Inverted Spectra without illusion

Brad Thompson

Department of Philosophy

Southern Methodist University

 

bthompso@smu.edu

 

Introduction

           

A natural view to hold regarding the phenomenal content of color experiences, one that I think has driven many present-day theories of color perception and phenomenal content, is that in color experience we are directly “presented” with the colors of things. Visual experiences somehow put us in a position from which we can “directly grasp” properties of the objects of perception. One way to think about this idea is in terms of the relationship between phenomenal character and phenomenal content. It is phenomenologically evident that in having a color experience, there is a distinctive sort of property that is before the mind. In fact, it is intuitive to suppose that the phenomenal character of a color experience simply consists in being aware of these properties. But it is also tempting to think that these properties are simply the colors of things in our environment. Or put another way, the properties that we are most directly aware of in color experiences are the properties that those experiences attribute to external objects. One articulation of this idea that has gained considerable currency in recent discussion of phenomenal content is in the idea that visual experiences are transparent.[1]

Many philosophers have also found it attractive to suppose that these properties, the properties that color experiences represent, are mind-independent physical properties. Usually, these properties are simply identified with the colors of things. For example, Byrne and Hilbert (1997) identify colors with types of surface spectral reflectances. Tye (2000) adopts a similar view. Jackson and Pargetter (1987) identify colors with the (perhaps highly disjunctive) physical properties that ground the disposition to appear colored.

Together, these two ideas motivate a certain view about the phenomenal content of color experience, one that I will call Standard Russellianism. In agreement with Sydney Shoemaker (1994), I will argue that Standard Russellianism is incompatible with the possibility of spectrum inversion without illusion. I will then turn to two defenses of Standard Russellianism against this objection. One, offered by Byrne and Hilbert (1997), attempts to show that those who accept the possibility of spectrum inversion without illusion are forced to misdescribe a similar case. The other, offered by Tye (2000), accepts the conceivability of spectrum inversion without illusion but denies its possibility. I will argue that both responses fail. As a consequence, one of the two intuitions that motivates Standard Russellianism is mistaken.

 

Russellian Theories of Phenomenal Content

 

            By having a visual experience with a certain phenomenal character, the world is presented to the subject of the experience as being some way. We can call this intentional content that is inherent in phenomenology “phenomenal content”. Phenomenal content is content that supervenes on phenomenal character. Necessarily, if two experiences have the same phenomenal character then they have the same phenomenal content. This is the defining feature of phenomenal content, as I will use the phrase.[2]

The intuition that perceptual experiences are transparent motivates a Russellian theory of phenomenal content. According to Russellian theories of phenomenal content, phenomenal content is purely extensional, consisting solely in reference to objects or the attribution of properties. Given that sameness of phenomenal character entails sameness of phenomenal content, Russellian theories of phenomenal content are theories that accept the following:

For any experience with phenomenal character r, there is some property p_r such that necessarily if an experience has phenomenal character r then it attributes p_r.

           The second intuition above, when combined with Russellianism, motivates what we can call “mind-independent physical-property Russellianism”. Mind-independent physical-property Russellianism holds that

For any experience with phenomenal character r, there is some mind-independent physical property p_r such that necessarily experiences with phenomenal character r attribute p_r.

For the sake of brevity, I will call mind-independent physical-property Russellianism “Standard Russellianism”.[3] Here I use the term “mind-independent” to refer to a property that does not essentially depend on any mental states or properties. I will call a property “mind-dependent” if it does essentially depend on mental states or properties.[4] I mean to include in the notion of a mind-dependent property such things as dispositions to cause certain types of mental states.

 

Spectrum Inversion

 

A problem for Standard Russellian theories of phenomenal color content comes from the possibility of spectrum inversion.[5] For instance, it might be that the color experience Jack has when viewing a red thing is phenomenally identical to the experience Jill has when looking at a green thing. It is conceivable that there could be two subjects who are spectrum inverted relative to each other.[6] In fact, it has been argued that it is in fact likely that there are actual cases of behaviorally undetectable spectrum inversion (Nida-Rumelin 1996).

