Let's step into the wayback machine and talk about some research that even the psychologists among us might not be aware of (I certainly wasn't). It seems that at the turn of the 20th century, many psychologists and psychophysicists (including the father of psychophysics, Gustave Fechner) were interested in aesthetics. Out of this interest came the idea, inspired by the work of Helmholtz and others, that geometrical figures had a certain "energy," which influenced the subjective quality of their combinations. For some reason, this idea reminded one psychologist, E. Bullough1, of what he believed to be a "decorative canon": "that dark colours should stand before light ones." Bullough believed that this canon was the result of different colors having different weights (which he saw as analogous to the energies of geometrical figures). Here's how he describes the weight of colors:
Supposing a wall 10ft. high were to be painted in two different shades of one colour, viz. pink and red, in such a way that both would be in equal proportions, the one occupying the space below the other; should the red be placed below, up to 5 ft. from the floor, and the remaining 5 ft. be painted pink, or should the position of the colours be reversed? It is pretty safe to say that the former course would be adopted. The effect of 5 ft. of red on the top of 5 ft. of pink would give the majority of people an uncomfortable feeling of top-heaviness, instability and lack of balance, which would be absent from the reverse arrangement of the colors. (p. 112)
In order to test whether the "decorative canon," and its implications for the weight of colors, had any objective basis, Bullough conducted the following experiment. He put together a bunch of figures like these:
He asked participants, among other things, which ones they liked better, and found that people preferred light colors on top and dark colors on the bottom, confirming his "decorative canon," and his belief that colors have weights, with dark colors weighing more than light ones.
Following up on Bullough's study, several researchers over the next few decades further explored the weights of colors. Most of these experiments used setups like this one2:
I know that's kind of small, but the two circles (labled G and H) represent holes behind which colored glass would be placed. The line between the represents a fulcrum, and participants would be asked to move a pointer on the line to the position at which they believe the fulcrum should be balanced. Thus if the color in one circle was perceived as heavier than the one in the other circle, participants would move the pointer closer to the heavier color. Using this method, Marion Monroe3 found:
- Overall, blue is heavier than red, red is heavier than green, and green is heavier than yellow.
- Brighter colors weigh less than darker ones.
- Increasing saturation makes colors appear to weigh less.
- Brighter backgrounds make dark colors look lighter (less heavy) and bright colors look heavier.
Monroe also discovered an illusion. When participants judged the midpoint of the line between the two circles, they tended to perceive it to be closer to the heavier color than to the lighter (less heavy) color.
Pinkerton and Humphrey4, using more sophisticated analyses, got slightly different results. Using the fulcrum method, they found that red was consistently the heaviest color, followed by blue, green, orange, and yellow, again confirming that darker colors are heavier than light ones, but they found no effect of brightness. Here are their average heaviness rankings (1 = heaviest, 4 = least heavy; from their Table 2):
Going with the latest and most sophisticated study, then, we can conclude that red is the heaviest color studied so far, and yellow the least heavy. As far as I can tell, no one has come up with anything resembling a plausible hypothesis for why some colors look heavier than others (unless you want to run with the "energies" thing), much less tested such an hypothesis. We do know that the weights of colors don't affect the weights of objects (e.g., red objects don't appear to weigh more than otherwise identical yellow objects)5, so the effect seems to be just about the color. That might provide a clue. Unfortunately, the literature trail seems to have gone cold with Pinker and Humphreys 1974 paper, so it may be a while before we learn anything more about color weights.
1Bullough, E. (1907). On the apparent heaviness of colors British Journal of Psychology, 2, 111-152.
2Monroe, M. (1925). The apparent weight of color and correlated phenomena. The American Journal of Psychology, 36(2), 192-206.
4Pinkerton, E., & Humphrey, N. (1974). The apparent heaviness of colours. Nature, 250(5462), 164-165.
5Payne Jr., C. (1958). Apparent weight as a function of color. The American Journal of Psychology, 71(4), 725-730.
This brings me to mind of color theory as it pertains to design, and seems to match fairly closely. The attention drawn by a shape has a lot to do with relative values (contrast) between objects, and perceived contrast is effected by contrast with the background they sit upon. One could probably easily get similar results if they were to test people as to whether they perceived darker colors as closer or more important in an image as they did with weight.
Forgot to mention: Different people see color fairly differently; even once you disregard colorblindness there is a large variation in the perception of color between people, and depending on the study size could easily account for the inconsistent red/blue results.
Found this, too:
The Effect of Color on Apparent Size and Weight
Carl J. Warden, Ellen L. Flynn
American Journal of Psychology, Vol. 37, No. 3 (Jul., 1926), pp. 398-401
Apparently there was once a real effort at practical (rather than theoretical) application.
Immediately upon reading this I began thinking of a possible explanation: Could the varying weight of color be related to deep-seated, ancient properties of our visual system that evolved as sort of orientation guides, whereby lighter colors-- i.e. yellow-- would most often appear "up"? It could be not so much an issue of weight as much as an issue of what color would naturally appear on ground level, which would appear in foliage/trees, and which would appear in the sky. This could be such a deeply wired mechanism that we don't even think about it that way, but rather assume that these colors have "weight".
I wonder if any philosophers arguing about the conceivability or possibility of inverted spectrum have discussed this issue (anyone know)? Of course the person who's color qualia are inverted with mine can also be inverted in the way they experience the "weight" of the color, but it looks like the weights do not map on to the color space so that it may be impossible to pull off a perfect color inversion without disrupting the weight inversion.
Brian, That's interesting. The explanation that Pinkerton and Humphrey suggestion in passing is that certain colors are seen as heavier because they are more important, for whatever reason. I wonder how you test that idea, though.
CA, yeah, in fact, there are some recent papers citing these studies in architectural journals, but no one doing new research, as far as I can tell.
Manny, I wonder how you would test that explanation, too.
Eddy, one of the first things I thought about, when I read the Bullough paper, was color qualia. I was hoping a philosopher would chime in.
Manny, I had the same intuition-- that's almost certainly what it is. It's not "weight" they've isolated, but "distance from assumed light source." This is consonant with the finding that lighter, more saturated colors are less "heavy" (colors get darker and dull the less light there is). If someone did a computer analysis of a bunch of pictures of outdoor scenes, I bet they'd find the areas of shadow showed this hierarchy of color tendencies. Borderline or random cases between colors might be where you'd expect the most individual variation. Think about it from a painter's perspective if it helps you: there's a lot more blue in a shadow than there is yellow.