Implicit and explicit systems differently predict possible dangers

Abstract

One strategy to address new potential dangers is to generate defensive responses to stimuli that remind learned threats, a phenomenon called fear generalization. During a threatening experience, the brain encodes implicit and explicit memory traces. Nevertheless, there is a lack of studies comparing implicit and explicit response patterns to novel stimuli. Here, by adopting a discriminative threat conditioning paradigm and a two-alternative forced-choice recognition task, we found that the implicit reactions were selectively elicited by the learned threat and not by a novel similar but perceptually discriminable stimulus. Conversely, subjects explicitly misidentified the same novel stimulus as the learned threat. This generalization response was not due to stress-related interference with learning, but related to the embedded threatening value. Therefore, we suggest a dissociation between implicit and explicit threat recognition profiles and propose that the generalization of explicit responses stems from a flexible cognitive mechanism dedicated to the prediction of danger.