Early Utterances Research Articles

How infants' utterances grow: A probabilistic account of early language development

Why are children's first utterances short and ungrammatical, with some obvious constructions missing? What determines the lengthening of children's early utterances over time? The literature is replete with references to a one-word, a two-word, and a later multiword stage in language development, but with little empirical evidence, and with little account for how and why utterances grow. To address these questions, we analyze speech samples from 25 children between the ages of 14 and 43 months; we construct distributions of their utterances of lengths one to five by age. Our novel findings are that multiword utterances of different lengths appear early in acquisition and increase together until they reach relatively stable proportions similar to those found in parents' input. To explain such patterns, we develop a probabilistic computational model, VIRTUAL, that posits an interaction between a) varying, increasing resources from various developmental domains and b) target utterance lengths mirroring the input. VIRTUAL successfully accounts for most of the empirical patterns, suggesting a probabilistic and dynamic process that is nonetheless compatible with apparent distinct milestones in development. We provide a new, systematic way of showing how developmental cascade theories could work in language development. Our findings and model also suggest insights into syntactic, semantic, and cognitive development.

A Usage-Based Approach to Pattern Finding: The Traceback Method Meets Code-Mixing

Usage-based approaches have become increasingly important in research on language acquisition and recently also in bilingual first language acquisition. Lexically specific patterns, such as What’s this? and frame-and-slot patterns, such as [I want X] play an important role in language acquisition scenarios. The ubiquity of such conventionalized chunks and frame-and-slot patterns supports the idea that children construct their early utterances out of concrete pieces they have heard and stored before. To investigate the emergence of patterns in children’s speech the traceback method has been developed, which accounts for the composition of utterances by relying on previously acquired material. Recently, the traceback method has also been applied to code-mixed utterances in bilingual children testing the assumption that bilingual utterances are structured around a frame-and-slot pattern in which the open slot is filled by (a) word(s) from the other language, e.g., [where is X] as in where is das feuer ‘where is the fire’. In this paper we want to present how the empirical use of the traceback method, and the general adoption of a usage-based theoretical perspective, can shed new lights on the study of bilingual phenomena, such as code-mixing.

The role of Function Words to build syntactic knowledge in French-speaking children

The question of how children learn Function Words (FWs) is still a matter of debate among child language researchers. Are early multiword utterances based on lexically specific patterns or rather abstract grammatical relations? In this corpus study, we analyzed FWs having a highly predictable distribution in relation to Mean Length Utterance (MLU) an index of syntactic complexity in a large naturalistic sample of 315 monolingual French children aged 2 to 4 year-old. The data was annotated with a Part Of Speech Tagger (POS-T), belonging to computational tools from CHILDES. While eighteen FWs strongly correlated with MLU expressed either in word or in morpheme, stepwise regression analyses showed that subject pronouns predicted MLU. Factor analysis yielded a bifactor hierarchical model: The first factor loaded sixteen FWs among which eight had a strong developmental weight (third person singular verbs, subject pronouns, articles, auxiliary verbs, prepositions, modals, demonstrative pronouns and plural markers), whereas the second factor loaded complex FWs (possessive verbs and object pronouns). These findings challenge the lexicalist account and support the view that children learn grammatical forms as a complex system based on early instead of late structure building. Children may acquire FWs as combining words and build syntactic knowledge as a complex abstract system which is not innate but learned from multiple word input sentences context. Notably, FWs were found to predict syntactic development and sentence complexity. These results open up new perspectives for clinical assessment and intervention.

The Effects of Age, Dosage, and Poverty on Second Language Learning through SparkLingTM in Infant Education Centers in Madrid, Spain.

The first 1000 days represent a unique window of opportunity for second language learning. In two recent studies we demonstrated that Spanish infants’ use of second-language (L2) English productive vocabulary and early utterances rapidly increased through the play-based, interactive and highly social SparkLingTM Intervention, which consists of an evidence-based method and curriculum stemming from decades of research on infant language development. Analyzing an expanded and more diverse sample of Spanish infants (n = 414; age 9–33 months) who received the SparkLingTM Intervention, this study examines the variability in L2 production, which was assessed via first-person LENA audio recordings. Infants’ age significantly and positively correlated with L2 production, demonstrating an advantage for older infants in the sample. While overall socioeconomic status (SES) was not related to L2 production, very young infants (under two years) who lived in high poverty homes showed faster increases in English production compared to peers who lived in moderate poverty homes. Infants’ attendance in the program (“dosage”) was also predictive of their L2 production outcomes. Infants across SES have the capacity to begin acquiring two languages in early education classrooms with SparkLingTM through one-hour/day sessions in social environments that engages them through frequent high-quality language input.

