Six children were visited in their homes every two weeks for 18 months from the age of six months in order to observe their developing comprehension and production of words. Controlled testing of the children's developing vocabulary was also carried out to provide a more detailed picture of the precise context in which a word was understood or produced. The results showed both similarities and individual differences among the children in patterns of early comprehension. For all six children there was a very close relationship between early production and comprehension of words: words that were contextually flexible in production tended also to be so in comprehension and words that were context-bound also tended to be so in both modalities although there was some evidence that the two types of word differed in the extent to which comprehension preceded production. All children were also similar in showing evidence of contextually flexible comprehension very early on. However, there were individual differences both in the rate at which comprehension developed and in the lag between comprehension and production. The significance of this pattern of development for accounts of lexical development is explored.

This study investigated overextensions in comprehension and production using a new method (the preferential-looking paradigm) in which children (N = 99, mean age (younger) = 1;9, mean age (older) = 2;3) were asked to find the referent that matched the label they were given. Both Real Referent (in which there was a match) and Anomalous (in which there was no match) trials were included, as well as nonverbal control trials. During the Real Referent trials, all children significantly preferred the matching puppet. During the Anomalous trials, children showed no preference with two of the labels (dog and cat); however, they did show a preference when ‘cow’ was requested but not available. There were no differences based on prior overextension performance in production. It is concluded that overextensions in production are not diagnostic of children's underlying semantic representations, and that anomalous trials in comprehension provide useful information concerning young children's lexical entries.

Theoretical arguments for considering production as a source of input for analysis (the OUTPUT-AS-INPUT hypothesis) are reviewed, and empirical evidence supporting this hypothesis is presented. The evidence consists of a longitudinal study of the developmental course of a self-created form, produced by one Dutch child. This form is the product of blending two words, wat and iets, to yield unitary wat-iets. In Dutch, independent wat and iets may each mean ‘some’ and/or ‘something’. Though wat-iets is not permitted by the language, and so does not occur in environmental input, the form stays in the child's repertoire for about 10 months (between 3;8 and 4;7) and is apparently subjected to processes of generalization: first the child treats wat-iets as a two-word frame that may be regularized, later as a unitary word that may be semantically extended. After the extension of wat-iets, independent synonymous wat and iets appear for the first time in the child's speech. It is argued that the child actually analysed his own creation.

Twelve children (five boys and seven girls, mean age 4;1) interacted with their mothers in two conversational settings: playing with a favourite toy and helping to prepare lunch or muffins. The percentage of two topic continuation processes – topic incorporation and collaboration – used by mothers and children was analysed. It was hypothesized that the help setting with a predetermined structure would elicit a greater percentage of topic incorporation by the dyads than the play setting. In fact, the play setting elicited a greater percentage of topic incorporation from mothers than the help setting. Children did not differ significantly in the percentage of use of topic incorporation between the two settings, but appeared to be equally proficient in incorporating topics in settings with and without a predetermined structure. The play setting also elicited more topic collaboration from the children than the help setting. While the help setting appeared to support collaborative action, the play setting permitted the dyads to be informative about the toys as well as related topics. Therefore, the two settings provided different opportunities for the dyads to continue topics.

The present studies investigated children's use of verb inflections to guide their initial mapping of verb meanings. Given that children initially apply the progressive -ing inflection to verbs denoting actions and the past -ed inflection to verbs denoting results of events, two studies were conducted to investigate whether children use these inflections during mapping of novel verb meanings. In both studies, subjects were taught novel verbs and were asked to extend those verbs to events in which the action or result differed from events used to teach the verbs. It was predicted that subjects would be less likely to extend verbs inflected with -ing to events with new actions and would be less likely to extend verbs inflected with -ed to events with new results. Eighteeen three- and five-year-olds and 24 adults participated in Experiment 1 in a between-subjects design that produced weak effects for the youngest subjects tested. Experiment 2 tested 16 three-year-olds and 19 five-year-olds in a within-subjects design and produced the predicted effect for three-year-olds, but not for five-year-olds. The findings are discussed in terms of their implications for early verb learning and regarding the use of the bootstrapping construct in language acquisition research.

Previous tasks have shown that preschool and early elementary school-children typically have trouble learning and identifying homonyms (Peters & Zaidel, 1980; Mazzocco, 1989). It is possible that a one-to-one mapping assumption or a lack of metalinguistic skills makes homonym learning and identification particularly difficult. In three experiments we examined a total of 60 three-year-olds' ability to pick out homonym pairs, and the extent to which they realize that although homonyms share a common label, they represent two different categories. In Experiment 1 subjects were asked to identify homonym pairs. In Experiment 2, homonym pairs and non-homonym pairs were labelled, then children were asked whether the pairs had the same name, and whether they were the same kind of thing. In Experiment 3 children were shown one-half of each of several homonym and non-homonym pairs, then asked to identify a name match and a category match from a set of pictures. From these experiments we conclude that children have the metalinguistic skills necessary to identify homonym pairs; moreover, they realized that homonyms represent two different categories. Finally, if children have a one-to-one mapping assumption, it is not strong enough to prevent them from acquiring homonyms.

