This study investigated the speed of short-term memory scanning in two groups of patients: patients with Parkinson's disease (PD) (primarily subcortical pathology) and patients with a presumptive diagnosis of dementia of the Alzheimer type (DAT) (primarily cortical pathology). Non-demented PD patients (M age = 58.3), mildly demented DAT patients (M age = 67.0), and normal control (NC) subjects (M age = 58.9) were compared to one another in order to determine whether the presence of slowed scanning differentiates subcortical from cortical pathology. The three groups did not differ significantly on the memory scanning task. Slowed scanning speed does appear to occur in some patients with PD, however. This may relate to age, duration of disease, or an interaction between these two. Despite normal scanning speed, most DAT patients required highly structured instructions to be able to carry out the item-recognition task, and many remainedunable to perform this task even with additional instructions.

Previous findings suggest that motoric response preparation cannot be initiated in parallel with memory scanning. In the present study, response preparation was initiated with the aid of a precue to examine whether such preparation can be maintained or continued while memory scanning is active. In Experiment 1, each trial began with a colored square indicating which hand might be needed to respond. A probe letter's memory set membership determined whether the primed response should be made or withheld. Lateralized readiness potentials were initiated by the square precue and continued to increase after letter presentation, suggesting that once response preparation had been initiated it was continued in parallel with memory scanning. Experiment 2 suggested that the difficulty of the concurrent memory search had little effect on the continuation of response preparation. The results support the view that motoric response preparation consists of at least two qualitatively distinct phases—initiation and continuation.