Comprehension test results from children acquiring Warlpiri as a first language in the Yuendumu community indicate that the canonical transitive sentence is one with an ergative-absolutive verb. We hypoth- esize that the acquisition process is influenced by a number of properties of the language, properties including the presence of homophonous forms for some of the case endings, the use of three different case frames, the optional ellipsis of core arguments, and the optionality of case markings in certain conditions.

In a between-subjects design, 105 children aged 2; 5–5; 5 enacted their interpretation of passive and inverted cleft (IC) sentences in which animacy was systematically varied. Two agent strategies were found at 3;0: dynamic agent and word order. For ICs, the word order strategy persisted at 4;0, but the dynamic agent approach was replaced by an animate agent strategy. For passives, 4-year-olds mostly used structural relations, either regardless of agent and patient animacy (50 % of children), or only for animate patients (30% of children). The latter finding is discussed in terms of topicalization priorities.

Subjects aged 3;6–9;0 were asked to judge sentences in which much and many modified prototypical and non-prototypical mass and count nouns, and to correct those sentences judged to be deviant. The experimental results indicate that children do not approach the co-occurrence conditions of much and many with various nouns from a semantic point of view, but rather from a morphosyntactic or surface-distributional one. Children learn the proper form that the noun must take in these constructions before they learn the proper choice of quantifier. In addition, they reserve many for use with plural count nouns long before they learn that much is restricted in direct noun modification to use with singular mass nouns.

H. Clark's hypotheses regarding the acquisition of spatial terms were tested using a task designed to elicit subjects' descriptions of different object arrangements. Children from 3 to 10 years and a group of adults were tested. There was some support for Clark's prediction that the perceptual difficulty of each dimension will affect the order of acquisition of the associated linguistic terms, in that the terms followed the predicted order (1) vertical, (2) horizontal-frontal, (3) horizontal-lateral, in the early years. However, it was concluded that since the horizontal-frontal terms lagged behind the others in later childhood and adulthood, perceptual difficulty alone was not an adequate explanation. In addition, no support was found for Clark's notion that linguistic terms associated with the ‘positive’ end of each spatial dimension will be acquired earlier than those associated with the ‘negative’ end.

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.

The interpretation of French statements involving the verbs savoir (‘to know’) and croire (‘to think’) was studied in a group of children aged from 4; 1 to 1. The experiment consisted in asking subjects to attribute four sentences (‘I know that P’, ‘I know that not-P’, ‘I do not know if P’, ‘I think that P’) to fictitious speakers (dolls placed in various situations of knowledge). Results enable us to infer levels of understanding, to locate the main difficulties in the processing of each statement, and to bring out recurrent confusions between the expressions. These results are interpreted as the expression of the processes contributing to the elaboration of ‘modalization’ in the child.

This study investigates the development of prepositions used in the expression of circumstances as adverbial adjuncts in the speech of three Spanish-speaking middle-class children. The study begins with the first appearance of one of these prepositions and the children are followed until age 4. In spite of the differences among the children concerning other aspects of their language development, what predominates in the present case is similarity. Both the forms and the meanings expressed turned up in much the same sequence in the speech of the three children. Yet when both form and meaning were seen together the results showed great variability; that is, each child began expressing a given circumstance by means of a different preposition. On the other hand, the circumstance most commonly expressed by means of new prepositions was place. The amount of polysemy increased with age, usually starting with a single meaning for each preposition. Expressions were found which seemed to represent intermediate stages between an early general category (co-occurrence) and the more specific semantic categories of space, instrument and company.

Forty monolingual Spanish-speaking Peruvian children ages 3; 6–5; 5 were given a search task to test comprehension of six locative phrases depicting the four spatial relations in front of, in back of, beside and under. Performance improved with age and was best for under (debajo de) and poorest on beside (al costado de, al lado de). There was no evidence that locatives lacking object part terms (delante de for in front of and al costado de for beside) resulted in poorer performance than locatives that included the object part labels (enfrente de, al lado de). Nor was there evidence to support the positive pole hypothesis that in front of items would be easier than behind. Performance was best when the children themselves were the reference object, next best with a fronted object, and poorest with a nonfronted object, suggesting that preschoolers may understand projective locatives with respect to the reference object's intrinsic orientation, rather than imposing their own point of view.

The comprehension of nine intonational contrasts involving different intonation-groupings, different nucleus placements, and different tones, was examined in 20 ten-year-old children and 20 adults. Fewer ten-year-olds than adults answered correctly on almost all tasks. Initial hypotheses concerning the order of acquisition among the three systems (grouping, nucleus placement, and tone) were in general not confirmed. But the distinction between context-wide and context-limited tonal meanings was predictive of acquisition; and a similar distinction based on local meanings rather than on the intonation patterns themselves is likely to apply to groupings and nucleus placements.

