Cyma Van Petten, Ph.D. - US grants

This proposal is aimed at expanding our knowledge of the psychological and physiological bases of language comprehension, primarily from scalp recordings of event-related brain potentials (ERPs). Current evidence suggests that the N400 component of the ERP is a physiological index of some aspect of word recognition reflecting the influence of semantic context at both the lexical and sentential levels. A late positive component of the ERP has been tied to the processes of memory encoding and retrieval, perhaps at a general level which encompasses both linguistic and extralinguistic knowledge. The relationship between the N400 and late positive components, and the psychological/physiological events underlying them, is underspecified at present and clarifying these relationships is also a goal of the proposal. Experiments 1 and 2 examine ERPs elicited by spoken words to clarify the interactions between lexical and sentence level semantics, the importance of word frequency in word recognition, and the timecourse of word recognition. Experiments 3 through 6 investigate the impact of other cognitive processes on language processing. Experiment 3 investigates the modification and reorganization of lexical knowledge by determining whether or not new or weakly-learned words produce different context effects than well-known words. Experiment 4 also investigates the structure of lexical knowledge and its relationship to memory retrieval processes by examining both semantic and phonological priming effects when subjects are in a tip- of-the-tongue state. Experiment 5 examines crosstalk between attended and unattended words under two different varieties of attention selection based on spatial location or color. This experiment will describe the relationship between the early attention-sensitive components of the ERP, thought to reflect filtering of unattended stimuli, and the later N400 component. Experiment 6 examines mnemonic processes during discourse comprehension in an attempt to better characterize the relationship between the late positive component and episodic learning or retrieval.

DESCRIPTION (adapted from investigator's abstract): Cognitive approaches to the problems of declining memory in the normally-aging population have proceeded largely independently of neurobiological investigations of age-related changes in brain structures and processes. The proposed research combines these two approaches in order to explore directly the relations between changes in neural structures and the cognitive/behavioral changes that occur with age. In recent years it has become evident that memory is not a unitary function, but is composed of different systems and processes, some affected by aging and brain damage and others not. Understanding the nature of these processes and their neural correlates will enable us to develop rehabilitation techniques that target specific areas of deficit and at the same time take advantage of specific areas of strength within an individual. Most studies of aging compare a randomly selected group of older adults to a group of young adults. Such comparisons, however, fail to take into account the heterogeneity of the elderly population. The investigators' research program specifically focuses on individual differences by categorizing older individuals according to their performance on groups of neuropsychological tests thought to tap functions associated with either prefrontal or medial temporal cortex. The composite scores of neuropsychological function are then used as predictors in all behavioral experiments and in electrophysiological studies that examine event-related potentials in related paradigms. A subset of the elderly individuals will also receive magnetic resonance imaging (MRI) of the brain regions thought to be implicated in different kinds of memory in order to confirm the relations between cognitive and brain function. Research will focus on the differences between memory for the content of an experience and memory for its context, often referred to as source memory. Older adults show a disproportionate deficit in source memory, which has been hypothesized to be due to declining frontal lobe function. Other research suggests that the medial temporal lobes are critical for the binding of various attributes of an experience into a coherent memory trace. The investigators' goal is to differentiate the contributions of these two brain regions to source and item memory and to document their differential impact on age related memory decline. The proposed experiments include several context/source manipulations--perceptual, spatial, temporal, and internal vs external--across a range of materials--verbal, nonverbal, auditory, visual, novel and familiar. In each study, comparisons will be made not only between older and younger adults, but also among subgroups of older adults according to their putative frontal and medial temporal functioning. Issues to be addressed include: the extent to which age-related memory deficits are attributable to problems of encoding, retrieval or both, the role of interference in source attributions, and the dissociation between explicit and implicit memory for source.

DESCRIPTION (provided by applicant): Three proposed experiments are directed at a better understanding of the role of prefrontal cortex in categorization under different strategies, with a focus on four cognitive processes: response inhibition, maintenance of stimulus features in working memory, manipulation of information in working memory, and directed search of long term memory. Although prefrontal cortex has traditionally been associated with rule induction and shifting from one rule to another, our initial event-related potential (ERP) study comparing multidimensional rule use, unidimensional rule use, and a multidimensional similarity strategy for categorizing novel visual stimuli showed similar large late prefrontal potentials for subjects using both of the multidimensional strategies as compared to the unidimensional rule. Proposed Exp. 2 examines the impact of maintaining multiple stimulus attributes in working memory (WM) on this late prefrontal activity, by manipulating the number of stimulus dimensions relevant to a specified categorization rule. The impact of rule dimensionality (hypothesized to influence the maintenance component of WM) is contrasted to a manipulation of rule complexity (conjunctive versus exclusive-OR), hypothesized to tax a theoretically separate aspect of working memory - the manipulation of information. Proposed Exp. 3 examines the interaction between WM and long-term memory representations of stimulus attributes relevant to category assignment, by comparing categorization judgments based solely on the contents of WM versus those that require retrieval of information from long term memory. Finally, Exp. 1 pursues a finding of the preliminary study that was unique to multidimensional rule use: an early prefrontal potential that was larger for stimuli close to the boundary between two categories as compared to stimuli far from the category boundaries, and thus more typical of their categories. We hypothesize that this frontal N2 component reflects inhibition of responses based on preliminary perceptual analyses when information consistent with an alternative, incompatible response is detected. Exp. 1 tests the hypothesis that one role of prefrontal cortex in categorization is to inhibit responses based on fast perceptual processes in order to allow slower analyses to run to completion when a rule-based judgment should be based on the slower analyses. The experimental design thus contrasts stimuli that are globally similar to the prototype (and individual exemplars) of one category although belonging to a different category according to the rule to stimuli for which global similarity and the rule both suggest the same response.