Chapter 5: The Manifest

By Mark Johnston
Princeton University


If other things and other people are to be manifest to me then they must have the natural capacity to appear as they are. That is to say that the manifest properties of things can be themselves the ways in which things appear. There is no need for a mental intermediary or representative, for a quale or a concept, to stand in for manifest properties. No, the idea that some of the properties of things are manifest is on its face the idea that they, and not their mental surrogates, figure in the relation of appearing.

To clarify this central idea of the ready intelligibility of manifest things it is useful to distinguish different uses of the term "appear". Sometimes when we speak of how things appear we are expressing a tendency to believe that they are a certain way. These are the so called epistemic uses of "appears," exemplified by cases like "It appears that the American people are enamored of divided government." Sometimes we are making a comparison between something which appears and a class of things which have a characteristic look or appearance, as when we say that Ted Kennedy looks like an unmade bed, or that the Windsors appear horsey. Hence the so-called "comparative" use of "appears". Sometimes we are registering the fact that we have misleading evidence that something is a certain way, as when we say of the half-submerged stick that it appears bent or separated into two pieces. Just to have a name we could call this the evidential use of "appears."

In talking of appearings, we are prescinding from such uses of "looks" or "appears." We are interested in how things are given to subject in experience, understood as prior to what the subject believes or takes himself to have evidence for. Appearings in this sense will ground the comparisons the subject is inclined to make on the basis of experience but they are independent of any such comparisons. (Cashing out the metaphor of grounding will occupy us in what follows.) It is this so-called "phenomenal" use of appears, which provides the idiom in which we are interested. Unlike Roderick Chisholm and Frank Jackson, who follow the Empiricist tradition in supposing that the ways in which things phenomenally appear are restricted to simple sensible properties like color, smell, taste, felt texture, and shape we shall impose no a priori constraint on the range of features which can phenomenally appear or be manifest features.

Moreover, in speaking of a phenomenal use of appears we are not supposing that this has anything to do with the enjoying of phenomenal feels or qualia, understood as mental features which manifest substances do not exemplify. As with any use of "appears" we want to hold in place the idea that manifest substances could indeed have the features which they appear to have. With those two qualifications of the tradition of talking about phenomenal appearings, we now turn to clarifying and motivating the idea.

Consider what has come to be called blindsight. Patients with a certain kind of damage to the visual system regard themselves as blind, i.e., no longer capable of having any visual experiences. Nothing it seems visually appears to them. However it turns out that when they are asked to make guesses about where things are in the setting before their eyes they guess more or less correctly. Some channel of information is still open so that they can form beliefs and acquire evidence about the setting before them without having visual experiences, without in our sense anything phenomenally appearing to them. However the ultimate scientific construal of blindsight turns out, its ready intelligibility testifies to our grasp of a notion of how things appear which is not epistemic, comparative or evidential.

Another way of introducing the notion of something phenomenally appearing to the subject is by way of the notion of the inverted spectrum. Begin with the one person case, in which a fiendishly clever "rewiring" of the major neural connections in the optic chiasm, a central node in the visual system, makes the sun now appear to me as bright blue things previously appeared and fresh grass now look the way that middle red things looked. We can it seems with equal legitimacy imagine a person wired up in this alternative way from birth. Such a person would learn to make color judgements as we do, or at least would give voice to the same belief expressing sentences in the same circumstances as we do, and so would not realize that there is a difference between how things appear to him and how they appear to the rest of us. The supposed difference in how things appear is precisely a difference in how things phenomenally appear. Whatever the ultimate philosophical fate of the inverted spectrum, its ready intelligibility testifies to our grasp of a sense of how things appear which is not epistemic, comparative or evidential.

A third route to the phenomenal appearings goes by way of a much discussed example of Frank Jackson's. Jackson has us image Mary, a brilliant scientist confined in a monochromatic room, in which no chromatic colors are exemplified. (She herself, we may presume, has porcelain skin, black Irish hair and coal dark eyes.) Ironically Mary specializes in the neurophysiology and psychophysics of vision and acquires all the physical information of what goes on when we see ripe tomatoes, or the sky and use terms like 'red" and "blue." She knows, for example, just which wavelength combinations from ripe tomatoes stimulate the retina and in exactly what ways and exactly the neurophysiological pathways by which such stimulations lead to the contraction of the vocal chords and the expulsion of the air from the lungs to produce the English sentence "The tomato is red." Nonetheless something new happens when Mary is released from her monochromatic room and sees red tomatoes for the first time. She acquires new information about the world, information which was not, or at least not obviously, exhausted by what all the physical information she possessed before. She may have known which colors things outside her room have and in particular that the first tomatoes she would see would be bright red. But for the first time the redness of the tomato phenomenally appears to her. Again, the final verdict on Jackson's example may not be in, but we readily understand that however rich Mary's epistemic position with respect to the colors of things is, she may well be deprived of any experience describable as things appearing red or blue or green to her. She may be in a position to describe how things epistemically, comparatively and evidentially seem or appear with respect to the chromatic colors, while nonetheless being deprived of the relevant phenomenal appearings.

In using the examples of the blindsight, the inverted spectrum and Jackson's Mary I do not mean to be endorsing the standard Cartesian or Empiricist purposes to which such examples are often put. In particular there is no necessary implication that in taking these examples as primarily concerned with phenomenal appearings, one is thereby committed to modeling phenomenal appearings as the enjoying of "qualia" or necessarily private properties of mental processes. Indeed the whole point of the anti-Cartesian account of appearing which we are now developing is to facilitate the claim that the ways in which manifest things (phenomenally) appear are the very same as their manifest and so publicly surveyable properties. In that sense, the trick is to endorse the Representationalist's criticism of the use of blindsight, the inverted spectrum and the case of Mary to motivate qualia, while at the same time repudiating any Representationalist reduction of phenomenal appearings to epistemic, evidential or comparative states of the subject. The good Representationist objection to treating being phenomenally appeared to as the enjoyment of qualia, is that this misconstrues what appearings are like for the subject. When something appears to a subject it is that thing and its manifest properties which are the focus of the subject's attention, not some private occurrence and its intrinsic mental features.

To be phenomenally appeared to is neither to enjoy qualia, nor to be in some representational state which "says" that the world is thus and so. (Remember our previous objections to each of the available models of appearing as the entertaining of a representational content.) To be phenomenally appeared to is to have a manifest property and a manifest substance present to one. So the right thing to say about Mary is not that she lacks color "qualia," nor that she has yet to entertain certain privileged visual representations of red, but that until she leaves her room and sees ripe tomatoes and the like, she doesn't know what red -- a publicly available property of external things -- is like. In that sense she did not have all the relevant physical information while in her cloister. Concerning a range of physical properties of ordinary physical things Mary did not know what those properties were like, because nothing with those properties had ever appeared to her.

Indeed there are law-like connections in nature that Mary is not in a position fully to understand. For example it is a law-like regularity that light of wavelength 690 nm is unique red, a red with no admixture of yellow or blue. But Mary does not know what unique red is like and does not know what exactly this truth of color science really involves. Because she has not enjoyed the right kind of phenomenal appearings Mary is ignorant of certain aspects of nature. (Similarly our ignorance of what it is "like to be a bat" is inter alia ignorance of what sonic hardness is like.)

