Pointing it out: processing faces in the absence of attention

Masked priming is a phenomenon in which subliminal stimuli modulate responses to subsequent visible targets. In congruence priming paradigms, subjects typically respond faster to congruent targets (i.e. of the same category as the preceding prime) than to incongruent targets. Such effects are generally only observed when the prime stimulus is attended. This is not the case for faces, which produce priming effects both when attended and unattended. But is face processing truly invulnerable to attentional modulation, or simply more robust to it than other stimuli? We hypothesised that congruence priming should be evident earlier when the face is attended, and tested this possibility using a reaching paradigm that indexes priming at a stage in which stimulus processing is still ongoing. Using this sensitive measure, we find converging evidence that the visual system is able to process masked faces in the absence of attention, and speculate on the nature of attentional effects on this processing.