The possibility of subisomorphic experiential differences


by Christopher D. Viger

Center for Cognitive Studies

Tufts University

11 Miner Hall

Medford, MA




Palmer=s main intuition pump, the Acolor machine,@ greatly underestimates the complexity of a system isomorphic in color experience to humans. The neuroscientific picture of this complexity makes clear that the brain actively produces our experiences by processes that science can investigate, thereby supporting functionalism and leaving no (color) room for a passive observer to witness subisomorphic experiential differences.


Palmer distinguishes simulating color experiences from actually having them by imagining a A>color machine= that actually processes information in light in the same way as people do and that responds as people typically do.@ [sect. 2.4, para. 3, my emphasis] Palmer=s thought experiment relies on his intuition that Ait would not be very difficult@ [ibid.] to construct such a machine. Palmer even offers a schematic; a few mirrors, prisms, cardboard masks, photocells, and computational circuits are up to the task. But the plausibility of such a construction being up to the task derives from grossly underestimating the task. People respond to colors with incredible subtlety and variety, including changes of mood, preference selections, and aesthetic judgments. On Palmer=s own terms the Acolor machine@ must respond likewise. But once the task is made clear it is also clear just how difficult it would be to construct such a machine, as any AI researcher will attest. All Palmer offers is the Afront end@ of such a machine, which indeed might not be too difficult to constructBbut no functionalist would declare that our eyes and, say, V1 (the Afront end@ of our visual system) have experiences. It is only by various mechanisms playing a functional role within a very complex system that experience can arise at all, and in such cases it is the entire system that is the experiencer. Only by presupposing qualia is there any motivation to attribute experiences to subsystems, which is the flaw with the color room thought experiment. The subsystem in the color room is not the locus of experience. Once we expose the requisite complexity for a system to be experientially isomorphic to humans vis-a-vis color experience, the force of Palmer=s thought experiment evaporates. Of course, the kind of machine he imagines would not have color experiences. It has nowhere near the complexity required to have experiences of any sort.

Consider the neuroscientific account of coming to have a color experience. AWhen an individual cone absorbs a photon, its electrical response is always the same, whatever the wavelength of the photon@ (Kandel, et. al., p.456). Individual cones are more likely to absorb photons of particular wavelengths depending on the pigment they contain, but the response of an individual cone does not tell the brain anything about color, because the response is stochastic. It is only the comparative strength of the responses by each entire subsystem of cones containing a particular pigment, subsystems large enough to exploit the stochastic responses of individual cones, by which color can be discriminated. AFor example, if an object reflected primarily light of a long wavelength, the response in the longer-wavelength cone system would be stronger than the response in the other system, and higher processing centers would interpret the object as being red or yellow@ (Kandel et. al., p.458). What emerges from this neuroscientific picture of color perception is that the brain interprets objects as having a certain color; our brains actively produce our color experiences. AVision is not merely a matter of passive perception, it is an intelligent process of active construction@ (Hoffman 1998, preface, p.xii). But then experiences admit no subisomorphic experiential difference, because the brain=s constructions are not the privileged possessions of the person whose brain is doing the constructing. For example, consider Palmer=s case of reversing pigments in the cones.

Palmer supposes that someone whose L-cones have the M-pigment and whose M-cones have the L-pigment would be Ared-green reversed trichromats@ [sect. 1.3, para. 5]. He notes, however [footnote 5], that red-green reversed trichromats could be the case only if there were some difference other than the pigments that L- and M-cones contain, otherwise all that would be reversed is the L- and M-cones themselves. But such a difference would have to be a functional difference, in that the distinct subsystems containing a specific pigment would have to play different functional roles within the larger nervous system. Then since the brain interprets color based on the relative strengths of responses from each subsystem of cones containing a particular pigment, a strong response by a particular subsystem would be interpreted as a certain color, in virtue of the functional role of that subsystem. In such a case reversing the pigments in the cones, but not the functional roles of the subsystems of cones, a red object would produce a strong response in the subsystem of cones that results in the brain interpreting the object as green. So the red object would be experienced as green. But it would also be reported as green, since the functional role of the subsystem leads to the object being interpreted as green, breaking the isomorphism between the original trichromat and the red-green reversed trichromat. If we further suppose that the red-green reversed trichromat would report the object as red and act behaviorally identical with someone who experienced it as red in order to preserve the isomorphism, we would also have to reverse the functional roles of the subsystem of cones. The point is simply that experiential differences that Palmer supposes to be subisomorphic actually presuppose functionalism in order to be coherent, in which case they are behaviorally detectable. The relational structure of experiences, up to isomorphism, entirely captures our experiences.

Paradoxically, what Palmer and other defenders of qualia require are internal states that could make no difference, no difference, not to emotional states, aesthetic judgments, preference selections, nor any of the myriad other ways that our color experiences matter to us. But if they can make no difference, it should make no difference to the defenders of qualia if there are such states.


Dennett, D. 1991. Consciousness Explained. Boston: Little, Brown & Company.

Hoffman, D. 1998. Visual Intelligence: How We Create What We See. New York: W.W. Norton & Company.

Kandel, E., Schwartz, J., and Jessell, T. 1995. Essentials of Neural Science and Behavior. Stamford, Conn.: Appleton & Lange.