Suppose that Jack is looking at a strawberry and that Jill is looking at a lime. The strawberry and lime could be such that, due to their being spectrum inverted relative to each other, Jack and Jill are having phenomenally identical color experiences.[7] It seems also that neither Jack nor Jill need be misperceiving the color of the fruit. After all, the having of an experience with a certain phenomenal character in response to the strawberry or lime is not simply a product of the nature of the fruit. It is in large part of product of Jack and Jill’s subjective constitutions: the way that their eyes work, the way that information received from the eye is processed in their brains, etc. There seems to be no grounds for saying that Jack’s way of perceiving red things, or instead Jill’s way, is the only correct way for red things to appear.

This can be made more vivid by imagining that half the human population is like Jack, and the other half is like Jill. By what criterion can it be rightly said that one half of the population misperceives the colors of things? Both those like Jack and those like Jill get around in the world just fine. And there does not seem to be anything about the relationship between the physical properties of objects and either group’s color experiences that would ground the claim that one group, rather than the other, accurately represents color. In considering the possibility of spectrum inversion, we seem to be confronted with a visual analogue to the case of something which tastes bitter to some and sweet to others (Shoemaker 1994b).

The possibility of “spectrum inversion without illusion” poses an immediate problem for Standard Russellianism. Jack and Jill have color experiences that are phenomenally the same, and their experiences are both veridical. But the external objects in their visual fields have different color properties. This entails that the shared phenomenal content of their color experiences cannot consist in the representation of color properties. And that is to say that Standard Russellianism for phenomenal color content is false.

            Suppose that Jack is looking at a strawberry and Jill is looking at a lime. Let the colors of the fruit be such that Jack and Jill are having phenomenally identical color experiences as they look at their respective fruit. If Jack and Jill are both having veridical color experiences, Standard Russellianism must be mistaken:

1.     Jack and Jill are having phenomenally identical color experiences.

2.     Jack and Jill are both having veridical color experiences.

3.     The objects in Jack and Jill’s respective environments have different physical colors.

\ The phenomenal content of color experiences cannot consist in the representation of physical color properties.

 

Byrne and Hilbert’s “Fred” Thought Experiment

           

Byrne and Hilbert (1997) attempt to show that endorsing the argument from the inverted spectrum as a challenge to Standard Russellianism forces one to then misdescribe another similar case. Their argument attempts to show that either premise one or premise two above must be mistaken.[8] Byrne and Hilbert have us imagine a case of intra-subjective spectrum inversion. But unlike cases described by Shoemaker (1994a), in which a subject undergoes a spectrum inversion over time, Byrne and Hilbert describe a case of spectrum inversion at a time. Their subject Fred has two eyes that are spectrally inverted with respect to each other. What presents a green appearance to Fred’s left eye presents a red appearance to Fred’s right eye. For example, if Fred looks at a strawberry with his right eye, it will present a red appearance. But if Fred looks at the very same object a moment later with his left eye, it will present a green appearance.

            The additional twist to Byrne and Hilbert’s story is that Fred has never had an opportunity to notice his unusual condition. That is because of two additional features of the thought experiment. First, the objects in Fred’s environment change colors on successive days. A strawberry that is red on one day will physically change so that it is green on the next day. Second, Fred only uses one of his eyes for each day. He alternates each day which eye he uses. If he uses his left eye one day, he will use his right eye the next day. The result is that a red strawberry, seen with his right eye, will present a red appearance. The next day, the strawberry will turn green. But Fred will on that day use his left eye, and thus the strawberry will continue to present a red appearance. Fred’s left and right eyes, or rather the experiences caused by his two eyes, are spectrum inverted with respect to each other. But due to the changes in his environment and his use of a single eye per day, this spectrum inversion is unnoticeable to him.

Byrne and Hilbert present their case of Fred as a challenge to those who would suppose that the inverted spectrum poses a threat to a thesis they call “NECESSITY”. NECESSITY is the thesis that for color experiences, sameness of phenomenal character entails sameness of color content, and that sameness of color content entails sameness of phenomenal character. Two experiences have the same color content, according to Byrne and Hilbert, just in case they represent the same color properties instantiated at the same locations. What Byrne and Hilbert call “color content” is thus a kind of Russellian content. The claim that two color experiences share phenomenal character if and only if they share color content amounts to the view that color experiences have Russellian phenomenal content.

