Colour Constancy and the Mind-Independence of Colour

According to the naïve realist, colours are mind-independent properties of objects that are distinct from their physical properties. In today’s post I outline the argument for the first part of the view: the claim that colours are mind-independent.

To say that colours are mind-independent properties is to say that their nature and existence does not depend on the experiences (or psychological responses more generally) of conscious subjects. This means that colours are not, for instance, dispositions of objects to produce colour sensations in suitably placed perceivers, as they are according to the dispositionalist (or ‘secondary quality’) view famously defended by Locke and traditionally popular among philosophers.

The line of argument that I develop in support of the claim that colours are mind-independent appeals to colour constancy. Colour constancy is the phenomenon whereby the colours of objects are perceived to remain constant throughout variations in the conditions under which they are perceived, and so throughout variations in the way that they appear as the perceptual conditions vary. A simple way of illustrating the phenomenon is to turn on a desk lamp in an already illuminated room. Turning on the desk lamp will bring about a noticeable change to the appearance of the desk and the objects on it, but the colours of the objects illuminated by the lamp will not appear to change.

Colour constancy is incredibly important for perceivers. Colours allow perceivers to distinguish, identify, and reidentify objects. And they allow us to do this in perceptual conditions that vary constantly: at different times of day, under natural and artificial illuminants, depending on whether objects are directly illuminated or in shadow, and against different backgrounds. In order for colours to play this role in our lives, we need to be able to perceive objects’ constant colours—colours that do not vary with the conditions under which we perceive them.

Colours are not the only properties that exhibit perceptual constancy. Paradigmatically mind-independent properties like shape and size do, too: we don’t perceive tables to change shape as we move in relation to them, and cows in the distance are not perceived to be smaller than cows that are nearer. This, I suggest, is no coincidence. Properties that exhibit perceptual constancy, like shape, size, and colour, manifest a distinction between appearance and reality: between the way objects instantiating the property (merely) appear in particular conditions, and they way that they really are – and, when everything goes well, can be perceived to be. Exhibiting this kind of distinction between appearance and reality is the mark of something that is mind-independent.

The problem for dispositionalist theories of colour—a problem noted by, amongst others, Mark Johnston and Mark Kalderon—is to account for the perceived constancy of colours throughout variations in the perceptual conditions. Take, for instance, a form of dispositionalism which holds that objects produce colour sensations that vary constantly as the perceptual conditions vary, and which identifies the ‘real’ colours of things with dispositions to produce particular kinds of sensations, in particular perceivers, in particular conditions (for instance, ‘normal’ perceivers in ‘normal’ conditions). A basic difficulty with this kind of view is that it seems to predict that the colours of objects should appear to be constantly changing with changes in the perceptual conditions—which they do not. It is true that views of this kind can make some sense of the idea of the ‘real’ colours of things. But ‘real’ here carries no metaphysical weight: what is identified as the ‘real’ colour is just one disposition to produce colour sensations among many. And, if nothing else, it is difficult to understand how that can be perceived to remain constant throughout variations in the conditions.

Fully developing this argument against the dispositionalist requires extending the argument to more complex forms of dispositionalism, defending the claim that we literally perceive colours, shapes, and sizes to remain constant—rather than simply believing that they do on the basis of constantly varying experience—and explaining the way in which things nevertheless do appear different as the conditions in which we perceive them vary. The claim that colours are mind-independent properties also needs to be defended against various forms of the Argument from Perceptual Variation, which uses variations in colour experience across conditions and perceivers to undermine the claim that there are colours that objects ‘really are’ (in any robust sense). I consider these issues in Chapters 2 and 3 of A Naïve Realist Theory of Colour. In tomorrow’s post, however, I want to move on to consider the second of the claims made by naïve realists: that colours are distinct from the physical properties of objects.


  1. I can see the appeal of realist ontologies, but I don’t see the appeal of naïve realism. For example, when it comes to colour constancy, experience is much more variable than this naïve realist account allows. Colours *do* change or are perceived differently under some conditions.

