Bridging temporal gaps between CS and US in autoshaping: insertion of other stimuli before, during, and after CS.

In four experiments pigeons were exposed to key-light illuminations separated from food delivery by 12-60 sec. Approach to the key light did not develop on conventional trace-conditioning arrangements but occurred consistently whenever some auditory or visual stimulus (a) filled the CS-US gap (serial conditioning) or (b) was always present except during the gap. Various comparison groups showed that this enhancement of conditioning cannot be mainly attributed to similarity between the CS and the added stimulus, or to spread of specific responses evoked by that stimulus, or to potentiation of CS's neural aftereffects by the extra stimulus. However, modifications of condition b in the final experiment revealed that CS approach was strong only when the stimulus present during the intertrial interval remained on until the termination of CS; if the stimulus ended at CS onset, conditioning did not occur. Although discriminability of CS-US gaps from intertrial periods seems necessary for conditioning to occur in the absence of close CS-US contiguity, the outcome of the final experiment indicates that such discriminability is not sufficient for conditioning. The results are primarily interpreted in terms of (a) possible second-order conditioning effects and (b) changes in the associative strength of the "local context" existing when CS appears, which may lead to superconditioning of CS.