The proponent of the inverted spectrum argument claims that Jack and Jill, who are spectrum inverted with respect to each other, both have veridical experiences of color. Byrne and Hilbert suggest that, for these philosophers, the story about content for the experiences Fred has from each eye should be analogous to the one the proponent of the inverted spectrum argument gives in the case of intersubjective spectrum inversion. Byrne and Hilbert have in mind those philosophers who would say that red appearances, had by Jill, represent physical greenness. And red appearances, had by Jack, represent physical redness. Likewise, such a philosopher ought to claim (according to Byrne and Hilbert) that red appearances for Fred's left eye represent physical greenness, and that red appearances for Fred's right eye represent physical redness.

Byrne and Hilbert finally have us imagine that Fred looks at a red raspberry. Normally, Fred only uses a single eye on a particular day. But this time, Fred views the raspberry with both eyes in a single day, looking first with his right eye and then with his left. The raspberry does not change in its physical properties (as normally happens between Fred’s switches in the use of his eyes). It seems that under these circumstances, the raspberry will present a red appearance when viewed with his right eye and then present a green appearance when Fred switches eyes. Corresponding to this change in phenomenal appearance, Byrne and Hilbert argue that for Fred the raspberry will appear to change colors from being red to being green.

Byrne and Hilbert claim that the thought experiment with Fred poses a challenge to those who would claim that two spectrum inverts, such as Jack and Jill, can both have veridical experiences of color. The Fred scenario is analogous to the inverted spectrum scenario. The advocate of the inverted spectrum argument, it is claimed, must say that the experiences from each eye are veridical. But if we accept that each of Fred’s eyes produces veridical perceptual contents despite their inversion, we cannot explain the fact that when Fred switches eyes the raspberry appears to change in color. According to the advocate of the inverted spectrum argument, Fred's experience with his right eye represents the raspberry as being red in virtue of presenting a red appearance. But Fred's experience with the left eye also represents the raspberry as being red, in virtue of presenting a green appearance. The raspberry cannot therefore be represented by Fred's experience upon switching eyes as changing color. Red appearances in the right eye represent objects as having the very same property that green appearances in the left eye represent. This entails that when Fred switches eyes while viewing the raspberry, his experience represents the raspberry as being the same color—red—throughout. The raspberry is not represented as changing in color.

The argument can be put as follows:

1.     The Fred scenario is analogous to the inverted spectrum scenario.

2.     If the Fred case is analogous to the inverted spectrum case, and if inverted spectrum without illusion is possible, then phenomenally red experiences from Fred’s left eye have the same color content as phenomenally green experiences from Fred’s right eye.

3.     If phenomenally red experiences from Fred’s left eye have the same color content as phenomenally green experiences from Fred’s right eye, then if Fred switches eyes during the day, his color experiences before and after the switch have the same content.

4.     When Fred switches eyes during the same day, his experiences have different content.

\ The inverted spectrum without illusion is not possible.

 

One way that an opponent might respond to Byrne and Hilbert’s argument is to deny premise three. Premise three assumes that those who endorse the possibility of spectrum inversion without illusion also hold that the content of color experience consists only in the representation of colors. But some, such as Sydney Shoemaker, have taken the possibility of spectrum inversion without illusion as an indication that color experiences have mind-dependent Russellian content. And so while Shoemaker might allow that spectrum inverts represent the same colors with phenomenally different color experiences, he would not claim that those experiences have the same intentional content. Accepting the Fred scenario as being analogous, such a view need not entail that Fred’s experiences before and after switching eyes have the same content.

Elsewhere I defend a Fregean theory of color content. On this view as well, premise three can be denied. Spectrum inversion without illusion is made possible on the Fregean view not by claiming that a phenomenally red experience had by Jack and a phenomenally green experience had by Jill have the same content. Instead, they represent the same colors but under different modes of presentation. Accepting the analogy with Fred, such a view can hold that the difference between Fred’s experience when viewing the raspberry through his left versus his right eye can be conceived of as a difference in the mode of presentation under which he represents a particular physical color.[9]

Byrne and Hilbert’s argument is thus only a challenge to those views that would accept the possibility of spectrum inversion without illusion while also endorsing certain forms of Russellianism. But I do not think the argument succeeds even against this more limited range of views. The problem is with premise one. The Fred scenario is a case of intrasubjective spectrum inversion. The ordinary spectrum inversion scenario is one of intersubjective spectrum inversion. And this makes the two cases relevantly disanalogous.

