Opposite Priming Effects on Identity vs. Category Recognition Require Conscious Awareness

How conscious and unconscious priming differentially modulate object recognition at different levels of abstraction (identity vs. category) remains incompletely understood, despite extensive research. We used a binocular rivalry paradigm with Continuous Flash Suppression (CFS) to manipulate conscious awareness with image or word primes while participants performed a name-picture verification task probing identity and category recognition for faces and animal bodies. Behavioral results revealed a striking dissociation: consciously perceived primes facilitated identity recognition but impaired category recognition. This effect was most pronounced with image primes, face targets, and right visual field presentations. Under unconscious priming, no such effects were observed. To elucidate the underlying mechanism, we applied a Drift Diffusion Model, which revealed that conscious priming selectively enhanced the efficiency of evidence accumulation and introduced a pre-decisional bias for identity decisions, with no reliable change for category decisions. Our findings demonstrate a double dissociation where conscious awareness is required for priming to exert robust and opposite effects on identity and category recognition. This finding challenges the view of priming as a uniformly facilitatory process, providing a new mechanistic framework for understanding how consciousness and abstraction level interact to shape visual perception. HighlightsO_LIConscious awareness is required for priming to exert robust effects on recognition C_LIO_LIThe priming aftereffects can be facilitatory or impairing depending on the level of recognition C_LIO_LIConscious priming facilitated identity recognition but impaired category recognition. C_LIO_LIDissociation of identity and category by conscious priming were strongest for image (vs. word) primes, face (vs. animal) targets, and right (vs. left) visual field presentations. C_LIO_LIDrift diffusion modeling revealed increased drift rate and pre-decisional bias, for identity but not for category decisions in conscious priming. C_LI