The ‘problem of the error-free progressive’ (Brown 5973) was reexamined. It was hypothesized that young children make errors with the progressive as part of a developmental process. An experiment designed to elicit present-tense verbs is described, using 36 pictures, illustrating 9 process and g stative verbs. Twenty-two children (4; 2 to 6; 6) were tested: all responded mainly in the present tense, and 10 out of 22 produced over-generalizations of the progressive to stative verbs. It is argued that resistance to the progressive is based on a semantic distinction that is modified subsequently in accordance with adult use.

On some views of language, when one uses a term to describe or label an object, a contrast is always implied. However, there is an important difference between nouns and adjectives in their contrastive use: adjectives imply a contrast between members of a single noun category, whereas nouns do not imply a contrast between members of a single adjective category. For example, describing a car as new suggests a contrast with other cars. However, labelling it as a car does not as clearly imply a contrast with other new objects. There is no logical necessity for this distinction – it is not captured by the literal meanings of words. Yet it is a conceptual distinction implicit in the use of these terms. In two studies, we tested preschool children's appreciation for this conceptual distinction, using both familiar and unfamiliar words. In Study I, 4-year-olds (but not 3-year-olds) were able to disambiguate the meanings of simple nouns and adjectives by assuming that the adjectives, unlike the nouns, implied a contrast between category members. In Study II, children selected pictures to match totally novel adjectives and nouns (e.g. the fep one, the skub). Relying on part of speech alone, they interpreted these unfamiliar nouns and adjectives differently. Implications for children's use of referential language and word-learning strategies are discussed.

The ability of Icelandic children (aged 4 to 8 years) to identify referents of gender-marked pronouns was tested under three conditions: (i) referents with natural/semantic gender, (ii) familiar referents with grammatical gender, (iii) unfamiliar referents with grammatical gender. Potential formal information about grammatical gender was held constant across all conditions. Children's correct identification of pronoun referents was most highly related to the familiarity of the referents and the availability of natural gender information, not to formal gender information. The results are discussed in light of contradictory findings from German. It is argued that subtle differences in the formal gender systems of.the two languages may underlie the different experimental results and may lead to different developmental patterns.

This study analysed the naturalistic productions of because and so by 96 children, vaged 3; 6–9; 6, vwhile vnarrating real, vpersonal events. Few semantic errors could be construed as evidence of confused thinking; none is a confusion vdescribed by Piaget. Of the vsemantically correct causal uses, 81 % encode psychological causality, mostly statements of other people's intentions. Other analyses revealed: (1) many of the relationships encoded are highly predictable, (2) virtually all causality occurred prior to the time of narration, (3) age trends are remarkably absent, (4) because and so are used in significantly different ways even by the youngest children, (5) only eight sentences showed a vreversal of the appropriate order of cause and effect. The discrepancy between the last finding and many other laboratory studies vwhich find many such reversals may be due to the fact that children are causally linking strictly successive events only 32 % of the time; for the rest, they are encoding events that partially or completely overlap in time.

Experiment I required first, third, and sixth graders to identify the actor or experiencer from active and passive sentences with actional and three subclasses of experiential verbs. Actional passives were understood better than experientials, with no difference among the subcategories of experiential verbs, and no effect of verb frequency or regularity. Response distributions for each verb type did not show two populations of subjects – one knowing and the other not knowing the passive – but approximately unimodal distributions whose mean increased with age. Experiment II studied preschoolers. Again, the actional passives were systematically easier than the experientials, and again, for neither verb type did the response distributions show two distinct subject populations, one competent and the other incompetent. The forms of the distributions cannot be wholly accounted for by fluctuating attention to cues to passivization, but indicate that many children have partial knowledge of the passive (increasing with age). A hypothesis about the nature of this partial knowledge accounts for the gradual acquisition and the difference between the verb types.

Third- and fifth-grade children, who had passed pretests of comprehension for words and sentence types used in the experiment, were tested on their ability to assign given/new roles in active, passive and cleft sentences controlled for stress. The experimental task involved a best-fit judgement that required the child to decide which one of two context sentences was the first part of a story continued by the target sentence. There was a significant difference in response to passive and cleft sentences, as opposed to actives. Despite a strong statistical tendency to perform correctly on passives and clefts, there were a substantial number of subjects who performed near chance in all sentence conditions. Five individual patterns of response to the task are identified. Strategies used by subjects in responding to the task, and the implications of the results for functionalist theories of language acquisition are discussed.