            When one supposes that spectrum inversion without illusion is possible, one must suppose that what color is represented by a phenomenally red experience depends on who is having the experience. Had by Jack, a red experience represents physical redness.[10] Had by Jill, a red experience represents physical greenness.

Byrne and Hilbert’s argument depends on the idea that those who endorse the possibility of spectrum inversion without illusion must also endorse the idea that each of Fred’s eyes can be the source of experiences which are the same in phenomenal character but differ in representational content. Fred using his left eye is meant to be analogous to Jack, whereas Fred using his right eye is meant to be analogous to Jill. But the analogy depends on a dubious assumption—that intrasubjective spectrum inversion and intersubjective spectrum inversion should be treated the same way.

To suppose that the Fred case is sufficiently analogous to the inverted spectrum requires supposing that what a color experience had by Fred represents depends on which eye is part of the causal source of the experience. But color experiences are not properly individuated by their causal origins in this way. One might insist that color content for a single individual is determined by phenomenal character quite independently of a particular token's causal origin. It is not plausible to think that the content of those experiences can depend on which eye is part of the causal chain of origin, or that two phenomenally identical experiences within a single subject can vary in content depending on causal origin.

Those who accept that color experiences have phenomenal content will certainly not endorse such a view, since phenomenal content is content that supervenes on phenomenal character. But neither will such a position be plausible for other types of theories of color content. Consider a philosopher attracted to an externalist theory of content, for example one on which color experiences represent whatever they causally covary with in the environment. Such a philosopher might endorse the possibility of spectrum inversion without illusion. She might claim that since red experiences for Jack covary with different properties than do red experiences for Jill, Jack and Jill’s red experiences have different color content. And she might even agree with premise two. Nonetheless, such a philosopher need not accept premise one. Her view does not entail that red experiences caused by the left eye represent physical greenness, while red experiences caused by the right eye represent physical redness. And so Fred’s two eyes need not be analogous to two individuals with inverted spectra.

Instead, the externalist about color content (as well as anyone else) might insist that when it comes to intentional content, color experiences are to be individuated by their phenomenal character. Red experiences for Fred might have vague or disjunctive content, since there are two distinct properties (physical greenness and physical redness) which causally covary with red experiences.[11] For example, a red experience might represent the property of being physically red on odd days or physically green on even days. After all, it is this disjunctive property that covaries with red experiences. All red experiences had by Fred, regardless of the day or which eye is used, will still have the same content.

 

Tye’s Response to the Inverted-Spectrum Challenge

 

            Michael Tye (2000) also tries to defend Standard Russellianism against the inverted-spectrum argument. Tye's response to this argument is two-fold. First, he appeals to his own teleological theory of content in order to give an account of how there could be qualia inversions within a population while one of the subgroups misrepresents color. According to Tye’s PANIC theory of consciousness, the having of an experience with a certain phenomenal character is identical to being in a certain kind of representational state. In particular, conscious experiences are “poised”, “abstract”, “non-conceptual”, “intentional contents”. Color experiences, on this view, are types of mental states that represent physical colors (which Tye further identified with surface spectral reflectances). Tye’s theory of representational content is teleological. A mental state represents a color by having as its function that it do so. More precisely, Tye holds that a sensory state S of a creature c represents that P = df If optimal conditions were to obtain, S would be tokened in c if and only if P were the case, and S would be tokened in these circumstances in c because P is the case (2000, p. 136).

Given Tye’s theory of consciousness, and the teleological theory of content that goes along with it, Tye can allow for the possibility of inverted spectra under particular circumstances. For example, his view allows that functional duplicates might nonetheless have inverted color qualia. This is because a state’s having the function of representing some color or other is not simply a matter of functional role, but is also determined by features of the subject’s environment and its evolutionary history. For example, one of the subgroups of a population may have an inverted-spectrum due to some kind of genetic defect which "crosses wires" in their visual system. Their color experiences come to track properties other than those for which they were designed, and on this basis they can be said to be misrepresentations. Tye also suggests that changes in the environment rather than changes in the species could result in misrepresentation, as when an organism is in an environment "importantly different" from the one within which the visual system evolved.