Thirty-six three-, four- and five-year-olds were asked to select labels for deceptive stimuli (e.g. for an eraser that looked like a pencil). Three types of labelling were investigated –simple (e.g. ‘is an eraser’); appearance-predicated (e.g. ‘looks like an eraser’); and reality-predicated (e.g. ‘is really and truly an eraser’). An age-related appearance-reality shift was observed in simple labelling (e.g. older children were more likely than younger ones to accept eraser and reject pencil as simple names for the pencil-eraser). This trend was robust over method and semantic domain, though weaker with object than with colour labels. As in previous research, older children were more likely than younger ones to map different appearance- than reality-predicated labels onto an item (e.g. to accept that the pencil-eraser looks like a pencil, but is really and truly an eraser); however, all age groups were reluctant to extend more than one name to a stimulus via a common predicate (e.g. to accept two reality-predicated labels for the same object). This one-label-per-predicate pattern was observed more frequently within reality than within appearance predicates; more frequently with colour than with object names, and with questions blocked by predicate than by name. It is argued that younger children maintained this pattern because of inflexible encoding, but that older ones did so because of better understanding of the appearance-reality distinction, greater reality dominance, and a Mutual Exclusivity bias.

The ability to convey the optimal amount of information during conversation is a fundamental aspect of language use. In this study the relationship between children's failures to produce unambiguous utterances and the mental effort demands of the communication task was investigated. Five-, six-, seven- and nine-year-old children performed a message production task and a finger-tapping task both separately and simultaneously. The decrease in finger-tapping frequency during the simultaneous performance was used as an estimate of effort demands of the message production task. Working memory capacity was assessed by means of a spatial memory test and an object features identification task. Children's intuitions about message adequacy were recorded in two message evaluation tasks. By age six children proved to be able to select the relevant information when they were explicitly asked to do so, indicating that effort demands of the communication task did not exceed their computational resources. However, results suggested that the relative effort requirements of the communication task decrease with increasing age. These findings support a performance theory of communication development in which effort demands are a determinant of children's message adequacy.

This study examines the role of cognitive development in children's use of pragmatic cues for anaphora resolution. Reported speech sentences like Minnie told Dorothy that she knew Superman are biased toward the matrix subject. This bias is claimed to depend on two conceptual shifts, first to the speaker's and then to the listener's perspective. 141 children aged 5;0–8;0 performed two tasks with biased and neutral sentences. In the Verbal task, they gave antecedent choices in response to a question (e.g.… that WHO knew Superman?). In the Puppet task, which prompts the perspective shift, they made a puppet say the reported speech portion (e.g. I/you know Superman). Violations of the pragmatic constraint decreased with age and task, consistent with the perspective-shift model. Parallel function effects in neutral sentences were weaker than in previous research on conjoined sentences, but similar to recent results for adults with these materials.

Previous work bas demonstrated that children understand sentences with actional verbs better than nonactional verbs. This ACTIONALITY EFFECT bas been reportee to be restricted to passives and to be independent of experimental context. The present experiment was conducted with 48 French-speaking children aged 5;0–7;11. The actionality effect was studied by systematically varying the voice of the test sentences and the voice of the interpretive requests. Pictures corresponding or not to the predicate—argument structure of the sentences were presented to the subjects, who were independently classified as visualizers or nonvisualizers, in order to investigate the relation between sentence actionality and mental imagery. The interaction between actionality, voice of sentence, and interpretive request revealed that the actionality effect depends on the type of task used in order to assess comprehension, and that it can be reversed in some conditions. Our results also suggest that the actionality effect is linked to mental imagery. Visualizers demonstrated better comprehension of actional sentences than nonvisualizers, whereas the reverse was true for nonactional sentences. Mental image may serve as a support for the computations involved in sentence comprehension.

The study examined whether the use of superordinate terms in children's definitions was predicted by relevance theory. Two hundred and six children aged five to ten years gave definitions for 16 basic-level words and four superordinate words from natural kind and artefact semantic domains. Superordinate terms were used more frequently when they supported more inferences. This was evidenced by their more frequent use in natural kind than in artefact domains, and more frequent use when the superordinate was itself defined by a semantically complex expression. When used, superordinates also usually occurred at the beginning of the definitional expression. It is argued that these findings reflect the speaker's intention to achieve optimal relevance, that is, to achieve maximum contextual effects with the least processing effort.

Fernald and her colleagues have argued for the universality of prosodic modifications in speech addressed to infants (e.g. Fernald, 1987; Fernald, Taeschner, Dunn, Papousek, de Boysson-Bardies & Fukui, 1989). An alternative proposal in Bernstein-Ratner & Pye (1984) and Pye (1986) argues that speech modifications to children are the result of a set of conventions that may vary from culture to culture. This note presents arguments against Fernald and in support of the cultural account of prosodic modifications to infants and children. First, in response to criticisms in Fernald et al., data are presented which defend the conclusion in Bernstein-Ratner & Pye that Quiché mothers show less prosodic modification to children than do mothers in other cultures studied to date. Second, data used to support the universal position in Fernald (1987) on adult identification of infant affective states are reinterpreted and shown to be equally interpretable as the result of a set of culturally transmitted rules.