IASCL 2024 Symposia

07/16/2024, from 10:30 AM to 12:30 PM , Room P131

Symposium: Assessment of Pragmatic Competence in Child Language.Methodologies & Research in Diverse Populations

Speakers: Maria Busch; Lisa Vössing; Laura Hüser; Luca Plachy

This symposium provides an overview of various methodological approaches for assessing pragmatic competence in children, especially considering diverse populations such as multilingual children, children with language needs and autistic children.

Conversation Analysis as a Methodical Approach to Assess Autistic Children’s Pragmatic Competence

Authors: Lisa Vössing;

“Autistic children are considered to experience a broad range of pragmatic language impairment that vary considerably between individuals (Paul et al., 2014; Volden, 2017). To assess these pragmatic challenges and (non-autistic) children’s pragmatic competence in general, a range of standardised tests, checklists and profiles exist (Adams, 2002). These assessments typically focus on the children’s performance in order to identify impairments. In contrast to this individual-focused perspective, other methodical approaches, e.g. Conversation Analysis, provide insights from an interaction-focused perspective (Wilkinson et al., 2020). The focus is on the interaction of interlocutors in a naturally occurring situation. Conversation Analysis thus allows to describe pragmatic competence based on authentic conversational interactions, but initially without an diagnostic aim.

This presentation aims to discuss Conversation Analysis as an approach stemming from qualitative social research that may contribute to the diagnosis of pragmatic competence in the field of speech and language therapy. To that end, the presented research is based on video recordings of autistic children (n=5, 9-13 years old) in authentic conversations at home and in a therapeutic setting. The analytical process is informed by Conversation Analysis and focuses on sequences in which pragmatic competence comes to the fore. Those findings on pragmatic competence will be presented and discussed in relation to other methodical approaches.”

Quantitative Longitudinal Study of Pragmatic Competence in Children with Special Language Needs

Authors: Laura Hüser; Dr. Markus Spreer

“In order to enable active participation of children with special language needs, it is essential to concentrate on language-systematic as well as pragmatic-communicative competence. Previous research suggests a negative relationship between impairments in structural language skills and pragmatic competence. For instance, affected children demonstrate reduced responsiveness to communicative requests (Andrés-Roqueta et al., 2021; Bishop et al., 2000). Moreover, language difficulties can influence communicative involvement and social interactions, particularly among peers (Janik Blaskova & Gibson, 2021).

A reliable diagnosis is crucial in developing a specific therapeutic approach that addresses relevant therapy goals. The DSM-5 has introduced Social (Pragmatic) Communication Disorder (SPCD) as a new diagnosis, highlighting the importance of pragmatic competence and potential corresponding disorders (William et al., 2017).

There are various methods to assess pragmatic competence such as observation, questioning, and standardised tests. The presentation focuses on the possibilities of diagnostics via standardised test procedures and quantitative research in the field of pragmatics and communication.

This research includes a longitudinal study conducted over a two-year period with four data collection points. The assessment was carried out at four primary schools in Germany, involving children with special language needs (N=~80 children). We address the following research questions:

a) How does children’s pragmatic competence develop during the first two years of school?

b) How are children’s language profiles related to their pragmatic competence?

By using standardised procedures, such as a questionnaire from the Clinical Evaluation of Language Fundamentals – Fifth Edition (CELF-5; Wiig et al., 2020), this quantitative study facilitates referencing to norm groups and adds to a wider understanding of the development of pragmatic and communication skills. The classroom represents an optimal research environment as it allows for the involvement of teachers and parents in monitoring the children’s development.”

Pragmatic Assessment in Multilingual Contexts – The Role of Heritage Languages

Authors: Luca Plachy; Maria Busch, Dr. Stephan Sallat, Dr. Matthias Ballod

“The field of intercultural pragmatics and interlanguage pragmatics has predominantly focused on adult learners (e.g. Keckes, 2015), leaving a gap in understanding of how multilingual children develop pragmatic competences. As pragmatics considers language use in context, it contains sociopragmatic and pragmalinguistic components (Félix-Brasfelder, 2021). While the assessment of pragmatic competences of multilingual children in Germany focuses mostly on tests, observations and interviews regarding the target language, the process still doesn’t take the heritage languages and thus the characteristics and resources of multilinguism sufficiently into account (Heller, 2019).

Multilingual children with typical language development have already developed complex pragmatic competences in their heritage language which include these sociopragmatic and pragmalinguistic components. Recognizing these resources in multilingual language development, this research project called MehrSelbst, funded by the German Federal Ministry of Education and Research, extends the assessment of pragmatic competences and pedagogical interventions by the children’s heritage language. Therefore, in addition to interviews with the children themselves and their educators on their pragmatic competences in German as a target language, the study also provides for interviews on the pragmatic competences in the children’s heritage language. The study includes children from grade 1, 4, 5 and 9 in Germany, which means children at the age of seven to fifteen.

These results are intended to serve as a basis for the development of intervention for pragmatic competences, that can be evaluated and applied for formal, informal and non-formal educational settings.”

Children’s Perspectives in Pragmatic Assessment – A Qualitative, Participatory Approach

Authors: Maria Busch; Dr. Stephan Sallat

“The assessment of pragmatic competences is primarily done through observations and interviews of caregivers and educators (Adams, 2002; Norbury, 2014). However, the subjective perspectives of the involved children themselves have hardly been included in applied and clinical linguistics, despite the fact that the UN Convention on the Rights of the Child calls for children’s perspectives to be taken into account in all matters affecting them (Roulstone & McLeod, 2011).

Gaining insights into children’s views of their own pragmatic competences and possible challenges is also relevant from a psychological perspective. Self-concepts, which are subjective beliefs and evaluative judgements one has about oneself and one’s competences, are interrelated with competence development, motivation, interest, among others (Trautwein & Möller, 2016). It seems promising to consider children’s perspectives and their evaluative judgements of their own pragmatic competences also in pragmatic assessment and research on developmental pragmatics.

The present study therefore addresses the following research questions:

1. How do children describe and evaluate their own pragmatic competence?
2. What methods can be used to collect and analyse the children’s perspective on pragmatic competences?
3. How can the children’s perspectives be included in participatory research on pragmatics and clinical reasoning in speech and language therapy (SLT)?

This study presents qualitative methodological approaches to collect and analyse children’s perspectives on their own pragmatic competences in the context of pragmatic assessment, focusing on conversational repairs and initiations in turn taking. The sample consists of monolingual and multilingual primary school students in Germany, with typical language development and developmental language disorders. Possibilities for incorporating the children’s perspective into research on pragmatic development and clinical decision making in SLT are derived from triangulating the children’s perspectives with those of caregivers and professionals in pragmatic assessment.”