The traceback method and the early constructicon: theoretical and methodological considerations

Abstract Usage-based approaches assume that children’s early utterances are item-based. This has been demonstrated in a number of studies using the traceback method. In this approach, a small amount of “target utterances” from a child language corpus is “traced back” to earlier utterances. Drawing on a case study of German, this paper provides a critical evaluation of the method from a usage-based perspective. In particular, we check how factors inherent to corpus data as well as methodological choices influence the results of traceback studies. To this end, we present four case studies in which we change thresholds and the composition of the main corpus, use a cross-corpus approach tracing one child’s utterances back to another child’s corpus, and reverse and randomize the target utterances. Overall, the results show that the method can provide interesting insights—particularly regarding different pathways of language acquisition—but they also show the limitations of the method.

Towards an inventory of prosodic contours produced by English-acquiring 2-year-olds: The use of H* and H + !H*

One viewpoint of early prosody is that utterances are highly simplified in comparison to adult models, in part, due to biological constraints such as incomplete control of pitch production (Lieberman, 1967; Snow, 2006). In contrast, research in Catalan and Spanish shows that early utterances (under 2 years) consist of the basic intonational categories (Prieto et al., 2012), including contours that would not be predicted under a biological approach. Our primary question asks what is the inventory of contours produced by American English-acquiring toddlers. Based on work showing toddlers to have near adult-like acoustic realizations of certain pitch accents (Thorson, 2015), we hypothesize that children will demonstrate sophisticated intonational configurations; here we analyze specific pitch accent varieties in toddler speech. Spontaneous utterances of four 2-year-olds were analyzed. Pitch accents were annotated using ToBI (Beckman & Ayers, 1997) and acoustically analyzed. Results reveal three varieties of ...

Verbal aspect in Serbian children’s language production

Previous studies have shown that an early usage of verbal aspect in children?s spontaneous production is regularly observed in Slavic languages. However, this leads to the question whether the bare presence of aspect in early utterances is firm evidence that a child has acquired it. In order to show that the acquisition of aspect in Serbian is a process that lasts for several years, elicited production was inspected in children that had already achieved an apparent progress in language (3?5 year-olds). The comparisons with adults were made and several indicators of development were observed: a) the distribution of aspectual forms in children?s and adults? language, b) the ability of functional usage of aspect in narrative, and c) the mastery of contrasting aspectual pairs. The findings provided the ground to propose three developmental stages assumed for the acquisition of Serbian aspect, and possible underlying mechanisms of development are discussed.

Two-year-olds but not younger children comprehend it in ambiguous contexts: Evidence from preferential looking.

Children use pronouns in their speech from the earliest word combinations. Yet, it is not clear from these early utterances whether they understand that pronouns are used as substitutes for nouns and entities in the discourse. The aim of this study was to examine whether young children understand the anaphoric function of pronouns, focusing on the interpretation of the pronoun it in English-speaking children at 1;6 and 2;0. We tested whether adults and children would prefer to look at a previously introduced vs. novel visual object depending on the argument form (it, the + noun, a + noun , or silence). Results demonstrate that, like adults, two-year-olds understand that it refers to a previously introduced referent. There is no evidence that this knowledge is established in children at 1;6. This suggests that some time between 1;6 and 2;0 children come to understand that it refers to a highly accessible referent introduced in the prior context.

Null Arguments in the Yoruba Child’s Early Speech

This paper studied null arguments in the early speech of the Yoruba child. We examined whether null subjects are more prominent than null objects and whether the usage of null subjects decreases with language development in the early utterance of the Yoruba child. We also examined the relationship between argument ellipses and finiteness. Minimalist Programme was adopted as the theoretical framework because of its provision for minimality in the processes of language acquisition. The database was made up of longitudinal naturalistic speech output of three Yoruba-speaking children between 15and 36 months of age. We discovered the preponderance of null argument-NPs at the initial stage; subjects-NPs were easily omitted while object-NPs were hardly omitted; the three Yoruba-speaking children began to make use of overt argument-NPs from fifteen months and gradually, null argument-NPs gave way. We found no direct relationship between the acquisition of finiteness and the end of null arguments.

