Reductive and non-reductive functional analyses of consciousness
The problems of qualia described in the last part of this refutation of functionalism affect all purely functionalist descriptions of consciousness. I will now draw a distinction between two kinds of functional analyses of consciousness and show how one of these kinds has been disproved empirically, as well as facing all of the logical problems described earlier.
A non-reductive functionalist analysis of consciousness would hold that, whilst functional roles fully determine the types of conscious mental state one experiences, they are not identical to types of conscious mental state (see David Chalmers, Facing Up To The Problem of Consciousness [1998] and The Conscious Mind [1996]). This would entail both that types of mental state cannot vary independently of their functional roles, so that experiences with different qualitative properties cannot have the same functional properties; and that mental state types either arise from or supervene on their functional roles.
A reductive functionalist analysis, however, would hold that types of conscious mental state are identical with their functional roles (again, see Chalmers [1996]). This position entails not just that (1) experiences with different qualia must have different functional properties, but also that (2) experiences with different functional properties cannot have the same qualia. My last entry showed how Jackson’s knowledge argument and Block’s problems of liberalism and chauvinism challenge (and, to my mind, effectively refute) thesis (1). Jeffrey A Gray and his colleagues at the Institute of Psychiatry have shown empirically that (2) is false (see their paper, Evidence Against Functionalism from Neuroimaging of the Alien Colour Effect in Synaesthesia [2006]).
Gray et al’s (2006) study followed on from an earlier paper (Implications of synaesthesia for functionalism: theory and experiments [2002]) which suggested that “coloured hearing” synaesthesia might contradict the thesis that “for any discriminable functional difference within a behavioural domain associated with qualia, there must be a discriminable difference between qualia.” Coloured hearing synaesthetes experience specific colours when they hear specific words, and some synaesthetes experience different colours to those described by certain colour words (for example, they might “see” red when they hear the word “blue,” and so on): this last phenomenon is known as the alien colour effect (ACE). If real, the ACE refutes reductively functionalist descriptions of consciousness because colour qualia are produced by two different functions (of hearing words and of seeing [coloured] surfaces).
Gray et al had already shown (in 2002) that synaesthetes who experienced the ACE were slower to name colours correctly in “Stroop tasks” (wherein the name of a colour is printed in ink of a different colour, and subjects have to name the ink colour, their reaction times being measured), a finding which was consistent with the idea that they face cognitive interference from other, synaesthetically induced and consciously experienced colours. In their 2006 study, Gray et al conducted simplified versions of these Stroop tasks whilst neuroimaging the brains of control subjects and synaesthetes with ACE. They confirmed their prediction that, if the ACE effect was real, the hippocampus in the brains of synaesthetes who experienced the effect would be activated (the hippocampus being the brain area dealing with kinds of conflict resolution), whereas in control subjects it was not. They also noted that, for synaesthetes with ACE, the two functions of hearing words and seeing surfaces in fact act in opposition to one another: so, not only do two functions produce one colour experience, but they create a conflicting experience. As they argue, “the linkage, in coloured hearing synaesthetes with the ACE, between a functional role and conscious experiences that adversely compete with it” is “equally incompatible” with reductive functionalist accounts of consciousness.
A functionalist may assert that this 2006 study simply shows that reductively functionalist descriptions of consciousness are incomplete, such that they apply to ordinary people but not to coloured-hearing synaesthetes; however, this response will not do. Unless functionalists wish to hold the absurd belief that synaesthetes are not conscious – and, even more unbelievably, have no mental states – then they must acknowledge that their account cannot completely describe mental states (at least, not conscious ones).
Conclusions
I have shown that purely functionalist descriptions of mental states (1) fail to capture what it means to understand an utterance or written response (Searle’s Chinese Room problem); (2) over-generalise (Block’s problem of liberalism) or over-specify (the problem of chauvinism) what counts as a mental state, but fail to capture it precisely; (3) wrongly characterise qualia as solely functional, when they are used to completely describe consciousness; and (4) have been empirically falsified when they identify conscious mental states with their functions. I believe these four problems are sufficient to refute all strong versions of functionalism both logically and empirically: indeed, to echo John Searle (in his paper, What’s Wrong with the Philosophy of Mind), “if you are still tempted to functionalism, you do not need refutation, you need help.”
In its analyses of mental states as purely functional intermediaries between sensory inputs and behavioural outputs, functionalism ignores something essential about them. As Thomas Nagel argues in his (2000) paper, The Psychophysical Nexus,
However complete an account may be of the functional role of the perception of the color red in the explanation of behavior, for example, such an account taken by itself will have nothing to say about the specific subjective quality of the visual experience, without which it would not be a conscious experience at all.
This does not in any way mean that there can be no acceptable physicalist description of mental states. Rather, just as Bertrand Russell found Frederick Copleston’s argument for the existence of God (as posted here) barely comprehensible because its assumptions and language were based in Mediaeval theology, so I believe the assumptions and language of functionalism will come to seem strange (or perhaps are beginning to seem so already) as neuropsychology develops.
As neuropsychology evolves further, it will almost certainly have to re- and re-re-evaluate what constitutes brain stuff; and, to quote Nagel again, these conceptual revolutions will require “the willingness to contemplate the idea of a single natural phenomenon that is in itself, and necessarily, both subjectively mental from the inside and objectively physical from the outside—just as we are.“ Meanwhile, we have the unsettling joy of not yet having an adequate theory of consciousness, and the concomitent duty and pleasure of pointing the way to possible futures in neuropsychology.