07/16/2024, from 10:30 AM to 12:30 PM , Room P300

Symposium: The intersection of socioemotional well-being and bilingual development – HaBilNet-sponsored symposium

Speakers: He Sun; Andrea C. Schalley; Adam Winsler; Ekaterina Tiulkova; Annick De Houwer

In this symposium, speakers from different countries will address how bilingual experience affects children’s language and social-emotional skills.

Harmonious Bilingual Experience and Child Wellbeing: A Conceptual Framework

Authors: He Sun;

The notion of Harmonious Bilingual Development has drawn increasing attention in recent years as it views children’s bilingual development in a wider scope, exploring the relations between bilingual experience and family members’ wellbeing. The current study operationalizes the concept through specific parental and child bilingual experiential factors and proposes a harmonious bilingual experience framework on bilingual children’s social-emotional wellbeing, addressing 1) the influences of parental bilingual proficiency and perceptions with regard to parental bilingual language use, 2) the impact of parental language use on children’s language use, literacy environment, and bilingual receptive vocabulary size, as well as 3) such bilingual experience and proficiency in relation to children’s wellbeing. Parents of 123 English-Mandarin bilingual children (4-5 years old; 61 girls and 62 boys) filled in questionnaires on their bilingual perceptions and home bilingual environment and reported on their own and children’s social-emotional behavioral skills. Children’s and parents’ social-emotional and behavioural skills were assessed with the Strengths and Difficulties Questionnaire (SDQ) (Goodman, 1997). Children’s English and Mandarin receptive vocabulary sizes were assessed with a locally designed picture matching task whose format is similar to the PPVT. Children’s nonverbal intelligence was assessed with Raven’s Colored Progressive Matrics. Our bilingual sample generally supported the proposed conceptual framework. Specially, parental English and Mandarin proficiency was found to significantly affect parents’ language use in the respective language. Parental language use also influenced children’s language use, literacy activity, as well as receptive vocabulary sizes in both languages. Finally, children’s bilingual experience was found to influence their own social emotional skills, after controlling for the impact of parental social emotional skills. Specifically, better societal language skills and more frequent bilingual reading were found to promote children’s prosocial skills, while more children’s books predicted lower children’s total difficulty level, including such as peer relationship problems.

Parental socio-emotional experiences and children’s bilingual development

Authors: Andrea C. Schalley; Susana A. Eisenchlas

“We focus on the relation between parents’ past and present socio-emotional experiences and their children’s bilingual skills development. We explore which parental experiential factors might impact on their children’s bilingual language acquisition outcomes, and consider the latter in regards to both reported productive and receptive skills.

The data are sourced from a transnational online survey we developed to examine factors impacting home language transmission. As theoretical framework and guide to the survey design, Spolsky’s (2009) language policy framework – including the three components of language ideologies, language practices, and language management – was applied in the context of the family and supplemented with affective and social components (Schalley & Eisenchlas, 2020).

Here, we specifically focus on the following factors: (i) parents’ own bilingual upbringing (if applicable), and in particular affective and social experiences when growing up and how these influence their own family language policy, (ii) parents’ current use of their languages within the family and in particular with their children, (iii) parents’ assessment of the importance of their languages, (iv) parents’ assessment of their own language identity, (v) parents’ emotional reactions to their children’s use or refusal of languages, (vi) parents’ reasons for transmitting or not transmitting their home languages, (vii) parents’ language anxiety levels in their languages, and (viii) parental beliefs on language ideologies in the society where they reside as well as parental beliefs towards multilingualism in general.

While the data collection and analyses are still ongoing, we expect to uncover evidence that specific parental socio-emotional experiences correlate with children’s language acquisition outcomes. Thus, one generation’s socio-emotional well-being might impact on the following generation’s bilingual development – an aspect that has to some extent been disregarded in research to date.”

Academic Benefits from Early Bilingualism: Elementary and Middle School Outcomes Associated with Ear

Authors: Adam Winsler; Gabrielle Norvell, Nadine Rozell, Tevis Tucker

Social skills, wellbeing, and academic performance are intertwined constructs in childhood, each contributing to the others. Earlier acquisition of English is generally associated with better academic performance for dual language learners (DLLs) in the U.S., but large-scale, prospective, longitudinal studies examining early English acquisition trajectories and how they relate to academic outcomes (accounting for relevant covariates) are rare. We explored how the year/grade in which DLLs acquired English proficiency relates to academic outcomes (grade retention, grade point average (GPA), reading and math test scores) in 5th grade and in 8th grade, controlling for gender, ethnicity, poverty, and school readiness skills at age 4. A large (N = 17,548), ethnically diverse (86% Latino, 10% Black, and 4% White/Other), and predominantly low-income sample (80% in poverty) of 4-year-old DLLs (mostly but not all Spanish-English) was followed longitudinally from their pre-K year through the end of middle school (8th grade in the US). Earlier acquisition of English, as defined by the school district as the student no longer needing English as second language services), predicted better performance on each of the 5th and 8th grade outcomes. Earlier proficiency in English was even more predictive of outcomes for students with initially high cognitive skill, Latino/Hispanic DLLs (compared to Black DLLs), and those not in poverty. Implications for practice and research will be discussed.

Social-emotional well-being among 5-year-old French-Russian children in France: links with input and

Authors: Ekaterina Tiulkova; Vanda Marijanović, Barbara Köpke

“This study explores the concept of Harmonious Bilingual Development (HBD) for Franco-Russian children with the aim to discern how input, parental factors, and children’s attitudes impact their social-emotional and behavioral skills (SEBS) and how these factors relate to their language proficiency. Both parent and child data were collected in France for 36 five-year-old children with Russian and French as their first languages.

A parent questionnaire and activity journal assessed parental attitudes towards bilingual education, children’s language exposure and the quality of language input (speakers, activities, media, reading). The Strengths and Difficulties Questionnaire was used to evaluate parent’s view of SEBS. Children’s language proficiency was determined on the basis of oral fluency measures in each language (mean syllable length, mean length of run, pause location) obtained through a picture-story retelling task and their attitudes were determined based on their affection for languages.

This presentation will link attitudes and input to fluency in oral productions considering variables such as parental education level, childbirth order, family size. The influence of attitudes was analyzed through a clustering method, underscoring the significance of both input quantity and quality in bilingual development. Children with limited exposure to Russian exhibited lower levels of SEBS, while those with higher exposure displayed higher levels. Parental attitudes were found to influence children’s perceptions, and children with positive language attitudes towards both languages exhibited superior SEBS. With respect to fluency measures, preliminary results suggest that children with higher quantity of input in Russian produced more syllables per run (interval between two pauses) and fewer pauses within utterances, and this in both languages. This study highlights the connection between children’s attitudes and their bilingual development. By advancing our understanding of HBD, this research advocates for supporting non-societal languages and promoting an inclusive linguistic environment for children, beneficial for their social-emotional well-being.”