The strategy here is to deny the possibility of inverted spectra without illusion, and to motivate the idea that one of the inverts is a misperceiver by noting that one of the inverts is a mutant. The mutant’s color experiences are caused by properties for which those states, which are representational, were not evolutionarily selected to represent. Tye is thus supplying us with a principled basis for saying that, for example, one half of a population of inverts misperceives color whereas the other half has veridical color experiences.

In response to this way of accommodating the possibility of spectrum inversion, one might argue that it remains conceivable that there could be two groups of individuals, perhaps of a different species, who are spectrum inverted. And in conceiving of such a scenario, we need not imagine that there is anything about one group rather than the other that would allow us to say that one group is a mutant group. Further, there does not seem to be any grounds for saying that one group rather than the other misperceives color.

            Tye concedes that inverted spectra without illusion remain conceivable in such cases. His primary response to the threat of inverted spectra without illusion is to accept that such scenarios are conceivable, but to deny that they are possible. Inverted spectra with illusion are possible, as in the above kind of case in which a mutation leads to color experiences that are no longer typically caused by their biologically normal cause. But inverted spectrum cases in which neither invert is a misperceiver of color is not possible, despite seeming to be so. Tye takes this line by noting that on his view representationalism, along with his particular theory of phenomenal content, is to be understood as an a posteriori thesis rather than a conceptual truth. Tye insists that qualia inversions without relevant biological differences (and consequently, given his teleological theory of content, no intentional differences) are conceptually possible but not metaphysically possible. The correct theory of phenomenal content entails that spectrum inversion without illusion is impossible, but since this theory is not knowable to us a priori, spectrum inversion without illusion cannot be ruled out a priori (and thus is conceivable). For all we know a priori, a red experience can represent physical greenness. This makes it conceivable to us that someone could be spectrum inverted relative to us without misrepresenting green things. But as a matter of fact, red experiences by necessity represent physical redness. And so as a matter of fact, spectrum inversion without illusion is not possible. Jack and Jill cannot both be veridical perceivers. One of them is a misperceiver, despite it being conceivable to us that they might both be veridical perceivers.

There is a large debate about the relationship between conceivability and possibility that is relevant to assessing the coherence of this position, and to determining when and in what sense it can be correct to say that something is conceivable but not possible (see Chalmers 1996, 1999; Yablo 1999). But I think that there are reasons independent of that general debate that show that this move on behalf of Russellianism about phenomenal color content cannot be successful. This response to the possibility of spectrum inversion without illusion is in conflict with the transparency thesis and the incompatibility of phenomenal color properties.

 

An inconsistent triad

 

            The allowing that spectrum inversion without illusion is conceivable is, I will argue, an unstable position for the Standard Russellian to adopt. The conceivability of spectrum inversion without illusion amounts to the conceivability that a single portion of an object could satisfy the conditions of satisfaction of a red experience had by one subject and the conditions of satisfaction of a green experience in another subject, simultaneously. But the transparency thesis that motivates Russellianism holds that we can know a priori from introspection that the conditions of satisfaction of color experiences are that the objects of perception possess the color properties that are manifest to us in experience. But it is also manifest to us that these properties, such as apparent greenness and apparent redness, are mutually incompatible. Together, these three theses form an inconsistent triad.

 

Transparency

 

Recall that the central motivation for Russellianism is the idea that perceptual experience is transparent. Representationalists like Harman and Tye have claimed that when we introspect on our perceptual experiences, we discover that the properties we find are not properties of our experiences, but rather properties that we represent the objects of our experiences to have. This is meant to be a point about the phenomenology of perceptual experience. And so it seems clear that the properties Tye and Harman have in mind are phenomenal properties. For instance, consider Tye’s remarks about his experience of the blueness of the Pacific Ocean:

Standing on the beach in Santa Barbara a couple of summers ago on a bright sunny day, I found myself transfixed by the intense blue of the Pacific Ocean. …[W]hat I found so pleasing in the above instance, what I was focusing on, as it were, were a certain shade and intensity of the colour blue. I experienced blue as a property of the ocean and not as a property of my experience (1992, p. 160).

