The Essence (?) of Color, According to Wittgenstein
The Essence (?) of Color, According to Wittgenstein


Wittgenstein’s view on the problem of what “essences” are can be demonstrated on the example of colors. His interest shifts from the question whether ascriptions of color are elementary to the description of the whole variety of the use of color names and ascriptions, via considerations about regular (phenomenological, that is: grammatical) laws of what is possible for (i.e. meaningful to say about) colors. The essence of color, if any, the only universal standing for the variety of color phenomena lies in the unity of the word (in this sense, Wittgenstein is a sort of nominalist), but the nature of this unity cannot and is not to be further explained.

Table of contents

    Wittgenstein’s treatise of the topic of colors can be seen as an interesting development of the view on the nature or essence of color (colors), but such development that ends with a considerable weakening (not to say deconstruction) of the conception of any essence.

    Wittgenstein was attracted to the question of colors in Tractatus (Wittgenstein 1993) where he deals the first time with the color exclusion problem. His conception of elementary propositions is such that any elementary proposition is true or false independently on any other elementary proposition (or all of them). This independence can be seen from the fact that a conjunction of two elementary propositions can be neither tautology, nor contradiction. This is not the case of the conjunction of two ascriptions of color to the same point in space and time (to say of some point that it is green and it is red, is a contradiction – see 6.3751). Hence ascriptions of color seem not to be elementary. Does it mean that the essence of color is to be found somewhere deeper that in what shows to us as “color”? Wittgenstein provides no clear answer. What is confusing here is the fact that color ascriptions serve to many empiricist philosophers (including Vienna Circle (but, presumably, not including Wittgenstein)) as notorious examples of a primitive observation.

    This problem becomes clearer and more insistent in the later texts, beginning with “Some Remarks on Logical Form” (1929). Wittgenstein discusses here possible ways of the analysis of color ascriptions. What is it that is ascribed when we say that something is “red”? The concept “red” seems to be not primitive, reducible. Where can one find the elementary propositions constituting the allegedly complex color ascription? Wittgenstein proposes an analysis into mathematized elements – that in a color ascription we ascribe n (certain number of) elements (quantities) of color (so that what we usually call “color” is a complex of such elements). However, there is a problem: since in mathematics any n includes also n-1, and n-2, then when we say (as an “elementary” proposition) that something possesses n elements (quantities) of “red”, it implies that it possesses also any lower number of these elements (and so all the lighter (or darker?) tones of the ascribed “color”). Which is counterintuitive – the essence of color thus cannot be analyzed this way, going under the surface of what we see as “color”. In this sense, and in opposition to what Wittgenstein says in Tractatus, color ascriptions are elementary. But on the other hand, there is the problem with their interdependence (any ascription of color excludes ascriptions of any other color). It seems that there are some types of elementary propositions that are interdependent. The logical form of our language is thus not uniform, it must respect the diversified shape of worldly phenomena.

    This quite strong phenomenological sketch (that the structure of phenomena influences and grounds the logical form of language) is quickly revoked in Philososophical Remarks (Wittgenstein 1964). But not so that language, previously seen as “realistically” based on worldly phenomena becomes now “arbitrary” (this is what Austin (1980) suggests). That language cannot be straightforwardly compared with the world, doesn’t mean that it doesn’t or needn’t respect its conditions (the world is still an environment whose claims and needs must language cope with, though it cannot be treated independently on language – compare Lance 1998 and his conception of language as a sport). Phenomenology now becomes identical with grammar. That is to say: the regular structure of the possibilities of experience (phenomenology) cannot be distinguished from the regular structure of what can be meaningfully said (grammar). How does this concern colors and their essence, if any? As for their essence, nothing changed much. Colors are still primary, elementary, irreducible, and their ascriptions are still interdependent (exclude each other). What is substantial for colors (for what “colors” are), the constitutive, normative relations among them, in this sense their “essence”, can be demonstrated by means of certain schemes.

    Wittgenstein introduces here the scheme of color-octagon, or two octagonal pyramids joint in their bases. The points of the octagon are red, violet, blue, blue-green, green, yellow-green, yellow and orange, the vertices of the pyramids are black and white. This scheme encloses the phenomenology, i.e. grammar of colors. It is normative, since the relations between concepts of colors (the laws of experience) are not liable to a subjective licence. Of course, the shape of the particular language is contingent, but for its respective speakers it is a priori. A contingent a priori (see Rorty 1991), pragmatically well-functioning.