    Experience *is* the interplay of world and mind. Excluding either aspect certainly does seem naïve.

    • Keith Allen

      Yes, you are quite right to stress that colour experience does vary with the conditions. It’s an interesting question exactly how to explain that. My preferred option appeals to apparent properties that vary with the illumination, although another option is simply to appeal our awareness of the environmental conditions in which colours are perceived. The thing that I do want to hold onto, though, is the thought that the colours themselves do not change. There is, I think, a difference between a change in the colour of an object (for instance, when you paint it) and a change in the way a coloured object appears (for instance, we you see it under different illumination conditions).

  2. Surely there is something odd here!

    If we accept that a painter ‘creates’ colours in his palette, just as the Paint utility on my computer allows me to ‘create’ any colour that I desire, then surely the ‘colour’ which is created has some physical property that I name as a particular colour irrespective of whether I am observing it at any moment or not.

    Moreover, on my computer each such ‘created’ colour even has a precise, objective, definition in terms the software code written by a programmer for the utility Paint.

    It is not, therefore, unreasonable to assume that, in principle, any of nature’s—assumed denumerable—possibilities of ‘colour’ can be so well-defined.

    How, then, could one deny that every fundamental particle appears—at the micro level—to have an associated property of ‘colour’ in this sense; and that that which our natural eyes (or our mechanical ‘eyes’) perceive (or measure) is the reflected result of the inter-actions of all such ‘colours’—of a composite of fundamental particles at the macro level—received in the presence of incident light.

    Whether or not such an associated ‘colour’ at the particle level is a constant property of the particle, or a derived property (such as velocity, position, etc.) that changes with time, is a matter that can only be postulated axiomatically either by an objectively consistent—coherent—theory of the particle’s properties, and verified, or conjectured upon, by unarguable objective evidence based on physical measurements.

    What seems incontrovertible from personal experience is that the colour of any object does change—albeit imperceptibly in most instances—over time.

    What then seems odd here is that it is not obvious whether the philosophical impulse that seeks a ‘theory of colour’ aims to:

    (i) investigate the nature of ‘colour’ as an objective property of a physical entity when measured by its visual impact upon a mechanical ‘eye’;

    or whether it aims to:

    (ii) investigate the neurological factors that determine how a living organism perceives individual—subjectively dependent and objectively undefinable—reactions to the visual impact of ‘colour’ upon an individual’s senses (whether directly through its ‘eyes’ or through direct stimulation of some corresponding regions in the brain).

    Surely the two are entirely different disciplines, each meriting independent investigation!

    • Keith Allen

      As I understand it, a philosophical theory of colour is (primarily) more concerned with i) than with ii). The philosopher is interested in the question of what colours are, and whether they exist. One of the ways of answering these questions is to think about whether there is anything in the objective world that corresponds (or corresponds sufficiently closely) to colours as we perceive them to be. My view is that there is, but these are not physical properties of objects. Question ii) sounds more like the kind of question that a psychologist would be interested in. So I agree that these are distinct questions, each of which merits investigation. (Although I would also add that an answer to ii) may bear on the answer i), in that understanding the psychological mechanisms underlying colour experience is sometimes thought to provide a reason for thinking that colours don’t really exist.)

  3. Joel Smith

    Hi Keith, I don’t really understand the argument against dispositionalism. Can you tell me more? As the context changes, the hypothesised sensations change. So if Red is the disposition to cause S sensations in C, and S* sensations in C*, etc. won’t it give the right result? The sensations change, whilst the perceived colour remains constant. Or are you thinking that the dispositionalist must think of colours as somehow looking dispositional, and that this is shown to be implausible by the phenomenon. Something like that might be going on in this remark (though I am perhaps misreading): “But ‘real’ here carries no metaphysical weight: what is identified as the ‘real’ colour is just one disposition to produce colour sensations among many. And, if nothing else, it is difficult to understand how that can be perceived to remain constant throughout variations in the conditions.”

