Replicating Synesthetic Stroop effects in Non-synesthetic Contexts

Synesthesia Stroop has been used as a method to attest genuine synesthetic grapheme-color associations, relying on automatic/interference processes similar to the original Stroop (1935) effect. Graphemes presented in colors either congruent or incongruent to participants’ synesthesia result in a Stroop-like effect. In an “ink naming” task, participants have to name the ink color displayed, whereas in a “retrieval” task, participants have to name their synesthetic color-associations. Here we investigate the robustness of the Synesthesia Stroop results. We hypothesized that for semantic associations (e.g., B is Blue), the congruency effect relies on the traditional Stroop mechanism (i.e., the initial letter triggering the automatic color name). We also investigated whether the memory retrieval processes can generate the congruency effect, rather than interference from automatized color associations. Eighteen synesthetes and 18 non-synesthetes completed four different variations of the ink naming and retrieval Stroop/Synesthesia Stroop. Exp 1 was a traditional Stroop experiment, in Exp 2 participants were presented only the initial letter of color words. Exp 3 was the usual Synesthesia Stroop experiment (performed on synesthetes only) and in Exp 4 we performed a Trained Synesthesia Stroop on both groups which was preceded by a 3-minute exposure to novel grapheme-color associations. Our results demonstrate that the usual Stroop and the Initial Stroop yield identical results in both groups (Exp 1 & 2). Second, results from the Synesthesia Stroop task was replicated in the Trained Synesthesia Stroop, for both groups (Exp 3 & 4) during the retrieval task, however, not in the ink naming task. Thus, we show that Synesthesia Stroop on semantic associations seem confounded by the traditional Stroop effect or semantic priming. We also reveal that the Synesthesia Stroop retrieval task may in fact not measure automatic synesthetic processes but rather memory retrieval whereby an ink-naming task appears a more reliable measure.