Models on the mechanisms underlying letter-colour synaesthesia diverge on a central question: whether triggered sensations reflect (a) quantitatively or (b) qualitatively deviant brain organization. We previously found evidence for (a) after observing synaesthesia-like letter-colour binding in adult non-synaesthetes following execution of a visual letter search task which employed likelihood manipulations of letter-colour pairings to implicitly train letter-colour associations [1]. The newly-formed associations were synaesthesia-like, since correlating with the synaesthetic-Stroop and showing the colour-opponency effect, present in synaesthetes [2]. This latter effect manifests as increased synaesthetic-Stroop interference when the real colour of a letter is opponent to the synaesthetic/associated colour, in line with involvement of visual/colour areas in letter-colour binding.
Here, we investigated the brain areas involved in the formation of these synesthetic associations. Based on [3] revealing reciprocal involvement of the PPC and dlPFC in distinct aspects of learning in a numerical conception task, we hypothesized that these two areas may also be differentially implicated in the learning of synaesthesia-like letter-colour associations by non-synaesthetes. Previous results would predict that while PPC supports processing of these associations after learning [see also 4], dlPFC shows a reciprocal function, i.e., suppressing these associations under normal conditions.
Using bilateral tDCS, we interfered with either PPC or dlPFC in two groups while they performed the letter-colour association task as in [1]. A third group performed the same task but without tDCS (control group). All three groups used non-opponent colour pairs for letter-colour association learning, to avoid ceiling effects. In comparison to the control group, dlPFC-stimulation significantly enhanced letter-colour binding. This enhancement was substantial, leading to interference between learned and real colours in the order of the colour-opponency effect (despite the use of non-opponent colour pairs). No such effect was observed with PPC stimulation.
This provides novel information regarding the network of areas implicated in the formation of automatic (letter-colour) associations. Synaesthesia-like binding of colours to letters by non-synaesthetes seems to be suppressed by dlPFC under normal conditions, and may be released by modulation of the dlPFC. We speculate that dlPFC-stimulation facilitates the formation of letter-colour binding, possibly by emphasizing relevant associations during task performance and/or by disinhibiting other brain areas (i.e., posterior parietal) involved in automatic, perceptual letter-colour binding.

synaesthesia; learning; tDCS;