It is the phenomenologically salient blueness—phenomenal blueness—that he claims to experience as being a property of the ocean. This claim about experience does not appear to depend on any support other than introspection. It is not a matter of theoretical reasoning about the nature of experience, nor does it depend on empirical considerations that go beyond the deliverances of introspection on the phenomenology of experience. Let p be a phenomenal property (such as apparent blueness) that contributes to the phenomenal character of a perceptual experience e, and that one is aware of when introspecting on that experience. The transparency claim, as a motivation for Russellianism, can thus be put as follows:

 

Transparency: It is a priori on the basis of introspection that e attributes p to the object of perception.

 

            It follows from transparency, of course, that in order for the experience e to be veridical, the object of perception must in fact have p.

 

The conceivability of spectrum inversion

 

            Intuitions that spectrum inversions without illusion are possible reflect our knowledge of our own color experiences and their phenomenal content. Consider a phenomenally red experience that one might have while looking at a fire hydrant. The experience, we will suppose, is veridical. When we imagine the possibility of spectrum inversion, we imagine that there could be a person who has a phenomenally green experience while looking at the very same fire hydrant at the very same time that one has a phenomenally red experience. And in conceiving that spectrum inversion without illusion is possible, one conceives of one’s own phenomenally red experience and another individual’s phenomenally green experience both being veridical representations of the fire hydrant.

            To grant that spectrum inversion without illusion is conceivable thus is not simply to grant that it is conceivable that Jack is a veridical perceiver of color and it is conceivable that Jill is a veridical perceiver of color. It is to suppose that Jack and Jill can both be veridical perceivers of color. And given that Jack and Jill, on this supposition, can have veridical but inverted experiences of a single object (such as the fire hydrant), the conceivability of spectrum inversion without illusion entails that it is conceivable that an object satisfy the conditions of satisfaction of a phenomenally red experience and the conditions of satisfaction of another subject’s phenomenally green experience simultaneously.

           

Apparent color incompatibility

 

            Transparency and the conceivability of spectrum inversion without illusion together conflict with a third thesis, one with at least as much if not more intuitive appeal. Transparency claims that it is a priori that the apparent redness and the apparent greenness of an experience are properties that an experience attributes to the object of perception. That is, transparency entails that it is a priori knowable on the basis of introspection that the conditions of satisfaction of phenomenally red and phenomenally green experiences are such that the object of perception must have apparent redness and greenness, respectively. The conceivability of spectrum inversion without illusion entails that it is conceivable that those conditions of satisfaction for a phenomenally red and phenomenally green experience be met simultaneously by a single object (and by the same region of that object). For these two claims not to lead to a contradiction, it must be the case that it is conceivable that a single object instantiate apparent redness and apparent greenness at the same place and at the same time.

            But it is manifest to us from introspection that apparent redness and apparent greenness are mutually exclusive properties. It is not in fact conceivable that an object have the redness and greenness that are apparent to us in visual experience at a single location and time. The claim that something cannot be both red and green all over has historically been a favorite philosophical example of an a priori truth.[12] Properly understood, I believe that it is indeed a priori that redness and greenness are incompatible.

Gilbert Harman, in a discussion of Bonjour’s In Defense of Pure Reason, offers the following criticism of the claim that something cannot be red and green all over. First, he stresses that just because something cannot look red and green all over does not mean that something cannot be red and green all over. He illustrates the point by having us imagine an object that looks red from one angle but looks green from another angle. It may be that the object cannot look red and green at the same time—but we still might want to say that the object is in fact red and green all over.

Harman also questions whether something couldn’t also look red and green all over. Perhaps it is a lack of imagination on our part that it might seem that something couldn’t look that way. And he cites research by Crane and Piantandia (1983) in which subjects reported that objects look both red and green at the same time.

