Information Used in Text Comprehension

If readers are to make sense of texts, they must develop mental models of the texts (Sanford & Garrod, 1998). In addition to comprehending novel words and complex syntax, they must use three kinds of information: content facts, content schemata, and text grammars (Kieras, 1985). Content facts are the simple propositions that are conveyed by the texts (e.g., facts about ants or facts about a character in a story). At this level, information does not have any superordinate organizational content. If students recognize the vocabulary words used to present the facts, they can comprehend the individual pieces of information. To gain meaning from the overall text, however, a student must have a content schema, or be able to organize a content schema from the facts presented in the text. A content schema represents a superordinate organization of a mass of possible content facts. For example, one can have a content schema for the social structure of ant or bee colonies, the metamorphosis process of caterpillars and tadpoles, or the activities at a birthday party. The speed of reading and comprehension of a text becomes easier when the reader possesses intuitive knowledge of the text grammar structure of a text (Kieras, 1985). A text grammar or macrostructure is a schema that represents a frequent organizational pattern of textual elements that is independent of specific content.

The role of schemata in text comprehension has been extensively studied (Anderson, 1994; Bartlett, 1932; Bransford, 1994; Kintsch, 1998; Rumelhart, 1980; Stein & Glenn, 1979; Van Dijk & Kintsch, 1983). Schemata are hierarchically organized sets of facts or information describing generalized knowledge about a text, an event, a scene, an object, or classes of objects (Mandler, 1984). (Note: Some authors use the term script to refer to an event schema—the stereotypical knowledge structures for common routines such as going to a restaurant, taking a subway, or going to a party [Beaugrande, 1980; Bower, Black, & Turner, 1979; Nelson, 1985; Schank & Abelson, 1977]. A script can be viewed as a specific type of schema.) Our schema knowledge enables us to behave appropriately in familiar situations, and when our schema information is applied to discourse (oral or written), it enables us to make the inferences necessary to comprehend the text—it enables us to read between the lines. If you have an elaborated schema or script for restaurants and you read the sentences, "John was hungry. He looked in the yellow pages," you would know that John may be intending to call a restaurant for reservations or to order a pizza—you would also know that he is not intending to eat the yellow pages. The ability to draw inferences is essential for critical and dynamic literacy. Although children who are poor comprehenders (despite adequate decoding skills) are less able than good readers to answer all types of questions about texts, they exhibit particular difficulty answering questions that require them to draw inferences (Oakhill & Yuill, 1996). In fact, when both good and poor comprehenders were able to refer to the text to answer questions, there was no difference between the good and poor comprehenders on literal questions. The availability of the text made little difference in the poor comprehenders' ability to answer the inferential questions. This deficit in inferencing may be related to lack of relevant schema knowledge, to difficulty in accessing relevant schema knowledge and integrating it with the text because of processing limitations, or to their being unaware that inferences are necessary.