If talk of what a physical property is like sounds a little odd then this simply testifies to the power exerted over our thought by the Cartesian/Galilean conception of physical properties as all of them quantities, i.e. properties about which detailed numerical comparisons and mathematical descriptions of their interrelations make sense but about which it makes no sense to ask "What are these properties (mass, distance, time, position, wavelength) like?" In making Mary an expert in neurophysiology and psychophysics, Jackson did not confer on her complete qualitative information about the physical world, but only the quantitative information with which those sciences are concerned. Thus the case of Mary is not a way of motivating a mind-body problem, e.g., "How could a publicly surveyable brain process instantiate private mental qualia?" Rather it raises the (less noxious) counterpart of the mind-body problem for non-mental, publicly available bodies, namely how can objects whose efficient causal transactions are able to be accounted for in purely quantitative terms, nevertheless instantiate the qualities which they appear to us to have. Remember that it is precisely our accusation against Descartes that his Projectivist denial of form -- the qualitative way in which bodies appear -- represented an "introjective error", the mentalization of the qualitative. Jackson's thought experiment serves to underline the indispensability of qualitative knowledge of the physical world. Only if we follow Jackson in taking for granted the mentalization of the qualitative will Mary's case even look like it provides the material for an argument for mind-body dualism. The so-called hard problem for physicalism, the problem of how to identify qualia with properties of brain processes, is a product of the Cartesian introjective error.

In a very witty polemical article "The Rediscovery of Light," Paul Churchland parodies the dualistic claims of Frank Jackson, John Searle and David Chalmers by showing their arguments that the physical description of the world leaves out what is distinctive about our mental life can be paralleled by arguments which show that what is distinctive about visible light is left out by electromagnetism.1 Electromagnetism is a purely quantitative description which provides a vocabulary and quantitative laws which suffice for the description and explanation of every causal transaction involving visible light. Churchland's parody is not completely successful. For one thing, I can imagine one of Jackson's cloistered scientists who is blind but knows all there is to know about electromagnetism, and who acquires new information when he is first provided with the gift of sight. Jackson would say that he learns what it is like to see light and this information concerning qualia is not exhausted by electromagnetism. Again, I think that this is a misdescription of what is learnt, the focus being upon what it is like to undergo the process of seeing light. What is learnt is what visible light is like, or what a particular kind of visible light is like. Electromagnetism does leave that out, but this does not point to any incompleteness in its account of light.

The completeness of a theory is determined by the point of the theory. Electromagnetism is a theory of causal transactions. Like any such developed theory it works precisely because it replaces the qualitative descriptions which sensing prompts with elaborate quantitative descriptions. Once again, there is no mind-body problem in play here but rather the general problem of the relation between the qualitative and the quantitative.


Is the Qualitative Reducible To The Quantitative?

A natural place to illustrate the difference between the quantitative and the qualitative is in the case of the colors. The idea of the colors as ordinary manifest properties is the idea that appearances which present the colors of things thereby show us what the colors are like. The central problem in the philosophy of color is to account for the relation between the colors so conceived and the quantitative surface properties which psychophysics identifies as determinative of how surfaces interact with light. Thus on Edward Land's theory of color each discriminable shade has associated with it a triple of designators, where a designator is the proportion of visible light of either long, medium or short wavelength reflected back by the shaded surface to the amount of light of that wavelength coming from a suitably determined surround. So a designator compares a disposition of the surface to reflect back light of a range of wavelengths to the corresponding disposition of the surround. How are triples of such quantitative light dispositions related to the shade qualities which we see?

I wish to consider five views: Color Physicalism or the identitification of colors with light dispositional properties, Projectivism about color with the projected properties being private qualia exemplified by experiences of color, Projectivism involving the entertaining of necessarily false content, Locke's view of colors as response-dispositional properties, and finally the view to be preferred, Hylomorphism concerning color. Color is here figuring as a paradigm secondary quality; after arguing for Hylomorphism concerning the colors, I shall then turn to the so called primary qualities, shape in particular.

First the two versions of color Projectivism. To suppose that in ordinary color experience we are systematically mistaking qualia or intrinsic features of my own private mental events for surface properties implausibly attributes massive category error to us -- mistaking a mental property for a physical one -- and indeed a massive category error we cannot correct. We simply cannot experience the world in vision (or indeed by way of touch) as a mere set of causes of qualia in us, causes whose natures remain unrevealed by the qualia itself. Sensory experience inevitably seems to be showing us something about how the world is. Hence the appeal of the representational version of Projectivism. On this view our experience represents the world as colored in just the qualitative ways it appears to be, i.e., in sensory experience we are representing the world as thus and so with respect to the qualitative colors. Projectivism now consists in the claim that these representation are globally false. The problem which disables this view is a generalization of the missing shade of blue problem urged against representational accounts of experience in Chapter X. Now the Projectivist hypothesis itself implies that all the shades are missing from the world, and probably necessarily so since the world is being conceived of as a world fully captured in quantitative terms. But how then do experiences get to have the necessarily false content which the Projectivist is now supposing them to have? Not, as we emphasized before, by way of any application of concepts in one's experience in the manner in which one deploys concepts in one's thought. And not by the Projectivist's hypothesis itself by way of any correlation with instances of the properties which it represents things as having.

Projectivism at least embodies a vivid sense of the disparity between how experience presents the colors as being and the physical quantitative features of surfaces which according to our best psychophysics causally determine the operations of our visual systems. In contrast, Physicalism about color attempts a straight out identification of the colors with such quantitative properties.


Colors As Qualities

The qualitative character of the colors, that which confers sense on such questions as "What is the shade puce like? Is it like chartreuse?" can be explicated in two ways. The first way is to concentrate on the relations of qualitative incorporation among the shades. We can see that purple has red and blue in it and that orange has yellow and red in it. But other colors are not in this way qualitaitve compounds. They are "unique". As C.L.Hardin puts it

When it comes to live physical candidates like (classes of) triples of designators, the distinction between unique and binary triples may not even make any sense. In any case even if color science found something answering to this binary/unique distinction at the physical level it would be a theoretical discovery appearing late in the history of mankind. Ordinary visual experience would not suufice to determine that G is unique and H binary. yet we know which colors are binary or unique simply on the basis of visual experience alone. So the colors are not to be identified with any of the Physicalist's candidates, for the colors have a qualitative structure manifest in visual experience.

Another way of bringing out the qualitative character of the colors and the difficulty it raises for Physicalism is by elucidating the family of similarity and difference relations among the shades. When we say for example that canary yellow is not a shade of blue, what we say comes to this: canary yellow is not as similar in hue as the various shades of blue are among themselves, i.e. the blues exhibit a degree of similarity in hue among themselves which no blue stands in to canary yellow. We are in a position to know such facts of qualitative similarity and difference simply on the basis of havng the relevant shades of things appear to us. These appearings show us the qualitative way each shade is and hence provide information that settles, determines or entails that canary yellow is not a shade of blue.

