The present research addressed the question of whether any change between the conditions of original preoperative learning and postoperative recovery would interfere with the availability and retrieval of the spared memories. The procedure was to train rats initially on a brightness discrimination for appetitive rewards and then postoperatively reinstate the appetitive motivation but require that the animals recover the brightness habit, or its reversal, to escape/avoid footshock. The postoperative availability of the spared memories of the preoperative learning was quantified by the occurrence of impaired postoperative reversal training. The results indicated that simply reinstating the preoperative motivation state is insufficient to allow the animal access to the preoperative memories. However, if the preoperative motivation state is reinstated and the animal is also provided rewards appropriate to that moii~ation, then the addition of another motivation and rewards will not prevent the retrieval of spared preoperative memories.The fact that visual decorticate rats are severely impaired when trained on the reversal of a preoperatively learned brightness discrimination demonstrates that this neocortical injury does not destroy the memories of the preoperatively learned behavior even though the lesioned animal is initially unable to perform the task (T. E. LeVere & Morlock, 1973, 1974. It would appear that lesions to other neocortical areas also do not destroy the memories of preoperatively learned behaviors. For example, somatosensory lesions can disrupt the performance of a learned tactual discrimination even though reversal training will demonstrate that the memory of the behavior is spared (Weese, Neimand, & Finger, 1973). Thus it would seem that an important question, if not the important question, for recovery of function theory is why the lesioned animal does not avail itself of memories that are spared. The present research was concerned with this question and what determines the postoperative availability of the spared memories of preoperatively learned behaviors.We believe that part of the answer to this question rests with the animal's attempt to compensate for localized neural injury by utilizing noninjured neural systems (see T. E. LeVere, 1980). If, for example, the visual decorticate rat is trained with visual/nonvisual compound cues, and thus provided the opportunity to compensate for its brain injury by using nonvisual cues, then it will perform significantly better than even normal rats but not recover anything of a preoperatively learned brightness habit (T. E.