Marta Kutas - US grants

The goal of the proposed research is to use information from event-related brain potential (ERP) recordings to analyze the special processes that are invoked when a strong context or expectation is violated. The utility of the ERP approach in this endeavor is attested to by the numerous studies which have described the sensitivity of the amplitude of the late positive "P300" complex to fluctuations in a person's expectations. Recent data from our laboratory, however, have indicated that not all unexpected events are followed by a P300. In particular, we have found that the infrequent violation of a semantic expectancy for a word at the end of visually sented sentence is associated with a late, monophasic negative wave (N400) which is distinct from the P300. We proposed that the N400 might reflect a delay in the ongoing processing triggered by the semantic incongruity and the subsequent backtracking to the point of interruption ("second look") and the "reprocessing" which attempts to make sense of the situation. The proposed studies are aimed at testing this hypothesis and at defining the conditions wherein some violations of expectancy are associated with the P300 complex and others the N400. The following questions will be addressed specifically: 1. How responsive is the N400 to variations in the parameters of the reading situation, such as (a) the probability of the semantic anomaly, (b) the amount or strength of context, (c) the serial position of the anomaly in the sentence, (d) the rate of word presentation and (e) the interval between the context and the anomalous word? 2. Is the N400 specifically associated with violations of semantic expectancy or will it occur following violations of well-known grammatical and/or lexical rules as well? If N400 waves are found in response to other linguistic anomalies, will their properties be specific to the type of violations? 3. Are semantic expectancies modality specific as some authors have contended, or modality independent? That is, will an N400 occur in response to a semantic discrepancy at the end of an aurally presented sentence in the same way as in the silent reading task? Moreover, will the "N400 effect" occur even when the semantic incongruity and the preceding context are presented in different modalities? 4. How dependent is the elicitation of the N400 upon the reading task in which it was originally observed? Would similar ERPs be obtained in other situations requiring a semantic analysis, e.g., the sentence-picture verification task of Clark & Chase (1972)? 5. Is the N400 uniquely related to the violation of linguistic rules, or will it appear in response to unexpected notes in well-known musical sequences? 6. And finally, are the N400 and P300 lateralized to the informed cerebral hemisphere in "split-brain" patients? What can we learn about the cognitive processes underlying the P300/N400 dichotomy by mapping and comparing their distributions in the normal and split-brain individuals?

DESCRIPTION (provided by applicant): This proposal is aimed at expanding our knowledge of the psychological and physiological bases of meaning construction from scalp recordings of event-related potentials (ERPs) to words. Current evidence suggests that the N400 indexes the combined influence of semantic context and long term memory on word processing. The proposed studies examine these relations and use the N400 to address specific questions about language comprehension. Expts 1-8 aim to reveal which aspects of context facilitate processing, and how, by examining interactions of different language cues and effects of constraint. Expts 1-2 examine whether lexically-based associative priming is maintained across clause and/or verb phrase boundaries. Expts 3-5 examine whether compositional semantic constraints (from quantifiers) can override lexically-driven processes. Expt 6 focuses on interactions between discourse-level constraints and background knowledge elicited by an individual sentence. Expts 7-8 focus on whether constraint effects operate only on lexical integration or also affect lexical/conceptual access per se and whether they arise at conceptual levels, lexical levels or both. Expts 9-16 focus on the interaction between the lexical information associated with verbs and people's knowledge of the temporal and causal structure of events. Expts 9-11 norm materials. Expts 12-16 examine the processing of verbs denoting events that are causally- and temporally-related and examine how verb aspect affects the activation of event-specific knowledge in long term memory. Expts 17-23 use visual half field presentation to examine how the two hemispheres make use of context in real time and how they do(n't) use various language cues to access concept and event knowledge from long-term memory. Expt 17 tests the two hemispheres' sensitivity to message-level constraint. Expts 18-19 examine the hypothesis that the two hemispheres use message-level information differently with the RH driven by plausibility & LH by prediction. Expts 20-23 explore the hemispheres' sensitivity to various language cues providing information about event-structure. Expts 20-21 use adjectival participles to determine whether verb inflections quickly provide knowledge about events (typical participants & their roles), and if the hemispheres differ in this. Expts 22-23 examine the hemispheres' sensitivity to event-location information in general and to the prepositions that cue such information in particular. Given our interest in real time language processing we also propose to collect eye movement data in several experiments, to get a better sense of when eye movement and ERP measures complement, overlap or are inconsistent with each other. Overall, the proposed studies are aimed at determining how we combine information from the linguistic stream and from semantic memory to make sense of language as it unfolds in real time.

