Language, Color Perception, and Mental Maps
post by jsbennett86 · 2011-09-10T17:48:05.443Z · LW · GW · Legacy · 13 commentsContents
13 comments
I came across this video today, and it got me thinking about the role of language in drawing mental maps. For those who don't want to bother watching it, the gist of the video is that the words we use to describe colors may, in fact, alter the way in which we perceive color. In the video, some African tribesmen are shown a circle of colored squares and asked to choose the one different from the others. In the first test, the squares are all green (in fact, almost indistinguishable shades of green), and though westerners have a hard time figuring out the odd one, the Himba get it easily. However, they cannot tell the difference between a bunch of green blocks and a very different shade of blue.
PZ Myers points out that this phenomenon in the video may be caused by different neurological wiring between westerners and this tribe. If that is the case, it seems more likely that their language would evolve to represent our experience of color, rather than vice versa. Either way, I can't help but feel that this sort of thing makes it more difficult to tell when you are talking about maps and when you are talking about the territory.
Consider the Blue/Green question here (or Blue/Black if we're arguing to the Himba). We're not hidden in a cave, unable to look at the sky. We both look at it, at the same time, and we say it is one color and the Himba say it is another. Who is right? We can try to dissolve this question by saying that, in fact, the sky is neither blue nor black and in fact only appears to us as one or the other because of how our brains model the sensory input from our eyes. But understanding this would not help the Himba see the blue square among the green ones any easier (nor would we westerners be able to see the odd green one among the greens). If this is a question of how our brains are wired, it suggests that we may have other differences in how we perceive the world (though my lack of imagination prevents me from imagining any right now). Can we overcome such things? Is it useful to try? Or would it just lead us to frivolous debates with people who literally cannot see things the same way as we do?
If, however, our brains perceive things in certain ways because of the words we use to describe them, this suggests we can better model reality by changing our language. By speaking in terms of biases and utility functions, we tend to do that here on LW. Can this be easily translated offline, into day-to-day interactions? If so, how difficult is it to change thinking patterns (or whole brain structures) by changing language? And how do you know if you are modeling reality more accurately (I can't see that making your beliefs pay rent helps any in the debate between blue and black skies)? How can you get others to understand what you're saying, and have it change their perceptions?
Or perhaps I'm just confused. Thoughts?
13 comments
Comments sorted by top scores.
comment by orthonormal · 2011-09-10T20:40:16.655Z · LW(p) · GW(p)
Differences in perception don't have to mean fundamental differences in anticipation. A good Himba scientist and a good Western scientist would expect the same distribution of wavelengths in light coming from the sky.
Replies from: jsbennett86↑ comment by jsbennett86 · 2011-09-10T21:00:22.476Z · LW(p) · GW(p)
Good point, and I didn't think of that when I said I couldn't see making your beliefs pay rent was of any use here. Of course, a Himba scientist and a Western scientist might still say, "We know the wavelength of the light diffracting off the sky. But is it blue or black?" This may just be a result of how an algorithm feels from inside.
comment by Armok_GoB · 2011-09-10T20:03:10.641Z · LW(p) · GW(p)
Ok, you get me thinking for minutes trying to figure out how many nanometers the peak of the frequency was and what RGB value and HSV value my mental example of "sky" uses. Then I watched the video, dead sure I'd be able to spot both differences easily, being surprised and frustrated when i couldn't, chose one and going all hindsight bias making me sure it was the right one, get surprised and confused when i were wrong, notice I were confused, and right now I'm damn close to actually grabbing the RGB value from the images and making some rationalization about how the RGB standard and monitors were designed by westerners and thus it's only indirectly I've gotten western colours. Another part of me is staring to wonder if I'm just hallucinating not seeing any bands in the rainbow to preserve my self image.
Up voted for making me think, and giving me some well needed actual practice of rationality.
... any help sorting this out? I'm kinda stuck.
