Divergent hypotheses exist concerning the types of knowledge underlying early bilingualism, with some portraying a troubled course marred by language delays and confusion, and others portraying one that is largely unremarkable. We studied the extraordinary case of bilingual acquisition across two modalities to examine these hypotheses. Three children acquiring Langues des Signes Québécoise and French, and three children acquiring French and English (ages at onset approximately 1;0, 2;6 and 3;6 per group) were videotaped regularly over one year while we empirically manipulated novel and familiar speakers of each child's two languages. The results revealed that both groups achieved their early linguistic milestones in each of their languages at the same time (and similarly to monolinguals), produced a substantial number of semantically corresponding words in each of their two languages from their very first words or signs (translation equivalents), and demonstrated sensitivity to the interlocutor's language by altering their language choices. Children did mix their languages to varying degrees, and some persisted in using a language that was not the primary language of the addressee, but the propensity to do both was directly related to their parents' mixing rates, in combination with their own developing language preference. The signing-speaking bilinguals did exploit the modality possibilities, and they did simultaneously mix their signs and speech, but in semantically principled and highly constrained ways. It is concluded that the capacity to differentiate between two languages is well in place prior to first words, and it is hypothesized that this capacity may result from biological mechanisms that permit the discovery of early phonological representations. Reasons why paradoxical views of bilingual acquisition have persisted are also offered.

This study investigated the use of ‘mama’ or similar sounds (collectively referred to as ‘MAMA’) by 75 infants less than six months of age. Parents were directed to listen for ‘MAMA’ sounds and to note the sounds made, the age of onset, whether the sounds appeared to be directed to any person or persons and whether they appeared to have a purpose. ‘MAMA’ began at a mode of two months, range two weeks to five months, was usually part of a cry, and was always interpreted as a ‘wanting’ sound. Most parents thought that the infant wanted some form of attention, but a minority thought it indicated hunger. Responses to a Structured Response Protocol indicated that some infants uttering ‘MAMA’ were satisfied if a favourite caretaker approached and paid attention to them while the remainder were satisfied if they were both paid attention to and picked up. The ‘MAMA’ cry appears to promote attention-giving behaviour by parents and other caretakers.

Descriptions of the development of prosodic and segmental tiers of children's phonological systems have been derived from investigations of the development of English. This paper provides a preliminary description of phonological tier development in Cantonese-speaking children. Eight children, (two each at 1;7, 2;6, 3;5, and 4;2 years) named 95 pictures. The data were analysed for word, syllable, onset-rime, skeletal, and segmental tiers. The results suggested a developmental order in the acquisition of hierarchical features. Decreasing order of accuracy of the tiers was word = syllable > onset-rime = skeletal > segmental. A model of feature geometry was adopted to describe the acquisition of features. An interesting finding is the way the laryngeal feature (aspiration) was combined with place contrasts one at a time rather than all at once.

This study had two major objectives: (1) to analyse the development of two morphological structures in Hebrew, one inflectional and the other derivational and (2) to examine the mutual contribution of morphological knowledge and learning the written code. In a longitudinal design, 40 children were tested twice, first in kindergarten (mean age: 5; 11) and again in first grade (mean age: 6;5), on two oral tasks – inflecting nouns for possession and deriving denominal adjectives – and one written task of writing a series of noun-adjective pairs. The derivational task was found to be harder than the inflectional task, both on the stem and the suffix level, attributable to its higher semantic opacity. In both oral tests, correct stem production when suffixed was related to the morphophonological level of stem change. Correlations were found between morphological and writing scores. Moreover, children who were more advanced in morphology in kindergarten progressed more in writing vowels from kindergarten to first grade, and those who were more advanced in writing in kindergarten improved more in derivational morphology with grade.