    • Keith Allen

      It depends in part on whether a complex disposition like this can literally be perceived, and I’m sceptical about that. But I also think, more generally, that it is difficult to identify the hypothesised colour sensations, because I think that perceptual experience is transparent. The general structure of the argument is that the view that colours are mind-independent best explains the phenomena. But it always possible for the dispositionalist to present their view as a kind of error theory, if they think that it is well-motivated on independent grounds.

      • Joel Smith

        The first two points seem to me to be rather different arguments against dispositionalism, and maybe they’re compelling. But I’m still not seeing the force of the constancy argument. Why is constancy best explained by mind-independence as you characterise it? I can see that it’s best explained by the view that colours are perceived as properies of objects (rather than as features of visual experiences), but that’s something with which the dispositionalist agrees. I don’t see why dispositionalism of the sort I mentioned would count as an error theory at all. According to it, things are as they seem (although, it perhaps wouldn’t be the case, on such a view, that the nature of colour can be read off the phenomenology – but that’s a different claim, it seems to me).

        • Keith Allen

          The dispositionalist needs to avoid understanding the constant property as the ground of the different dispositions to produce different sensations in different conditions, as that then begins to sound like the view that colours are mind-independent. Equally, though, it may be that the form of dispositionalism you are suggesting is modified to allow for genuine perceptual constancy — in which case the problems with the view lie elsewhere.

  4. The argument from colour constancy shows that colours behave like mind-independent properties for a given individual, but it doesn’t seem to me to impact on the orthodox scientific position that the properties we see as colour are mind-independent, but the way we see these properties as colour is species- and individual-dependent.

    • Keith Allen

      Hi David, this is a good point and isn’t something I’ve been able to discuss here. And perhaps I should add that it is a thought that many philosophers are attracted to, too. What I try to do in Chapter 3 of the book is respond to this line of objection, and show it is possible to hold onto the mind-independence of colour at the same time as acknowledging that colour perception varies, to some degree between individuals but more dramatically between members of different species. Particularly in inter-species cases, I’m tempted towards the view that members of species perceive different (mind-independent) properties of objects.

  5. Raul Villuendas

    Colors are concepts that emerge from our observational “SYSTEM”.
    It is our visual and cognitive system that generates the concept of color. So as any physical complex system of information management it generates concepts that help interacting with objects and their inherent physical properties.

    Any complex systems like a convolutional neural network ( generates concepts like the “circolarity”, the “spirallity”, the eye, mouth, and color as well (redness, yellowness, etc.) in case of visual recognition neural networks.
    So concepts related to the visual recognition emerge within our complex visio-cognitive system when interacting with the object. And then they become available for other cognitive system in our brain to use them (like language).

    This explains for example that different visual systems like the one from bees generate a different concept of color as their cognitive system interact differently with the property of objects of reflecting the electromagnetic spectrum.

    This makes clear as well the ontology of qualia, demonstrating that qualitative properties of things do emerge from the auto-organization of complex systems, what is to say, they emerge from auto-organized-quantity.

    Work of Stuart Kauffman and recent books like “Auto-organizzazioni” by Alberto F De tOni, Luca Comello and Lorenzo Ioan can better illustrate the idea.

  6. I find it odd that you take colour constancy as evidence for the mind-independence of colour. On the contrary, I see it as evidence for an unconscious mental process which uses the relative colours of objects in the visual field to infer a partial sense of the ambient light spectrum and to “see” the colours of the various components in the light of that inferred spectrum (ambiguous interpretation of “in the light of” being intended). That this process is not uniquely defined is of course to be expected. And its mind-dependence would seem to be well illustrated by the varying robustly held responses of different minds to the famous striped dress conumdrum (see eg

    • Keith Allen

      Thanks for the comment Alan. I think it’s true that constancy isn’t perfect, and there can be situations in which we aren’t able to identify the colours of things. I also agree that we can be aware of the way that scenes are illuminated — for instance, we are often able to recognise shadows and regions that are illuminated by different light sources. But identifying the colours of things is important to us as perceivers, and I don’t think it is just relative colours that matter. We perceive things under conditions that vary constantly, and so the relative colours would be just too variable. The dress raises the issue that there are differences in colour experience between different subjects. Personally, I’m not convinced that some of these differences are as great as they are sometimes supposed to be. In the case of the dress, the robost variation concerns a representation of the dress, rather than the dress itself — I don’t think there was any disagreement about what colour it was when seen in normal conditions. But these issues do require careful consideration, as there is certainly an important challenge here.