A five-year research program is described which is aimed at clarifying the effect of normal aging on the organization, speed and efficacy of mental operations in language processing and memory. Event-related potentials (ERPs) will be recorded from the scalp of normal human subjects between the ages of 20 and 90 in response to linguistic materials and combined with behavioral measures. The overall goals are to establish a data base and to test hypotheses about the impact of normal aging on semantic and phonological analyses, encoding and retrieval aspects of memory for language materials, and speed of categorization in the visual and auditory modalities. We plan to approach these goals via the hypothesized relationship between the N400 component and semantic/lexical search mechanisms, and the P3 component and the duration of stimulus evaluation processes. The specific aims of the separate experiments are: (1) To use the relationship between reaction time and P3 latency for target detection in oddball paradigms of varying complexity to test the hypothesis that aging reflects a generalized slowing of CNS functioning as opposed to specific deficits for certain mental operations and materials. (2) To investigate semantic priming effects in three different tasks (lexical decision, pronunciation, delayed letter search) in the same group of subjects in order to determine whether aging effects are task independent. Moreover, to use these tasks to test the hypothesis that aging is more detrimental to attentional that automatic aspects of language processing. (3) To use tests of recognition and cued recall for congruous and incongruous words to determine the extent of information on initial presentation. (4) To examine whether aging has similar or different effects on semantic versus phonological processes. This research project will refine electrophysiological measures of important cognitive functions in relation to language processing and memory organization in normal adults throughout life. This normative data base should be of value for the future assessment of young and old individuals with disorders of language processing.

The Cognitive Science Department at UCSD is developing a new program of training with the goal of unifying research in brain science, behavioral studies, and computational theories and models. Although still in the early stages of development, the program already shows real promise. The advanced graduate students are producing insightful and publishable papers that cut across these disciplines. The program's three years of experience leads to confidence that the multidisciplinary approach to the study of cognition will significantly enrich our understanding of the biological and computational bases of cognition. These traineeships will enhance the educational experience of students by permitting them undistracted time and energy to attend to the rigorous coursework and experimental research required by this new discipline. The goal of this grant is to attract intellectually diverse and multiply-equipped students, with special attention to minorities, women, and others who have been under-represented.

