Publications

Abstract

Continuous vector word representations (or
word embeddings) have shown success in cap-turing semantic relations between words, as evidenced by evaluation against behavioral data of adult performance on semantic tasks (Pereira et al., 2016). Adult semantic knowl-edge is the endpoint of a language acquisition process; thus, a relevant question is whether these models can also capture emerging word
representations of young language learners. However, the data for children’s semantic knowledge across development is scarce. In this paper, we propose to bridge this gap by using Age of Acquisition norms to evaluate word embeddings learnt from child-directed input. We present two methods that evaluate word embeddings in terms of (a) the semantic neighbourhood density of learnt words, and (b) con-
vergence to adult word associations. We apply our methods to bag-of-words models, and find that (1) children acquire words with fewer semantic neighbours earlier, and (2) young learners only attend to very local context. These findings provide converging evidence for validity of our methods in understanding the prerequisite features for a distributional model of word learning.

Abstract

Research on infants’ online lexical processing by Fernald, Perfors, and Marchman (2006) revealed substantial individual differences that are related to vocabulary development, such that infants with better lexical processing efficiency show greater vocabulary growth across time. Although it is clear that individual differences in lexical processing efficiency exist and are meaningful, the theoretical nature of lexical processing efficiency and its relation to vocabulary size is less clear. In the current study, we asked two questions: (a) Is lexical processing efficiency better conceptualized as a central processing capacity or as an emergent capacity reflecting a collection of word-specific capacities? and (b) Is there evidence for a causal role for lexical processing efficiency in early vocabulary development? In the study, 120 infants were tested on a measure of lexical processing at 18, 21, and 24 months, and their vocabulary was measured via parent report. Structural equation modeling of the 18-month time point data revealed that both theoretical constructs represented in the first question above (a) fit the data. A set of regression analyses on the longitudinal data revealed little evidence for a causal effect of lexical processing on vocabulary but revealed a significant effect of vocabulary size on lexical processing efficiency early in development. Overall, the results suggest that lexical processing efficiency is a stable construct in infancy that may reflect the structure of the developing lexicon.

Abstract

We report on two experiments that investigated the acquisition of the Tagalog symmetrical voice system, a typologically rare feature of Western Austronesian languages in which there are more than one basic transitive construction and no preference for agents to be syntactic subjects. In the experiments, 3-, 5-, and 7-year-old Tagalog-speaking children and adults completed a structural priming task that manipulated voice and word order, with the uniqueness of Tagalog allowing us to tease apart priming of thematic role order from that of syntactic roles. Participants heard a description of a picture showing a transitive action, and were then asked to complete a sentence of an unrelated picture using a voice-marked verb provided by the experimenter. Our results show that children gradually acquire an agent-before-patient preference, instead of having a default mapping of the agent to the first noun position. We also found an earlier mastery of the patient voice verbal and nominal marker configuration (patient is the subject), suggesting that children do not initially map the agent to the subject. Children were primed by thematic role but not syntactic role order, suggesting that they prioritize mapping of the thematic roles to sentence positions.

Abstract

Humans associate abstract words with physical stimulus dimensions, such as linking upward locations with positive concepts (e.g., happy = up). These associations manifest both via subjective reports of associations and on objective performance metrics. Humans also report subjective associations between colours and abstract words (e.g., joy is linked to yellow). Here we tested whether such associations manifest on objective task performance, even when not task-relevant. Across three experiments, participants were presented with abstract words in physical colours that were either congruent with previously-reported subjective word-colour associations (e.g., victory in red and unhappy in blue), or were incongruent (e.g., victory in blue and unhappy in red). In Experiment 1, participants' task was to identify the valence of words. This congruency manipulation systematically affected objective task performance. In Experiment 2, participants completed two blocks, a valence-identification and a colour-identification task block. Both tasks produced congruency effects on performance, however, the results of the colour identification block could have reflected learning effects (i.e., associating the more common congruent colour with the word). This issue was rectified in Experiment 3, whereby participants completed the same two tasks as Experiment 2, but now matched congruent and incongruent pairs were used for both tasks. Again, both tasks produced reliable congruency effects. Item analyses in each experiment revealed that these effects demonstrated a degree of item specificity. Overall, there was clear evidence that at least some abstract word-colour pairings can systematically affect behaviour.