07/16/2024, from 10:30 AM to 12:30 PM , Room P104

Symposium: Neural tracking of speech in the developing brain: paths forward

Speakers: Melis Çetinçelik; Tineke Snijders; Sergio M. Pereira Soares; Anika van der Klis; Katharina Menn

This symposium investigates infants’ neural tracking of speech, various factors influencing it, and its development in the early years, focusing on the relationship between individual differences in neural tracking and language development.

Relating infants’ neural tracking of speech to word segmentation and language development

Authors: Tineke Snijders; Fleur M.H.G. Vissers, Iris C. Schmits, Anne Kösem

“Infants use rhythmic information as an important cue for word segmentation. One possible mechanism for this might be neural tracking of speech: the synchronization of neural responses to rhythms in speech. In this talk, we will discuss two longitudinal studies relating neural tracking to word segmentation performance, as well as to later vocabulary development.

In the first study, 108 Dutch infants were followed in multiple sessions. In an EEG session at 7.5 months, we assessed infants’ neural tracking of speech by looking at speech-brain coherence (SBC) in response to book-read child-directed speech. At 9 months we tested their word segmentation in a head-turn paradigm, comparing looking times to familiarized versus novel words. SBC at 1.5-1.75 Hz (~stressed syllable rate) over left-frontal electrodes was significantly related to the word segmentation familiarity effect. At the conference, we will additionally discuss novel results relating infants’ neural tracking and word segmentation to their later vocabulary development (18,24,31 months).

In the second EEG study, 65 9-month-old Dutch infants listened to rhythmic and non-rhythmic speech in which bi-syllabic low-frequency words were repeated. Sentences had a consistent alternating strong/weak syllable pattern with a rate of 3.2 Hz. The non-rhythmic condition was created by speeding up or slowing down speech across several syllables. Rhythmic speech resulted in more SBC than non-rhythmic speech, over bilateral frontal electrodes for the syllable rate (3.2Hz), and over left frontal electrodes for the stressed syllable rate (1.6Hz). SBC at the syllabic rate was significantly related to infants’ word segmentation performance, as shown by their word familiarity ERP effect. We will additionally present new results relating SBC at 9 months to further vocabulary development at 27 months.

Together, the experiments inform about the functional relevance of neural tracking of speech in infants for their word segmentation ability and further language development.”

Infants’ neural tracking of multimodal speech and its relationship with vocabulary development

Authors: Melis Çetinçelik; Antonia Jordan Barros, Caroline F. Rowland, Tineke M. Snijders

“Infants’ early experiences with language frequently occur in face-to-face interactions with caregivers. In these interactions, infants receive multimodal language input from both the auditory speech signal and the speaker’s face, including eye gaze and rhythmic movements of the lips. While it has been demonstrated that both eye gaze and visual speech cues facilitate word learning and speech processing, the effects of these cues on infants’ neural tracking have not been examined. Neural tracking refers to the time-locking of neural oscillations to rhythmic units in speech and may be a neural mechanism that allows infants to use the temporal regularities in speech for speech processing.

In two EEG studies, we tested the role of eye gaze and visual speech cues on infants’ neural tracking of speech. 32-channel EEG data was recorded from Dutch 10-month-olds while they watched videos of a native Dutch speaker reciting passages in infant-directed speech (IDS). Speaker’s eye gaze (direct vs. averted) was manipulated in experiment 1, and visual speech cues (fully audiovisual vs. mouth region occluded) in experiment 2. Speech-brain coherence was calculated to assess infants’ neural speech tracking at the stress and syllable rates (1-1.75 and 2.5-3.5 Hz respectively in our stimuli), as amplitude modulations are especially marked in these frequencies in IDS. In both experiments, infants tracked the speech envelope at the stress and syllable frequencies, but contrary to our predictions, this was not modulated by the audiovisual cues, neither by eye gaze, nor visual speech cues. Furthermore, infants’ neural speech tracking at 10 months was related to their receptive and expressive vocabulary size at 18 months.

Overall, our results suggest that infants’ speech tracking is not necessarily impaired when visual speech cues are not visible or when the speaker’s gaze is averted, and may be a potential mechanism for successful language acquisition.”

Neural tracking and language development in early life: A longitudinal approach

Authors: Sergio M. Pereira Soares; Tineke M. Snijders, Caroline F. Rowland

“One of the most fascinating recent discoveries in neurolinguistics is that neurons’ oscillatory patterns synchronize with external signals such as speech. Especially in early child development, this is of great importance to better understanding language acquisition milestones, as e.g., the ability to segment continuous speech into smaller linguistic units. Studies on infants have found that modulations of infant directed speech amplitudes are particularly strong around certain frequency ranges, corresponding to stressed syllables, syllables, and phonemes. This early entrainment seems to indicate that infants are able to neurally track speech. Thus, this raises the hypothesis that the alignment of neural activity to speech-relevant amplitude modulations might sustain language development. Indeed, the initial literature in this field seems to suggest a link between tracking and early language acquisition. This link might result from infants with greater tracking having an advantage for language acquisition, or alternatively that infants with better language are better at tracking speech.

Here, we want to assess whether individual trajectories in the development of neural tracking of speech relate to individual differences in language skills and if/how environmental factors (e.g., SES) influence these effects. We predict that at different ages, different rates might drive tracking mechanisms. We also expect that more favourable environmental conditions will modulate these effects.

We are collecting EEG data from monolingual Dutch infants while they listen to child-directed stories at three different timepoints in the first year of life (6, 9, and 12 months). Furthermore, we collect early language development data in the second year of life. Speech-brain coherence data will be longitudinally compared between the three sessions and regressed to concurrent language skills at the beginning of their second year of life (N=70, expected by April 2024). Mediation analyses will be performed to explore the role environmental factors might have on individual language development.”

Longitudinal trajectories of infants’ neural tracking of nursery rhymes and their implications

Authors: Anika van der Klis; Melis Çetinçelik, Katharina Menn, Tineke M. Snijders, Caroline Junge

“Speech contains rhythm, expressed at nested frequencies. A recent discovery is that brain rhythms align with the rhythms of surrounding speech, which even infants can already achieve. Importantly, the degree to which infants exhibit this coherence can be linked to their language acquisition. For example, cross-sectional research demonstrates that the neural tracking of stressed syllables in nursery rhymes (at a rate of 1-3 Hz) at 10 months (but not 14 months) predicts future vocabulary development. These findings raise several questions: What are the developmental trajectories in neural tracking across different frequencies (stressed syllable < syllable < phonological rate) from infancy into toddlerhood, and is neural tracking of each rate equally important at each age?

To address these questions, we turn to the YOUth cohort, an ongoing longitudinal cohort study following Dutch children from preterm to childhood, aiming to capture meaningful individual differences. EEG measurements were collected at three time points (5 months, 10 months, and around 3 years of age). During these assessments, children viewed clips in which two female actors alternately sang Dutch nursery rhymes. Language outcomes were also assessed at the final wave using the Peabody Picture Vocabulary Test (PPVT).

