A dissociation of the effects of control and prediction

Abstract On three different tests we assessed the ability of a feedback stimulus presented during inescapable shock exposure to mimic the effects of control over shock offset. We found that animals exposed to inescapable shock with a feedback stimulus and animals that received escapable shock demonstrated equivalent levels of fear conditioned to the shock context; moreover, these two groups showed less fear than animals exposed to inescapable shock without feedback. This pattern of group differences was observed also on a shuttlebox escape test. However, on an appetitive noncontingent test we observed that feedback did not affect responding. On this test animals previously exposed to inescapable shock with and without feedback demonstrated an equivalent pattern of responding that was different from that of the escape group. This dissociation of the effects of control and feedback lends support to the notion of independence of action for control and prediction.