This paper examines naturalistic adult-to-child speech produced by 15 Italian middle-class mothers to determine which specific patterns characterize linguistic input to children at 1;4 and 1;8. Since Italian is a PRO-DROP language, we expect that adult-to-child speech will show a bias towards a more salient semantic and morphological significance of verbs relative to nouns. We expect that verbs will more likely occupy the sentence-initial position, and have more morphological inflections relative to nouns. Mother-to-child speech was coded for type and token frequency, utterance position, and morphological variation of nouns and verbs. The results confirm our predictions. Namely, Italian-speaking mothers produced verb types and tokens more frequently than noun types and tokens, they placed verbs more frequently than nouns in salient utterance position, and they morphologically marked verb stems more than noun stems.

Rispoli (1999) suggests that previous studies arguing for a contingency between the case of subject pronouns and the presence/absence of verbal agreement in the acquisition of English (e.g. Schütze, 1997) suffer from methodological problems, and presents new data that fail to support earlier findings. I show that Rispoli's methodology unnecessarily biases his study against finding the predicted contingencies: it fails to take account of children's productive lexical inventory of pronoun forms. As a result, syntactic versus morphological sources of error fail to be distinguished. I explain why this distinction is crucial within the AGR/Tense Omission Model, and clarify its predictions.

Astington (1988) found that seven- to nine-year-olds often fail to distinguish between promises and predictions when judging the utterances of characters in simple stories. Instead, these children attend only to the outcome of the story (i.e. whether the promised event occurred) when deciding whether a promise has been made and, to a lesser extent, when deciding whether the speaker is responsible for the outcome. The purpose of the present study was to examine whether seven- to nine-year-olds (a) vary their judgements of responsibility according to the reason that the promised action was not completed, and (b) recognize that an unfulfilled promise is a promise regardless of whether the speaker's failure is unavoidable or intentional. Seven-year-olds, nine-year-olds, and adults were asked to make promise and responsibility judgements for two story types: stories in which the promiser intentionally failed to fulfil his or her promise and stories in which an unforeseen event prevented the promiser from fulfilling the promise. Participants at all ages assigned responsibility correctly across both story types. In making promise judgements, however, the seven-year-olds' decisions about promises reflected a misguided attention to the outcome of a promise or the obstacle to its fulfilment. The nine-year-olds recognized that an unfulfilled promise is a promise but only when there was a clear reason for the speaker's failure to fulfil his or her obligation. We suggest that children consider only sincere promises to be instances of promising and make inferences about speaker sincerity by looking to external factors in the communicative context.

Two experiments investigated why preschool children sometimes produce multiple words for a referent (i.e. POLYNOMY), but other times seem to allow only one word. In Experiment 1, 40 three- and four-year-olds completed a modification of Deák & Maratsos' (1998) naming task. Although social demands to produce multiple words were reduced, children produced, on average, more than two words per object. Number of words produced was predicted by receptive vocabulary. Lexical insight (i.e. knowing that a word refers to function or appearance) and metalexical beliefs (i.e. that a hypothetical referent has one label, or more than one) were not preconditions of polynomy. Polynomy was independent of bias to map novel words to unfamiliar referents. In Experiment 2, 40 three- and four-year-olds learned new words for nameable objects. Children showed a correction effect, yet produced more than two words per object. Children do not have a generalized one-word-per-object bias, even during word learning. Other explanations (e.g. contextual restriction of lexical access) are discussed.

Children's ability to distinguish between the text, WHAT WAS SAID, and the intentional structure, WHAT WAS MEANT, was interrogated by means of verbatim and paraphrase questions in two types of discourse, narratives and nursery rhymes. Three- to seven-year-olds participated (n = 119, mean age 5·1). There was an interaction between the type of discourse and the younger children's ability to separate wording from intentional structure. In the narrative form they had difficulty rejecting true paraphrases when asked to focus on wording, while in the nursery rhyme form the difficulty was accepting a true paraphrase when asked to focus on intention.