The Physicalist account of color has it that the shade canary yellow is the light-dispositional property which standardly explains the canary yellow appearances. (Think of it as a triple of designators, if you like.) Mutatis mutandis for the various shades of blue. Suppose that color science ends up discovering this: the light- dispositional property which standardly explains the canary yellow appearances and the various light-dispositional properties which standardly explain the various appearances as of the shades of blue are not, when taken together, as similar among themselves as are the various non-dispositional properties which standardly explain the various appearances of the shades of blue. On the simplest version of the Physicalist account, this would be the discovery that canary yellow is not a shade of blue, i.e., not to be counted among the blues.3

But is it really a matter of scientific discovery that canary yellow is not a shade of blue? No: such similarity and difference principles surely have a different status. We take ourselves to know these principles just on the basis of visual experience and ordinary grasp of color language. No one had to wait until the end of the second millennium A.D. to find out whether or not canary yellow is a shade of blue.

That, of course, is just a first move against the Physicalist account. The Physicalist should be allowed to answer that indeed it is not a matter of scientific discovery that canary yellow is not a shade of blue. Rather, he might say, such a principle, along with other similarity and difference principles, must be held true as a condition on any family of properties deserving the color names. So the principle that canary yellow is not a shade of blue turns out to be relatively a priori after all. More exactly what is a priori is a biconditional: P deserves the name "canary yellow" just in case (i) there is a light dispositional surface property P standardly responsible for the appearances as of canary yellow things and (ii) this property stands in the right similarity relations to other standardly explanatory categorical properties.

On the proposed theory, a given property turns out to count as canary yellow only if a complex similarity condition on that property and a host of others is discovered to hold. For example the candidate properties to be the blues have to show a natural or genuine similarity among themselves, a similarity which they do not share with the candidate to be canary yellow.

Suppose color science discovers this condition holds along with the other similarity and difference conditions which the Physicalist treats as central. Then some physical properties turn out to be canary yellowness, teal, turquoise, sky blue and so on. And particular things turn out to have these properties. But what then gives one the right to say that there are canary yellow things is not simply visual perception and the very general background beliefs which inform visual perception, but also and crucially, recherche facts from color science. That is, on this version of the Physicalist account one is not justified in believing that some things are canary yellow unless one knows that color science finds that among the causes of our experiences of color are physical properties which stand in certain complex similarity and difference relations. For this is a central precondition which this version of the account lays down on any property characterized in color science turning out to be canary yellowness, and hence on particular things turning out to be canary yellow. The result of such an account is that the colors are not available to us in perception in a way that immediately leads to justified belief.

The conclusion for which we are aiming is this: once the Physicalist account is adjusted to accommodate the similarity and difference conditions it will imply that we are not justified simply on the basis of visual perception and the background beliefs which characteristically inform perception in believing that say, Zinka the canary, is canary yellow. For we are evidently not justified simply on this basis in supposing that the light-dispositional surface causes of our visual experiences exhibit the relevant similarities and differences. The Physicalist thus fails to account for the colors as manifestly there, pervading the surfaces of manifest substances, and so open to view in a way that immediately leads to justified belief.
However, to successfully argue that we must engage with a complication familiar to epistemologists. This is the idea that by a convenient "failure of deductive closure" we could still be perceptually justified in believing that there is a property, canary yellow, had by Zinka even though we are not perceptually justified in believing that any property satisfies the similarity condition for being the property canary yellow. Whatever the general merits of the idea that one need not be justified in believing all the deductive consequences of what one is justified in believing, the idea of failure of deductive closure has its limits, and it can be shown that the conclusion for which we are aiming cannot be plausibly evaded by an appeal to a convenient failure of deductive closure. For on the present version of the Physicalist account, the requirement that a host of light-dispositional similarity and difference relations hold is not just a collateral consequence of there being colors in general and canary yellow in particular. Instead, the present account has it that the claim that there are colors is conceptually equivalent to the claim that the light-dispsotional surface properties standardly causally responsible for our experiences as of colored things exhibit the required similarities and differences.

The relevance of this last point may be brought out in the following way. Imagine a sophisticate who took the alleged conceptual equivalence to heart and found himself therefore hesitating in concluding just on the basis of the way Zinka the canary looks that Zinka is canary yellow. "Zinka certainly looks the way something would have to look to count as canary yellow" he thinks "but we must wait and see if color science discovers the similarities and differences required for there to be such a property as canary yellow." Given his lucid understanding of the Physicalist account of canary yellow, the sophisticate would not be justified in concluding just on the strength of perception that there is such a property as canary yellow. Hence he is not justified just on the strength of perception in taking Zinka or anything else to be canary yellow. Yet on the present account the sophisticate has the correct understanding of the concept canary yellow. So we in our turn can hardly be justified in concluding just by looking that Zinka (or anything else) is canary yellow. For we gain no global advantage with respect to justification by failing to be conceptually lucid. Thus the Physicalist does fail to account for the colors as manifestly there, pervading the surfaces of manifest substances, and so open to view in a way that immediately leads to justified belief. .

To be sure, there are well known cases in which more empirical knowledge would put one at a comparative disadvantage with respect to empirical justification -- cases in which one "knows more by knowing less" -- and we can invent conceptual analogues of such cases.4 However such cases never show the kind of global disadvantage with respect to justification from which our sophisticate suffers. If conceptual lucidity is not enough in itself to produce a global epistemic disadvantage then given Physicalism about color we can be no more perceptually justified in believing that things are canary yellow than is the conceptual sophisticate. Conclusion: the Physicalist account either fails to give our qualitative knowledge of the colors the right status or it fails to capture the fact that the colors of things are perceptually manifest in a way that leads to immediate justified belief.

Lockean Dispositionalism Concerning Color

The same argument which shows that the colors are not to be indentified with light-dispositional properties also shows that the colors are not to be identified with the micro-physical bases of such light-dispositions. These would be heterogeneous disjunctions of the various micro-physical properties which ground a shade- determining triple of designators. The same problem arises: the similarity and difference relations which are definitive of the colors, need not apply to the candidate micro-physical properties. Requiring that they apply if any set of micro-physical properties are to turn out to be the colors makes the existence of the colors a scientific conjecture, not something manifest to us in ordinary perception.

Contrast the Lockean account of colors as response- dispositions, power to produce certain experiences in us. That account secures the right status for the similarity and difference principles, allowing for example that we can know just on the basis of perception and ordinary understanding of the color terms that canary yellow is not a shade of blue. To see this let us reduce the problem to its simplest form: take teal and turquoise. They are similar color properties. Indeed they are essentially and intrinsically similar. That is to say teal and turquoise exhibit a kind of similarity that is not a similarity in the other properties to which they are related, nor a mere similarity in their causes and effects, nor a similarity in the properties upon which they supervene. Rather, the similarity between teal and turquoise with which we are concerned is to be found in any possible situation no matter how their instances, effects or contingent relations with other properties (including lawlike relations) vary. This is what I mean to focus upon by saying that teal and turquoise are essentially and intrinsically similar. Suppose one could spell out the nature of teal and the nature of turquoise, i.e. the higher-order features these properties have in any possible situation. Then that specification of features would list some common features of teal and turquoise. That is the way in which teal and turquoise are similar. They are not similar simply in virtue of being (even nomically) related to similar consequences or similar bases. They are similar in virtue of what they essentially and intrinsically are.

