The Influence of Knowledge-Based Inferences on The Reading Time of Expository Text

Abstract

This chapter examines inference generation during the comprehension of expository text. There are many subtypes of expository texts that span across many topic domains. The chapter focuses on passages that describe mechanisms found in nature or technology. To understand passages, the reader must access knowledge about concepts that are explicitly mentioned in the passage and must generate at least some knowledge-based inferences. The cognitive system stores world knowledge in the form of both generic knowledge structures (GKS) and specific knowledge structures. Nodes in the network are constructed on the bases of the explicit text, the GKS's, and specific knowledge structures. Three broad classes of knowledge-based inferences are associated with scientific mechanisms. They are states, future events, and past/concurrent events. The knowledge-based inferences do have an impact on reading time for expository text.