Data collection is complete (>500 children participated). In the coming months, we will analyze the trajectories of neural tracking at the three frequencies. We will also compare the predictive value of neural tracking at each frequency rate for each age group. Our study seeks to confirm and expand upon existing findings by examining a larger group of neurotypical children and sampling at three time intervals into early childhood. Additionally, we aim to explore the implications of these neural tracking patterns for language acquisition at a later age than ever sampled before.”

Phonological Acquisition Depends on the Timing of Features

Authors: Katharina Menn; Claudia Männel, Lars Meyer

“The infant brain is characterized by slow electrophysiological activity, which limits initial processing abilities to environmental information that is slow. Nevertheless, infants start acquiring the phonemes of their native language during their first year of life—which is paradoxical, given the short lifespan of phonemes. We here show that phoneme acquisition hinges on the timing of phonological features, which for instance specify phoneme class, manner, or place of articulation. While individual phonemes alternate quickly with an average duration of ~50 ms in infant-directed speech, these features often span sequences of multiple subsequent phonemes, thus fitting infants’ slow processing.

We traced the emergence of feature-based phoneme representations that are known to govern speech processing in the mature brain. We collected electroencephalogram (EEG) data from a final sample of 66 children aged 3 months to 4.5 years while they listened to stories in their native (German) and an unfamiliar language (French). Categorical processing of features was assessed using EEG deconvolution (Temporal Response Functions). Our cross-sectional analysis uncovers a gradual developmental increase in neural responses to native phonemes, but not to non-native phonemes. Critically, infants seem to acquire those features first that extend over longer time intervals—thus meeting infants’ slow processing abilities. Shorter-lived phoneme features are added stepwise, with the shortest acquired last. Post-hoc analyses indicate that a feature’s similarity to pitch contours best predicts its age of acquisition, indicating that phonological acquisition may extend upon neural entrainment to prosody.

Our study highlights the role of electrophysiological maturation in shaping early language acquisition. We suggest that the duration of the acquired units is critical to the learning trajectory. We also suggest that parental speech adaptations might temporally align them to their infants’ processing abilities.”

07/16/2024, from 10:30 AM to 12:30 PM , Room P217

Symposium: Developing Chinese as a first language in US, Hong Kong and Beijing

Speakers: Ziyin Mai; Jingyao LIU; Mengyao SHANG; Arthur KAN; Virginia Yip

Three newly constructed open-access corpora documenting development of Chinese in preschoolers in three sites and four studies comparing different aspects of grammatical development based on the corpus data

Acquiring Chinese in US, Hong Kong and Beijing: three new corpora and three verbal structures

Authors: Ziyin Mai; Mengyao Shang, Jingyao Liu, Shanshan Yan, Stephen Matthews, Virginia Yip

“This paper presents key architectural features of three recently completed child Mandarin corpora and an analysis based on the data: The Child Heritage Chinese Corpus (CHCC) and Hong Kong Child Mandarin Corpus (HKCMC) document the development, maintenance and shift of language dominance in US and Hong Kong children (n = 4 in each site), all acquiring Chinese and English in largely “one context-one language” input models. The US children come from varying Chinese heritage backgrounds, whereas the HK children were addressed primarily in Mandarin at home and exposed to English through intensive English-medium pre-schooling. Both corpora recorded adult-child interaction sessions in unstructured toy-play activities regularly and longitudinally over two or three years (starting age from 1;7 to 4;11), with either Chinese or English as the designated language in a session.

The Beijing Child Mandarin Corpus (BJCMC) was constructed to address the absence of systematic multimedia corpora of monolingual child Mandarin at preschool age and the practical need for a monolingual reference point for our bilingual children. BJCMC maximally matches HKCMC in age range, SES, and recording setup and recorded 48 children aged from 3;0 to 6;9 cross-sectionally. Transcription and data analysis are in progress.

Our preliminary analysis found comparable grammatical complexity between the US and HK children between 2;10 and 4;8. The US children begin to lag behind the HK children after 4;8, presumably due to dramatically reduced input and experience of Mandarin in the US context. Closer examination of three complex verbal constructions (ba-construction, resultative verb compounds and postverbal prepositional phrases) is in progress. These verbal structures, which were shown to be vulnerable in school-age heritage bilinguals in our previous studies (Mai et al., 2021, etc.), are predicted to show early signs of vulnerability in our bilingual pre-schoolers, compared with the Beijing monolinguals.”

Marking exclusive focus in Mandarin-English bilinguals: syntactic, discourse and prosodic means

Authors: Jingyao LIU; Ziyin Mai

“Interface properties have been shown to be vulnerable to cross-linguistic influence (CLI) and input reduction, and are thus the locus of delayed acquisition in bilingual children. Our study examines the development of two focus particles (FPs) – only in English and zhi(you) in Mandarin in four Mandarin-English bilingual preschoolers in the Hong Kong Child Mandarin Corpus (HKCMC).

Only and zhi(you) express exclusive meaning by identifying the intended focus within its c-commanding scope and negating the overt or covert contextual alternatives to the focus (e.g., wo ZHI mai shui “I-ONLY-buy-water (but nothing else)”). Focus manifestation in only and zhi(you)-utterances display systematic similarities and differences in syntactic preference, discourse sensitivity and prosodic marking. Experimental studies have shown that focus identification presents great challenges to monolingual children in both languages (Notley et al., 2009), but so far little research has examined spontaneous only/zhi(you)-utterances in naturalistic settings.

We extracted 434 only/zhi(you)-utterances produced by the bilinguals from the corpus. Results showed an overall target-like performance in syntactic positioning and semantic association of the FPs. Interestingly, the children were able to spontaneously employ discourse means to provide contrastive contexts, disambiguating the sentences in terms of focus interpretation. Nevertheless, our bilinguals did not use prosodic cues to indicate focus in either language, even in obligatory contexts where contrastive prosodic stress is required in English. Overall, our results reveal early acquisition of syntactic, semantic and discourse properties and late acquisition of prosodic properties attached to only/zhi(you) FPs in bilingual children, highlighting uneven development across different linguistic features within one complex structure. To what extent this is attributable to the bilingual acquisition conditions is being evaluated through comparing the HKCMC children with monolingual baselines in both Mandarin (Beijing Child Mandarin Corpus/BJCMC) and existing English baselines in CHILDES.”

Noun-Modifying Clause Constructions in Mandarin-English bilingual children across contexts

Authors: Mengyao SHANG; Ziyin Mai, Stephen Matthews, Virginia Yip

“This study investigates the naturalistic production of Mandarin Noun-Modifying Clause Construction (NMCC) in Hong Kong bilingual children and US heritage children. As an under-studied construction, NMCCs share structural similarities with Relative Clauses (RCs). However, NMCCs do not necessarily involve a syntactic gap as typical RCs do but solely rely on a semantic-pragmatic link between the head noun and the modifying clause, as in (1) where ka ‘card’ is associated pragmatically rather than syntactically with [women kai youxi] ‘we start the game’.

