8.30.00 2831 words; dcole@d.umn.edu

Inverted Spectrum Arguments

David Cole

UMD

Formerly a spectral apparition that haunted behaviorism and provided a puzzle about our knowledge of other minds, the inverted spectrum possibility has emerged as an important challenge to functionalist accounts of qualia. The inverted spectrum hypothesis raises the possibility that two individuals might think and behave in the same way yet have different qualia. The traditional supposition is of an individual who has a subjective color spectrum that is inverted with regard to that had by other individuals. When he looks at red objects, this individual has the qualia normally produced in others by blue objects. And when presented with a blue object, this individual experiences qualia that most persons experience only when presented with red objects. And so forth - the Invert's color spectrum is the inverse of normal; there are systematic inter-subjective differences in qualia.

However (the thought experiment continues) the Invert can make all the color discriminations that a normal person can make - unlike those handicapped by color blindness. And since he learned color terms by being presented with objects and color samples in the shared physical world, his use of language and his other color related behavior agrees with normal individuals. Finally, his color proximity judgments will be the same as a normal person's. Thus this individual has an in principle undetectable condition - his nonstandard qualia cause no functional difference. And so it seems that the particular qualia he has - and by extension that any of us have -- have no essential functional role. Qualia are epi-functional.

Color Inverted Earth

An interesting recent elaboration of, and variation on, the inverted spectrum possibility is presented in Ned Block's (1990) "Inverted Earth". Inverted Earth is just like Earth, except that the colors of everything are all changed around. Complementary colors on the color wheel are swapped, in just the way that appearances would be changed if one were wearing spectrum inverting spectacles on earth: grass is red there, the sky is yellow, etc. In addition, on inverted earth the color vocabulary is also inverted - they call their yellow sky "blue", their bright red grass "green", and so forth.

Now suppose mad scientists render you unconscious, implant color-inverting lenses in your eyes, change your body pigment so that it will look normal to you upon awakening, and then transport you to Inverted Earth. When you wake on Inverted Earth, you notice NO difference. As Block says, "'What it's like' for you to interact with the world and with other people does not change at all." (683 in B F G). "So once 50 years have passed [during which time the "causal groundings" - the reference - of your color terms shift to those standard on inverted earth], you and your earlier state at home would exemplify...a case of functional and intentional inversion together with same qualitative contents -- the converse of the inverted spectrum case. This is enough to refute the functionalist theory of qualitative content and at the same time to establish the intentional/qualitative distinction." (ibid)

But how does a thought experiment refute an empirical psychological theory? The Inverted Earth scenario, like any thought experiment, makes several assumptions. One of these, I believe, is questionable, and amounts to rejecting functionalism at the outset. Block supposes that the inhabitants of Twin Earth have different qualia than Earthlings, but that these qualia have the same functional role as the corresponding qualia of Earthlings. For example, Twin Earthlings look at the sky, have a sensation like those an Earthling has when looking at something yellow, but then go on to call it "blue", and say that it is similar to the color of a mountain lake, fountain pen inks, the background of the field of stars on the American flag, Paul Newman's eyes, and so forth. OK. But in addition, if functional identity is preserved, they must have all the same affective reactions. They must find this color as relaxing and soothing as an Earthling finds blue, they must judge that a deep blue colored object (that is, subjectively yellow) is darkish, shades off into black, is cool, and so forth. Now if the qualia they are experiencing and which prompt them to say "deep blue" are exactly the same as a saturated yellow produces in a normal Earthling, is it really plausible that this quite different qualitative experience could be to them just as an experience of saturated blue is to us? Or, if the experience of the qualia is just the same in all its associations and effects, is it plausible to believe the Twin Earthlings really have different qualia?

Block assumes that a stimulus, such as light of certain wavelengths, will always produce the same qualia in beings with human physiology, no matter how the functional roles of those qualia are changed. But that is just what is at issue! To assume this is just to assume that qualia cannot be accounted for functionally. If we reject this unsupported assumption, it is an open question whether the inhabitants of Inverted Earth will have the same qualia that an Earthling would in their place. The qualia normally produced by red objects are associated with blood, fire, danger, agitation and excitement, fight or flight. The Inverted Earth scenario assumes that an Inverted Earthling would have different qualia with just those associations and functional role. Not only is this assumption at issue, I do not believe that it is plausible. In the remainder of this paper, I will consider other qualia inversion scenarios in support of this claim.