Replies from: jsbennett86, saturn↑ comment by jsbennett86 · 2011-09-10T20:44:46.274Z · LW(p) · GW(p)
I was sure that I had picked out the different square, only to find I was wrong. Looking back, I can't see any difference, really (and I suspect the original one I saw was due to a shadow on the screen). Nevertheless, the scientists do say that there is a difference between the square the Himba picked and the others. Some people can see weird things, like the polarization of light, so it's not a stretch to imagine they are more sensitive to different aspects of light than we are, and less sensitive to things like hue. I really wish the clip explained what the difference they were seeing was.
If the Himba were to design a color wheel, I wonder what it would look like.
comment by DanielLC · 2011-09-10T19:27:06.186Z · LW(p) · GW(p)
People tend to group things together. Thanks to using words a lot, we tend to group things by the words we use to describe them.
If we change how we refer to biases, this will change what we consider to be the same bias or different biases. It will not help us notice them more. Talking about them more will, but that's a different effect.
Replies from: jsbennett86↑ comment by jsbennett86 · 2011-09-10T20:19:50.346Z · LW(p) · GW(p)
True. New words help to highlight new distinctions, but you have to have a distinction for the word to apply to before it is useful. Otherwise, it either refers to nothing at all or the same distinction that another word illustrates.
Asking what color the sky is may show cultural differences (for example, many many cultures use the same words for blue and green, though they're perfectly capable of pointing out the difference between the two), but the demonstration with the colored squares suggests something different is going on with the Himba.
So the Himba can make some distinctions that we cannot (as opposed to do not), and vice versa. We each have words to describe those distinctions (and I'm sure the Himba could explain why, with their own vocabulary, why that one green square is different from the other ones that look exactly the same to us). The more I think about it, the more likely it seems that brain patterns change the language than vice versa.
Why did the scientists in this program seem to think it was the other way around, I wonder?
Replies from: DanielLC↑ comment by DanielLC · 2011-09-11T00:52:52.372Z · LW(p) · GW(p)
but you have to have a distinction for the word to apply to before it is useful
All colors are distinct. It's a lot easier to just put them into several groups. This has been shown to be how people work. The idea that they do this with colors before they compare ones that they're currently looking at is surprising, but it's not impossible. A large portion of your brain is used for understanding what you see, so there's a lot of room for those shortcuts to take place, and a lot of benefit to having them.
Why did the scientists in this program seem to think it was the other way around, I wonder?
I suspect it's because of the other studies that they haven't mentioned. Perhaps they looked at tribes genetically identical to the Himba, and noticed that they notice colors based on their language. Perhaps this is just the most noticeable one, and other studies were with genetically identical groups. Perhaps they just know a lot more about how the brain works in general. I suggest reading Conjunction Controversy (Or, How They Nail It Down).
comment by XiXiDu · 2011-09-10T18:51:53.603Z · LW(p) · GW(p)
The name of the color of the sky is an agreed upon label that aids communication. The inability to discern a blue sky from a night sky, or a green square from a blue square, is simply a problem with either your sense organs or the processing capabilities of your brain. The map–territory relation is a metaphor used to highlight the fact that reality is that which, when you stop believing in it, doesn't go away.
Replies from: jsbennett86↑ comment by jsbennett86 · 2011-09-10T20:25:57.261Z · LW(p) · GW(p)
While I was writing this post, I changed where I was going with the map/territory thing, and that led to me confusing myself. I'm not sure what I was thinking originally, other than that it seemed quite significant at the time.
comment by atucker · 2011-09-10T18:05:30.713Z · LW(p) · GW(p)
We can try to dissolve this question by saying that, in fact, the sky is neither blue nor black and in fact only appears to us as one or the other because of how our brains model the sensory input from our eyes.
That seems pretty much right to me.
I think that language basically changes the resolution with which you can distinguish between different shades of color. Humans are able to detect differences in color which they aren't able to store and recall offhand. The colors have different wavelength(s?) of light, which human language and perception roughly, but not completely, maps to.
If, however, our brains perceive things in certain ways because of the words we use to describe them, this suggests we can better model reality by changing our language.
This is also probably true. NVC is supposed to make you more empathetic and respectful by changing the way that you talk about things, and as far as I can tell, non-negligibly changes my emotion-related thought processes. Specifically, I now try to focus more on needs and getting those fulfilled, rather than directly reacting to what people say.