We seek converging evidence from cognitive studies of non-demented patients with Parkinson's disease (PD), electrocortical event-related potentials (CEPs), and functional magnetic resonance imaging (fMRI) to test our interactive neurocognitive model of core cognitive processes and executive resources in comprehension.
Specific Aim 1 manipulates executive resources (working memory, strategic planning, inhibitory control) in ambiguous sentences. PD patients' impaired sentence comprehension will be related to limitations in specific executive resources. Resource-related slowing of CEPs will be seen in PD for the same material. fMRI in young subjects with this material will recruit interactive neural networks for sentence processing: left ventral inferior frontal cortex (vIFC) and left posterolateral temporal cortex (PLTC) for core language processes, and specific cognitive resources in left dorsal IFC (dIFC), prefrontal cortex, striatum, and right PLTC. To compensate for age- and disease-related resource limitations, healthy seniors and PD patients will up-regulate resource-related networks, but we expect no change in the core sentence processing network.
Specific Aim 2 tests a material-neutral deficit for rules that depends on implicit memory. We examine regular and irregular morphology in verbs and nouns, and assess non-linguistic concept acquisition mediated by implicit- or rule-based learning. PD patients will show a material-specific deficit for rules in verbs. fMRI in young subjects will recruit left vIFC only for regular verb morphology, and dIFC for decision-making resources. d]FC will be up-regulated in aging and PD.
Specific Aim 3 assesses the generalizeability of our model to prosody comprehension. PD patients judge acoustically simple and complex prosody stimuli at baseline and during a secondary task. Restricted resources will limit PD patients' comprehension of complex prosody. fMRI in young subjects will recruit orbital frontal and dIFC only for complex prosody, and dIFC will be up-regulated in aging and PD. Our data support a componential neurocognitive architecture consisting of dynamically interactive networks modified to process sentences depending on available resources and relative demand.
Showing the most recent 10 out of 56 publications