Luminance Inversion

Colors are the favorite qualia in philosophical arguments about qualia, functionalism, and consciousness. I suspect that this is because the functional roles of different colors differ in very subtle ways, so subtle that they are easily overlooked, and so inversion arguments become tempting. But color is just one aspect of experience, even of visual experience. Luminance or brightness is another property of visual images. Accordingly, we can imagine a person with inverted luminance: what looks bright and white to us appears dark and black to him, and vice versa. Now try to suppose that there is no functional difference between the experience of a person with normal vision, and one who has inverted luminance. A scene that appears bright to us will appear dark and murky to him - yet he must report that things are well illuminated and distinct. How can this be possible? Where we report a scene as bright and cheerful, he will experience darkness, yet report the same. Surely there is reason to suspect that he cannot be having the very same qualitative experience that I have, when I survey a dark, murky, gloomy scene, if such qualia in him cause reports that the scene is bright, clear, distinct and cheery.

Pitch inversion

Vision is not the only modality that might have inversions. Musical sounds have an experienced loudness and pitch. Taking pitch first, we can then suppose that there is an individual who has inverted pitch, relative to normal humans. What you and I hear as thundering bass, the pitch invert (call him "PI") hears as a high pitched siren. What you and I hear as the high pitched screaming alarm of a smoke detector, PI hears as fog horn. Except, of course, PI's use of language is not overtly any different than ours - when he has the subjective experience of high pitched squeaking, he calls it "thundering bass", and when he experiences a low boom, he remarks on the shrillness. We all agree that tubas are capable of much lower notes than are piccolos. Is it possible that PI's inverted qualia have exactly the same functional roles as ours. It must be if qualia are not determined by functional role. What evidence is there that qualia are not determined by functional role? Why, thought experiments involving twin planets and inverts. However reflection suggests that there cannot be Pitch Inverts who are qualitatively distinct but functionally identical to ourselves. The first blush plausibility of the possibility of the inversion fades as we start to consider the richness of the experience. For one thing, on the face of it, PI will have quite different experiences of timbre. Timbre is a very important part of auditory experience; timbre is determined by harmonics. Since harmonics are multiples of a fundamental frequency, high pitched tones can have fewer audible harmonics than can low pitched tones. Toward the limits of the upper limit of auditory perception, sounds can have no audible harmonics at all. However, low frequency sounds can have very rich harmonic composition. We can discern a variety of timbres based on the same fundamental, and can experience the shift in timbre as this harmonic composition changes. For the invert, however, there can be no experience of harmonic composition for such notes. He cannot experience the change from sine to square to saw tooth waveform of a 100hz note - these all involve the experience of harmonics.

Other acoustic phenomena are frequency dependent - the experience of beats is an example. Beats are produced when two simultaneous tones differ slightly in frequency. The number of beats per second is the (absolute) difference of the frequency of the two tones. Low frequency notes then will produce slow beats as the frequency is changed by a given percent, high frequency tones will produce rapid beating. Normal humans can discern up to about 7 distinct beats per second. It is not clear how invert's experience could be like a normal person's. As low frequency notes shift, the beats "emerge" as the tones change audibly in pitch. But high frequency notes will have beats emerge even when there is no audible difference in pitch - a feature exploited in tuning instruments, by relying on beats to get more accurate tuning than could be accomplished by relying on perceived pitch differences alone. The beats are determined by the absolute difference in frequency, but, as is typical of much perception, pitch perception is logarithmic. The beats will emerge from the shifting tones in a different way for a normal hearer and for a pitch invert. More importantly, auditory experience has a kinesthetic aspect. High tones are associated with tightenings - of vocal cords, increased lung pressure, diaphragm tension, and so forth. Low notes are low in the throat, produced by relaxings, larger air flow, and so forth. These associations most likely permeate all experience of pitch, not just that of tones the perceiver is producing or contemplating producing. The vast richness of these myriad connections and associations is an important part of what gives experience its character. The conscious brain is an unbelievably vast network, with connections across multiple areas, sensory and motor. Part of the experience of a sound may be shaped by subliminal experience of what is required, what it is like, to produce similar sounds. That experience will not be the same for PI. While it might be possible to attempt to avoid this objection by supposing the kinesthetic qualia to be inverted as well, the point is that qualia are not raw and simple, they are embedded in a vast network of connections and associations.