    A bit later The Big Typescript (Wittgenstein 2000) Wittgenstein makes the scheme a little more complicated. He tries to distinguish between so called basic (primary) colors – red, blue, green, yellow, and the other four, that are “mixed” colors. The octagon (or the double pyramid) is replaced by the color-circle, where the basic colors are fundamental (within their continuum the “pure” color is identifiable as a point), whereas mixed colors are not identifiable as points and represent only a continuum. Wittgenstein is led to this distinction by the different status of color mixtures. As he shows, the mixture of red and yellow is not a mixture in the same sense, as the one of violet and orange. The latter one just doesn’t produce the color which stands in the circle between the constituents (i.e. red). That is to say: all colors are not of the same kind (or the relations among colors are not always the same or symmetric). What is even more disquieting is Wittgenstein’s consideration about the exclusive ascriptions: of course, to say that something is red and that it is green doesn’t make sense (in a sense), but an average speaker needn’t necessarily feel it this way. What is decisive for the conclusion whether something makes sense or not, is whether any speaker can (feel that she/he can) use the “sentence” meaningfully in some situation. If she/he can, then philosophy cannot forbid it to her/him. It is linguistic practice, not philosophical generalization that decides what does make sense and what doesn’t. The essence of color is expressed in grammar, i.e. meaningful use, and even if it includes that ascriptions of colors exclude each other, it doesn’t mean that anyone cannot make the exclusive conjunction meaningful. In this sense, the essence of colors can seem “illogical” (in the usual, everyday sense of the word “logic”).

    This gap becomes much wider in The Brown Book. Whereas previously the four-polar color-circle was the ultimate authority; for example for the conclusion that any red-green combination doesn’t make sense, here Wittgenstein presents another perspective: If for example some social class (“patrician”) is characterized by red and green clothes, the combination red-green will be perfectly meaningful, in the sense of “patrician”. An analogous example is: if some culture doesn’t have a common name for our “blue” and calls dark blue “Oxford” and light blue “Cambridge”, these people’s answer to the question what Cambridge and Oxford have in common will be: Nothing (see Wittgenstein 2005, p. 134f). Of course, this sense of color combination is quite different from the problematic idea of one point in space time having two different colors, or the one of “reddish-green” color (which is such “in itself”, so to speak). Hence, the purpose of these counterexamples of “patrician” colors or the distinction Cambridge/Oxford is not to refute the older statements about the color exclusion. The notion of what are the constitutive relations among (i.e. phenomenology of possibilities of) colors, hence, what is the essence of colors, is only broadened this way. It is not easy just exclude anything from the essence of color (from what is meaningful to say about colors and relations among them, in whatever sense – all this belongs to their “essence”, as Wittgenstein sees).

    These examples, though fictitious, introduce relativistic questions: is it possible that various people or rather various cultures have various systems of colors? And can we decide which system is “true”? For now, Wittgenstein answers nothing. Later, he will admit the possibility, but with certain (to so speak Davidsonian) limitations; but the decision, if any, will have to be done otherwise than by a straightforward comparison of the color concepts with colors “in reality”.

    Wittgenstein then had left the topic of colors for more than ten years, and returned to it in Remarks on Colors (Wittgenstein 1992), his response to Goethe’s Farbenlehre which incited his great interest. The main purpose of Goethe’s analysis of colors is to provide a criticism and alternative of Newton’s optical experiments. For Goethe, the nature of colors in general cannot be conceived by one optical experiment, unjustly generalized. White doesn’t consist of all the rainbow-colors, except of the context of light fraction. A color-theorist, claims Goethe, must respect the variety of color laws and relations among them, which differ from context to context. If there is any medium within which what is essential for colors is available, it is the medium of our experiencing (Erleben) – which includes the regular impact of colors and their combinations on the perceiver, as well as all the conventional (allegoric, symbolic etc.) constituents of the meaning of colors (Goethe 2003).

    Wittgenstein’s late return to colors, inspired by Goethe, proves his slight weakness with respect to the temptation of phenomenology (for the problematics of Wittgenstein’s “phenomenology” see Gier 1981 or Kienzler 1997). However, he is well aware of the disparate character of the “essence” of colors. Either “phenomenologically”, or “grammatically”, one cannot find a simple, unite “essence”.

    The central question he asks – and the central problem he sees – here is the one of the “sameness” of color. He discusses several problematic examples: 1) We call “red” both the autumn leaves and some red clothes – however, “in a sense”, it is not the same color. Actually, all the things we call “red” can seem quite different (and the difference is not only the one of light/dark). 2) One can paint both “white” things and “illumined grey” things (things usually conceived and seen as such), using the same palette color. 3) When one paints a dark room in the full light, how can she/he then compare the colors of the painture painted and seen in the full light with the colors of the room seen in the dark?