Although it has long been observed that some children incorporate unglossable syllables into their early utterances, it has been difficult to integrate these ‘fillers’ into theories of language acquisition. Because they straddle boundaries between phonology and morphosyntax, and between pragmatics and lexicon, they do not fit neatly into linguists' notions about ‘modules’ of language. Fillers have been reported in quite an array of languages, and yet they seem to be more common among learners of some languages than others. Even when language is held constant, children seem to vary immensely as to whether they produce fillers at all. With more researchers reporting fillers in more languages, it seems time to (1) review what we now know about fillers; (2) propose a reasonably unified set of criteria for identifying them; and (3) suggest an approach that will promote their further study.

In her note, Ann Peters distinguishes between two types of fillers: (1) premorphological, whose presence in the utterance is motivated by purely phonological considerations (the child attempts to produce an adult utterance at the prosodic level, or approximates the prevalent rhythmic pattern in the language); and (2) protomorphological, which function as place-holders for grammatical morphemes and eventually differentiate into various grammatical morphemes. I would like to suggest that there may also be a third type which sometimes appears when the child begins to generalize over a set of related words. Just as protomorphological fillers can provide valuable information about how children acquire function words, these ‘generalized’ fillers may offer insights into how children form categories of function words. I will illustrate this phenomenon with data from the Naomi corpus in the CHILDES database (Sachs, 1983).

As Peters points out in this and other work, the treatment of filler syllables continues to be problematic. To the extent that theoretical linguistic approaches to language acquisition have addressed the problem of morphology at all, this has been almost entirely in the domain of syntax. Thus, the phonological and prosodic characteristics of filler syllables in children's early speech have been largely ignored, the assumption being that the appearance of grammatical morphemes is ‘triggered’ by the maturation of syntactic representations (e.g. Radford, 1990). If this were the case, we would not expect to find crosslinguistic differences in the timing of the acquisition of determiners (Lleó & Demuth, 1999), nor the variable appearance of grammatical morphemes in some contexts and not others (Demuth, 1994, Gerken, 1996). These cases of crosslinguistic and within-language variation call for a more comprehensive account of early morpheme acquisition which can explain their variable presence or absence.

In her new and again significant paper on fillers, Ann Peters does us the honour of referring to our three-stage model of premorphology, proto-morphology and morphology proper (or modularized morphology), especially in regard to fillers (cf. also Kilani-Schoch & Dressler (2000a, b), where ample data analyses are presented).

Peters' note is valuable as an overview of current knowledge about filler syllables, in part because it clarifies how much we do NOT know about these elements. The note also points to the potential value of filler syllables for theoretical accounts of early language development, including contrasting predictions derived from nativist accounts against those that Peters describes as ‘constructivist.’ Characterizing the nature and function of filler elements could be very helpful for distinguishing these accounts, particularly if the focus is on identifying changes that might reflect a transition from phonological to morphological functions. As Peters indicates, differentiating these functions is unlikely to be easy. Indeed, it will be particularly difficult if Peters is correct in believing that the morphological categories are constructed; that position predicts a gradual transition from phonological to morphological functions and thus a period during which filler syllables are neither phonological nor morphological but, instead, somewhere in between.

Peters has made an excellent case for the importance of ‘filler’ syllables during the early stages of children's language development. In this commentary, I discuss some of the implications of filler-syllable use for related areas of child language research. The comments are based on observations made during collaborative cross-linguistic studies of both normally developing children and children with language disorders acquiring Italian (Leonard, Bortolini, Caselli, McGregor & Sabbadini, 1992; Bortolini & Leonard, 1996; Bortolini, Caselli & Leonard, 1997; Leonard & Bortolini, 1998), Hebrew (Dromi, Leonard & Shteiman, 1993; Leonard & Dromi, 1994; Dromi, Leonard, Adam & Zadunaisky-Ehrlich, 1999; Leonard, Dromi, Adam & Zadunaisky-Ehrlich, 2000), and Swedish (Hansson, Nettelbladt & Leonard, 2000).