(1) [MODIFYING CLAUSE] de HEAD NOUN

[ women kai youxi] de ka

We start game de card

‘the card for us to start a game’ (Sophie 6;10)

Given that NMCCs are pervasive in spoken Mandarin but are rarely studied in the acquisition literature, our study focuses on the non-RC type of NMCCs in child Mandarin across two acquisition contexts in 8 children from Child Heritage Chinese Corpus (CHCC) and the Hong Kong Child Mandarin Corpus (HKCMC).

We compared 420 NMCCs produced by the 8 bilingual children with 347 NMCCs produced by monolingual children from 7 existing corpora in CHILDES. We found that non-RC type NMCCs are produced less frequently by the bilinguals, possibly attributable to influence from English, in which structural equivalents of NMCCs are rare. Production of non-RC type NMCCs is further reduced in the heritage bilinguals (children in CHCC) than the Hong Kong bilingual children in HKCMC, who had more access to Mandarin in the larger environment. The generally reduced production of non-RC type NMCCs in bilinguals may be caused by reduced input in Mandarin, especially for child heritage speakers acquiring Mandarin as a minority language in English-dominant contexts. Our findings contribute to the growing literature on the diversity of bilingual development in different contexts.”

Verb-particle and directional verb constructions in Mandarin-English bilinguals

Authors: Arthur KAN; Ziyin Mai, Stephen Matthews, Virginia Yip

“This study presents the first multifactorial analysis of verb–particle constructions (VPCs) in English and directional verb constructions (DVCs) in Mandarin in a dataset comprising bilingual children raised in two types of “one context-one language” conditions. English VPCs allow two word orders: VOP (put the book down) and VPO (put down the book), except when the object is pronominal (VOP: put it down vs. VPO: *put down it). The DVC in Mandarin is a structural equivalent of VPC in English, which strongly prefers the VPO order. Despite extensive research into cross-linguistic influence (CLI) in bilingual acquisition, how CLI is modulated by amount of input is still not well understood. Given the cross-linguistic differences, we predict bi-directional CLI effects interacting with input conditions in English VPCs: the VPO order should be more prominent in children with more input in Mandarin.

We extracted 1937 VPCs from two Mandarin-English bilingual child language corpora (Child Heritage Chinese Corpus, Hong Kong Child Mandarin Corpus) and English monolinguals in the Manchester Corpus. Word order (VPO vs. VOP) was fitted to a generalized logistic mixed-effects regression, which includes Group (US bilingual, HK bilingual, English monolingual), MLUw, Language dominance (English-dominant vs. Balanced vs. Mandarin-dominant) and Type of object (lexical vs. pronominal) as fixed factors. Results indicate a significant difference in proportion of VPO between the bilinguals and the English monolinguals, rather than between the two groups of bilinguals, indicating general differences between monolingual and bilingual learners regardless of fine-grained bilingual input conditions. In addition, an interaction effect between population and object type was found, suggesting shared knowledge by all groups in distinguishing between lexical NPs and pronominals. Analysis of DVCs is in progress with DVC tokens being extracted from the two bilingual corpora and the new Beijing Child Mandarin Corpus.”

07/16/2024, from 10:30 AM to 12:30 PM , Room P429

Symposium: Language and communication in children with neurodiverse pragmatic profiles

Speakers: Núria Esteve-Gibert; Elena Castroviejo; Zuriñe Abalos; Isabel Martín-González; Albert Giberga; Nadia Ahufinger; Clara Andrés-Roqueta

We bring together studies that use different methodologies to understand how linguistic, cognitive, or learning difficulties impact the pragmatic skills of children with DLD and ASD profiles.

Narrative coherence in autistic and typically developing children

Authors: Zuriñe Abalos; Begoña Vicente & Elena Castroviejo

“This study investigates the frequency of different types of causal Rhetorical Relations (RRs) produced by 40 autistic and 53 typically developing (TD) Spanish-speaking children, matched in verbal mental age (6;0-11;11 years). Since causality plays a pivotal role in constructing coherent narratives, identifying differences in the acquisition of causal RRs between the typical and atypical groups could contribute to enhancing our comprehension of the development of narrative coherence among autistic children. Specifically, we examine the frequency of explicit/implicit (1), objective/subjective (2) and relevant/irrelevant (3) causal RRs produced by children in a narration task (Frog, where are you?).

(1) a. The jar broke because the dog fell. (Explicit)

b. The jar broke. The dog fell. (Implicit)

(2) a. The boy got mad at him because he smashed the jar. (Objective)

b. They’ll buy another jar, because it’s broken. (Subjective)

(3) a. He was scared because the bees were chasing him. (Relevant)

b. They yelled her name, because she may already have a name. (Irrelevant)

In addition, narratives are coded following the Narrative Scoring Scheme (NSS), to obtain a more complete understanding of children’s macrostructure (coherence) skills. Previous works examining discourse production in autism have revealed persistent challenges in organizing coherent narratives, establishing causal connections among events, and providing relevant information. Considering these findings, we hypothesize that autistic children will (i) show a lower frequency of causal RRs overall, particularly exhibiting fewer explicit, subjective and relevant causal RRs, and (ii) score lower in the NSS. Furthermore, whereas a linear developmental trajectory is expected within the TD group concerning the frequency of causal RRs (and NSS scores) (Sah, 2015), we hypothesize that the autistic group will exhibit a non-linear trajectory, given that narrative coherence extends beyond mere structural language skills.”

That child is a grasshopper! ASC children’s processing of novel metaphor

Authors: Isabel Martín-González; Kristen Schroeder, Camilo R. Ronderos, Elena Castroviejo, Ingrid L. Falkum, and Agustín Vicente

“Autistic young children seem to develop metaphorical capacity at a slower pace than TDs (Van Herwegen & Rundblad, 2018) but reach levels close to TDs in offline measures like multiple choice. However, they still present atypicalities in processing measures (Vulchanova et al. 2019). We developed a combined paradigm to explore both offline (picture selection) and online (eye-tracking) performance in ASC young children (3 to 10 years old) when comprehending novel metaphors; considering their verbal mental age/chronological age and balancing world knowledge/reducing uncertainty when comparing them to their TD peers. This paradigm was used to study TD’s development of metaphorical skills.

In our task, children had to select one picture out of 4 options: the correct one, the competitor, and two distractors. Children heard: Grasshoppers jump a lot. That {child, animal} is a grasshopper. Which one is it? In the literal condition, they would hear “animal”, and in the metaphorical condition, “child”. The pictures were the same in both conditions (in this case, a jumping child would be correct in the metaphorical condition and a grasshopper would be the competitor). Children’s proportion of fixations to the metaphorical image vs. the literal image (i.e., their preference in terms of attention) was measured for two critical time-windows: the target word region (grasshopper within the metaphor) and the question region (which one is it?).

