Non-random associations of graphemes to colours in synaesthetic and non-synaesthetic populations

Abstract

This study shows that biases exist in the associations of letters with colours across individuals both with and without grapheme-colour synaesthesia. A group of grapheme-colour synaesthetes were significantly more consistent over time in their choice of colours than a group of controls. Despite this difference, there were remarkable inter-subject agreements, both within and across participant groups (e.g., a tends to be red, b tends to be blue, c tends to be yellow). This suggests that grapheme-colour synaesthesia, whilst only exhibited by certain individuals, stems in part from mechanisms that are common to us all. In addition to shared processes, each population has its own distinct profile. Synaesthetes tend to associate higher frequency graphemes with higher frequency colour terms. For control participants, choices are influenced by order of elicitation, and by exemplar typicality from the semantic class of colours.

View correction statement:

Corrigendum

Acknowledgments

This research was supported by the first author's British Academy Postdoctoral Fellowship (PDF/2001/345). We would like to thank Matthew Whalley and Ganesh Gandhi for their assistance in collecting some of the data reported here.

Notes

An anonymous reviewer has asked us to point out that our focus on graphemes gives rise to an item list (n = 36) that is smaller than some studies (where the synaesthesia is triggered by words; e.g., n = 80 in Ward et al., 2005). We tested consistency for all 26 letters, but restricted ourselves to the numerals 0–9 simply because grapheme–colour synaesthetes often report that numbers 10-and-above are coloured as a combination of their constituents (e.g., 25 is the colour of 2 and 5). The crucial fact, however, is that our list is equally sized for both synaesthetes and controls and that synaesthetes perform significantly better—even with the longer time interval.

It is likely that control participants generate a list of candidate colours for each letter, with the candidate list ordered—at least to some extent—according to ease-of-generation. While Forced-choice controls are obliged to state a colour for every letter (working steadily through their candidate list) Free-choice controls are free to reject all candidates at any given point, and might even return to the top of the candidate list for the next selection. This may explain why Free-choice controls show an influence of ease-of-generation in the overall frequency of colours generated, without an effect of presentation order.