If teal and turquoise were categorical micro-physical properties then any essential and intrinsic similarity between them would have to be a similarity in some higher-order micro-physical respect. What we know simply on the basis of perception is not sufficient to know that there is such a similarity.

However, if teal is essentially the disposition to manifest a certain appearance Te and turquoise is essentially the disposition to manifest the appearance Tq then teal and turquoise will be essentially and intrinsically similar if these two manifestations are similar. That these dispositions have similar manifestations is a fact available to us in visual perception. For it is evident in visual perception that the appearance Te is similar to the appearance Tq. That these manifestations are more similar to each other than either is to the manifestation of the disposition canary yellow is also a fact available to us in visual perception. By a simple extension of these considerations, the fact that canary yellow is not a shade of blue, i.e. the fact that canary yellow is not as similar to the blues as the blues are among themselves, is guaranteed by the claim that these properties are dispositions and by the evident fact that the appearance of canary yellow is not as similar to the appearances of the blues as those appearances are among themselves.

Notice that the different status of color similarities on the Physicalist and Lockean accounts derives exactly from the central difference between the two accounts. It is precisely because the Physicalist account treats the color appearances as merely the standard effects of the microphysical properties it identifies as the colors that the account cannot allow for perceptual knowledge of intrinsic and essential similarities among the colors. On the Lockean account the color appearances are not merely the standard effects of the dispositions whose manifestations they are. Since they are also the manifestations cited when attributing the relevant dispositions, we know something intrinsic and essential to these dispositions when we know their manifestations.

What then is the difficulty with the Lockean account? It faces a decisive dilemma: either it simply recapitulates in a different terminology the Projectivist problematic or it is unhelpfully circular. Let me explain. If the colors are none other than dispositions to visually appear to us in certain ways, how are we to think of these appearances?

One option is to think of appearances as the enjoying of qualia. The problem is that we do not experience color in anything like the way in which we experience, say, nausea, i.e. as a kind of inner event we undergo as a result of the influence of something external, say rotting and noisome meat. The noisome meat has the disposition to produce an "inner experience" of nausea in us; nausea being, along with pain, the best candidate to be understood as the enjoying of qualia. But we do not mistake our experience of nausea for a surface property of the meat. Conceiving of visual experiences as the enjoying of qualia thus implies that we are systematically mistaking intrinsic features of mental events for properties of surfaces.

A second option is to think of visual experiences as the entertaining of false representational content, i.e. a content to the effect that the surface one is seeing is red*, where being red* is a non-dispositional surface property distinct from the property of being red which the Lockean is treating as a disposition. However, this is itself a very unattractive variant on Projectivism. If visual experience represents surfaces as red* and they are in fact not red* but red then visual experience is systematically false. Visual experience simply misrepresents the world. The dispositionalist gloss on this, namely that surfaces are disposed to prompt us to entertain such false content, is a sideshow. The colors are not properties manifest to us in visual experience.

If the colors are to be manifest to us in visual experience then the very properties which surfaces visually appear to have must include, along with the shapes, the colors themselves. So if red is a disposition to produce a visual experience of a certain kind of surface the name of the kind of surface in question must be none other than "Red". If Lockean dispositionalism is to be anything other than a variant on Projectivism then red must be treated as the disposition to appear red, green as the disposition to appear green, and so on. This however is to render the dispositional account ungrounded in a way that deprives it of any interest. Red is the disposition to appear to be disposed to have the disposition to appear... and so on without terminating, and so without associating any definite property with red or green or indeed any color. The destructive dilemma against Lockean Dispositionalism is that it either simply recapitulates Projectivism or it is empty.

In developing this dilemma against Dispositionalism I have oversimplified. I have written as if there were only two ways to regard appearances and so dispositions to appear. Either they are the enjoying of qualia as on the Sense Datum Theory or they are entertainings of content as the Intentionalist would have it. But of course there is a third conception which is to be preferred to both. Appearances are really appearings, i.e. the coming to hold of a relation between a subject, a property, an object. So under certain lighting conditions and from a certain viewing position the cricket ball appears round and red to me. On this view of appearances, which we argued to be the right view and the right way out of the problem of acquaintance, we can state the dilemma this way: are the relevant properties which figure as elements in appearings color properties like being red or not? If not, Dispositionalism is just a "polite" variant on Projectivism, Projectivism with a fig leaf if you like, since the colors are not manifest in visual experiences. If so, then contrary to Dispositionalism, the colors cannot be dispositions on pain of an infinite regress.



The right conception of the colors, and indeed of all sensible properties, is a Hylomorphic conception which treats these properties as qualities constituted by the causally relevant underlying physical quantities. So as a first approximation (to be refined in the wake of the discussion of color relativity below) the colors are qualitative properties constituted by surface light- dispositions of the sort explored by psychophysics. As such they do not compete for the same explanatory space as the light- dispositions. As psychophysics make clear, all of the detailed physical functioning of our visual systems relevant to the surfaces of things appearing to us is causally determined by light-dispositional properties of the surface (on Edward Land's theory, triples of designators).

The light dispositions do not visually appear to us. The activation of an appropriate light-disposition is the necessary condition at the level of material process of the surface appearing anyway to us. Other things being equal, it may also be causally sufficient. However the property of the surface which appears will be a definite shade property; Chinese red or canary yellow or whatever the case may be.

Just as on the side of the subject a certain physical process involving the visual system must take place for anything to visually appear, so on the side of objects certain physical light dispositions must be present if the object is to be capable of visually appearing. Once we admit all this and thereby grant psychophysics complete hegemony in its own explanatory space, we are still free to maintain that in virtue of such quantitatively characterizable physical features and processes what appears are not quantities but qualities: the colors and the shapes of ordinary experience. Color and shape make up the visible form of objects, light dispositions and surface lattice structures are the material properties which constitute this form.

Where do the Lockean dispositions -- the dispositions to appear this or that color or this or that shape -- come in? Since we now have an independent characterization of the colors and shapes there is no harm in admitting these dispositions understood not as shapes or as colors but as dispositions of colored and shaped objects to appear as they are in those respects. As well as being red, an object will typically have the disposition to appear as it is with respect to color, i.e., to appear red. The fact that red objects typically have that disposition is part of what it is for red to be a manifest property. We can imagine special cases in which something will have a color but not have the disposition to appear that color. One such case is the radiation zone, the region inside the sun immediately surrounding its core. From the conjectured physical character of its contents the radiation zone is thought to emit spectral red light. It is thus conjectured to be a radiant red region. However the radiation zone is encased within the convection zone, which is exceedingly hot; so hot that it is physically impossible for any sighted being to pass through it and so see the radiation zone. On any reasonable account of dispositions, including the one provided in Chapter X, it is implausible to regard the radiation zone as having the disposition to appear radiant red.

Such cases serve to highlight the fact that colors are not dispositions to appear colored, but it is also important that such cases are atypical and so do not involve what happens for the most part. For the most part the colors of things are manifest, i.e. things have the disposition to appear as they are with respect to color.


Are Shapes And Sizes Manifest Properties?

Do ordinary objects have the disposition to appear as the very shape they are? If we conceive of shapes as what I will hereafter call "quantitative shapes" i.e., precise quantitative properties exhaustively described with full numerical precision in analytic geometry, then it will follow that the shapes of manifest substances are not themselves manifest, i.e., the substances in question are not disposed to appear as they are with respect to shape.

