Despite their shifting shape, even my 5-year old niece can distinguish between the Barbapapas because of their distinct colours. If they didn’t vary so much in their pink, black, red, green, yellow and blue, following a Barbapapa story wouldn’t be so much child’s play anymore. But how can we characterize color? How do we distinguish between blue Barbarix and green Barbalala, and the close hues of Barbarix and violet Barbarella?
Although the light spectrum, of which colour consists, is continuous, we distinguish discrete categories of colour – red, yellow, green, blue and some of the more outlandish, like “aubergine”, though some people contest it being a colour at all. Traditionally, the debate on how these colour categories arise has focused mostly on whether there are actual biological constraints to these colours, or whether colours are a construct, arising from culture and communication. A recent paper in PNAS attempts to tackle this question from the side of neuroscience, asking how colours are represented in the brain.
The researchers used the functional MRI adaptation method to identify brain regions that categorize color. fMRI idetects the levels of blood oxygen, the adaptation method looks at whether adaptation occurs in a brain region. When a group of neurons responds to a stimulus, their response becomes toned down when exposed to this stimulus repeatedly. But when a different stimulus is presented, a fresh group of neurons is excited and the fMRI response is not toned down. By comparing the fMRI response, we can determine whether the same or different groups of neurons respond to two stimuli.
To test which brain regions respond to color, the researchers showed trial participants blocks of either green or blue. They presented only one hue of green, but three different hues of blue. The blocks of colors were presented in pairs, and every color occured with every other color. So this design looks at both color categories – green vs. blue – and the difference in hue between colors – light blue vs. dark blue. In the fMRI analysis, the researchers looked for neurons that respond only to the category difference green vs blue, and not to hue differences light blue vs. dark blue. These neurons would be the ones encoding color category. Importantly, the participants did not have to make any judgements on the color itself – whether blocks are similar, different, or what their names are. They instead focused on a different task, pressing a key when a lighter patch was presented. Only after the fMRI were they asked to name the colors. This study design means that participants weren’t actively categorizing color, or thinking about their names. Intrinsic category judgements – whether a block is green or blue – are so not confounded with an activation of brain regions involved in language processing.
In this study, the left and right middle frontal gyrus (MFG) showed a strong response to differences in category, but not in hue. Color is here encoded purely as a category, as green vs blue, and this encoding is automatic – the participants did not have to think about the color categories for the MFG to show different responses to the colors. So what is color? The researchers here found no evidence for color category encoding in brain regions classically thought to be involved in languague processing. Also, categorization appears not to occur at the stage color vision or visual processing. So color is not – at least not only – a concept of language, or constrained by our vision. Instead, the MFG distinguishes between colors, and places them in categories. Distinguishing between blue Barbarix and green Barbalala is hard-wired into our brain. Though whether “aubergine” is a color or not remains to be tested.
Original research paper by Bird et al.: http://www.pnas.org/content/early/2014/02/27/1315275111.full.pdf+html