This study investigated the development of knowledge about the presuppositions of cognitive verbs that take sentential complements. The verbs included factives, which presuppose the truth of their complements, and nonfactives, which carry no such presupposition. Three tasks assessed children's ability to (a) assign truth values to complements according to the presuppositions of the main verbs; (b) select verbs to describe people's mental states; and (c) state the presuppositions of the verbs in definitions. The results indicated that the presuppositions of the factives know, forget, and remember and the nonfactive think are not learned until age 4. Believe, which has factive and nonfactive properties, is mastered after age 7. The children's performance differed across tasks due to variations in processing requirements.

Samples of the communicative behavior of a group of higher-level mentally retarded adults engaged in conversation with peers were examined for indications of mature linguistic competence, specifically, grammatical morpheme and complex sentence use. The findings confirmed the expectation that the eventual level of mastery of these aspects of linguistic competence in higher-level retarded individuals is relatively high. Evidence for a normal developmental progression in the mastery of the grammatical morphemes was also forthcoming. In an analysis of individual complex sentence structures, no relationship was found between relative frequency of use of different types of complex sentences and presumed order of acquisition. However, subjects' ability to combine complex sentences did appear to be related to the presumed order of acquisition, although other factors may have also contributed to this relationship. Unexpectedly, a significant negative correlation was observed between relative frequency of complex sentence use and an estimate of conversational communicative competence. A possible reason for this finding was discussed.On the basis of recent reviews (Rosenberg, 1982(Rosenberg, , 1984, it appears that language (phonological, syntactic, morphological, semantic) development in the retarded begins later, proceeds at a slower rate, and reaches a lower level of competence than it does in CA-matched nonretarded individuals, although the course of language development and the linguistic system acquired are similar in these two populations. Moreover, even when matched on MA, the retarded still tend to perform at a lower level than the nonretarded on certain aspects of language, in particular, grammatical morphology and complex sentences. However, conversational communicative competence (e.g., the ability to take turns, to recognize speech-act obligations) in the retarded, at least when such competence is assessed in situations where cognitive demands can be assumed to be minimal, appears to exceed their mastery of aspects of linguistic knowledge. Indeed, Abbeduto and Rosenberg (1980), on the basis of the results of a study of a

This study examined the communicative behavior of mildly retarded adults engaged in conversation with peers. Contemporary models of pragmatic interaction were applied to samples of triadic naturally occurring conversation. The subjects made few errors in turn-taking; moreover, the rule system involved was consistent with that posited for nonretarded adults. They recognized those illocutionary acts that obligated them to respond as well as the specific responses required. That they were actively involved in information exchange was indicated by the observation that the majority of their turns were responses to the preceding turns of others, even when under no obligation to respond. Individual differences on most measures were observed, and the measures appeared related to each other, but not to IQ. Unexpectedly, the subjects produced few indirect speech acts.

The purpose of this study was to test how information-processing load affects the writing process (through thinking-aloud reports) and the story written. Information-processing load was increased by having subjects write to an ending sentence with more content constraints. Secondary reaction times were synchronized with thinking-aloud statements to yield a measure of cognitive effort for the components of the writing process and for the overall task. A high information-processing load led to lower rated story coherence, but not to lower rated quality. A high load did not increase overall cognitive effort, but changed the distribution of processing time, with more reviewing earlier. Results suggest that a high information-processing load altered the distribution of writing processes, which resulted in lower story coherence.

This study tested the hypothesis that the recall of verbal material (critical material) accompanying semantically well-integrated (SWI) sentences will be superior to the recall of verbal material accompanying semantically poorly integrated (SPI) sentences. Complex sentences were constructed which contained two underlying sentences: a matrix sentence and an embedded sentence. Under the SWI condition, one underlying sentence was an SWI string, while under the SPI condition one was an SPI string. The critical material (identical for both levels of semantic integration) was contained in the second underlying sentence. The location of the critical material (i.e., whether it was the matrix or the embedded underlying sentence) was varied. A list-learning study-test procedure was used with 5 trials. The results indicated superior recall for the critical material under the SWI condition, and were interpreted in terms of a storage hypothesis.