    All these examples show that it is not at all easy to state how can two things have the same color, how to compare it, and what does this “sameness” mean. The universal, unum versus alia, seems here to be nothing more than one word standing against all the disparate phenomena. But it would be a philosophical error to search for some one thing (in whatever sense of “thing”) hidden behind the one word (“craving for generality” – Wittgenstein 2005, p. 17ff). In this sense, Wittgenstein seems to be a kind of nominalist – the universal shared by all the particular things is a word, nomen. But there is no further analysis of what this universal word capturing the “essence” is. The universality of the word means nothing more and nothing less than the universality of use (just the fact we use the one word in all the different contexts). And that we know that something is red, cannot be further explained (the only possible explanation is that we have learned English – see Wittgenstein 1958, § 381).

    The relations among colors become still more diversified. In one context (optical) colors differ: some can be seen-through, and some cannot (white, black, brown); in another context (colors of a paper) all the colors are of the same sort. However, philosophy shouldn’t try to explain away these differences and reduce them on a simple essence and simple essential relations among colors, but on the contrary to try to conceive as many such differences as possible. The essence of colors lies in the meaning of the words for colors; there is no better (in fact no other) way how to conceive the “essence” of colors than by a description of this variety.

    As for the relativistic problems with alternative systems of colors, Wittgenstein introduces two types of anti-relativistic argument. One of them is so speak Davidsonian (cf. Davidson 1974): in order that we are able to state that something is a concept of color, though differing from our concepts and not quite understandable for us, it must be somehow akin to our concepts. We must always have some auxiliary evidence to discover whether something is a concept of color (a bit like the evidence of whether someone is a good tennis or chess player which doesn’t require that the author of the judgment is himself/herself a good tennis or chess player). After all, we have no better criterion for being a color than that it is one of our colors. The other argument is: if we are to decide between two different conceptions (lists) of the primary colors (one of them includes green among them, one of them considers it as a mixture of blue and yellow), we must look at which one of them works better in practice. I.e. which one of them enables us to fulfill more tasks (or more complicated tasks). Wittgenstein thinks, which is not without problems, that the conception of four basic colors is better in this sense. But whatever is the answer here, the only acceptable relativism is the relativism of systems that are akin and that function equally well in practice.


    1. Austin, James 1980 “Wittgenstein's Solutions to the Color Exclusion Problem”, Philosophy and Phenomenological Research 41, 142-149.
    2. Davidson, Donald 1974 “On the Very Idea of a Conceptual Scheme”, Proceedings and Addresses of the American Philosophical Association 47, 5-20.
    3. Gier, Nicholas F. 1981 Wittgenstein and Phenomenology, Albany.
    4. Goethe, J.W. 2003 Farbenlehre, Stuttgart.
    5. Kienzler, Wolfgang 1997 Wittgensteins Wende zu seiner Spätphilosophie: 1930-32, Frankfurt a.M.
    6. Lance, Mark 1998 “Some Reflections on the Sport of Language”, in James Tomberlin (ed.), Philosophical Perspectives, 12: Language, Mind, and Ontology, Oxford.
    7. Rorty, Richard 1991 Contingency, Irony and Solidarity, Cambridge.
    8. Wittgenstein, Ludwig 1929 “Some Remarks on Logical Form”, Proceedings of the Aristotelian Society, Suppl. vol. 9, 162-171.
    9. Wittgenstein, Ludwig 1964 Philosophische Bemerkungen, Oxford.
    10. Wittgenstein, Ludwig 1958 Philosophische Untersuchungen, Oxford.
    11. Wittgenstein, Ludwig 1992 Werkausgabe Band 8, Frankfurt a.M.
    12. Wittgenstein, Ludwig 1993 Tractatus logico-philosophicus, Praha.
    13. Wittgenstein, Ludwig 2000 The Big Typescript, Wien.
    14. Wittgenstein, Ludwig 2005 The Blue and Brown Books, Oxford.
    Ondřej Beran. Date: XML TEI markup by WAB (Rune J. Falch, Heinz W. Krüger, Alois Pichler, Deirdre C.P. Smith) 2011-13. Last change 18.12.2013.
    This page is made available under the Creative Commons General Public License "Attribution, Non-Commercial, Share-Alike", version 3.0 (CCPL BY-NC-SA)


    • There are currently no refbacks.