The proposed research is aimed at advancing our understanding of the processes involved in reading. We are particularly interested in inference, where we understand inference to include such relatively simple processes as finding the referent of an anaphor as well as the more complex processes involved in logical reasoning. We have proposed that inference involves activating the to-be-inferred information (whether it be information previously mentioned in the text or information in the reader's general knowledge), connecting the to-be-inferred information in relation to the information currently being processed, the storing the connected information into the memory representation of the text. We believe that both forward and backward inferences are described by the processes of activation, connection, and storage, and that these processes require time for execution. We propose experiments to investigate each of these aspects of three kinds of inference: inferences about reference, inferences about the meanings of concepts, and inferences about the meanings of events. The task employed in many of the proposed experiments is item recognition. In previous research, we have shown that the amount of priming in item recognition can be used to measure the relative activation of concepts and the relative distances between concepts in memory. The experiments are designed to increase our knowledge of the basic processes of reading. While most of the proposed experiments will use college students as subjects, several experiments are designed to investigate developmental issues with second, fourth, and sixth grade subjects. Specifically, the experiments will investigate development of abilities to organize information in text, to understand anaphoric references, and to understand contextually determined meanings.
Showing the most recent 10 out of 14 publications
