Gordon H. Bower - US grants

The research objective is to describe how memory operates in helping adults understand and remember descriptions of human actions and interactions. These abilities are central to social cognition and adaptation. Three interrelated perspectives on narrative understanding will be investigated: (1) the way readers or listeners understand a human drama by explaining the characters' actions in terms of their goals and plans, as these occur as part of narrative schema which require goals, complications, actions, and resolutions; (2) how readers use attribution rules to infer the motives and causes of actions from the narrative situation and the nature of the actor, with memory biased so as to serve the interests of the reader and the story-character he identifies with; and (3) the way readers assimilate and elaborate a text using their schematic knowledge of common situations and activity scripts. The unifying metaphor is that readers act like intuitive psychologists, trying to understand characters' actions according to their fitting into a plan to achieve a plausible goal, trying to understand the source of goals and relationships among conflicting goals and competing plans. We hypothesize that readers use their intuitive theory of intentional action to make strong goal-based predictions about what types of events will occur in a story. These predictions influence how readily particular narrative events will be understood. Also, narrative elements that fit into major constituents of the narrative schema, such as complication and resolution, will be judged as more important, will occur in summaries and be recalled more than schema-irrelevant material. We will also refine and test schema theory by examining the properties of new, arbitrary schemas we have people learn. This research combines interests of memory theory, natural language understanding systems, and social psychology.

This research is an attempt to apply a new kind of computational model, connectionist or parallel-distributed-processing models, to a number of complex, higher-level cognitive phenomena. These models are important, because a number of investigators claim that they are anatomically and physiologically more plausible than other sorts of models, and it is important to see how far they can go in accounting for the behavior of humans in the sorts of cognitive experiments that psychologists typically run. The first set of phenomena, involving attention and configural learning, includes the observation that over the course of discrimination learning, people come increasingly to "pay attention to" relevant cues, those that predict validly the correct response, and to ignore irrelevant cues. The second set of phenomena include generalization enhancement and distributed representations. The third set of phenomena involves representation issues, the issue of how different featural representations of stimuli should be mapped onto connectionist models. All of these issues will be approached with a combination of modeling and experimentation.

biomedical equipment purchase;

DESCRIPTION (Adapted from Applicant's Abstract): This research investigates a fundamental aspect of language understanding, namely, the way people construct and use mental models of situations being described in discourse. The mental model is like an internal "theater stage" that a reader or listener constructs in imagination to represent the locations, objects, characters, actions, and causal mechanisms described in a text. Readers/listeners use their developing model to interpret later statements in a discourse, to resolve ambiguities and references, to direct inferences, to learn new information, and understand how it fits into their prior knowledge. Imparting an accurate situational model is typically the main goal of instructional discourse as well as plot-centered narratives. When describing a mechanism, scene, or narrative, writers use various linguistic devices to foreground certain concepts and referents, thus placing them in the focus of attention within the reader's situational model. This attentional focusing importantly influences the way comprehension proceeds. Research has shown that memory representations of objects in focus within the mental model become highly activated, thus facilitating later references to them or retrieval of information about them. Furthermore, activation of memory objects occurs according to their proximity to the focus in the mental model. The proposed research investigates the properties and consequences of this focus of attention as it is moved around within a mental model. Shifting the focus when following characters' movements leaves a trail of activation on memory-objects along that path in the model, thus enabling rapid retrieval and reference to these objects. The proposed experiments will examine the analog, imagistic nature of the mental model, varying the properties of the spatial displays, the characters' movements, and the reader's perspective on them. The theoretical issues addressed by these experiments include how mental models are maintained, updated, and manipulated in people's working memory, whether and to what extent the models have spatial-imagery properties, and how they are constructed and oriented around a particular station-point or perspective.

The Purpose of this program is to train graduate students for research in cognition, with special attention to analysis of skills in perception, learning, memory, judgment, language, reasoning, problem solving, and decision making. Trainees will be predoctoral students working towards a Ph.D. in psychology. Through formal courses, specialized seminars, and apprenticeship research, students are trained in past and current work in experimental cognitive psychology, theoretical advances, and new developments in experimentation. The training program generally requires four years beyond the Bachelors degree, and follows the requirements of the Ph.D. program of the Psychology Department. From the beginning, students work closely with one or more faculty members, to select a set of courses and develop a research program best suited to their scientific interests and goals. In-depth survey courses in sub-areas of cognitive psychology are supplemented with state-of-the-art seminars and complemented by courses in related areas, such as neuroscience, linguistics, education, human factors, and artificial intelligence. Students also have opportunities for applied research in laboratories within the University and nearby research institutes with which we have close affiliations. Funds are requested for six (6) full time predoctoral trainees per year. Approximately 85 students apply each year for admission to the cognitive program; we normally enroll 5 or 6, selected for their outstanding promise for a productive career. Many of our graduates take academic positions in teaching and research but a significant number elect positions in applied settings, in research or development of products designed to supplement or increase human effectiveness in perception, learning, memory, reasoning, and decision making.