation, whatever its mechanism, is an important concern for physicalists, Kim claims, for three reasons.  Firstly, our understanding of human agency and our moral practice requires that it be our beliefs and desires that cause our actions, not mere physical happenings.  [1]  Secondly, human knowledge requires mental causation (since it is predicated upon our having been in appropriately causally linked “cognitive-relations” with external objects and having reasoned from those relations).  Finally, in order for psychology to serve as a useful descriptive enterprise requires effective causation by mental states/properties – to say that ‘anger’ caused an action is to posit a real thing with causal efficacy (9 – 10).

The second obstacle for physicalists is consciousness, specifically the problem of “how [there can] be such a thing… in a physical world, a world consisting ultimately of nothing but bits of matter distributed over space-time behaving in accordance with physical law” (7).  Consciousness seems less immediately troublesome for the physicalist, but is vitally important in certain contexts.  Ethics, for example, makes much of the distinction between those things with consciousness and those without and even the average person finds consciousness to be highly valuable for the access it provides to things like sunsets, flavours, etc.  So consciousness matters rather more than even some philosophers (Dennett) suppose and is not, in any case, explained simply by compiling a list of “psychoneural correlations” (10 – 13).

The Problem of Mental Causation:

Kim believes that a “minimal physicalism” requires supervenience (that is, all physicalisms must include it but are not necessarily identical to it), which he defines as “the claim that what happens in our mental life is wholly dependent on, and determined by, what happens with our bodily processes” (13 – 14).  Supervenience cannot do the work physicalists want it to, however, because of several of their other assumptions; such as the

  1. “[principle of] causal closure of the physical domain.  If a physical event has a cause at t, then it has a physical cause at t” (15), and the
  2. principle of causal exclusion.  If an event e has a sufficient cause c at t, no event at t distinct from c can be a cause of e (unless this is a genuine case of causal overdetermination)” (17), and the
  3. principle of determinative/generative exclusion.  If the occurrence of an event e, or an instantiation of a property P, is determined/generated by an event c – causally or otherwise – then e‘s occurrence is not determined/generated by any event wholly distinct from or independent of c – unless this is a genuine case of overdetermination” (17).

These assumptions inexorably lead to problems for mental causation on supervenience theses.  M (a mental event) is thought to cause M’ (mental causation), but M’ instantiates because of/is generated by P’ (the physical property upon which it supervenes).  Only one of M or P’ can be the cause of M’ (by appeal to principle 2 above) and, since P’ is sufficient for the instantiation of M’ (whatever else happened in the past), the only way to resolve this problem is by claiming that M causes M’ by causing P’.  Physicalists should reject this, however, for two reasons.  Firstly, it is an instance of cross-domain causation (of a sort excluded by principle 1).  Secondly, M has its own supervenience-base P that is sufficient to cause M’, since it is sufficient to cause P’.  Here, then, is a case of overdetermination of P’ (by both P and M), but in order to avoid it we must abandon mental causation (else we are just iteratively pushing the problem into an infinite regress).  So we have preserved lawful relations between the mental and the physical, but have robbed the mental of all causation, letting all the action take place at the level of the bases of supervenience (19 – 21).

The problem of mental causation is summed up for Kim like this:

Causal efficacy of mental properties is inconsistent with the joint acceptance of the following four claims: (i) physical causal closure, (ii) causal exclusion, (iii) mind-body supervenience, and (iv) mental/physical property dualism – the view that mental properties are irreducible to physical properties” (21 – 22).

And since physicalists cannot reject (i) or (iii) without losing physicalism, and (ii) seems a reasonable metaphysical principle, only (iv) can be given up – reduction of the mental to the physical seems to be in order, but…

Can We Reduce Qualia?

We must be clear about what it is to reduce something.  Kim wants to avoid using Nagel’s bridge laws reductionism [2], finding them inappropriate for their vacuity (they could explain many dualisms) and for reasons he promises to elucidate in a later chapter.  Instead, Kim’s model of reduction of a mental property (e.g. pain) is as follows.  First, we must functionalize the property in question – that is, identify the role it is thought to play, the mechanisms that bring it about, and its probable consequences.  Then, we may identify its ‘realizers’, the things that are necessary for its manifestation (C-fibers, neural activation patterns, etc.).  Kim addresses the matters of multiple realizability, specific realizers, loss of singular ‘pain’ concepts, but none of this is particularly germane here – the upshot is that should functional reduction work, we will have saved the causal powers of pain’s realizers, and mental causal efficacy.  It should be noted, however, that this will only work if mental properties actually are functionally reducible – so are they (22 – 27)?

Kim says that intentional and cognitive properties are, but phenomenological properties (qualia) are not.  To make this case there is no need to resort to “anything as esoteric and controversial as the ‘zombie’ hypothesis much discussed [in philosophy] recently” – rather, all that is required is the modest metaphysical prospect of qualia inversion (to be discussed later in his book) (27).