Abstract

The computations involved in statistical learning have long been debated. Here, we build on work suggesting that a basic memory process, chunking , may account for the processing of statistical regularities into larger units. Drawing on methods from the memory literature, we developed a novel paradigm to test statistical learning by leveraging a robust phenomenon observed in serial recall tasks: that short‐term memory is fundamentally shaped by long‐term distributional learning. In the statistically induced chunking recall (SICR) task, participants are exposed to an artificial language, using a standard statistical learning exposure phase. Afterward, they recall strings of syllables that either follow the statistics of the artificial language or comprise the same syllables presented in a random order. We hypothesized that if individuals had chunked the artificial language into word‐like units, then the statistically structured items would be more accurately recalled relative to the random controls. Our results demonstrate that SICR effectively captures learning in both the auditory and visual modalities, with participants displaying significantly improved recall of the statistically structured items, and even recall specific trigram chunks from the input. SICR also exhibits greater test–retest reliability in the auditory modality and sensitivity to individual differences in both modalities than the standard two‐alternative forced‐choice task. These results thereby provide key empirical support to the chunking account of statistical learning and contribute a valuable new tool to the literature.

Abstract

Humans vary in almost every dimension imaginable, and language is no
exception. In this article, we review the past research that has focused on
individual differences (IDs) in first language acquisition. We first consider
how different theoretical traditions in language acquisition treat IDs, and
we argue that a focus on IDs is important given its potential to reveal the
developmental dynamics and architectural constraints of the linguistic system.
We then review IDs research that has examined variation in children’s
linguistic input, early speech perception, and vocabulary and grammatical
development. In each case, we observe systematic and meaningful variation,
such that variation in one domain (e.g., early auditory and speech
processing) has meaningful developmental consequences for development
in higher-order domains (e.g., vocabulary). The research suggests a high
degree of integration across the linguistic system, in which development
across multiple linguistic domains is tightly coupled.

Abstract

Statistical learning (SL) has been a prominent focus of research in developmental and adult populations, guided by the assumption that it is a fundamental component of learning underlying higher-order cognition. In developmental populations, however, there have been recent concerns regarding the degree to which many current tasks reliably measure SL, particularly in younger children. In the current article, we present the results of two studies that measured auditory statistical learning (ASL) of linguistic stimuli in children aged 5–8 years. Children listened to 6 min of continuous syllables comprising four trisyllabic pseudowords. Following the familiarization phase, children completed (a) a two-alternative forced-choice task and (b) a serial recall task in which they repeated either target sequences embedded during familiarization or foils, manipulated for sequence length. Results showed that, although both measures consistently revealed learning at the group level, the recall task better captured learning across the full range of abilities and was more reliable at the individual level. We conclude that, as has also been demonstrated in adults, the method holds promise for future studies of individual differences in ASL of linguistic stimuli.

Abstract

Much of Lieven’s pioneering work has helped move the study of individual differences to the centre of child language research. The goal of the present chapter is to illustrate how the study of individual differences provides crucial insights into the language acquisition process. In part one, we summarise some of the evidence showing how pervasive individual differences are across the whole of the language system; from gestures to morphosyntax. In part two, we describe three causal factors implicated in explaining individual differences, which, we argue, must be built into any theory of language acquisition (intrinsic differences in the neurocognitive learning mechanisms, the child’s communicative environment, and developmental cascades in which each new linguistic skill that the child has to acquire depends critically on the prior acquisition of foundational abilities). In part three, we present an example study on the role of the speed of linguistic processing on vocabulary development, which illustrates our approach to individual differences. The results show evidence of a changing relationship between lexical processing speed and vocabulary over developmental time, perhaps as a result of the changing nature of the structure of the lexicon. The study thus highlights the benefits of an individual differences approach in building, testing, and constraining theories of language acquisition.

