Remarks on Colour

I've had this book for so long (without ever having read it) that the price on it is $2.95. So it's an earlier edition than the one here.

Color had at one time been a kind of standard philosophical topic. The empiricists had deemed color a "secondary quality" of objects, something that arises and is dependent on our perceptual interaction with objects rather than a "primary" quality of objects themselves. Goethe's Theory of Color, in the context of advances in optical theory, had provoked the question of how physical explanations of color relate to the experience of color.

What especially interests Wittgenstein is what always interests him -- what is the status of our color concepts? Is the fact that white is not included among the show more primary colors, or that we never speak of "bluish yellow" reflective of facts in the world, or are they more like the logic of our linguistic behavior? If the latter, are they changeable? The Remarks on Color come from the same time as the remarks in On Certainty, just before Wittgenstein's death from prostate cancer. Those questions about the grounding of apparently empirical statements were very much on his mind.

One of the discussions reading the book stimulated for me goes against the idea that precision is a requirement for "understanding" or for competence, as if the application of concepts (like our color concepts) were governed by rules held up against our experience of reality. People are generally very competent at telling the colors of objects, but very few could give you definitions of colors or say precisely where one color ends and another begins. And even the definitions and boundaries that some could give would be questionable. Defining a color in terms of a wavelength of light, for example hardly captures what is meant by the color. Pointing to a spot along a spectrum and saying, "Here is where orange ends and yellow begins" will always seem declarative, not factual.

Surely some of this is what the empiricist philosophers were getting at by treating color as a "secondary quality" of objects . But it won't do just to say that, well, color is an imprecise thing because it isn't completely objective. Other qualities -- hardness, softness, slickness, . . . we operate confidently with all of these but we can't set boundaries to them or recite confident definitions.

All of this has to do with a kind of autonomy of speech, something overdone in the postmoderns, maybe, but valid nevertheless. The application of terms (or better maybe, the practice of language) isn't bound by objects and their properties in any simple way.

But then some (thank God, only some) will see all of this as a "problem" -- confusion, disorder, imprecision, lack of clarity. And they will propose to solve the problem with precise definitions and strict boundaries. Such things, they say, are needed for clarity and accuracy in our concepts. All that precision and clarity are really needed only in extraordinary circumstances, as when a painter color-matches paints. The extraordinary circumstances don't reveal a flaw in the ordinary circumstances.

Worse yet, others (like traditional AI researchers) will say that, whether we can articulate it or not, we do have an implicit, precise understanding of what colors are and where one color stops and another begins. It's as though it must be that way because they can only imagine it that way. Failures of the imagination are hardly the key to truth.

Wittgenstein also discusses color-blindness, and since I am color-blind, I always get caught up by the topic -- people who are curious about color-blindness often ask me "what do you see?" as if that were a question I could make sense of in a way that would make them understand something. Suppose I ask them "What do you see?" They say, "I see red." So do I . . . when I see red.

Here's one way to explain it, maybe. Suppose you are sitting with an artist, one with a great sensitivity to colors, and you have a bunch of red color samples in front of you. The artist sees many different reds, while you see red in each one. To you they just seem so close together as to be indistinguishable. To her, they seem different. Then imagine it's you and me sitting in front of the color samples, but this time they aren't all red. To you, they seem different. To me, they don't. It's not a matter of seeing something different from color-sighted people; it's a matter of not seeing distinctions that others see, just as you don't see distinctions that the artist sees. Don't know if that works, but . . . wth. show less

edição bilíngue alemão - português