If the only shapes are the quantitative shapes then physical objects could not have any of the ordinary shapes which immediate perceptual belief attributes to them. Projectivism, whether polite or impolite, will then be true of shape. To see that this is so, consider how things are with our best physical models of objects. Any glimpse into the chemistry of surfaces teaches that in so far as the physical objects we sense have quantitative shapes, those shapes are not anything like the relatively regular shapes which immediate perceptual belief attributes to objects. No names exist for the irregular shapes which would tightly enclose the "boundaries" of the molecules making up the surface of any physical object. Nor would there be much point in introducing such names since each such shape is likely to be completely idiosyncratic. Nor should we rest content with the thought that, for example, a table top can nonetheless be roughly square, where "roughly square" is a name for a range of quantitative shapes which are geometrically alike -- as it might be squares, rectangles with similar sides, rhombuses, rhombuses with a few kinks in their sides and so on. None of these shapes are to be found in objects. Indeed if an object had any such shape it would have dimensionless edges and apexes. (Being infinitely "sharp," it would cut us to pieces!) The candidates to be the real physical shapes of objects are so complex relative to the shape names which perception encourages us to apply that the upshot of supposing that the shapes we sense have to be quantitative shapes is that we habitually misperceive shape and wildly misperceive it in a way that makes untenable the idea of perceiving the quantitative shape up to some approximation of it.

Consider this analogy: approaching a villa you see its huge iron gates awaiting you. From the distance the gates appear to you to be solid but in fact they are intricately worked in a dense open pattern. Only when you get quite close to the gates can you see through the iron work. Clearly, from far away you mis-perceived the shape of the pattern in the gate, or worse, you failed to perceive it at all. It was not visible to you. (You perceived no approximation of the pattern. You just couldn't see it.) So also with the shapes of objects interpreted as quantitative. You don't perceive any such shapes. They are too small-scale to be visible.

Nor can we feel or touch such shapes. Consider the case of running your hand over the spines on a porcupine. You will have no sense of the details of the surface texture of the porcupine's skin beneath its spines. Thanks to the spines and the size of your fingers the skin is intangible. So also with the shapes of objects interpreted as quantitative. They are too small-scale to be made out by something as gross as a finger tip. We can neither see nor touch them. Nor do we see or touch approximations of them. Seeing the gate as solid is failing to see the internal structure of the gate. Feeling the porcupine as prickly is failing to feel its smooth skin beneath. Who would want to speak of approximations here?

Someone might take this as showing that real shapes are not manifest properties because they are so small-scale, but that can be exposed as an unstable conclusion, ready to collapse into the bizarre consequence that objects have no shape or size. For once we move from a relatively crude chemical picture of surfaces as complexes of molecules to the model of a surface provided by quantum chemistry the very idea of a physical surface and of the real physical shape and size of an object begins to evaporate. We have instead a complex field of force which is operating at places where ordinary perception reveals no object but only space. It is not that the field of force of a bigger object is more extensive than the field of force of a smaller one. It is rather that it has slightly different "gradients" or patterns of falling off as we move away from the notional center of gravity of the object.

Despite all this, it seems deranged to conclude that material objects have no shape and no size. We must have gone wrong in supposing that shape could only be quantitative shape as described by chemistry and physics. We need another conception of how manifest physical objects can be cubical, spherical, cylindrical and so on. As we shall see, on that conception the shapes and sizes of manifest substances will turn out themselves to be manifest. The manifest shapes and sizes are qualitative in a sense which I shall now explain.


Qualities and Quantities Again

Consider that most neglected of important distinctions in ontology, i.e. the distinction between red and being red. As Jerry Levinson points out there is a simple yet decisive argument for making the distinction: red could be the color of Sam's XJS V12 convertible, but it is simply ungrammatical to say that being red is the color of Sam's XJS V12 convertible. Compare "Sam is my friend." and "Being Sam is my friend." What are we to make of that? "Being red" is the name of a property, a condition exemplified by many red things. Things get to be red by having the quality red pervade dominant salient parts of their surfaces. What is the import of saying that red is a quality and not a quantity. Levinson gives the most plausible account. (I differ with him in only one detail which will turn out to be crucial for present purposes.) Qualities, e.g. red, charity, vivacity, flatness, squareness, bigness, unlike their corresponding properties, admit of comparisons of degree and amount. We can speak of one thing as more red (or colder) than another but not of it being more being red (or of it being more being cold). We can speak of something being very vivacious but not of being very being vivacious. We can speak of an abundance of charity but not of an abundance of being charitable. A person can possess some tenacity but not some being tenacious. It is as if we think of qualities as kinds of non-material ("formal"?) stuff which things can possess to various degrees and in amounts. Things get to be in the corresponding states or conditions, viz., of being red, of being cold, of being charitable, of being tenacious by partaking to some degree and in some amount the non-material stuff in question. Hence as a rule the quality and the corresponding property will be co-instantiated. Maybe that is why we seldom care to distinguish them.

Qualities and their associated properties differ with respect to whether they are primarily particular to their bearers. To be cherry red is to have some cherry redness pervading your surface. This means that different objects can be cherry red, i.e. exhibit a common condition or property, thanks to having different particular cherry reds pervade their different surfaces. The qualities are particularized, the associated properties are universals common to many. I might say "That red (pointing to the surface of my car) is the product of much weathering." I couldn't say "That property of being red is the product of much weathering." Whereas it at least makes sense to talk of qualities as if they were particulars, the property of being red is a universal common to many particulars.

Now the ordinary names for shapes and sizes behave like quality names. (Here I differ from Levinson.) Just as something can be more red or less red than something else it can be more or less square or round or cylindrical or big for a Italian greyhound than something else. The logic of ordinary shapes and sizes is not so much the logic of a non-material stuff but that of non-material containers for such "stuffs" as warmth, weight, solidity. And the ordinary idea of the surface of an object is not the Physicalist idea of a set of slices slice across a force gradient declining from the notion center of gravity of the object. Instead it is the idea what the barrier color of the object pervades. By having its quality stuffs "contained" by a cylindricality of a certain size and having its barrier color pervade its cylindricality, an object counts as having the property of being cylindrical and being a certain size. If the object has no barrier color but exhibits a transparent volume of color, as with a cylinder of pink ice, then the color volume like the coldness and hardness qualities of the ice will be contained within its shape.

The qualitative shapes form a family and just as with the colors there are unique hues like red and yellow in terms of which we can characterize the qualities of binary hues like orange there are basic qualitative shapes which cannot be explained in terms of the other shapes. Plausibly included among these are flat, curved and pointed. Just as Jackson's Mary had to have experience of red to know what it was like there is no way of knowing what these shape qualities are like without seeing or feeling them. A knowledge of analytic geometry might be held to provide knowledge of their strict quantitative counterparts, but it is hard to see how such knowledge could be gained except by way of an idealized refinement of our actual experience of things with qualitative shapes; written words and diagrams for example. Spatial imagination like color visualization depends essentially upon experience of at least some of the qualities in question.