Abstract

A core question in language acquisition is whether children’s syntactic processing is experience-dependent and language-specific, or whether it is governed by abstract, universal syntactic machinery. We address this question by presenting corpus and on-line processing dat a from children learning Mandarin Chinese, a language that has been important in debates about the universality of parsing processes. The corpus data revealed that two different relative clause constructions in Mandarin are differentially used to modify syntactic subjects and objects. In the experiment, 4-year-old children’s eye-movements were recorded as they listened to the two RC construction types (e.g., Can you pick up the pig that pushed the sheep?). A permutation analysis showed that children’s ease of comprehension was closely aligned with the distributional frequencies, suggesting syntactic processing preferences are shaped by the input experience of these constructions.

Abstract

Background: There is variability in the language of children with autism, even those who are high functioning. However, little is known about how they process language structures in real time, including how they handle potential ambiguity, and whether they follow referential constraints. Previous research with older autism spectrum disorder (ASD) participants has shown that these individuals can use context to access rapidly the meaning of ambiguous words. The severity of autism has also been shown to influence the speed in which children with ASD access lexical information. Aims: To understand more about how children with ASD process language in real time (i.e., as it unfolds). The focus was the integration of information and use of referential constraints to identify a referent named in a sentence. Methods & Procedures: We used an eye-tracking task to compare performance between young, high-functioning children with autism (HFA) and children with typical development (TD). A large sample of 5–9-year-old children (mean age = 6;8 years), 48 with HFA and 56 with TD participated; all were attending mainstream schools. For each item participants were shown a display of four images that differed in two dimensions. Each sentence contained an adjective and noun that restricted the choice from four to two (the target and competitor), followed by a prepositional phrase (e.g., the blue square with dots); this added modifying information to provide a unique description of the target. We calculated looking time at the target, the competitor and the two distractors for each 200 ms time interval as children processed the sentence and looked at the display. Generalized estimating equations were used to carry out repeated-measures analyses on the proportion of looking time to target and competitor and time to fixate to target. Outcomes & Results: Children in both groups (HFA and TD) looked at the target and competitor more than at the distractors following the adjective and noun and following the modifying information in the prepositional phrase more at the target. However, the HFA group was significantly slower in both phases and looked proportionally less at the target. Across the sample, IQ and language did not affect the results; however, age and attention had an impact. The older children showed an advantage in processing the information as did the children with higher attention scores. Conclusions & Implications: The HFA group took longer than the TD group to integrate the disambiguating information provided in the course of processing a sentence and integrate it with the visual information, indicating that for the ASD group incremental processing was not as advanced as for children with ASD, and they were less sensitive to referential conventions. Training for young children with ASD on the use of referential conventions and available contextual clues may be of benefit to them in understanding the language they hear.

Abstract

Research with adults and older children indicates that verb biases are strong influences on listeners’ interpretations when processing sentences, but they can be overruled. In this paper, we ask two questions: (i) are children with Autism Spectrum Disorder (ASD) who are high functioning sensitive to verb biases like their same age typically developing peers?, and (ii) do young children with ASD and young children with typical development (TD) override strong verb biases to consider alternative interpretations of ambiguous sentences? Participants were aged 5–9 years (mean age 6.65 years): children with ASD who were high functioning and children with TD. In task 1, biasing and neutral verbs were included (e.g., eat cake versus move cake). In task 2, the focus was on whether the prepositional phrase occurring with an instrument biasing verb (e.g., ‘Chop the tree with the axe’) was interpreted as an instrument even if the named item was an implausible instrument (e.g., candle in ‘Cut the cake with the candle’). Overall, the results showed similarities between groups but the ASD group was generally slower. In task 1, both groups looked at the named object faster in the biasing than the non-biasing condition, and in the biasing condition the ASD group looked away from the target more quickly than the TD group. In task 2, both groups identified the target in the prepositional phrase. They were more likely to override the verb instrument bias and consider the alternative (modification) interpretation in the implausible condition (e.g., looking at the picture of a cake with a candle on it’). Our findings indicate that children of age 5 years and above can use context to override verb biases. Additionally, an important component of the sentence processing mechanism is largely intact for young children with ASD who are high functioning. Like children with TD, they draw on verb semantics and plausibility in integrating information. However, they are likely to be slower in processing the language they hear. Based on previous findings of associations between processing speed and cognitive functioning, the implication is that their understanding will be negatively affected, as will their academic outcomes.

