The acquisition of connective forms and the meaning relations between connected clauses in the development of complex sentences is described for four children from two to three years of age. The major results of the study include the developmental interactions between syntactic connectives and meaning relations, and between these interactions and the discourse environments in which they occurred. The first syntactic connective the children learned, and, was the most general: semantically, and was used to encode conjunction with all of the different conjunction meaning relations in the order Additive < Temporal < Causal < Adversative. Other connectives were semantically more specific, and were learned subsequently with different syntactic structures in the order Conjunction < Complementation < Relativization. These results are discussed in terms of FORM, relative linguistic complexity; CONTENT, the intersection of form with conceptual and semantic factors affecting acquisition; and USE, discourse cohesion.

A Reichenbach-type theory is proposed that accounts for a child's early system of temporal reference and relates it directly to the adult system. This theory makes predictions different from those of a competing theory, a strong decentring hypothesis. Experimental results and data from spontaneous speech are examined, and the predictions of the first theory are borne out: young English-speaking children express notions of temporal ordering as well as aspectual notions.

Natural speech of 32 young children from 2; 0 to 3; 1 (1·97 to 6·38 in MLU) was studied for evidence of use of coordination (by and). Analyses of the structure of coordinative expressions reflected development in structural variants as well as amount of coordination. Analyses suggested that early forms of coordination in natural speech reflected the same constraints which had been suggested by previous study of these children's language through an experimental elicited imitation task. A basis for a developmental theory of acquisition of coordination is sketched.

This paper evaluates Lust's (1977) study of coordination in child language, both on its own merits and in the light of my own research. First, particular design weaknesses in Lust's study are noted which render her results questionable. In addition, it is argued that the hypotheses proposed by Lust fail to make any substantive predictions about the types of errors which children make with coordinate structures. Finally, it is shown that Lust's hypotheses cannot account for the results of my own comprehension and non-imitative production experiments. It is concluded that Lust cannot account for a wide range of phenomena concerning coordination in child language, and consequently an alternative set of hypotheses is proposed.

This research explored overextension in the early vocabularies of six children, followed in a language diary study from 1; 0 to 1; 8. Results indicated that only one-third of the first 75 words acquired by each child were ever overextended. A small set of high-frequency, early acquired words accounted for a disproportionate number of overextensions. Overextensions were classified into three types: categorical overinclusions, analogical overextensions and predicate statements. Four types of information served as the bases for word applications: perceptual, action-functional, affective and contextual. The use of words to denote associative complexes of a well-organized, systematic character was discussed as a characteristic form of early word usage.

The speech of four two-year-old children growing up bilingually in a German-speaking community was studied for periods varying between five and nine months. An analysis of their language mixing revealed an initially higher rate of mixing which diminished with a growth in language development as measured in MLU. The data suggest that the children were at various stages in a gradual process of language differentiation thus providing support for the one-system theory of bilingual acquisition. An examination of the distribution of lexical substitutions by part of speech revealed that nouns were most frequently substituted by all children; however, more function words were substituted than content words overall.

Five French-speaking middle-class couples and their male only-children were tape-recorded separately at home while interacting verbally in a free-play, a story-telling, and in a family meal situation. The children's ages ranged from 1; 6 to 3; 0. The speech of the fathers, mothers, and children was transcribed and analysed for its semantic, syntactic, and language-teaching aspects. The results indicate that paternal speech displays the same simplification processes usually found in maternal speech to young children. Paternal speech, however, also contains some linguistic patterns at variance with those found in maternal speech. It is hypothesised that maternal and paternal speech may be complementary in their influence on language development in the children.

Fifty-two children aged 2; 5 to 6; 3 acted out the meaning of reversible passive and inverted cleft sentences. The younger children overgeneralized two strategies that led to passive sentence reversal. In order of frequency, they are: (1)‘agent + action’ (the noun preceding the verb being interpreted as agent even when it is the second noun), and (2) ‘order-of-mention’ (the first noun being treated as agent). Whereas younger children tended to use an individually consistent approach to both structures, older children were more likely to vary their approaches when they did not comprehend a structure.

This paper presents a study of children's ability to recognize the presuppositions of sentences with factive and nonfactive verbs, and a review of recent research on the acquisition of factive presuppositions. In the present study, subjects in grades K, 2, 5 and 8, and adults, were asked to judge the truth of the complement clauses of sentences of the form ‘Doctor Fact (verb) that the ball is (is not) blue.’ The task was presented in the context of a television quiz show, with subjects as putative contestants. A tendency was found among younger subjects to overgeneralize the scope of negation of complement verbs into the main clauses of factive sentences. A gradual increase in comprehension of factive and nonfactive verbs was found, extending into adolescence. In a comparison of the present results with those of several recent studies, marked differences were found in the age of acquisition of factivity. These differences are discussed in terms of variations in task requirements and assessment techniques. It is concluded that acquisition proceeds on a verb-by-verb basis, rather than as a unitary process.

It is argued that previous assessments of children's knowledge of the hard to see type of construction were confounded by a variety of extra-linguistic factors. Therefore, the relatively delayed age of comprehension previously reported (6½–8 years) may have been due to younger children's deficiencies in extralinguistic skills. In the present study, with these extralinguistic complications eliminated, the passing age was found to be 5 years, and even 4-year-olds evidenced considerable knowledge of the target structure. Other findings were: variation in sentence difficulty as a function of the syntactic and/or aspectual character of the verb; high test–retest reliability at all levels of performance; and a necessary-but-not-sufficient empirical relation between comprehension of the target construction and the passive. Results are discussed in terms of their implications for the acquisition of this particular structure and for the general problem of detecting linguistic competence from performance.

The ‘fis phenomenon’ and similar anecdotal evidence, which appears to give insight into child phonology, and also certain experiments designed to reveal the nature of homonymy in child language, are discussed with reference to an approach which distinguishes between eight different types of possible phonetic episode. It is concluded that the anecdotes and experiments are less instructive than is often supposed.

We regret that the review of Tatiana Slama-Cazacu, Dialogue in children, published in Vol. 6, No. 2 (June 1979), pp. 394–6, failed to point out that the original date of publication of this work was 1961. The date given, 1977, was that of the English translation. We apologize to Professor Slama-Cazacu for any false impression given by this editorial ommission.