Abstract knowledge of word order by 19 months: An eye-tracking study

ABSTRACTWord order is one of the earliest aspects of grammar that the child acquires, because her early utterances already respect the basic word order of the target language. However, the question of the nature of early syntactic representations is subject to debate. Approaches inspired by formal syntax assume that the head–complement order, differentiating verb–object and object–verb languages, is represented very early on in an abstract, rulelike format. In contrast, constructivist theories assume that it is initially encoded as lexicalized, verb-specific knowledge. In order to address this issue experimentally, we combined the preferential looking paradigm using pseudoverbs with the weird word order paradigm adapted to comprehension. The results, based on highly reliable, coder-independent eye-tracking measures, provide the first direct evidence that as early as 19 months French-speaking infants have an abstract representation of the word order of their language.

The acquisition of auxiliaries BE and HAVE: an elicitation study*

Auxiliary syntax is recognized to be one of the more complex aspects of language that children must acquire. However, there is much disagreement among researchers concerning children's early understanding of auxiliaries, with some researchers advocating a relatively abstract grammar as the basis for auxiliary syntax, while others view the acquisition of auxiliary syntax as the gradual accumulation of linguistic knowledge, initially tied to individual lexical items. To investigate the status of children's early knowledge of auxiliary syntax, two studies were carried out. In study 1, 28 children (M= 3;1) were tested for their use of the auxiliaries BE and HAVE in declaratives, while in study 2, 19 children (M=3;3) were tested for their use of these auxiliaries in questions. Although overall error rates were low, there were differences between BE and HAVE in the proportion and types of errors observed in declaratives and questions, and some individual children showed very high error rates. The implications of these findings for different models of auxiliary syntax in children's early utterances are discussed.

The role of illocutionary operators in the emerging grammars of bilingual children

In this paper, we bring together recent discussions about the role of function words in emerging monolingual grammars with the discussion about role of these words in mixed utterances in emerging bilingual grammars. Recently several linguists have addressed the question of determining the organizing principles of early syntax in developing grammars of monolingual children. Van Kampen (this volume), Tracy (2002), among others, note that in early two-word utterances certain “function” words occur very frequently and play a special role: they appear as illocutionary operators and seem to have a bootstrapping function with respect to the development of syntactic structure. In this article, we consider spontaneous production data from three bilingual Dutch/French children and seek to answer the question whether the role of “function” words is the same in these bilingual developing grammars as in monolingual grammars. We show that this indeed appears to be the case. We then examine the claim by Deuchar and Quay (2000) that “function words” are mixed more readily than other words. We argue that our data do not present convincing arguments in favor of such a hypothesis.

Bootstrapping in the acquisition of word stress in Brazilian Portuguese

This paper deals with the acquisition of word stress in Brazilian Portuguese. In the course of acquisition, children use several strategies to mark stress prominence before the adult algorithm of primary stress is used productively. I argue that the prominences found in children´s early utterances do not reflect word stress but prominences of a higher prosodic level. In other words, children use the stress information available in higher prosodic domains as a cue for the acquisition of the algorithm of primary stress.

Prosodic Characteristics of Early Multi-Word Utterances in Italian Children

The capacity to relate two or more words into single intonational contours is an important phenomenon in the transition to early multi-word utterances, but there are few systematic investigations of the related acoustic properties. This study analysed the prosodic characteristics of early multi-word utterances produced by a sample of Italian children with the aim of verifying which prosodic models are detectable in early word combinations. The study focuses on the occurrence of the final syllable lengthening phenomenon. The 32 participants were 13-14 months old at the start of the study. Vocabulary development was assessed monthly up to approximately 200 words. Results show that the capacity to link two or more words in a single intonational contour is not fully developed during the production of the first multi-word utterances, and that final syllable lengthening, even if developed independently, is partially influenced by the development of intonational models.

Filler syllables: what is their status in emerging grammar?

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.