Abstract

Humans appear to rely on spatial mappings to describe and represent concepts. In particular, conceptual cueing refers to the effect whereby after reading or hearing a particular word, the location of observers’ visual attention in space can be systematically shifted in a particular direction. For example, words such as “sun” and “happy” orient attention upwards, whereas words such as “basement” and “bitter” orient attention downwards. This area of research has garnered much interest, particularly within the embodied cognition framework, for its potential to enhance our understanding of the interaction between abstract cognitive processes such as language and basic visual processes such as attention and stimulus processing. To date, however, this area has relied on subjective classification criteria to determine whether words ought to be classified as having a meaning that implies “up” or “down.” The present study, therefore, provides a set of 498 items that have each been systematically rated by over 90 participants, providing refined, continuous measures of the extent to which people associate given words with particular spatial dimensions. The resulting database provides an objective means to aid item-selection for future research in this area.

Abstract

Variability in children's language acquisition is likely due to a number of cognitive and social variables. The current study investigated whether individual differences in statistical learning (SL), which has been implicated in language acquisition, independently predicted 6- to 8-year-old's comprehension of syntax. Sixty-eight (N = 68) English-speaking children completed a test of comprehension of four syntactic structures, a test of SL utilizing nonlinguistic visual stimuli, and several additional control measures. The results revealed that SL independently predicted comprehension of two syntactic structures that show considerable variability in this age range: passives and object relative clauses. These data suggest that individual differences in children's capacity for SL are associated with the acquisition of the syntax of natural languages.

Abstract

Languages are strong markers of social identity. Multiple features of language and speech, from accent to lexis to grammatical constructions, mark speakers as members of specific cultural groups. In the current article, we present two confederate-scripted studies that investigated the social effects of the Australian hypocoristic use (e.g., uggie, uni, derro)—a lexical category emblematic of Australian culture. Participants took turns with a confederate directing each other through locations on a map. In their directions, the confederate used either hypocoristic (e.g., uni) or standard forms (e.g., university). The confederate’s cultural group membership and member prototypicality were manipulated by ethnic background and accent: In a highly prototypical in-group condition, the confederate had an Anglo-Celtic background and Australian English (AusE) accent; in a low prototypical in-group condition, the confederate had an Asian background and AusE accent; and in the out-group condition, the confederate had an Asian background and non-AusE accent. Hypocoristic use resulted in significantly higher participant-rated perceived common ground with the confederate when the confederate was an in-group but not an out-group member, which in some instances was moderated by in-group identification. The results suggest that like accents, culturally significant lexical categories function as markers of in-group identity, which influence perceived social closeness during interaction.

Abstract

The current study investigated the influence of a play-based curriculum on the development of pretend play skills and oral language in children attending their first year of formal schooling. In this quasi-experimental design, two groups of children were followed longitudinally across the first 6 months of their first year at school. The children in the experimental group were attending a school with a play-based curriculum; the children in the control group were attending schools following a traditional curriculum. A total of 54 children (Time 1 Mage = 5;6, range: 4;10–6;2 years) completed standardised measures of pretend play and narrative language skills upon school entry and again 6 months later. The results showed that the children in the play-based group significantly improved on all measures, whereas the children in the traditional group did not. A subset of the sample of children (N = 28, Time 1 Mage = 5;7, range: 5;2 – 6;1) also completed additional measures of vocabulary and grammar knowledge, and a test of non-verbal IQ. The results suggested that, in addition to improving play skills and narrative language ability, the play-based curriculum also had a positive influence on the acquisition of grammar.