Other writings     Drawings gallery

Exotic Qualia, Functionalism & Martian Zombies

Can functionalism ever escape Exotic Qualia objections?

I shall formulate a meta-argument encompassing all Exotic Qualia problems and argue that while Lewis and Horgan might successfully escape certain guises of the problem neither eradicates it completely. I then suggest the only promising defence of functionalism therefore is a Chalmers-approach.


I take functionalism as the position that mental states are states that play a specific causal role in regards to their causes and effects, to other mental states, and to the behaviour of the individual.

In other, perhaps simplified, words what it is for an individual to be in a mental state, ‘pain’ being a worryingly common example, is for that individual to be in a mental state that for him fulfils the ‘pain-role’ – e.g. is caused by touching a hot stove, distracts him from other mental states, makes him go ‘ouch,’ and has the effect of him withdrawing his hand.

Opponents and proponents of functionalism alike seem to agree that functionalism’s biggest problem is qualia. I say ‘problem’ in the singular since I shall argue that it is really one problem in many guises.

‘Qualia’ is a peculiar word in that it is much easier to understand than to explain. To return to our previous example; aside from the mental state of pain fulfilling role there is also an immediate quality to pain, which is not to be found in an objective description of said role. In our example we offered a clinical description from a God’s eye perspective. Most likely our description would look quite different had the recipient of said pain been us. For one, we can imagine it would contain a significantly increased amount of profanity in addition to being first person.

We described pain from the outside looking in but is it not at least as legitimate to describe pain from the inside looking out? Although this might seem an odd way to do philosophy, given that the subject matter is the attempt to explain mind, we should hardly want to explain it away. While we want to explain how the mind works we should preferably also respect the fact that there is a very real way in which it is like something to be said mind.

The common charge against functionalism then is that it does not properly respect qualia. This is usually shown by imagining bizarre thought-experiments, which I shall refer to as Exotic Qulia – e.g. madmen who feel pain when doing mild exercise, people who see red when the rest of us see blue, zombie-people who behave exactly as we do even though they have no mind, or increasingly unlikely contraptions – an intricate system of pipes and valves mimicking the functions of a brain, Chinese people mimicking neurons by signalling one another via walkie-talkies etc.

Inverted Qualia; a common Exotic Qualia example.

It is interesting to note the difference in the bizarreness of these thought-experiments and its tendency to increase the longer they fail to persuade functionalists of their erroneous ways. However, there is a meta-argument against functionalism underlying everything from the Madman to the Chinese Nation Brain:

1.      If functionalism is true and inasmuch as A and B are functionally alike then A and B are in the same mental state.

2.      However, it is possible that A and B are functionally alike yet in different mental states (qualia).

3.      Therefore, functionalism is false.

(Adapted from Byrne, 2010)

I left out that A and B have to be in sufficiently similar situations; we do not consider the difference between A and B mental states relevant if A is skiing and B is leading an exodus out of Egypt. The rhetorical trick to this argument is to pick intuitively appealing examples of A and B and, depending on the level of bizarreness, perhaps one might want to substitute P2 with ‘However, even though A and B are functionally alike they are obviously different in mental states since B is possibly just any hunk of junk whatever (borrowed from Searle, Dennett, 1982)!’

The logically astute reader might already see something amiss in the argument. We cannot derive the falsity of functionalism from the possibility in P2 unless there is a necessity to be found somewhere in P1. Consider; if pigs are mammals, pigs do not lay eggs. However, there is no inconsistency in both positing the possibility of egg-laying sows while still retaining the factuality of pigs being mammals. One might wonder why I refrained from simply including necessity. Given that P1 contains at least two conditionals I am not entirely sure what the scope of the necessity ought to be. Perhaps I could work it out. However, all is moot when considering it is not obvious the functionalist should grant any particular necessity scope in P1 or any at all.

