What supplies the concepts in causal inferences in story comprehension? This question was examined in 5 experiments. Elementary school children and college students listened to stories containing a "premise" sentence describing a character's intent in initiating a series of actions, followed by an unexpected "outcome." After each story, the subjects were asked inference questions about the reason for the outcome (Experiments 1-4) or asked to explain the outcome (Experiment 5). In the various experiments, the availability of object concepts for an inference was manipulated by providing 0 or 2 clues preceding the outcome, additional filler information that diluted the clue information, a mention of an object in the premise or no mention, and an object title for each story or no title. The results suggested that the availability of an object concept in sentences prior to the outcome was a critical determinant of whether an object inference occurred, especially for children. Given availability, the thematic prominence of the object influences whether the object will be used in an inference.

Common ground is contextual information shared by a listener and speaker that enables the listener to convert an ambiguous utterance to an informative communication. Four experiments examined young children's understanding of the common ground in interpreting ambiguous referential utterances. Kindergarten and 2nd-grade children and college students were read short vignettes containing statement, joint activity, status common ground, and an ambiguous or informative utterance about a display of 4-6 object drawings. The subjects were asked (a) whether the listener knew which object to pick (Experiment 1), (b) to pick an object themselves or choose "none" (Experiment 2), (c) the source of the listener's knowledge in the context or utterance (Experiment 3), and (d) whether a designated object was the "right one," the one the speaker "meant," or one the speaker "could have meant" (Experiment 4). Even the kindergarten children used statement information effectively in interpreting ambiguous utterances, and all groups had difficulty using status information.For most developmental researchers, the most important acquisition in the development of referential communication skills concerns informative utterances. An informative utterance is a statement that uniquely identifies an object in a referential field of objects. The utterance is informative because the object can be identified on the basis of the words in the utterance alone, without reference to the context of utterance.The results of many studies have established that children younger than 7 or 8 years of age often do not appreciate the properties of such utterances as speakers (Sonnenschein & Whitehurst, 1984) or listeners (Ackerman, 1981;Robinson & Whittaker, 1985;Sonnenschein, 1988). As speakers, for example, the referential utterances of younger children often are ambiguous in that the utterance refers to two or more objects in a perceptual array. Similarly, as listeners, younger children often do not distinguish between ambiguous and informative utterances in picking a referential object. Furthermore, if the children do identify the appropriate object on the basis of other information in the context, the children usually are unable to distinguish between the source of information in the utterance per se or that in the context (Beal & Flavell, 1984;Robinson, Goelman, & Olson, 1983). The theoretical importance attached to the adequacy of the information in the utterance alone makes the informative utterance the canonical referential speech act (cf. Clark & Carlson, 1982) for most developmental researchers.However, in many referential situations, speakers may use ambiguous utterances to accomplish communicative goals. For an ambiguous utterance, the information in the utterance aloneWe thank the children and staffs of Kemblesville, Wilson, and Pulaski Elementary Schools. We also thank three anonymous reviewers for their excellent comments, and the undergraduates who contributed to the research.

What supplies the concepts in causal inferences in story comprehension? This question was examined in 5 experiments. Elementary school children and college students listened to stories containing a "premise" sentence describing a character's intent in initiating a series of actions, followed by an unexpected "outcome." After each story, the subjects were asked inference questions about the reason for the outcome (Experiments 1-4) or asked to explain the outcome (Experiment 5). In the various experiments, the availability of object concepts for an inference was manipulated by providing 0 or 2 clues preceding the outcome, additional filler information that diluted the clue information, a mention of an object in the premise or no mention, and an object title for each story or no title. The results suggested that the availability of an object concept in sentences prior to the outcome was a critical determinant of whether an object inference occurred, especially for children. Given availability, the thematic prominence of the object influences whether the object will be used in an inference.

No abstract

No abstract