Abstract
Previous research suggests that conscious and unconscious processing follow different time courses. Early studies showed that primes’ impact on responses to subsequent targets survives only short prime-target intervals for subliminal primes, but lasts longer for supraliminal primes. However, this dissociation may reflect differences in stimulus strength rather than in conscious perception. Later studies showed that with liminal primes, for which conscious and unconscious processing can be compared under the same stimulus conditions, unconscious priming occurs only for fast responses, whereas conscious priming remains stable across the RT distribution. This dissociation may reflect decay of unconscious representations. However, it could also reflect that fast responses, unlike slow responses, rely on fast-and-dirty processing that is particularly sensitive to unconscious information. Finally, these dissociations were reported for semantic priming and whether they hold for lower-level prime representations is unclear. Here, we used the liminal-prime paradigm and measured repetition priming, response priming from learned stimulus response associations and semantic response priming. We forced participants to rely on fast, intuitive processing by imposing a short response window, and manipulated prime-target intervals. Both response priming effects quickly waned for unconscious primes and generally increased for consciously perceived primes. By contrast, repetition priming was independent of conscious perception and increased with prime-target interval. These findings suggest that conscious perception prevents decay of the prime’s impact on response selection. However, the different pattern observed for repetition priming calls for further research on the longevity of unconscious effects that are not mediated by response priming.