Predictability and controllability: Differential effects upon contextual fear

The independence of action of controllability and predictability has recently been questioned by research demonstrating that the effects of control over shock termination can be mimicked by feedback stimuli when a contextual fear measure is used. This suggests that the varied effects of controllability, particularly controllability of shock termination, may result not from controllability per se but from predictability of shock absence. The present experiments address this issue by examining whether controllability and predictability similarly affect contextual fear under several parametric conditions. In Experiment 1, control over shock termination was found to reduce contextual fear at an earlier point in training than prediction of shock absence. Experiment 2 demonstrated an effect of controllability under conditions in which the feedback effect is precluded. Experiment 3 examined the possibility that the group differences observed in the above experiments could be due to a potential difference in the conditionability of the response-produced stimulation and the external feedback stimulus. The outcome of this study makes it unlikely that this is the case, since no evidence of overshadowing of the feedback stimulus was observed on a test of its associative strength. These experiments suggest that the effects of controllability may not be reducible to those of predictability. Furthermore, they have important implications for theoretical proposals concerning the effect of feedback on contextual fear.