Agrammatic aphasics demonstrate sensitivity to syntactic structure alt
hough they generally fail to make adequate use of structural informati
on in producing and comprehending sentences. Under the 'mapping hypoth
esis' (cf. Schwartz et al. 1987), the agrammatic performance pattern i
s taken to reflect impaired mapping between grammatical constituents (
subject, object) and thematic roles (agent, theme). We describe a trea
tment programme ('mapping therapy') directed at the remediation of map
ping operations, and report results of a study designed to assess the
effects of this intervention on comprehension and production. Eight ch
ronic non-fluent aphasics were trained to identify the verb, agent, an
d patient/theme in sentences presented in a combined written-spoken fo
rmat. Three types of sentences were used: active transitives with acti
on verbs (Type A); active transitives with experiential verbs (Type B)
, and non-canonical sentence types (e.g. passives) with action verbs (
Type C). The first phase of training utilized Type A sentences, the se
cond Type B, the third Type C. Generalization to the untrained sentenc
e types was tested by means of multiple-baseline/generalization probes
. Generalization to standard comprehension and production tasks was as
sessed through a comprehensive Language Assessment Battery. Acquisitio
n and generalization profiles differed across subjects. However, follo
wing training most patients showed improved scores on predicted measur
es of sentence production, and two patients also showed improvement on
syntactic comprehension tests. The best outcomes were seen in patient
s with relatively pure agrammatism. Those with more severe and more co
mplicated impairments had poorer outcomes, in part, we believe, becaus
e of the resource demands of this particular training task.