[1] Kim does not elucidate why this should concern us, familiar as he presumes his audience to be with the philosophical literature, but the broad idea is probably something like this – if what one does happens without reference to things like his desires or beliefs, then his actions no longer seem to be for reasons, and, if not for reasons, then are no more meaningful, praiseworthy or blameworthy than a tree falling in the woods.  Moral responsibility cannot obtain since it is dependent upon the reasons-responsiveness of people and, on this conception, people would not be reasons-responsive.

[2] Wherein a property has been reduced when it is appropriately connected with some “nomologically coextensive property [or properties] in the base domain” by means of these bridge laws (22 – 23).


Naturalism: Either Vacuous or False

We hear talk of ‘naturalistic’ metaphysics, ethics, science, and epistemology, but it is never quite clear just what work it does when prefixed to any of these – or at least, it is not clear to me.  Now, in practise ‘naturalistic’ can often be taken as a code-word that stands in for something like ‘atheistic’ or ‘non-supernatural’, and that is fine (so far as it goes), but if naturalism is nothing besides atheism or anti-supernaturalism, what point is there in using that word instead of one of the others?  Unless, of course, its use is meant to disguise the actual content of the view in question (whatever motivation  there might be to do so, I’m sure I don’t know what it is) or because, by design or accident, it helps views to slip past the critical faculties of reasoners without sounding alarms.  This is not so outlandish – think about the effect that terms like (to choose two) ‘Christian’ or ‘Democrat’ have on people’s faculties of critical reasoning.  And as a concept naturalism lends itself quite well to this sort of role for, despite its being appealed to so very, very frequently, it is sometimes hard to sort out just what its supposed contents are.  Personally, I think the concept is either vacuous or false and so should be junked.

This is a strong claim, but I think it to be the right one.  Thinking about what naturalism might mean, it clearly has something to do with the denial of supernatural entities or phenomena – but in what does that denial consist?  On its simplest plausible construal, naturalism is the claim that there is no such thing as ghosts, say, because ghosts are super-natural entities that don’t observe the natural laws of the universe.  This is true, of course – there are no ghosts – but it is a perfectly meaningless statement.  Imagine for a moment that we lived in a world like our own in most relevant respects (for example, gravity has the same force and effects on matter, etc.) but with the exception that this world contains ghosts.  Obviously, the ghosts must in at least some ways causally interact with that world, else we couldn’t know of their existence.  Moreover, the ghosts would exhibit certain regularities in their existence, behaviours, and powers.  They must, since everything that exists does – in fact, it is due to the exhibited regularities of things that we can agree that there are such things as cocker spaniels and mousetraps and that they are different sorts of things.  Since the ghosts would causally interact with the world in regular ways, we would be able to postulate and identify mechanisms and laws [1] by which such interaction is possible.  So in this world ghosts exist and are governed by the laws of that world, which makes them a natural part of that world – naturalism would be every bit as true in that world as in our own, even though it is a world that possesses ‘supernatural’ entities!

A critic could object, of course, that this is simply a counterfactual scenario and that of course he doesn’t mean that ghosts are compatible with some postulated set of natural laws, but that there are no supernatural elements that are allowed by our set, the actual physical laws.  But what exactly might he mean by ‘the actual set of physical laws’? 

If by this he means to say the set of laws as revealed by a future ideal science[2], then he is again saying nothing of substance – if, on the way to our ideal science, we discover that telekinesis is real and governed by natural laws, then it would be (and always have been) part of our ‘naturalistic’ universe, skeptics be damned!  Moreover, if our ideal science is the litmus test of what counts as ‘natural’ then, since it accurately describes the actual laws of the universe, of course there could be no ‘supernatural’ phenomena – such phenomena would simply be those that are impossible under that set of laws.  If, on the other hand, our critic means by the actual set of physical laws those that are revealed by our current best physics, two things fall out of this.  Firstly, he should not properly be called a ‘naturalist’ – he is, more accurately, a physicalist.  Secondly, if he is a physicalist, then he is almost certainly wrong.  Honest scientists are rather open about the fact that modern physics is not an ideal science and, consequently, that cherished theories may have to be altered or abandoned on our way to such an ideal science.

If my reasoning is correct, then, ‘naturalism’ is either vacuous or false and can be usefully disposed of without doing harm to any philosophical or scientific enterprise.


[1] Even if such laws/mechanisms were found to be ‘fuzzy’.  Fuzziness is not a thing that should bother the denizens of an age comfortable with quantum mechanics.

[2] For the sake of the argument, let’s suppose that an ideal science is capable of fully explaining all possible phenomena within the world to which it pertains.  I don’t know that this is true, but that is beside the point here.