Recognizing the unknown: Motor-response execution reflects the availability of positive evidence during recognition

Recognition is a core component of cognition, yet how we categorize a stimulus as unknown remains an open issue. In this study, we tested whether, during recognition, motor-response execution is affected by the presence vs absence of positive evidence for the to-be-recognized stimulus. Participants performed a visual lexical decision task with five categories of stimuli, varying along a word-likeness gradient: High-frequency words, low-frequency words, letter-substitution pseudowords (derived from real words by changing one letter), pseudowords (non-existing words not resembling any words), and illegal strings of letters. The effect of stimulus category was separately examined in the premotor and motor component of the response. Premotor times (and RTs) varied systematically with word-likeness: The more extreme categories of high-frequency words and illegal letter strings yielded the fastest responses, whereas the more ambiguous categories produced slower responses. In contrast, motor time, which indexed motor-response execution, was insensitive to word-likeness, but faster for stimuli providing positive evidence about their lexical status (i.e., words and illegal letter strings), compared to those that did not (i.e., pseudowords and letter-substitution pseudowords). These results point to a differentiation in the decisional dynamics modulating premotor and motor time, with motor-response execution being selectively sensitive to the presence vs absence of positive evidence, rather than to the more general dimension of word-likeness. More broadly, the findings highlight a differentiation in the processes underlying memory-based recognition depending on the presence vs absence of positive evidence, particularly with respect to the implementation of the corresponding action.