Inverted loudness

Inverted subjective loudness presents similar problems. Although apparently a logical possibility, a subject who experienced low intensity sounds as very loud, and loud sounds as we experience whispers, seems hardly conceivable, past the initial bald statement of the logical possibility of such inversion. Silence, for the loudness invert, would be deafeningly loud, yet the subject, if functionally identical to us, would report that it was restful, an excellent opportunity for thinking things through, and perhaps even so quiet one could hear a pin drop. The supposition that the character of qualia is independent of functional role seems to make no sense here. Why should this not be a general feature of qualia, whether this dependence on functional role is immediately apparent or not?

Radical Synesthesia and Transesthesia

Synesthesia is a condition where a perceiving subject experiences qualia typically associated with another sense modality. For example, heard sounds might be accompanied by an experience of flashing lights. Let us then imagine an extreme case: suppose that there could be a victim of synesthesia who had visual stimuli instead of auditory stimuli. For example, such a radical synesthete might locate sound sources in a visual field, and different sounds might be experienced as varying colors and luminance. Imagine a modality transposing radical synesthete, with all qualia swapped between two modalities, a condition we might call "transesthesia". The vision-audition transesthete will 'see" sounds and "hear" sights, that is, experience (only) visual qualia when hearing, and experience (only) auditory qualia when seeing. Let us suppose that the qualia have the functional roles normally associated with the sensory modality (and thus the visual qualia function as auditory qualia do in normal subjects). The entire qualia spectra are transposed across these two sensory modalities.

The transesthete's visual qualia for sound will be located in a visual qualia field. Unlike the ordinary visual field, this will somehow have to extend behind the head - we hear sounds behind us. Multiple distinct sounds can come from a single location - as when listening to a monaural recording of a symphony concert or crowd noises. Somehow this will be experienced by the tranesthete as colored shapes all at the same point.

The transesthete's qualia for vision will be sounds. Distinct timbres might represent different colors; loudness might represent brightness, and so forth. The transesthete will somehow have the functional equivalent of a foveal region. Looking at a complex brightly colored scene will presumably be loud and cacophonous - yet will not distract the transesthete any more than gazing out over a busy plaza or a sunlit rippling wooded stream distracts a normal perceiver. What will the unison of octaves represent? Chords? The experience of subjective octaves and chords will have to produce no discernible difference between the transesthete and the normal experiencer - how is that possible? What will it be like to have auditory experience with the spatial resolution of vision (such that one can read letters by sound, or follow a circuit diagram, or knit, or see the subtle smile of the Mona Lisa)?

Considering the alleged possibility of a transesthete raises many questions about the connection between experience and judgment and behavior. These cannot be fully explored here - but I think it is clear that it is reasonable to suspect that a transesthete may in fact be impossible. But then the character of qualia is linked to their functional role, a role that ultimately leads to overt behavior, including linguistic report. And one way of accounting for this link is to suppose that functional role determines the character of qualia.

Inverted hedonic spectrum

As a final inversion possibility, consider someone who experiences pleasure when injured, and pain under circumstances where others experience pleasure. However, the sensations produced by injury, though very pleasant, have the functional roles that pains have in normal individuals. They are aversive, and so a primary effect is to make the subject do what is required to avoid them. He moans and cries, whimpers and cringes, when pleasures are great. The hedonic invert regards the possible occurrence of pleasures with dread. However, he seeks and looks forward to pains.

The impossibility of hedonic inversion is apparent. States that located in the body image and are strongly aversive, are pains and cannot be pleasures. We can be sure that others have pains because they have such aversive states. Furthermore, the senselessness of the claim that there could be intrinsically attractive pains or aversive pleasures is not epistemic - I cannot conceive in my own case having such "pleasures" or "pains". The problem here has to do with the functional essence of these qualitative states.

Inverted Earth Redux

Consider an individual with topologically inverted qualia - what looks up to me looks down to him, his visual field is inverted. Suppose further, as before, he has learned to react to his inverted visual qualia so as to behave, linguistically and otherwise, just as those with normal visual qualia behave. But far from being just a thought experiment, this possibility has been investigated empirically. George Stratton wore special prism-containing inverting goggles for an extended time. At the end of this disorienting ordeal, Stratton (1897) adapted sufficiently to be able to walk around more or less normally -- and he reported that the world ceased to appear upside down. The qualia changed with change of functional role. Related results are reported by Kaufman (1974), using distorting lenses (distorted lines came to look straight, with adaptation, replicated by Festinger et al. (1967). And McCullogh (1965) reports similarly mutable qualia, using color-altering contact lenses.

The real world empirical results support functionalist accounts of qualia, and run counter to the intuitions pumped by inverted spectrum arguments. However even here, we can counter those intuitions by considering a broader range of qualia inversion possibilities.