Indeed David Lewis states that ‘the concept of pain as Armstrong and I understand it is a nonrigid concept. Likewise the word “pain” is a nonrigid designator (Lewis, 1983).’ According to Lewis then, in all the possible worlds in which the concept of pain exists, ‘pain’ does not necessarily refer to the same bare-feel mental state – nor does it have to as long as it refers to the mental state that happens to be playing the pain-role. Should that not be the end of it? Is Lewis not free now to claim that while ‘pig’ could refer to egg-laying creatures in an imagined world in our actual world all pigs just so happen to be mammals?

If it were that easy there should be no reason for Lewis to be concerned with Exotic Qualia; yet he clearly is as evidenced by his claim that ‘[a] credible theory of mind needs to make place both for mad pain and for Martian pain (ibid.).’ A reason for his concern is found in the fact that he takes the possibility of Exotic Qualia seriously:

Nonrigidity might begin at home. All actualities are possibilities, so the variety of possibilities includes the variety of actualities. Though some possibilities are thoroughly otherworldly, others may be found on planets within range of our telescopes (ibid.).

Lewis then reintroduces our meta-argument in the guise of his hypothetical madman who feels pain (qualia) when doing mild exercise and responds to qualia-pain by thinking about mathematics. A note of interest is that while Lewis saw a boon to functionalism in its ability to account for the qualia of hydraulic-brained Martians, depending on pre-philosophical intuitions one might just as easily consider said Martians arbitrary ‘hunk-of-junk’ assemblages counting against functionalism.

Lewis attempts to account both for madman and Martian by combining the view that ‘pain’ is the mental state fulfilling a certain causal role and that ‘pain’ is identical to some matter of physics. Says Lewis ‘If the state of having neurons hooked up in a certain way and firing in a certain pattern is the state properly apt for causing and being caused, as we materialists think, then that neural state is pain. But the concept of pain is not the concept of that neural state (ibid.).’ Lewis could have avoided confusing me if he would just make a clearer semantic distinction between role-pain and qualia-pain. Lewis’ madman is in qPain even though he is not, relative to himself, in rPain because his neural pattern is the same as ours when we are in rPain.

However, it should now become clear that if Lewis takes seriously the possibility of Exotic Qualia, then we can just reintroduce the problem at the qPain level. Suppose the madman and I have identical neurological patterns. We are both in role-pain, we are both in neuro-pain, but who is to say whether we are both in qualia-pain? After all, the charge against functionalism was not that it does not properly respect neurology but that it does not properly respect qualia. One might be inclined to think that the thought of A and B being functionally alike as well as physically alike verges on the impossible – as some philosophers do (Byrne, 2010). However, if that is the case then a credible theory of mind should not need to make place for mad pain.

Terence Horgan sees the same problem as do I with Lewisian functionalism:

…it is just self-evident, I submit, that the qualitative content of Jill’s experience when she looks at grass is an absolute, intrinsic feature of her mental life – not a feature that is implicitly population-relative. There is, absolutely and non-relatively, something it is like for Jill when she looks at grass.

In other words, any fancy explanation involving rPain or even nPain does nothing to alleviate us of Exotic Qualia problems inasmuch as it does not tell us something convincing about qPain. As initially promising as Horgan’s criticism of Lewis is, I should say that in the end his solution does not differ all that much from Lewis. Horgan differentiates between phenomenal pain (qPain) and functional pain (rPain) wherafter he proposes an identity relation between qPain and ‘neurophysiological state-types’ (nPain) and that rPain should be construed functionally. I shall not get into too much detail here but suffice it to say that while Horgan might be successful in dealing with Inverted Qualia he himself admits that his theory is vulnerable to Absent Qualia objections.

The prospect of Martians who are functionally similar to us but who either lack qualia altogether or else have dramatically different qualia, raises with a vengeance the traditional problem of other minds. If partial functionalism is correct, how could we ever tell whether Martians have qualia?