How exactly do we get from the manifest shape qualities, the seen and felt "containers" of the other manifest qualities of things, to what I have been calling the quantitative shapes? By increasing refinement of the comparisons of degree which are indicative of our talk of qualities. So, this surface is flatter than that, still a third is very much more flat than the first compared to the second. Measuring devices and a little geometric theory soon encourage the notion of the perfectly flat as a limiting idealization of the process of finding a surface flatter than any yet considered. Similarly with the perfectly sharp edge, or the perfect cube or cylinder. Then we can describe such shapes by way of algebraic formulas describing an exact volume in an abstract three dimensional mathematical space. By then we have idealized away from the qualitative shapes of actual experience which seem by comparison vague or ill-defined.

The next step is to model the shapes of ordinary objects as approximating to this or that quantitative shape and likewise model there movements as exact velocities and their heavinesses as exact masses. We then have a simple kinematic model of objects which presents itself as a better than the ordinary conception because it has shed ill-defined descriptions in favor of mathematically precise descriptions which allow for further refinements of our abilities to predict the movements and interactions of bodies.

But suppose that reality is inherently imprecise, at least by the standards of the idealizing models which we abstract from our experience of the manifest. Then our precise models are not a revelation of an underlying reality but just devices for the purposes of more refined prediction and control. To set the artefacts of such refined models over against the manifest as an underlying reality obscured by subjective appearances, as the Projectivist does, is simply to invert the order of intelligibility.


Modeling With Quantitative Shapes

Now speaking of the quantitative shapes, we can say that the edge of a table top is approximately or roughly square on the grounds that it looks, feels and measures square, and that it interacts with things of other shapes as something square would, at least as such interactions are described in that primitive kinematics which we practically master as we come to learn our way around in the world. It is within such a primitive kinematics and its more sophisticated descendants that the attribution of the quantitative shapes to objects plays its main role. In claiming that the cricket ball is a hard, solid, one-pound, spherical object flying through the air at about a hundred miles per hour, and meaning thereby to refer to the quantitative shapes, weights and speeds, one is modeling the cricket ball in a primitive kinematics which describes the interactions of middle-sized objects.

This gives a hint as to how to handle "roughly square", "roughly spherical" etc.. In saying that the cricket ball is roughly spherical, roughly a pound in weight and traveling at roughly one hundred miles per hour, we could mean something that is true and relevant about the relevant quantitative properties, even though nothing physical is spherical, weighs exactly any definite quantity we like to mention, or travels at any speed we are tempted to ascribe to things. (I assume that most of the weights and speeds which could truly be ascribed without qualification to things are irrational quantities whose specifications would require non-terminating decimals. An electron cannot obey the instruction "Adagio" and not because it goes too fast.) We could be saying that when it comes to the rough and ready purposes of negotiating our environment of middle-sized objects by way of a kinematic model of that environment, the cricket ball can be successfully modeled as a spherical, one pound object with a speed of one hundred miles per hour. Here the quantitative shapes along with the other kinematic quantities of speed and weight are concepts employed in a kinematic model.

What is the relation between such quantitative properties and the qualitative shapes, speeds (fast, slow, largo, adagio, presto etc.) and weights (heavy, light, etc.) which things appear to have? A visual experience of the cricket ball as spherical is not an experience of it as appropriately modeled as spherical in a kinematic theory of the ball and our common physical environment. After all, the experience could be, in the sense established in Chapter X, non- conceptual. Nor is the content of one's immediate perceptual judgement that the cricket ball is spherical that it is successfully modeled as such. Immediate perceptual judgement does not say this, since like experience it concerns itself with the manifest qualitative shapes. But in an important sense it shows this. That is to say, when someone has an immediate perceptual belief to the effect that the cricket ball is spherical and finds the actions this belief prompts to be successful then it is a reasonable bet that the cricket ball is successfully modeled as such in a rough and ready kinematic theory suited to negotiating our interaction with our environment. Immediate perceptual beliefs about the shapes and speeds and weights of things concern and typically hold true of the qualitative shapes. The arising of such beliefs and their success in guiding action testifies to the fact that there are successful kinematic models of what is perceived in terms of precise quantitative counterparts of those qualitative properties.

Whether these models are any of them also true mirrorings of physical reality is something on which I maintain an official silence.
Once again, the defense of the manifest should not depend on settling the issue between the Scientific Realist and the Constructive Empiricist. The enemy of the manifest is not Scientific Realism per se but rather the associated Scientism which supposes that any mode or object of understanding must be reduced to efficient causal explanation in purely quantitative terms.

The Argument from Relativity.

Afficionados of perception will be chomping at the bit, for our defense of the manifest has so far neglected a central datum of perception, the relativity of sensory qualities. For some, this above all is the motive for thinking of the manifest as merely appearance. Hume for example took this to be the crucial consideration in favor of a Projectivist account of the so-called secondary qualities

The fundamental principle of [the modern philosophy] is the opinion concerning colors, sounds, tastes, smells, heat and cold; which it asserts to be nothing but impressions in the mind, derived from the operation of external objects, and without any resemblance to the qualities of the objects. Upon examination, I find only one of the reasons commonly produced for this opinion to be satisfactory; viz that derived from the variation of these impressions, even while the external object, to all appearance, continues the same.5

The considerations concerning relativity can get started in a variety of ways. Here is one way. Other perceivers, e.g., anomalous trichromats, the birds and the bees, are blamelessly different from us, and so do not misperceive our manifest colors. Other colors are manifest to them. The same facts tempt Projectivists to assert that colors are not in the object but are mental specters which we mislocate on the surfaces of objects. The birds, the bees and the anomalous trichromats enjoy different mental specters from ours. The main thing to be said for Projectivism is that it is an interpretation of the facts of relativity.

As Hume suggests, to assay the facts of color relativity we do not need to resort to other species, or to human exceptions such as anomalous trichromats. Perfectly ordinary human perceivers in perfectly good though different lighting conditions can see the same things as having different colors. So from your side of the inlet the sea looks grey and from mine the same body of water looks green. (This is of course a case of volume color but similar cases could be produced for both surface and radiant color.) You have a veridical experience of the sea's greyness. I have a veridical experience of the same body of water being green. So, on the assumption that the colors are manifest properties of surfaces or, in this case, of volumes, the part of the sea which we each look at is both grey and green.

If we are Hylomorphists about color then our conception of what is taking place is this: the light-dispositional property of the relevant volume of sea-water constitute both the greenness and the greyness of the sea. The light-disposition is the "matter" for these qualitative forms. Now our respective visual experiences are ex hypothesi veridical. So when I look at the sea it is the greyness and not the greenness of the sea which explains and determines the character of my veridical experience of the sea's being grey. When you look at the same part of the sea it is the greenness and not the greyness of the sea which explains and determines the character of your veridical experience of the sea's being grey. This seems an odd kind of selectivity, especially if we think of the greenness and the greyness of the sea as monadic or intrinsic properties of the sea, i.e. properties whose holding of the sea have nothing to do with how things other than the sea are. For, if the greyness and the greenness are in this way intrinsic properties and are each constituted by the light-dispositional properties of the surface then whatever the greyness is causally responsible for, so also is the greenness, and vice versa. After all, the causal responsibility of the greenness in generating an appearance simply amounts to its constituting property, viz. the light disposition, being causally responsible by way of some material process for generating the neural basis of the appearance.