To be included in the study, the ASC children had to present IQ within the typical range and be fluid speakers.

We plan to conduct our analysis using R software to build a mixed effects model with the following general structure: response ~ Conditiongroup + (1+Condition || Participant) + (1+Conditiongroup || Item). The model will have the same structure for the eye tracking data.”

The use prosodic and gesture information to process pragmatic meaning in DLD

Authors: Albert Giberga; Alfonso Igualada, Nadia Ahufinger, Mari Aguilera, Ernesto Guerra, Núria Esteve-Gibert

“Previous studies show that the combination of linguistic prosody and body gestures aid children and adults in accessing pragmatic meanings (Armstrong et al., 2018; Crespo-Sendra et al., 2013; Morett et al., 2021). Here we investigate whether children with Developmental Language Disorder (DLD) are more likely to benefit from prosodic and gestural cues than Typically-Development (TD) children when processing pragmatics. Because structural language abilities impact on pragmatics comprehension (Katsos et al., 2011), when structural language is compromised the presence of prosodic and multimodal cues can be particularly beneficial to help access pragmatic information.

A total of 39 children with DLD and 39 TD children aged 5 to 10 participated in a visual-world eye-tracking task. We manipulated two factors: the type of meaning to be processed (literal interrogative sentences; nonliteral indirect requests ˗ the first being less pragmatically complex than the second), and the presence of multimodal cues (prosodic cues; prosodic+gesture cues; no-enhancement). We predicted that the accumulation of multimodal cues would help children with DLD to choose the target image faster and with more accuracy, compared to TD children, especially when processing pragmatically-complex utterances.

Offline results revealed that prosodic as well as prosodic+gesture cues help children in processing less-complex interrogative utterances (p > 0.001; p > 0.001, respectively), with no interaction with participants’ group, while prosodic+gesture cues were especially useful when children with DLD process more-complex nonliteral requests (p > 0.001). Online eye-tracking data revealed that children with DLD divide their looks between target and competitor images more so than TD children (them being more accurate in looks to target), but no significant difference in the timing of these looks. This study informs about the types of cues that can help children with language impairment have successful communicative interactions.”

The role of emotional prosody to identify emotions: an eye-tracking study with children with DLD

Authors: Nadia Ahufinger; Mari Aguilera, Ernesto Guerra, Llorenç Andreu, Mònica Sanz-Torrent, Oriol Verdaguer, Coral Mayo

“Previous studies have suggested that prosody plays a central role in emotional comprehension. Moreover, emotional prosody comprehension is essential for social communication, but it is still unknown whether children with DLD exhibit emotional prosodic comprehension deficits to identify the emotions of others. Studies examining semantic and prosodic cues with eye tracking showed that prosodic meanings can be overridden by semantic cues when linguistic information is task relevant. This pilot study investigated the effects of emotional prosody to identify emotions in Catalan-Spanish children with and without DLD.

Participants were 22 children with DLD and 22 typical development children (sex and age paired; Mage = 9,7; 10 girls per group). The eye-tracker-software iMotions was used for two experiments. Four images (database of Karolinska Directed Emotional Faces) of the same person expressing different emotions (happy, sad, angry, scared, neutral) were presented. Participants were asked to listen to some sentences (subject, verb and predicate). In Exp, the emotions were said explicitly (e.g., “Laura is happy”) and in Exp2 the emotions were said implicitly driven from the verb (e.g., “Cristina escapes from the dog”). Half of the trials were prosody-semantically congruent, and the other half were prosody-semantically incongruent. Participants had to click on the expression face that matched with the oral cue.

Preliminary results showed that behavioral and eye tracking results showed for both groups that the semantic target was significantly preferred in both congruent and incongruent conditions. In congruent conditions, accuracy was better than in incongruent conditions. TD children outperformed children with DLD in terms of faster and larger preference for the target in both experiments.

Emotional prosody is a key dimension for emotional understanding ability, especially in population with DLD. Evidence in the emotional field is needed to design and implement a comprehensive intervention combining language and psychological aspects.”

Measuring receptive pragmatic skills in children with Neurodevelopmental Disorders with PleaseApp

Authors: Clara Andrés-Roqueta; Raquel Flores-Buils & Alfonso Igualada

“Pragmatic skills allow children to use language with social purposes. Nevertheless, pragmatic difficulties are observed in children with Autistic Spectrum Disorders (ASD) and Developmental Language Disorders (DLD), among other disorders (Saul et al., 2023). Andrés-Roqueta and Katsos (2017) pointed out that there are few assessment instruments that offer quantitative and qualitative data about pragmatic skills, and they are usually focused on specific aspects. In parallel, digital instruments are expected to compensate for executive function, communication and motivational difficulties when assessing children with developmental disorders (Lorah et al., 2015).

The aim of this study is to test whether PleaseApp (Andrés-Roqueta et al., in press) is an adequate and reliable instrument to assess pragmatic age-appropriate skills in children with ASD and DLD. This digital instrument assesses eight receptive pragmatic skills of children from 5 to 12 years-old: figurative language, narrative, reference, indirect speech, visual and verbal humour, gesture-speech integration, politeness, and complex intentionality.

Methods. A group of typically developing (TD) children aged 5-12 years, and two groups of children with ASD and DLD were recruited. Psychometric properties were examined by confirmatory factor analysis within the TD group, and between group comparisons (age differences and inter-group differences with clinicals groups) were examined with a one-way Anova and a Tukey’s test was performed.

The eight levels presented an adequate fit and good reliability. Moreover, significant age differences were observed within the TD sample in five out the eight levels. Finally, when comparing clinical groups to TD group, results revealed a lower performance of both groups (ASD and DLD) in the eight pragmatic components of the PleaseApp. PleaseApp is presented as a good tool for assessing age-appropriate receptive pragmatic skills in children with ASD and DLD, and it also helps to clarify the specific problems that a child has to better plan adapted interventions.”

07/16/2024, from 10:30 AM to 12:30 PM , Room P200

Symposium: Bilingualism and autism: Children’s abilities across linguistics domains and via varied contexts

Speakers: Tamara Sorenson Duncan; Gonzalez-Barrero Ana Maria; Megan Gross; Natalia Meir

This collection of research discusses key domains of language development and its acquisition in varied contexts for a growing population: bilingual children on the autism spectrum.

