Colour is generally regarded as a property of objects in the world, or at least of the light that is reflected from objects. In fact there is no colour out there it is constructed by our eyes and brains. Light waves that are usually associated with green for example, may be experienced as yellow or grey according to what else the brain is experiencing at the time, what it expects to see, and what it has just seen.
This initial classification of a colour is made by retinal neurons. The brain then continues to sort the incoming stimuli into various colours according to a classification system which is partly innate and partly learned.
Imagine a band of colour blending smoothly from red at one end to blue at the other. Now imagine dividing the band into thousands of tiny slices, each one so thin that the difference between one and the next along is so slight as to be imperceptible.
Now think of a person naming the colour of each slice in turn, moving along the spectrum from one end to another. If they start at the red end, they would name the first slice, obviously, red. Then, because they would be unable to detect any difference between the slices, they would name the next one red too. And the next, and the next, and the next
Right up to the blue end.
In fact, of course, we dont do this. Rather at some point in the exercise we make a decision to switch colour categories. However, the decision would not be based on a perceived difference between one slice and the next because there would not be one rather it would reflect a shift from one conceptual category to another. We dont see the dividing line between one colour and the next we impose it.
Some languages divide the visible spectrum into many more colour categories than others. English speakers, for example, divide the colour space into eight categories (red, blue, green, pink, purple, orange, yellow and brown), whereas the language pf the Berinmo tribe, in Papua New Guinea, has only five colour names for the same range. But the difference between the two cultures is not just that they use different words for the same colour qualia rather that they actually experience different things when they look at the same colour.
A study by Jules Davidoff, of Goldsmiths College in London, found that Berinmo speakers do not just have a relatively crude way of describing colour differences, they also perceive fewer distinctions. This suggests that having a language-based concept for a particular colour may be necessary in order for us to see it as distinct from another. It could be that the relatively impoverished colour lexicon of the Berinmo tribe reflects some difference in the colour part of their visual sensory cortex so that it is not equipped to make fine colour distinctions. An alternative interpretation is that the Berinmo have not developed the language-based concepts required to make the distinctions, so that although their colour neurons register subtle colour changes in the same way as English speakers, these changes are not represented at a higher cognitive level and therefore do not become conscious.
