One-trial overshadowing: Evidence for fast specific fear learning in humans

•Fast specific fear learning occurs when a single conditioning experience produces situation-specific fear.•Fast specific fear learning emerged in verbal ratings, skin conductance, and fear-potentiated startle.•Fast specific fear learning correlated marginally significantly with trait and state anxiety.•Fast specific fear learning is more consistent with configural compared to elemental models of associative learning.•Failures in fast specific fear learning may result in overgeneralization.

Adaptive defensive actions necessitate a fear learning system that is both fast and specific. Fast learning serves to minimize the number of threat confrontations, while specific learning ensures that the acquired fears are tied to threat-relevant cues only. In Pavlovian fear conditioning, fear acquisition is typically studied via repetitive pairings of a single cue with an aversive experience, which is not optimal for the examination of fast specific fear learning. In this study, we adopted the one-trial overshadowing procedure from basic learning research, in which a combination of two visual cues is presented once and paired with an aversive electrical stimulation. Using on-line shock expectancy ratings, skin conductance reactivity and startle reflex modulation as indices of fear learning, we found evidence of strong fear after a single conditioning trial (fast learning) as well as attenuated fear responding when only half of the trained stimulus combination was presented (specific learning). Moreover, specificity of fear responding tended to correlate with levels of state and trait anxiety. These results suggest that one-trial overshadowing can be used as a model to study fast specific fear learning in humans and individual differences therein.