This proposal is aimed at expanding our knowledge of the psychological and physiological bases of memory primarily from scalp recordings of event-related potentials (ERPs) to verbal and pictorial stimuli in a variety of implicit and explicit memory tasks. Current evidence suggests that a late positive component between 300 and 800 msec may index conscious recollection while several earlier components may be related to perceptual priming. We plan to test these working hypotheses. The resulting data should further our knowledge as to the independence, interdependence, and relative timing of the processes underlying perceptual priming, conceptual priming, and performance on explicit tests of memory. The proposed studies continue our investigations of word priming (perceptual and conceptual) and recognition memory using manipulations that differentially affect measures of one but not the other. We will also extend these questions to pictures. Several experiments will also utilize patients with memory disorders as we think that such a combined approach is necessary to fully understand memory mechanisms and to validate the proposed links between memory-related processes and their brain substrates as indexed by ERP measures. By altering the modality of presentation from study (auditory) to test (visual), Expt. 1 will investigate the hypothesis that an early occipital negativity elicited by words in a lexical decision task is an index of perceptual priming. Expt. 2 will examine the specificity of this effect by replicating Expt. l employing an auditory lexical decision task. Expt 3 will examine the modality independence of the recollection template by examining the ERPs elicited during an auditory lexical decision with interleaved recognition judgements following a levels-of-processing study manipulation. Expt. 4 would isolate the ERP components associated with perceptual priming in the word stem completion task by altering both the font and typecase from study to test. Expt. 5 attempts to tease apart perceptual and conceptual priming effects by examining ERPs to visual word stems during completion and cued recall following study in either the same (visual) or a different (auditory) modality. Expts. 6 & 7 utilize the method of opposition in a word stem completion paradigm to separate facilitation due to priming from that due to intentional retrieval. To that end, Expt. 6 will vary the amount of attention directed at words during study while Expt. 7 will present words too briefly to allow identification (Expt. 7). Expts. 8 & 9 will examine perceptual (for same token as study) and conceptual (for token different from study) ERP priming effects on picture fragments prior to their identification. Expt. 10-Il use amnesic patients and elderly controls to further validate our ERP indices of priming and memory tested explicitly.

A five-year research program is described which is aimed at clarifying the relationships among normal aging, executive (suppression, shifting of focus, consolidation) and non-executive (activation, storage, decay) working memory (WM) operations, and language comprehension. It has been suggested that along with age comes a reduced WM capacity which is manifest not only in a greater tendency to forget but also in an increased difficulty making sense of structurally complex sentences (comprehension), in switching from one idea or action to another quickly (attentional switching), and in ignoring and/or suppressing that which is irrelevant (inhibition). We detail a series of experiments using both performance and electrophysiological measures to test these proposals. Event-related brain potentials (ERPs) will be recorded from the scalp of normal human subjects between the ages of 20 and 90 as they process linguistic and non-linguistic materials. The proposed experiments include manipulations of the different subcomponents of WM and/or long term memory with the specific aims of: (1) extending our findings on the effects of aging on sentences with embedded relative clauses from reading to speech; (2) determining whether the elderly do indeed have problems with language comprehension because of the demands that sentence processing makes on consolidation processes in working memory; (3) testing the hypothesis that the elderly will differ from the young in anaphoric reference to the extent that pronoun resolution is implausible, thereby requiring inhibition of a prior expectation; (4) examining the hypothesis that inhibitory processes become slower and/or less efficient with advancing age by investigating negative priming with homophones in sentences and during a lexical decision task; (5) testing the hypothesis that the elderly have a harder time ignoring and inhibiting irrelevant stimulus attributes using a generalized "2-back" working memory paradigm with non-language materials as well; (6) testing the proposal that because of less efficient inhibitory mechanisms, older individuals will show larger "fan effects" than younger individuals in verifying probe items related to a series of propositions that they have been taught. This research project will refine electrophysiological markers of important cognitive functions in relation to language processing and organization of working and long-term memory in normal adults throughout life. These data can, in principle, offer guidelines on the best structure for language materials aimed at the elderly. Moreover, this normative data base should be of value for the assessment of young and old individuals with disorders of aging and language processing.