English L2 morphosyntactic development in bilingual autistic and neurotypical children

Authors: Tamara Sorenson Duncan; Johanne Paradis, Olivia McMurtry & Lydia Samis

“Understanding bilingual development among autistic children requires examining the influence of both their autism and their bilingual environment. Among monolinguals, whether autistic children show relative strengths in morphosyntactic development compared to other linguistic subdomains is debated (i.e., the so called “form vs. meaning” debate; Naigles & Tek, 2017). Among neurotypical bilingual children, length-of-L2-exposure, age and vocabulary size can be associated with individual differences in L2 morphosyntactic development (Paradis, 2023). Accordingly, this study asked: (1) Are neurotypical bilinguals more advanced than same-age autistic bilinguals in L2 morphological accuracy and use of complex sentences? (2) Are L2-exposure, age, and L2-vocabulary size related to children’s morphosyntactic development similarly for both groups?

The English-L2 morphosyntax of autistic (n=23) and neurotypical (n=26) children, (mean=6;10; range=4;7-9;6) from diverse L1 backgrounds was examined using transcribed and coded conversational language samples. Morphosyntactic variables were use of complex sentences and accuracy with tense and non-tense morphology. Length-of-L2-exposure and age were gathered via parent questionnaire and vocabulary size was measured using the PPVT. Linear regression modelling with Group (neurotypical, autistic), L2-exposure (months), Age (months) and L2-vocabulary (raw scores) as predictor variables revealed: (1) The autistic group used fewer complex sentences and were less accurate with non-tense morphemes than the neurotypical group; no group differences emerged for tense morphemes. (2) For both groups, children with more L2 exposure used more complex syntax, older children were more accurate with tense morphology, and children with larger L2-vocabularies used more complex syntax and had greater accuracy with both morpheme types. Overall, our findings show that autistic bilingual children can have lower morphosyntactic abilities than neurotypical bilingual age peers into the school years. Nevertheless, the same individual difference factors predicted variance in morphosyntax for both groups. We conclude that bilingual development in autistic children shares characteristics with autistic monolinguals and with neurotypical bilinguals.”

Lexico-Semantic Skills of Bilingual Autistic and Non-Autistic Children

Authors: Gonzalez-Barrero Ana Maria; Gabrielle Morin

“There is a growing number of autistic children who speak two or more languages. Although there is an increasing interest in the language development of these children, more research has focused on one of their languages. In this study, we used a verbal fluency task to examine the language and cognitive skills of bilingual autistic and non-autistic children. Verbal fluency tasks provide key information on lexico-semantic skills (Shao et al., 2014) and they also tap into two challenging areas in autism: language and executive functioning. Our objective was to examine the temporal dynamics of a verbal fluency task including bilingual children’s two languages.

Twenty-four bilingual school-aged children (12 autistic and 12 non-autistic) participated in the study. Children were given 1 minute to name all the elements they knew for a given category. Items produced in the bilingual children’s dominant and non-dominant language were compared. Groups were closely matched on key variables that influence verbal fluency performance (e.g., age, NVIQ). Responses produced in each language spoken by the child (i.e., dominant and non-dominant) were coded based on the specific time in which they were produced.

Results showed that bilingual autistic children were faster and produced more words in their dominant relative to their non-dominant language. We are currently analyzing the temporal dynamics of bilingual non-autistic relative to bilingual autistic children. Results from this analysis will provide insight into the cognitive profile of bilingual autistic children. For instance, if it is found that autistic children are slower at producing words compared to their non-autistic peers, future research could explore further the processing styles of this population and how they relate to theories of slowing processing speed in autism (e.g., Zapparrata et al., 2022). These results can also inform future evidence-based interventions by tailoring them toward the processing style of bilingual autistic children.”

Language choices of bilingual children diagnosed with autism: Monolingual & bilingual interlocutors

Authors: Megan Gross; Lavender Probasco, Nancy Garcia & Rudolph Lucier

“Although research on bilingualism and autism has expanded considerably in the past decade, most studies continue to focus on skills in the societal language and group comparisons with monolinguals, rather than examining bilingual development across both languages and the variability that exists within bilingualism. In particular, there has been limited work among autistic children on code-switching, a common bilingual language practice that serves important pragmatic functions (e.g., Yu, 2016). In the current study, we examined language choices, including code-switching, by bilingual children with an autism diagnosis when interacting with both monolingual and bilingual conversation partners in relation to their parent-reported language environment at home, in school, and in therapy.

To date, four Spanish/English bilingual children with a diagnosis of autism, ages 5-7, have participated in structured and unstructured interactions with monolingual and bilingual adults over Zoom. Structured interactions included a scripted-confederate dialogue task in which children described pictures with three different partners: a bilingual who code-switched frequently, a Spanish-speaking monolingual, and an English-speaking monolingual. Unstructured tasks included interactive story games across multiple sessions with the same bilingual experimenter.

The children’s families had different regional backgrounds (El Salvador, Mexico, Peru, Spain) and lived in different parts of the U.S. All four children attended English-only schools, although three received at least some therapy services in both languages. Preliminary analyses revealed variable patterns, including some children who code-switched mostly during unstructured tasks and others who code-switched primarily during a structured task or not at all. Given the variability among participants, we will present an in-depth profile of each child’s observed language use, as well as family language practices and experiences with code-switching. Although these individual profiles do not allow for broad conclusions, they will illustrate the variability that exists within both autism and bilingualism and inform factors to examine in future work.”

Different paths to bilingualism in Autism Spectrum Disorder (ASD): Natural and Unexpected

Authors: Natalia Meir; Iris Hindi

“Several case studies have provided evidence of unexpected language acquisition among individuals with ASD (Smith & Tsimpli, 1995; Zhukova et al., 2021). Unlike other bilingual children with and without ASD, who acquire both languages through interactions with family and community members, some children with ASD achieve second-language learning through non-interactive multimedia. This study examines unique profiles of children with ASD who acquire a new language that is not the ambient language of their environment and explores the reasons and implications of this exceptional phenomenon of multilingualism in ASD.

We recruited three groups of children aged 4-10: (1) children with typical language development, born to English-speaking homes, who acquired English as their Heritage Language and Hebrew as the Societal language (BiTLD, n=20); (2) children of the same profile with ASD (BiASD-HL, n=12); and (3) children with ASD who learned English via non-interactive multimedia (BiASD-IT, n=14).

Background questionnaires were administered to all children. Children with ASD were assessed using ADOS-2 (Lord & Rutter, 2012) to determine ASD severity. Morpho-syntactic skills were measured using LITMUS Sentence Repetition tasks (Marinis & Armon-Lotem, 2015). Extensive task batteries tested children’s verbal and non-verbal theory of mind (ToM) skills. All children were tested in English and in Hebrew. Children and parents were interviewed regarding the mode of acquisition and language use preferences.

No group differences were observed in morpho-syntactic performance in English and Hebrew. Verbal and non-verbal ToM abilities varied across groups, with higher performance in the BiTLD group. Follow-up analyses revealed that the main reason for learning a new language in the BiASD-IT group was to enhance Internet activities.

