Abstract
The nature of binding in grapheme-color synesthesia addresses important issues regarding neural basis of synesthesia and visual awareness in general. Previous work with priming tasks have shown the effects of synesthetic colors, but whether synesthetic color is bound to object type or token remains unclear. Furthermore, recent data have cast doubt on whether the synesthetic priming effects are specific to synesthetic experiences. We evaluated type- and token-based color-letter binding using a task combining a synesthetic priming task and the object reviewing paradigm. A participant viewed a preview display with two achromatic letters, followed by a linking display. Then, a color patch appeared at one of the preview letter locations, and observers named the color as quickly as possible. The color matched with the synesthetic color of the preview letter at the patch location (Same-Object condition), of the other preview letter (Different-Object condition), or of a letter not presented (No-Match condition). Type- and token-based bindings are reflected by difference between No-Match and Different-Object conditions, and between Different-Object and Same-Object conditions, respectively. For control participants, letter-color associations were trained, before the object reviewing task. Although mean response times failed to reveal any significant differences between synesthetes and control participants, RT analysis with ex-Gaussian distribution revealed a distinct pattern in synesthesia group that response facilitation in Gaussian component, presumably reflecting stimulus encoding, exclusively reflected type representations. In contrast, the control group showed response facilitation in Gaussian component dominated by token representations. Thus, color-letter associations in “associator” type synesthetes are type-based without binding to object token, consistent with their subjective reports. Contrary to recent failures in showing differences between synesthetes and non-synesthetes, combination of type-token distinction and response time distribution analysis could indicate that color sensations in synesthetes are not simply the extreme form of normal associations, and cannot be attributed to demand characteristics.
Grants-in-Aid from JSPS (#21300103) and Global COE “Revitalizing Education for Dynamic Hearts and Minds” from JMEXT.