[unreadable] DESCRIPTION (provided by applicant): We propose renewal of our interdisciplinary training program in Language, Communication, and the Brain. The program emphasizes new technologies and new theoretical frameworks in cognitive science and neuroscience (e.g. advances in neural imaging, electrophysiological and behavioral studies of real-time language processing, computer simulations of language learning and breakdown). The proposed program pulls together the expertise, ideas, populations, and technologies that are available in abundance across this community, and places them at the disposal of young scientists interested in the mental and neural mechanisms that underlie language learning, language use, and language disorders. The program for the next five years will be headed by an Executive Committee of seven senior scientists, including the program director (E. Bates), and directors of six research and training components; (1) Communication Disorders (B. Wulfeck), including studies of adult aphasia and childhood language disorders; (2) Psycholinguistics (D. Swinney), including studies of real-time language and processing in adults, and studies of language learning in children; (3) Multilingual and Comparative Language Studies (M. Polinsky), a new component emphasizing studies of multilingualism (processing, learning, disorders), loss and relearning of "heritage" languages in immigrant populations, and comparative studies across typologically distinct language groups; (4) Neural Network Studies of language (J. Elman), with an emphasis on simulations of language learning and language breakdown under a range of different assumptions about the structure of the system and the context of learning and loss; (5) Event-Related Brain Potential (ERP) (M. Kutas); and (6) Functional Magnetic Resonance Imagining (M. Sereno), two components to capture two complementary neural imaging techniques, applied to normal children and adults, and to children and adults with neurological impairments and/or behaviorally defined communication disorders. All trainees specialize in (at least) two of the six areas (a major and a minor), and receive some exposure to all six areas through laboratory rotations, coursework, and activities within the UCSD Center for Research in Language. Courses and laboratory rotations are offered by a larger faculty of scientists at UCSD, San Diego State University and the Salk Institute for Biological Studies. We request eight predoctoral and three postdoctoral trainees per year. Predoctoral students apply through Cognitive Science, Linguistics, Psychology, or the SDSU/UCSD Joint Doctoral Program in Communication Disorders, and meet all requirements for those departments. Postdoctoral trainees apply directly to the Center for Research in Language. The proposed increase of three trainees (two predoctoral, one postdoctoral) is justified by growth in the last five years in three areas: the new UCSD Neural Imaging Center, the Center for Human Development (a new research unit to study neural to cultural aspects of human development), and new opportunities in multilingual/cross-language research. [unreadable] [unreadable]

DESCRIPTION (provided by applicant): A five year research program is proposed which is aimed at clarifying the extent to which the organization of semantic memory and semantic analyses, as used in language comprehension, are indeed relatively spared from the normal aging process, as is commonly believed. We assess this belief by examining effects of aging on different subprocesses; more specifically, by (1) comparing and contrasting processing of different modes of sensory input that refer to the same concepts--words and pictures; (2) examining two distinct contributors to semantic processing in language-the semantic integration of an item within its context and the various semantic constraints that accrue as context taps semantic memory to make sense of incoming stimuli prior to an item's appearance; (3) comparing the whole brain's response to word & picture processing when one hemisphere has a slight processing headstart; (4) examining individual differences in brain/behavioral measures in tasks tapping processing speed, inhibition, & working memory, an anti-saccade task accenting sensorimotor inhibition, a reading span task, and a sentence congruity reading rate task. To these ends we analyze event-related brain potentials (ERPs) recorded from the scalp of (brain) intact humans between 18-90 yrs of age and concomitant behavior as they process words or pictures. Twenty four experiments manipulate stimulus (visual, pictorial) and context modality (verbal, pictorial), the nature of relations among stimuli (categorical, antonymic), the probability that a particular item will be encountered (cloze probability), whether or not and, if so, how well an item fits within its context, how constrained and thus predictable an item is and at what level (lexical, semantic), visual field of input, and task, or combinations thereof. The specific aims are to determine whether or not and if so, how, normal aging changes what lexical, semantic, and contextual cues the two hemispheres are sensitive to and how each uses them to access semantic memory and to make sense of language. In so doing, we test (1) our hypothesis that though everyone benefits from context, most older adults, unlike younger ones, do not use it to predict features of upcoming stimuli, though greater verbal proficiency may compensate for this, (2) Cabeza's hypothesis that older adults are more likely to use both cerebral hemispheres to perform cognitive functions that were in their younger years the primary responsibility of just one, and (3) links between well-known ERP effects and age-related changes in processing speed and cognitive functions. This research project will refine ERP markers of important cognitive functions in language processing and memory use throughout the lifespan. This normative database will be of value for the assessment of adults of all ages with semantic/language processing disorders.