The problem is this: a consequence of the forgoing is that for all those perceivers on the one side of the inlet who see the volume of water in question as green, the greyness as well as the greenness of the volume is causally relevant to their having the experience they do, and yet the greyness does nothing to determine the character of their experience. Mutatis mutandis for the perceivers on the other side of the inlet; the greenness of the volume of water in question is causally relevant to their having the experience they do, and yet does nothing to determine the character of their experience. The puzzle is then how, given that both are equally causally relevant, the greyness of the water or alternatively the greenness of the water can ever selectively determine the character of a perceiver's experience and so be manifest properties. Call this the Selection Problem, a problem which arises immediately from (i) the facts of color relativity and (ii) supposing that colors are monadic properties constituted by the light dispositions (or micro- physical properties) of surfaces. Projectivism thus seems the more consistent position: there are no external colors to either determine the character of a perceiver's experience or to be causally relevant in the production of that experience.

To save the very idea of the colors as manifest properties of external things in the face of the problem of color relativity, we need a third position, distinct from Projectivism and the view that colors are intrinsic qualities of objects.


Relational Qualities

The colors, and this is so also for the sensible qualities in general, are relational qualities, i.e., qualities whose holding of surfaces, volumes and light surfaces depends in part on how other things, viz., perceivers and viewing conditions are. A dispositionalist reconstrues these relational qualities as dispositions to produce experiences of qualities, in the relevant perceivers under the relevant viewing conditions.

Nonetheless, it is possible to take relational qualities at face value. Consider Sabine, the first born of three children, Sabine, Venatius and Jules. Sabine is the only girl in the family. Her being the only girl depends upon how she is and on how her siblings are. It is a relation whose holding depends upon the qualities of its bearers. Likewise with Sabine being a different sex from Venatius. No dispositional account of these relational qualities seems plausible. <This is a bad example, I would very much appreciate a better one.>

Colors are relational qualities where the relata are surfaces or volumes or light sources, kinds of perceivers and kinds of viewing conditions. Visual experience does not encourage the immediate perceptual belief in the relational nature of the colors. At least this is so for the steady colors if not for the highlights or shimmering colors.

A basic phenomenological fact is that we see most of the colors of external things as "steady" features of those things, in the sense of features which do not alter as the light alters and as the observer changes position. (This is sometimes called "color constancy".) A course of experience as of the steady colors can easily be taken as a course of experience as of light-invariant and observer-invariant properties, properties simply made evident to appropriately placed perceivers by adequate lighting. Contrast the highlights: a course of experience as of the highlights reveals their relational nature. They change as the observer changes position relative to the light source. They darken markedly as the light source darkens. With sufficiently dim light they disappear while the ordinary colors remain. The highlights and those shimmering colors we see on credit card holograms and the backs of compact disks wear their light- and observer-dependent natures on their face.

The other side of color constancy or the manifest stability of the steady colors is the fact that once we leave the normal range of lighting or consider the visual systems of other animals, we must countenance the same variation and instability in the steady colors as we observe in the shimmering colors. It is not that the steady colors are monadic while the shimmering colors are relations. It is rather that a central function of the visual system, viz., to track ordinarily stable and salient surface properties for the purpose of identifying and re-identifying objects, is achieved only if most surface colors are stable over normal conditions. Hence color constancy and the collateral fact that a course of experience as of the colors does not reveal their relational nature.

By treating colors as relational qualities we can accommodate the obvious facts of variation in our experience of even the steady colors. For example, there are red patches of color on the pages of many magazines. It is utterly implausible to treat every such case as an illusory appearance of red, even though red is not one of the four colors (magenta, cyan, black and yellow). It would be equally strange to deny what the closest viewing of these red patches reveals: that these patches are made up of small magenta and yellow dots, that therefore these patches are motley in color. How can the same patch be red and motley yellow and magenta? Relationalism to the rescue: the patches are red for the naked human eye from a normal reading distance and motley yellow and magenta to the closest view.

Nevertheless, a color Relationalist can consistently find some truth in many remarks about "real" colors. Chromatic lights are said to obscure the real colors of patches viewed under them. The Relationalist avoids one kind of invidious distinction between a cloth's looking pinkish-blue in daylight and the same cloth's looking simply pink under pink light. For the Relationalist, both are equally veridical colors. But the second color is, as things ordinarily go, the color associated with the more transient and interrupted appearance of the cloth. If we mean by "real color" the least transient veridical color then daylight and ordinary indoor light typically allow the "real" colors of things to shine forth.

Nor need the Relationalist count every appearance in respect of color a veridical appearance. To illustrate the point, consider the perfectly consistent fancy of a surface which is red but which also has another surface property which acts via electromagnetic radiation on the visual cortex to produce the stable appearance of a green surface before one. Even though the surface looks green under the condition of having the strange electromagnetic effect emanate from it the Relationalist need not count this appearance as veridical. The immediate perceptual belief it would naturally prompt, viz., that the surface is green, is a false belief. For after all the material process which mediates the appearance of a green surface, the operation of the electromagnetic field directly on the visual cortex, is widely non-standard, bypassing as it does both light and the eyes. While the Relationalist can count many superficially conflicting appearances as veridical, some ways of producing appearances will make only for color illusions.


The Selection Problem Revisited

How does the claim that colors are relational qualities help with the Selection Problem? Recall the case of the sea seen from different sides of the inlet. The same body of water appears grey to those on one side and green to those on the other. If we suppose that green and grey are monadic or intrinsic properties of the water constituted by the water's light dispositions, and also accept, what anyway seems plausible, that these colors get to be causally relevant in the generation of experience due to the fact that the light disposition which constitutes them is causally relevant in the generation of the neural process which constitutes the experience then we face an odd consequence. When those on the one side of he inlet see the water as green, the greenness of the water is causally relevant in generating the experience, but so is the greyness of the water, since after all the light disposition of the sea constitutes both properties. Yet, and this the problem, only the greenness of the sea determines the character of the experience in question, i.e. the sea appears green to them and not grey.

It will be obvious to the reader that the assumption that is driving the problem is the assumption that green and grey, because they are monadic properties of the water, are each wholly constituted by the light disposition of the water, so that the one color is causally relevant in the generation of an experience only if the other color is. Once we think of colors as relational qualities then we have a way of undermining this assumption. The relational qualities which are the greyness and the greenness of the sea are not wholly constituted by the light disposition of the water. There are two other sorts of properties which constitute these relational qualities. These are (i) the properties which determine the nature of the visual system of the kind of perceiver the water is green or grey for, and (ii) the properties which determine the viewing condition under which the sea is grey or green. Clearly, it is the second class of properties which differ in the case at hand.

