Verbal Working Memory and Verbal Learning: Word and Pseudoword Learning in a Case of Working Memory Deficit

Over the past years there has been increasing evidence that patients with pathological impairments of "verbal working memory" according to Baddeley still have the ability of verbal learning. This capability comprises, amongst others, learning of associations between meaningful words (paired-associates learning). After presentation of word pairs (e.g., Rohr-Tor, Boot-Schiff), the patient/control subject in the following trial is asked to respond to the first word of a pair with the second one ("cued recall"). However, the learning aptitude does not to the same degree extend to associations between words and pseudowords (pronounceable nonwords) (e.g., the pair Weg - Mant). These findings indicate that verbal working memory is also engaged in certain types of long-term learning, namely the learning of new vocabulary. We tested this hypothesis by performing paired-associates learning tests including a patient with outstanding impairments of verbal working memory (digit span of 3.6) and a group of matched control subjects. As a result, in tests comprising real words the patient showed learning abilities within the range of control subjects. However, in tests with one- and two-syllable pseudowords his performance remained poor. Unexpectedly, we also found dissociations in the performance of the control subjects. Furthermore, one control subject did not show improved learning performance for two-syllable pseudowords. But in contrast to our patient, approximations to target words and false assignments were observable. We interpret the dissociation between word and pseudoword learning in the patient as an affirmation of our hypothesis. The dissociation in our control subjects can be attributed to a utilization of different learning strategies. The absent learning aptitude in one control subject is interpreted as a problem in the capability of assignment and not as a general learning deficit.