Remarks on Colour, Ludwig Wittgenstein (365 Books, Day 238)

[rushthatspeaks]

I had to read this book after having read Derek Jarman's Chroma. Chroma is not about the theory of color, but is a set of anecdotes, associations, uses, properties, and appearances of colors-- and having read this I now see why Chroma is structured the way it is. The Wittgenstein is about the logic behind the concept of color, and the idea that Wittgenstein evolves is that color is, except in very narrow scientific circumstances, entirely contextual.

So Chroma is Jarman going through and giving you every little bit of context he can think of about various colors, so that maybe you can see something of what he sees in one. Fair enough.

As for why color is entirely contextual.

This is complex, and it doesn't help that the book is in draft, since as with most Wittgenstein it was posthumously assembled from things picked up off the floor of his room (I mean this literally). So it is a repetitive and unstructured read, with a lot of things described by the editor as having been crossed out thrown in for the sake of completeness, and strange punctuation, and so on.

But as nearly as I can make it out.

There is an obvious difference conceptually between the idea of the color 'gold' and the idea of the color 'yellow', yes? Gold shines or glitters, has a patina. Wittgenstein proceeds to demonstrate that there is a similar kind of conceptual difference between the idea of 'white' and the idea of other color words such as 'yellow'. How is this provable?

By reference to the concept of transparency. You can think of a yellow transparency, like a pane of amber glass. You can think of a green or red transparency.

Now think of a white transparency.

Now that you've finished stubbing your mental toe, you can see that white is conceptually different from yellow. There is some difference in how we think about white and spatial depth from the way we think about yellow and spatial depth. What's the difference?

Well, what's the exact description of the difference between gold and yellow? Without reference to science. I mean the conceptual difference.

I don't know and Ludwig doesn't either. Just, there is one.

Also, how do you tell if something is white in the first place? In one light, a thing may appear white. In another, gray. In another, light pink. Which appearance is correct? There are even different colors of white, because a thing that is entirely white can have tints and highlights. So white fades into other colors. It's debatable when something is 'really' white. And if it's white, you use this one conceptual thing, which I just explained that we don't understand, and if it isn't, you don't. So how do you know whether you are invoking in your mind the concept of white, and in what sense?

There is no actual logic behind the use of color-words, is what Ludwig is trying to tell us here. There is only a set of agreed-upon contextual definitions, and we do not know what lies behind those. What we mean by color-blindness is a physically based inability to learn to apply the contextual color definitions that other people use-- which is why a color-blind actor can act the part of a color-sighted person with utter convincingness, because the color contexts are provided by the script when they are plot-relevant.

As a result, Ludwig suggests that we could create systems of color harmony, as we have musical harmony, which are based entirely on arbitrary rules of context, things like 'only use x shade of red when there is no orange', and that the results would be perfectly nice-looking as long as the rules were internally consistent, no matter what the rules might be.

He also believes that color-sighted people ought to be able to imagine a set of color-concepts that bear the same relation to their own as the concepts that color-sighted people understand bear to color-blind people. You can get a color-blind person to believe that you can reliably tell the difference between a red and a green apple by sight, even if they can't do it. So color-sighted people should be able to come up with color-ideas of the same kind, things that other people could conceivably distinguish. I think thinking about it this way helps me a lot when I think about other cultures' color-words, things like the Japanese aoi or Welsh glas.

Ludwig is also despairing about the human race's general illogicality, irrationality, and inability to come up with anything sensible, but then, when isn't he. And the images are lovely-- some of the things that he says when trying to come up with a white transparency or a grey flame make me want to learn to paint. (Why are all color theorists wrong about grey. Seriously. Wittgenstein thinks grey can't be luminous. What is this I don't even. But the rest of it all seems sound.)

So, a dense but rewarding little book, as expected, and well worth putting up with the fact that not only do you have to read each paragraph sixteen times, several of the paragraphs turn up about that often in the draft. Ah well. If somebody ever assembles a book this good off my bedroom floor, I should only be so lucky.

Comments

[rax]

...damn/thank you, now color theory is on my summer reading list. *sighs, buys more books*

[q10.livejournal.com]

my familiarity with color studies comes mostly from a different perspective, but i'm pretty sure that the aoi/glas analogy is different from what's being gotten at with the color-blindness point, or at least is different from the case of color-blindness. i'm going to switch to a slightly simpler Russian/English example, though. the standard claim (i haven't checked with a Russian speaker) is that English word blue covers roughly the union of Russian синий and голубой. crucially, though, an English speaker with normal color vision, when presented with a синий sample and a голубой sample in the same light, can always tell that they are not identically colored. (or, put another way, if you have two otherwise identical objects one of which is painted синий and one of which is painted голубой, an English speaker with normal color vision can notice when you swap one out for the other, even if their vocabulary may limit their ability to describe what's changed.)

this is different from a case of genuine color-blindness which is characterized by the inability to make perceptual color distinction that other people can (controlling for context). that is, there are people for whom, if you pick the right red and the right green you can swap out the green-painted object for the red-painted object and if their head is turned they are incapable of figuring out that anything has changed. this is pretty different from the case of varying color vocabulary across languages. if you have normal color vision, it's easy enough to imagine being on the other end of this (in fact with enough work and a willingness to find just the right ‘color blind’ people, you can reverse things in the real world). imagine somebody approaches you with two identical cans of paint, labeled ‘A’ and ‘B’. to you, both appear to be totally identical yellows. this person however insists that they're completely different colors, and promises to be able to tell you, by visual inspection, whether a given painted object was painted out of can A or can B. you paint some things, and you can't for the life of you tell the difference, but this person, by visual inspection alone, consistently can. any test you can devise, they can always tell the difference (as long as the light is good). when you look at things under the same light that they do, you see no difference whatsoever.

there's no principled reason (either from a metaphysical point of view or from a point of view of what we know about the physics of color and color perception) why any of us couldn't in principle find ourselves in this situation. and this is the situation that's directly analogous to the situation of the color-blind person who can't distinguish the green and red objects. but it's not analogous to the cross-linguistic case, because you can reliably tell the blue paint labeled ‘синий’ from the blue paint labeled ‘голубой’ (in the sense that if somebody paints objects with the two, you can, assuming good light, say which can each object was painted from).

[asakiyume.livejournal.com]

I've really, really enjoyed yours and sovay's posts about this book. At that remove, at least, I've already enjoyed the book very much. The ideas he brings up are stimulating to think about.

And yes, what is it about grey?!

[redbird]

I think part of the problem with gray is that people don't look at the sky very often, or think about what they see.

Ages ago, when I was spending a lot of time on rec.arts.sf.fandom, one of the ways of saying that someone else was being profoundly unrealistic and wrong was "What color is the sky on your planet?" And I got to thinking, well, what color is the sky on my planet? Blue? Gray? Black? Look at it a while, and the answer to "What color is the sky on your planet?" is "All of them." Not all the colors that can be; rather, our eyesight evolved on this planet, so all the colors it sees are colors that are here. (Not all the colors that are here: there are colors bees see that we are blind to.)

[helivoy.livejournal.com]

There's a very interesting chapter on color in Guy Deutcher's Looking Through the Language Glass. Among other things, it finally explains Homer's "wine-dark sea" which I've wondered about ever since I read the epics.