  7. Bill Skaggs

    What makes this stuff difficult, I think, is that a color is actually two distinct things, one mind-dependent and the other mind-independent. On one hand, a color is a “quality”, which is an entity that resides in the mind/brain. On the other hand, a color is a physical property determined by the way the surface of an object interacts with incident light.

    What makes this confusing is that our brains have evolved such that the mind-dependent and mind-independent entities are very strongly correlated — this is what we mean when we speak of “color constancy”. But we will never get this stuff straight until we realize that we are dealing with two logically and philosophically quite distinct things — a quality and a physical property — that just happen to be very strongly correlated at an empirical level.

    • Keith Allen

      This is an interesting way of looking at it, but one of things I want to do in the book is see if it is possible to resist. For reasons I touched on at the end of the week, I’m not inclined towards thinking of colours as qualities that exist in the mind or the brain. For one thing, it doesn’t seem to fit well with reflection on our experience. Now it may be possible to explain this away, perhaps due to the strength of the correlation. But if you can explain it without appealing to some kind of error theory like that, then I think there’s something to be said for that.

  8. Andrew Werth

    I find some of these attempts at philosophical explanations of color to be rather maddening (though admit that it may just be my inability to follow the argument completely and keep the terms straight).

    Consider a “simple” object, such as a Post-It Note piece of paper that we would normally describe as “yellow.” This object has a reflectivity based upon it’s makeup and shape that determines what percentage of light is reflected at various points along the color spectrum. Depending upon the distribution of wavelengths of light incident on the object, the reflectivity of the light of the object, what other light is being reflected on surrounding areas, and the particular workings of the eye that receives such light (e.g., “normal” distribution of cones, no damage to visual cortex, etc.) a viewer perceives the paper as having a certain color. We can define colors based upon agreed upon standards (when such-and-such “standard” light bounces off an object that reflects light just so into a standard observer, we call this “yellow”). If you want to call the reflectivity curve it’s color, that’s fine, but that’s not the complete explanation for how the color is perceived.

    I find talk about whether colors “exist” to be problematic, in that “exist” is a poorly defined word for certain types of phenomena. Does “digestion” exist? (Can you show me a digestion?) Does a “home run” (in baseball) exist? (If so, can one pinpoint the precise time frame for which it exists?) I think certain events “happen” and we describe them as such but because they describe processes that occur over time they don’t have an “object”-like existence. This is how I think of consciousness and colors. Consciousness doesn’t exist in the sense of it being a material thing in the world; consciousness *happens* in a suitably structured brain under the right conditions. Similarly, colors “happen” when a brain is tweaked in just the right way (usually by light hitting the retina, etc., but also perhaps by stimulating just the right neurons by other means).

    When I look up in the sky and see a rainbow, what is the object whose “mind-independent properties […] are distinct from their physical properties”? (I would say that there is no object — I am receiving wavelengths of light separated by passing through droplets of colorless water.) When I look at the newspaper and see an area of color in a photograph that appears “orange” to me, is that the color of the newspaper? Upon closer inspection, it seems to be made up of dots of reds or magenta and dots of yellow. The color seems to me to be dependent upon not just the light and my retina but also how far away I am from the object. Isn’t the color I perceive mind-dependent, and specifically determined by the exact configuration of light hitting the newspaper, bouncing off at a certain angle, being integrated by my eye based upon my distance from the paper, etc.?

Comments are closed.

Back to Top