Language choice in the earliest utterances: a case study with methodological implications

In this case study we investigate how early a developing bilingual exposed simultaneously to English and Spanish can make appropriate language choices. We propose two methodological requirements for studies of this kind: (1) detailed records of the development of the child's lexicon; and (2) data collection in more than one language context. Our own study relies on detailed records of the child's cumulative vocabulary from the first word at ten months, and on weekly audiovideo recordings in both English and Spanish contexts from age 1;3. Analysis of the data shows a strong tendency for the language of the child's utterances to match that of the context at ages 1;7-1;8.

One vs. two systems in early bilingual syntax: Two versions of the question

This paper identifies two versions of the question as to whether there is a single initial system in the syntax of developing bilinguals. Version 1 asks whether there are early mixed utterances, and if so, attributes this to a single initial system. Version 2 asks whether the utterances containing words from one of the child's languages exhibit the same syntax as the utterances containing words from the child's other language. We argue with reference to our own data (from an English-Spanish bilingual from ages 1;7 to 1;9) that Version 1 is not tenable because of the paucity of lexical resources when the child begins to produce two-word utterances. However, we argue that the early two-word utterances in our data do seem to exhibit a single rudimentary syntax, based on a predicate-argument structure found in all utterance types, mixed and non-mixed. We then argue in relation to Version 2 of the question, that it can only be answered once the child's utterances can be identified as language-specific in the two languages, and that this is not possible before the emergence of morphological marking. This is illustrated by an analysis of our data from ages 1;8 to 2;3. We argue that language-specific morphology allows us to identify the language of the utterances in our data and to see evidence for the appearance of two differentiated morphosyntactic systems.

Development of phrases in the early utterances of children and cross-fostered chimpanzees

Project Washoe and its sequel with Moja, Pili, Tafu, and Dar, simulated with infant chimpanzees the conditions in which the language of human children develops gradually and-piecemeal into the language of human parents. This article traces patterns of growth and development in the early utterances of children and chimpanzees. The evidence for continuous processes and variables contradicts the yes-no, either-or Aristotelian logic of philosophical linguistics that has prevailed for so long. What the children and chimpanzees actually say supports, instead, a view of common laws and continuity that is much more compatible with modern natural science.

Functional Categories in Child L2 Acquisition of French

Vainikka and Young-Scholten (1994; in press) argued that adult language learners initially project verb phrases and that subsequent projection of functional categories is input driven, with inflectional phrases emerging before complementizer phrases. They assumed that the same is true for first-language acquisition (Clahsen, Eisenbeiss, and Vainikka (1994)). On this account, one would expect child second language (L2) learners to lack functional categories in the early stages of language learning. In this article, we argue against this position using data drawn from a longitudinal corpus of spontaneous production data from two English-speaking children who started attending French kindergarten at age 5. Our data suggest that functional categories and their projections are present in the earliest utterances available. Relevant data include the productive use of determiners, inflection, case marking, subject clitics, wh-questions, and correct negative placement. Changes were largely quantitative rather than...

Language Mixing in a French-Dutch Bilingual Child

Child bilingualism has been a domain of growing interest in the last few years. A central question in research concerns the differentiation of the two languages in the developmen-tal process: do children develop two separate language systems from the very beginning or do they start with a combined system? In this discussion, aspects of word order play an essential role. Radford (1986) has compared early child utterances with so called "small clauses". In small clauses, word order would be relatively free due to the fact that children have not yet acquired the concept of case marking which puts constraints on word order. In this assumption, word order would not be expected to be differen-tiated in the first stages of the two languages of the bilingual child. Others however (Meisel, 1989, Frijn & de Haan 1994) have suggested that word order from the two-word stage on is almost invariably correct and in line with parameter settings in the adult language. At first sight, the utterances of the French-Dutch bilingual child that we study do not support one or the other of these two views unambiguously. Despite the fact that French is a head initial, SVO language and that the majority of the utterances of the child are in accordance with this parameter setting, utterances with SOV order and other Dutch-like word orders do appear in her French with a certain frequency. In our discussion we will show that, while the early (S)OV patterns can probably be explained by the absence of a fully fledged functional projection IP in the child's grammar, this cannot account for these patterns in later phases. The persistent presence of OV patterns in the French utterances - that are (although very rarely) encountered in French monolingual children as well - seems to be caused, then, by the continuing Dutch input that may very well be the factor that "pushes up" the production of [XP V] patterns in the child's French.