Horgan proposes we simply ask the Martians whether they think there is more to mentality than is functionally definable and on the basis of their answer infer whether they have qualia. Inasmuch as Horgan is right in his assessment of his own theory’s vulnerabilities and asking is his only solution, I feel perfectly justified in proclaiming that Horgan makes no headway in defeating our meta-argument. The whole point of the ‘other minds’ problem is that zombies would be behaviourally, functionally, and conversationally indistinguishable from non-zombies despite it not being like anything to be them. A Martian zombie functionally inclined to do so would insist just as vehemently as anyone else that it has qualia.

I couldn't find Martian Manhunter as a zombie.

I have argued that neither Lewis nor Horgan manage to satisfactorily contend with the meta-argument and it is unclear to me whether mucking around with theoretical minutiae ever could. For this reason I find David Chalmers’ approach much more promising. In ‘Absent Qualia…’ Chalmers argues for what he calls the principle of organizational invariance; that experience is invariant across systems with the same fine-grained functional organization. In terms of our meta-argument Chalmers only concedes an empirical necessity to P1 enabling him to grant the logical or metaphysical possibility in P2 without any problem. If we are talking logical or metaphysical modality, Chalmers would deny P1 and affirm P2, whereas with nomological modality he would affirm P1 and deny P2.

Chalmers offers two arguments against the possibility of different Exotic Qualia following the structure ‘If x is nomologically possible then so is y, but we have good reason to think y is impossible, therefore we have good reason to think x is impossible.’ I shall focus on the Dancing Qualia argument since Chalmers deems it applicable to all Exotic Qualia. Suppose A and B are functionally alike but one has a usual carbon brain and one has a silicon brain. Suppose, for the sake of reductio, that A and B differ in their mental states. If that were true we could construct a series of intermediates between A and B that are gradually less carbon and more silicon. The implication being that the experiences of the intermediates would, without their awareness, become less A-like and more B-like – e.g. less red, more blue, or less conscious, more zombie.

Suppose now that we only turned part of A’s carbon brain into silicon but retained A’s original brain-piece as an inactive backup. Suppose we gave A a switch to flick between using the carbon and silicon pieces. A would then be able to flick between A-qualia and B-qualia at will but without ever being aware of any difference. Since we have good reason to think such a scenario absurd – after all, if such switching is possible how do we deny it happens all the time? – then we have good reason to think it absurd that functional equivalence between A and B does not nomologically necessitate qualia equivalence.


I have argued that since functionalism cannot completely eradicate the meta-argument, the promising defence should be to deny the possibility of Exotic Qualia. Ideally functionalists should like to demonstrate a logical or metaphysical impossibility, not just a nomological. However, if mere conceivability is sufficient for affirming possibility (controversial) it is unclear to me how any theory could ever meet such a demand. After all, there is nearly no limit to what can conceivably be imagined and I doubt there is any logically necessary entailment between qualia and anything whatever. Chalmers, self-admittedly, does not give us a conclusive reason to deny Exotic Qualia but his reasons are good enough to let me sleep at night and at least he is on the right track. There is something to be said for a stronger Chalmers-type argument but regrettably it lies outside my capabilities.



  • Lewis, David, ‘Mad Pain and Martian Pain,’ in Philosophical Papers, Vol. I, (Oxford University Press, 1983, pp. 122-30)
  • Chalmers, David J., ‘Absent Qualia, Fading Qualia, Dancing Qualia,’ in Thomas Metzinger (ed.), Conscious Experience, (Ferdinand Schöningh/Imprint Academic, 1995, pp. 309-31)


  • Horgan, Terence E., ‘Functionalism, qualia, and the inverted spectrum,’ Philosophy and Phenomenological Research Vol. 44, No. 4, June 1984


  • Dennett, Daniel C., ‘The Myth of the Computer: An Exchange,’ The New York Review of Books Vol. 29, No. 11, June 24, 1982


  • Byrne, Alex, “Inverted Qualia”, The Stanford Encyclopedia of Philosophy (Spring 2010 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/spr2010/entries/qualia-inverted/>.
Share and Enjoy:
  • Facebook
  • Twitter
  • Digg
  • StumbleUpon
  • Reddit

Leave a Reply