There is now room for the following position. When those on one side of the inlet experience the sea as green the causally relevant properties are (i) the light disposition of the water, (ii) the properties which determine the nature of their visual systems and (iii) the properties which determine the nature of their viewing condition. It is these three sorts of properties which together constitute the greenness of the water for them from that viewing condition. Mutatis mutandis for the greyness of the water for the others from their viewing condition. If as we are supposing the water looks different to the two groups because of the difference in their viewing conditions then the greyness of the water and the greenness of the water will be differently constituted properties, differing precisely when it comes to (iii) above. So it will not follow that the greenness is causally relevant only if the greyness is. Moreover the greenness of the water will count as causally relevant when and only when it determines the character of the experience; likewise for the greyness. The Selection Principle is dissolved once we treat the colors as relational qualities, constituted not only by the light dispositions of the colored things but also by the underlying physical properties which fix the nature of the viewing condition and the nature of the relevant kind of visual system.

Hylomorphism is the correct view about the colors, it is just that because the colors are relational qualities the constituting bases for the colors are more complex than we earlier allowed.


The Mixing Problem

Once the facts of color relativity are assimilated, there is a second problem which besets the view that colors are intrinsic properties of surfaces and favors Relationalism instead. Return to the case of the sea water which is both grey and green. Typically when things are both grey and green they are greyish green or greenish grey, but this it seems need not be true for the sea. Or take the case of pointillistic realization of red by way of motley yellow and magenta. Usually when things are both red and yellow they are orange. But there need be no orange on the magazine page which is both red and yellow in the same place. Why not? Why is there no "mixing" of these different colors had by the same surface?

The Relationalist has an easy answer. We get mixing and the resultant composite hues only when the colors in question are alike in respect of both the class of perceivers they are for and the conditions under which they appear. That is why the greyness and the greenness of the sea do not mix and why the redness and the yellowness on the magazine page do not mix. Here is one more reason to treat the colors as relational qualities.

Treating the colors as relational qualities still allows for a Direct Realism concerning color. When we experience the colors of things those very properties, not their mental or conceptual surrogates, are elements in the appearing of the things to us. Perhaps the resultant position does not deserve the name of Naive Realism, since naive perceivers need have little if any inkling that the colors appearing to them are relations rather than monadic properties of surfaces.

So we arrive at a Direct Realism which is not Naive: the colors are manifest qualities of surfaces, but they are relational qualities. The same result applies for the same reasons to each class of so- called Secondary Qualities; the smells, the felt textures, the tastes and the sounds. Things come out slightly differently with the Primary Qualities but not in a way that makes for any deep distinction.


Is There A Contrast With Shape?

There are likewise relational shapes, e.g., being rhomboid from here, being oval from there. But a natural view is that there are also non-relational shapes which along with the principles of perspective explain things having the relational shapes they do. So the very same tile can appear rhomboid from here and simply square, while a modernist tabletop can appear rhomboid from there and simply rhomboid. (Imagine being there involves looking down on the rhomboid table-top.) The relational shape is not a concept nor a mode of presentation of the non-relational shape. Nor is it a feature of a sense datum. It is just another property of the object manifest in experience.

Berkeley, the reader will remember, has a very respectable argument that seems to force relativity on all shapes even if we are convinced that something's being rhomboid from here can often be explained by the principles of perspective and its being square. Berkeley has us imagine mites which he supposes to see those small scale shapes which are relevant to their negotiating their world, shapes of things which we do not see and which are at least superficially at odds with the shapes we do see. Again, the structure of the argument from relativity is here the same as in the case of color; we imagine blamelessly different and in their own way equally able perceivers which experience different manifest properties from us, properties which are somehow at odds with the properties we perceive. Still the idea of the real non-relational shapes seems to allow for a criterion of getting it right with respect to shape.

Some suppose they should resist the relativization of shape by siding with the mites. The mites after all can make out shapes which we cannot discern, they are better perceivers of local shape than we are. They see what we need a microscope to see, the mountains and crevasses which make up the surfaces we ordinarily take to be flat. That at least is the thought which must be deployed if we are to resist the relativization of shape. It is unclear whether that thought can do the work it is deployed to do. For we can imagine that Berkeley's mites have littler mites "on em and so on ad infinitum." Well, if not ad infinitum, at least to the point where things are so small scale that the very idea of the physical shapes of the things in question begins to give way to the idea of the shapes they would be properly modeled as having in an appropriate physical theory, strict quantitative shapes which nothing material could be seen or felt to have. Before we reach that point there is presumably a range of discerning mites who are very plausibly counted as blamelessly different in what they see with respect to shape. The need to relativize will emerge here at the level of the most detailed and small scale of the perceptible shapes. So why resist relativisation as between us and Berkeley's troublesome mites?


Coeval Reciprocity

One feature of relational qualities is that they require the existence of their relata and of what they supervene upon. So Sabine's being a different sex from Venatius requires both the existence and sexual differentiation of Sabine and Venatius. The qualitaive relation, understood as that particular sexual difference which holds between Sabine and Venatius thus came into existence only when Sabine's younger brother came upon the scene. The corresponding point about the relational qualities which are the colors and the shapes may initially come as something of a shock. After a while, it simply seems a confirmation of what we have been calling the Ecological View.

The colors for us and the shapes for us are part of a manifest world for us which emerges with us as we and our visual systems evolve: the realm of color for us is genuinely for us, our aesthetic, practical, and ethical business is normatively to be transacted within such a world. The colors for the birds are genuinely "for the birds." In being relational the colors are not thereby "subjective" in the sense of being relations to mental states. Rather they exhibit the coeval reciprocity that is typical of a kind of animal and its environment. The two make a whole which evolves in respect of which properties are manifestly there for the animal in question.




1. "The Rediscovery of Light" Journal Of Philosophy XCIII 1996.

2. C.L. Hardin Color For Philosophers (Hackett Press, 1985) p.66.

3. Or at least it would be that discovery if the similarity in question was plausibly identified with a similarity in respect of hue.

4. Suppose that I am a waif brought up in a monastery. As a matter of strict monastic rule no married male is allowed to enter the gates of the monastery. I have a highly predictive stereotype associated with the concept of a bachelor inhabiting the monastery, viz. that of a male inhabiting the monastery. Now as a result of monkish gossip it is widely but wrongly suspected that Brother Bernard, who we all know to have been a man of the world, has a wife or two in Tuscany. I know of the gossip and accept it, but because of my stereotype of a bachelor as a male inhabiting the monastery, I adhere to my true and well-grounded belief that Brother Bernard is a bachelor. In rhetoric class I am finally taught the definition of "bachelor" and I conclude that Brother Bernard is not a bachelor. As a result of conceptual lucidity I know less by knowing more.

Nonetheless the crucial point is that even in such cases being conceptually lucid does not confer a global disadvantage when it comes to justified applications of the concept in question. When it comes to applying the concept bachelor outside the monastery, where there are indeed married males, I am not disadvantaged as a result of what I learned in the rhetoric class. Nor is this so within the monastery when it comes to the bachelorhood of the other monks beside Bernard. And there are many possible cases in which thanks to my newly acquired conceptual lucidity I would be better placed when it comes to having justified belief about who the bachelors are. The thing that it is hard to believe about our color sophisticate is that in every actual and possible case he is worse off than we as a result of his conceptual sophistication, so that he never could be justified simply on the basis of perception in judging the colors of things, while we are almost always so justified. This is not made any easier to believe by examining cases in which one ignores a misleading defeater thanks to a conceptual mistake.

5. A Treatise of Human Nature Book 1, part iv, "Of The Modern Philosophy"