Pitch-color associations under auditory illusion

Individuals in the general population tend to map pitch height onto lightness (Marks, 1974; Ward et al., 2006; Wicker, 1968), as well as to specific hues (Hamilton-Fletcher et al., 2017), but not onto chromaticity (Ward et al., 2006). Critically, while Marks (1974) was not able to determine whether the pitch-lightness association was absolute or relative, Hamilton-Fletcher and colleagues (2017) reported that from 880Hz and above pitches are associated to yellow hues. We collected data from 6734 participants, who were asked to report the color that instinctively had the best fit to pitches ranging from C4 (261Hz) to B4 (493Hz). In the first block all participants heard the pitches in a random order once. This was followed by a second block where half of the participants heard the same pitches in an ascending order, and the other in descending order. Ascending and descending scales are supposed to create an auditory illusion by which the final pitches of the range are perceived as more extreme than they truly are. This illusion would allow us to evaluate to what degree lightness and color associations to pitch are modulated by bottom-up or top-down processes. We find that the ascending pitches were assigned higher lightness values than randomized ones, and oppositely the descending pitches were assigned lower lightness values. Comparatively, chromaticity values were not affected by the illusion. Regarding pitch association to the yellow color, we found that while it is assigned to an average of 419Hz when pitches are presented in a random order, it is associated to an average of 410Hz during ascending presentation and to an average of 440Hz during descending presentation. These results show that both pitch-lightness and pitch-associations are subject to auditory illusion, supporting that pitch-color associations are relative rather than absolute.