Abstract
The long-awaited PALPA by Kay et al. (1992) is all that it claims to be, but no more than this. As a set of resources for clinicians already familiar with its cognitive neuropsychological theoretical background it provides controlled stimuli and tasks which allow for systematic replication of case studies, baselining for individual cases, and for making comparisons with appropriate control subjects (as recommended by Shallice in 1979). As the first major contribution to aphasia assessment for a number of years it represents a major landmark in the clinical field. However, both the excitement we have felt in awaiting its arrival, and its relative novelty, should not, we feel, obscure the observation that there are important restrictions on the validity and usefulness of the PALPA. In this paper we have focused on these restrictions, firstly considering those which relate to information processing, and secondly, considering those which relate to our fundamental understanding of language.