Verbal short-term store-rehearsal system and the cerebellum. Evidence from a patient with a right cerebellar lesion.

We describe an 18-year-old patient who underwent surgical removal of the right cerebellar hemisphere for the presence of a neoplastic lesion. After surgery, the patient's neuropsychological examination was normal except for a transient selective verbal short-term memory (STM) impairment characterized by reduced verbal digit span and rapid forgetting of verbal material. An extensive examination of the patient's deficit was performed in order to identify which of the two components of STM (phonological short-term store and/or rehearsal system) was impaired. The functional locus of the deficit was identified at the level of the phonological output buffer, a component of the rehearsal system, as suggested by the pattern of results obtained, namely: the improvement of the digit span seen with pointing compared with the verbal response; the advantage of auditory over visual presentation of digits; and the lack of a phonological-similarity effect with visual presentation of letters. On the other hand, the functioning of the phonological store was demonstrated by the normal amplitude of the recency effect in free recall of words and by the phonological-similarity effect with auditory presentation of letters. Our finding is consistent with previous functional (PET) studies showing the involvement of the right cerebellum during tasks requiring silent recirculation of verbal information. We conclude that the cerebellum takes part in the planning of speech production at a level that does not require an overt articulation.

The phonological short-term store-rehearsal system: patterns of impairment and neural correlates.

Two left brain-damaged patients (L.A. and T.O.) with a selective impairment of auditory-verbal span are reported. Patient L.A. was unable to hold auditory-verbal material in the phonological store component of short-term memory. His performance was however normal on tasks requiring phonological judgements, which specifically involve the phonological output buffer component of the rehearsal process. He also showed some evidence that rehearsal contributed to the immediate retention of auditory-verbal material. Patient T.O. never made use of the rehearsal process in tasks assessing both immediate retention and the ability to make phonological judgements, but the memory capacity of the phonological short-term store was comparatively preserved. These contrasting patterns of impairment suggest that the phonological store component of verbal short-term memory was severely impaired in patient L.A., and spared, at least in part, in patient T.O. The rehearsal process was preserved in L.A., and primarily defective in T.O. The localisation of the lesions in the left hemisphere (L.A.: inferior parietal lobule, superior and middle temporal gyri; T.O.: sub-cortical premotor and rolandic regions, anterior insula) suggests that these two sub-components of phonological short-term memory have discrete anatomical correlates.

The production of specific and generic associates of living and nonliving, high- and low-familiarity stimuli in Alzheimer's disease.

The goal of this paper was to address three issues relevant to the semantic-lexical deficit of Alzheimer's disease (AD) patients: (1) the nature of the deficit (contrasting the loss of information with the difficulty of intentional access hypothesis), (2) stimulus familiarity effects, (3) semantic category effects (contrasting living with nonliving categories). Sixteen patients affected by AD and 11 matched control subjects were given a naming task, a word-picture matching task and a generative associative naming task, constructed by using as stimuli the same 40 items. Stimuli were either living or nonliving items of high, medium, and low familiarity. Responses given on the generative associative naming task were classified as generic (poorly informative) or specific (more informative). As expected, AD patients were poorer than controls both in naming and in word-to-picture matching tasks. In the generative associative naming task, AD patients tended to produce fewer specific than generic associations with respect to controls. Category effects had a significant influence on performance only on the naming task, but not on the word-to-picture matching or on the generative associative naming task. Stimulus familiarity, on the other hand, strongly influenced performance both in AD patients and in controls: naming and comprehension were better for high familiarity items with respect to low familiarity items. For generative associative naming, the tendency of AD patients to produce fewer specific than generic associations was mainly significant for low familiarity items.