ALGORITHMIC AND HEURISTIC STRATEGIES IN COMPREHENSION OF COMPLEMENT CLAUSES BY PATIENTS WITH PARKINSONS-DISEASE

Language ability in patients with idiopathic Parkinson's disease (PD) and normal control subjects (NC) matched on age, sex, education and socioeconomic status (SES) was investigated. The two groups of subjects were tested on eight sentence types in Greek in the form of main and complement clause with eight matrix verbs. These matrix verbs were ask (ask information), promise (commissive meaning), tell1 (order, command) and tell2 (give information) in sentences with no semantic constraints, and confess, sell, trust and scold in sentences with semantic constraints (implicit causality). The results show that language ability, despite relative preservation is significantly impaired in PD patients as compared to that of NC. More specifically, syntax with semantic constraints was the most effective independent variable to classify PD patients and NC subjects into two distinct groups according to a Logistic Regression Analysis. To restrict the algorithmic process in sentence comprehension, PD patients seem to make use of the minimal distance principle (MDP) and the ''experiencer constraint'' heuristic strategies. Possible similarities in language behavior between PD patients and aphasics, in general, are suggested.