Language use statistics and prototypical grapheme colours predict synaesthetes' and non-synaesthetes' word-colour associations

Abstract

Synaesthesia is the neuropsychological phenomenon in which individuals experience unusual sensory associations, such as experiencing particular colours in response to particular words. While it was once thought the particular pairings between stimuli were arbitrary and idiosyncratic to particular synaesthetes, there is now growing evidence for a systematic psycholinguistic basis to the associations. Here we sought to assess the explanatory value of quantifiable lexical association measures (via latent semantic analysis; LSA) in the pairings observed between words and colours in synaesthesia. To test this, we had synaesthetes report the particular colours they experienced in response to given concept words, and found that language association between the concept and colour words provided highly reliable predictors of the reported pairings. These results provide convergent evidence for a psycholinguistic basis to synaesthesia, but in a novel way, showing that exposure to particular patterns of associations in language can predict the formation of particular synaesthetic lexical-colour associations. Consistent with previous research, the prototypical synaesthetic colour for the first letter of the word also played a role in shaping the colour for the whole word, and this effect also interacted with language association, such that the effect of the colour for the first letter was stronger as the association between the concept word and the colour word in language increased. Moreover, when a group of non-synaesthetes were asked what colours they associated with the concept words, they produced very similar reports to the synaesthetes that were predicted by both language association and prototypical synaesthetic colour for the first letter of the word. This points to a shared linguistic experience generating the associations for both groups.

The spatial mapping of concepts in English and Mandarin

“…In a similar vein, in recent work, we have found evidence that participants' explicit associations between concepts and colour (Goodhew & Kidd, 2017) also correspond to those associations that produce congruency effects on implicit behavioural responses. This is further evidence that participants do indeed have explicit access to how concepts are represented, and in particular, how they are related to physical dimensions such as space, thereby validating the approach we adopted here.…”

“…Furthermore, language-use statistics have also been found to explain variance in other systematic associations, such as those between concepts and colours (e.g. happy and yellow) (Goodhew & Kidd, 2017). Given the large degree of overlap between the spatial mapping of concepts in English and Mandarin observed in the present work, this framework would predict similar language-use biases (e.g.…”

Bliss is blue and bleak is grey: Abstract word-colour associations influence objective performance even when not task relevant

“…The third piece of evidence to suggest that humans may encode abstract word-colour associations which manifest on objective task performance is our previous work, which showed that synaesthetes (i.e., individuals who have sensory associations manifesting as atypical perceptual experiences) and non-synaesthetes alike report explicit subjective associations between abstract words and colours (Goodhew & Kidd, 2017). We also found that these could be explained by language use statistics.…”

“…We also found that these could be explained by language use statistics. That is, consensus associations between words and colours (e.g., joy-yellow and sorrow-blue) were reflected in systematic cooccurrence between these words (e.g., between word joy and word yellow), as quantified by latent semantic analysis (LSA) (Goodhew & Kidd, 2017). However, Goodhew and Kidd (2017) relied on subjective reports of associations provided by participants.…”

“…That is, consensus associations between words and colours (e.g., joy-yellow and sorrow-blue) were reflected in systematic cooccurrence between these words (e.g., between word joy and word yellow), as quantified by latent semantic analysis (LSA) (Goodhew & Kidd, 2017). However, Goodhew and Kidd (2017) relied on subjective reports of associations provided by participants. While it has been established that concurrent-inducer pairings influence behaviour for synaesthetes (Dixon, Smilek, Cudahy, & Merikle, 2000;Mattingley, Rich, Yelland, & Bradshaw, 2001), the goal of the present study was to determine whether these subjectively reported associations also influence objective task performance in non-synaesthetes, even when they were not relevant to the task.…”