Before assessing these challenges to the apriority of color incompatibility, we should clarify exactly what is being claimed to be a priori. The claim that has the most plausible right to a priori status is the claim that nothing can be two particular shades of color at the same time and place. Shades are determinates, as opposed to color categories (like being a shade of red) which are determinables. And here, colors should be understood as the qualities that are phenomenally present to us in color experience.[13]

Understood in this way, the claim that it is a priori that colors are mutually exclusive remains a thesis that the imagined defender of Standard Russellianism must deny. But the claim is clearly not refuted by the research of Crane and Piantandia. Crane and Piantandia’s subjects reported that they experienced something as being reddish-green. They did not perceive a surface as having two distinct determinate shades (a shade of red and a shade of green). Rather, they experienced a single shade of color that was best described as “reddish-green”. So this experiment might show that in some sense redness and greenness are not incompatible, in the sense that there is a determinate shade of color that is both red and green. But it does not show that something can be two determinate shades at the same time. Further, one might very well question whether such an experience could ever be veridical. Even if one could experience something as being two shades, it doesn’t strike me as unreasonable to suppose that any such experience is necessarily partly a misperception.

Harman’s example of something that looks red from one angle and green from another can also be questioned. To further elaborate on his example, a surface with ridges such that the object depicts two different things depending on how you orient it.[14] In this sort of case, there is a level of spatial specificity at which a location on the object has only one, not two, colors. Different portions of the surface are visible from different angles, while others are invisible. There is no single portion of the surface that has two shades of color.

This example is not much different in kind from talking about the color of a region of colored newspaper print. Something will look purple from a distance, but upon closer inspection the “purple” region is composed of smaller red and blue dots. Whether it is correct to say that the area of the newsprint is in some sense purple is an interesting question. But such a case does not provide an example of something being two determinate shades simultaneously at a single location identified with the same degree of specificity.

Other kinds of examples that the Standard Russellian might appeal to here undermine rather than support Standard Russellianism. For example, alexandrite is a gem that looks a strikingly different color depending on the lighting conditions.[15] Under white light, alexandrite looks a grayish purple. Under candle light, the same stone looks red. And it may be true that we do not want to say that any particular color experience had under one particular set of lighting conditions is veridical, to the exclusion of all other color experiences had under other lighting conditions. But insofar as this is what we would be inclined to say, we will be moved to one of the following views: 1) that color experiences do not represent stable mind-independent properties, 2) that phenomenal content does not consist solely in the representation of such properties, or 3) objects can change colors depending on how they are illuminated. Such cases do not support the idea that objects can have multiple determinate shades of color simultaneously.

It seems to me that it is a central function of color vision to provide perceivers with information about the ways things in his or her environment differ. When looking at a red thing next to a green thing, it isn’t just that one is presented with an aspect of the red thing and an aspect of the green thing. The green and red things are not merely represented differently by my experience. The two objects appear to actually be different with respect to their color. It doesn’t seem to be compatible with the veridicality of my experience that the two objects in fact be precisely the same color-wise, as they would be if they were both red and green all over.

In summary, we are robustly aware of certain properties when we introspect on color experiences. Our awareness of them is such that it is a priori that they are mutually incompatible. But the transparency thesis claims that these are the properties that our color experiences attribute to external objects. But if this is so, then it should not be conceivable that two experiences (such as a phenomenally red and a phenomenally green experience), representing two of these distinct properties, could both veridically represent the same portion of an object at the same time. Thus, if the transparency thesis is true then spectrum inversion without illusion should not be conceivable. It is conceivable—so the transparency thesis must be false.

Transparency, the conceivability of spectrum inversion without illusion, and the incompatibility of apparent color properties are not mutually consistent. The Standard Russellian might deny transparency, but then the view lacks its primary motivation. He might deny that we can know a priori that phenomenal color properties are exclusive of each other, but this would be to deny what appears to be a datum from introspection that is at least as compelling as transparency. Finally, the Russellian might deny that spectrum inversion without illusion is even conceivable. But that is to concede that the present reply to the inverted spectrum argument fails. And it remains to be explained why spectrum inversion without illusion is not conceivable, despite seeming to be so.

 

Conclusion

 

            Given the possibility of spectrum inversion without illusion, what ought we conclude about phenomenal content? One conclusion would be that color experiences do not have phenomenal content. An anti-representationalist, such as Ned Block, might describe cases of spectrum inversion by simply appealing to non-intentional features of experience—qualia (1996). On such an account, qualia determine the phenomenal character of an experience. Jack and Jill have experiences with different phenomenal character when looking at a ripe tomato in virtue of tokening qualia of different types. Neither Jack nor Jill need misrepresent the tomato, since the tomato is not represented as having this feature of their visual experiences. It might be that Jack and Jill both represent the tomato as being red—but the phenomenal characters of their respective experiences are not fully determined by this representational fact.

