Purchase this article with an account.
Yoana Kuzmova, Jeremy Wolfe, Anina Rich, Angela Brown, Delwin Lindsey, Ester Reijnen; PINK: the most colorful mystery in visual search. Journal of Vision 2008;8(6):382. doi: https://doi.org/10.1167/8.6.382.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Desaturated red is called “pink”, a “Basic Color Term” (BCT, Berlin & Kay, 1969). In contrast desaturated blues and greens have names like “lilac” or “pale green” which are not BCTs in English. Does this distinct linguistic status reflect a special visual status? For example, we asked, would pink targets be comparatively easy to find in visual search? Observers searched for a desaturated target (e.g., pink) among saturated (e.g., red) and achromatic (white) distractors. We picked saturated distractor hues at 9cd/m2, equidistant in CIELAB color space from the “white”, 50cd/m2 Illuminant C distractors. Desaturated targets fell midway between the saturated and white distractors in CIELAB color space. Search was much faster and somewhat more efficient when stimuli were in the reddish/pink range than in any other hue range. The magnitude of this advantage for pinkish targets was very large (hundreds of msec), and extended beyond categorical “pink” to include search for “peach” among orange and white. This suggests that the linguistic term, “pink”, does not itself mediate the effect. We have replicated the result with colors chosen in a similar manner in other color spaces (CIExyY, RGB), and using equiluminant stimuli or heterogeneous distractor hues. In all cases, search for desaturated red and orange hues was significantly more efficient than search for any other desaturated target. What are the sources of this robust effect, if not the linguistic status of “pink”? Perhaps the linguistic term imperfectly reflects an underlying specialization in visual selective attention that favors desaturated reds and oranges, but the basis of the specialization is not yet clear. Stimuli with large R-G signals are generally found faster than those with smaller R-G signals, whereas the relation between RT and Tritan signals is non-monotonic. More speculatively, it is interesting that preferred targets seem to be skin tones.
This PDF is available to Subscribers Only