            But intuitively, color experiences do have phenomenal content. It seems that, in virtue of having visual experiences with a certain phenomenal character, the world is presented to the subject as being some way rather than other ways. Furthermore, this intentional feature of visual experience is shared by phenomenal duplicates. When I imagine a creature who is having a visual experience that is phenomenally just like the one that I am having, I thereby imagine that the way the world appears to her is just like the way the world appears to me.

            Standard Russellianism identifies these “ways of appearing” with mind-independent physical properties that objects appear to have. The possibility of inverted spectra without illusion shows that this view is unsatisfactory. The way red things appear to Jack is the way green things appear to Jill. But there is no mind-independent physical property that is plausibly represented by such experiences and that red things and green things have in common.

            There are two alternative theories of phenomenal content that promise to accommodate inverted spectra. One possibility is to accept Russellianism but deny that phenomenal content consists in the representation of mind-independent physical properties. This sort of view has been defended by Sydney Shoemaker (1994, 2001).[16] A second alternative, which I develop and defend elsewhere, identifies phenomenal content with a kind of Fregean rather than Russellian content.[17] On this view, color experiences do not involve “directly grasping” those features of objects that color experiences represent. Rather, ways of appearing are modes of presentation of represented properties. Like Standard Russellianism, this view holds that color experiences represent mind-independent physical colors. Jack and Jill both represent red things as having the very same mind-independent physical properties. But they do so under different modes of presentation.

References

 

Block, N. 1996. Mental paint and mental latex. In E. Villanueva, ed., Philosophical Issues, 7. Atascadero: Ridgeview.

 

Bonjour, L. 1997. In Defense of Pure Reason. Cambridge University Press.

 

Byrne, A. and Hilbert, D. 1997. Colors and reflectances. In A. Byrne and D. Hilbert, eds., Readings on Color, vol. 1. MIT Press.

 

Byrne, A. and Hilbert, D. (eds.) 1997. Readings on Color, vol. 2. MIT Press.

 

Chalmers, D. 1996. The Conscious Mind. Oxford University Press.

 

Chalmers, D. 1999. Materialism and the metaphysics of modality. Philosophy and Phenomenological Research 59:473-93

 

Crane, H. and Piantandia, T. 1983. On seeing reddish green and yellowish blue. Science 221:1078-9.

 

Dretske, F. 1995. Naturalizing the Mind. MIT Press.

 

Harman, G. 1990. The intrinsic quality of experience. Philosophical Perspectives 4:31-52.

 

Harman, G. 2001. Rational insight versus general foundations. Philosophy and Phenomenological Research, 63:657-63.

 

Heil, J. 1988. Privileged access. Mind 97: 232-45.

 

Moore, G. E. 1903. The refutation of idealism. Mind 12.

 

Nida-Rumelin, M. 1996. Pseudonormal vision: An actual case of qualia inversion? Philosophical Studies 82:145-57.

 

Shoemaker, S. 1994. Self-knowledge and “inner-sense”. Philosophy and Phenomenological Research LIV:249-314.

 

Shoemaker, S. 1996. Intrasubjective/intersubjective. In The First-person Perspective and Other Essays. Cambridge University Press.

 

Shoemaker, S. 2001. Introspection and phenomenal character. Philosophical Topics.

 

Thompson, B. manuscript. Shoemaker on phenomenal content. http://faculty.smu.edu/bthompso/shoemaker.pdf

 

Thompson, B. manuscript. Senses for senses.

 

Tye, M. 1992. Visual qualia and visual content. In T. Crane, ed., The Contents of Experience. Cambridge University Press/

 

Tye, M. 1995. Ten Problems of Consciousness. MIT Press.

 

Tye, M. 2000. Consciousness, Color, and Content. MIT Press.

 

Yablo, S. 1993. Is conceivability a guide to possibility? Philosophy and Phenomenological Research 53:1-42.