The findings demonstrate that paths to language acquisition in ASD can vary: children can acquire their two languages through interactive input at home and in educational settings, as well as through non-interactive media.”

07/16/2024, from 10:30 AM to 12:30 PM , Room P301

Symposium: Morphosyntactic difficulties in young children with DLD: milestones, measurement and interventions

Speakers: Rob Zwitserlood; Anouk Scheffer; Luisa de Heer; Gerda Bruinsma; Ellen Gerrits

This symposium discusses grammatical growth in Dutch young children with DLD, different procedures for goal setting and measuring therapy effect, and several grammatical intervention approaches researched in various age groups of children with DLD.

Morphosyntactic development of 3- to 6-year-old Dutch children with DLD

Authors: Anouk Scheffer; Brigitta Keij, Britt Hakvoort, Esther Ottow-Henning, Ellen Gerrits, & Frank Wijnen

“Background:

Most children with a developmental language disorder (DLD) have morphosyntactic difficulties. This is suggested to be the most important clinical marker of DLD. However, whether this is valid for young children with DLD, who are in the earliest stages of their morphosyntactic development, is unclear. This study investigates the complexity, diversity, and accuracy of the morphosyntactic repertoires of 3 to 6-year-old Dutch children with DLD. We compare their morphosyntactic repertoires to those of typically developing (TD) children matched on language level.

Method:

Language samples of 59 children (29 children with DLD and 30 TD children) were analyzed on three dimensions; morphosyntactic complexity, diversity, and accuracy. The TD children and DLD children were language-matched on their morphosyntactic development using the levels of the Dutch version of the Language Assessment, Remediation, and Screening Procedure (TARSP). The children with DLD (age range 2;7-5;4 years, mean age 4;1 years) were older than the language-matched TD children (age range 2;0-3;9 years, mean age 2;9 years).

Results:

Children with DLD are comparable to language-matched TD children in the accuracy and diversity of their morphosyntactic repertoires, but they produce less complex utterances. In the presentation we will also discuss an additional analysis on another group of 30 children with DLD. For these children, we explore the growth patterns of the morphosyntactic complexity, diversity, and accuracy of their utterances over a period of three months.

Conclusion:

The results indicate that children with DLD lag behind in their grammatical complexity as compared to language-matched TD children. In the presentation we will describe the morphosyntactic skills of Dutch children with DLD and what our results mean for clinical practice.”

Language sampling analysis and production tasks in preschoolers with DLD: A match made in heaven?

Authors: Luisa de Heer; Iris Duinmeijer, Inge van Dijke, Lisanne Geurts, Anouk Scheffer

“Introduction:

Morphosyntactic difficulties are a core symptom in children with Developmental Language Disorder (DLD). In order to determine treatment goals for this language domain it is vital to have a good understanding of children’s linguistic abilities. Language sampling analysis (LSA) is often seen as the gold standard in language assessment, as it reflects abilities in a natural situation. At the same time, language production tasks are frequently used for assessing morphosyntactic skills because they can target specific structures. This presentation delves into the question whether LSA and language production tasks are complementary in the assessment of morphosyntactic skills of young children with DLD.

Methods:

A cohort of 33 preschoolers with DLD (age range 2;10-3;9 years; mean age 3;6 years) were tested with LSA and a newly developed grammatical production task. LSA was performed using the Dutch version of the Language Assessment, Remediation, and Screening Procedure (TARSP). The grammatical production task covered both syntactic and morphological structures known to be difficult for children with DLD. The occurrence of structures and the identified morphosyntactic developmental stage were compared between these two methods.

Results:

Preliminary results from six preschoolers show that LSA and the production task produce different results in all participants. The production task seems to demonstrate greater sensitivity in detecting the use of structures such as negations, while LSA seems more adept at identifying others structures, such as the use of the copula. In our presentation, we will present the results of the total sample.

Conclusions:

LSA and language production tasks seem to be complementary in assessing morphosyntactic abilities. In our presentation, we will discuss how LSA and language production tasks can be combined to tailor goal setting and evaluation of grammatical interventions for children with DLD to their morphosyntactic abilities.”

Effect of a focused stimulation intervention on morphosyntax of children with DLD

Authors: Gerda Bruinsma; Frank Wijnen, Ellen Gerrits

“Background:

Children with Developmental Language Disorder (DLD) struggle to acquire morphosyntactic rules, which limits their ability to communicate. While efficacy studies have demonstrated positive effects of interventions on morphosyntax in controlled research settings, there is a pressing need to assess intervention effects in real-world clinical practice.

The Dutch ‘Language in Interaction Therapy’ (LIT) aims to incorporate effective elements into a practical therapy approach to stimulate morphosyntax. LIT is based on functional language use within natural contexts, employing an implicit learning approach using focused stimulation, recasting, and production elicitation. We studied the effect of LIT on morphosyntactic development in children with DLD by comparing LIT to usual care speech and language therapy.

Method:

Using a time-series design, fourteen children with DLD received LIT, while another fourteen were in a non-equivalent usual care control group (age range 4;1 – 5;9 years). All children attended schools for special education. The study spanned one school year (40 weeks) with measurements every 4-6 weeks. Initially, both groups received usual care, but the LIT group’s therapy was temporarily replaced with LIT targeting morphosyntax for a 12-week period, while the control group continued their usual care addressing various therapy goals. Outcome measures included Mean Length of Utterance (MLU) and morphosyntactic complexity, assessed through spontaneous language samples during play and story retelling, as well as performance on a sentence repetition task.

Results:

Preliminary results indicate that MLUs and complexity increased in both groups. The LIT group did not show significantly more progress than the control group. There was substantial inter-individual variation. We will also present results of additional analysis exploring these differences in relation to both child and speech-language therapist (SLT) characteristics.

Conclusion:

Children with DLD receiving intervention show improvements in morphosyntax. Implementing effective elements from existing efficacy studies does not necessarily increase effectiveness.”

Effectiveness of a serious game for grammatical therapy in Dutch school-aged children with DLD

Authors: Rob Zwitserlood; Ingrid Singer, Annemarie Kerkhoff, Ellen Gerrits

“Background:

Serious game ‘Bouke Bouwt’ to enhance morphosyntax in children with developmental language disorder (DLD) was developed in co-design with researchers, speech and language therapists (SLTs), game designers, and children with DLD. This metalinguistic therapeutic game uses various game-mechanics to keep children motivated. Children are motivated to play games, but SLTs feel uncertain whether games for morphosyntax therapy are effective.

Method:

Twenty-four children with DLD (mean age 8;3 years) participated. Seventeen children were multilingual. All children visited mainstream schools and had SLT in private practices. Using a single-case design, we compared eight weeks of regular grammatical therapy (control condition) with eight weeks of therapy using the game (experimental condition). Both conditions contained home assignments. The weekly 20-minutes sessions targeted various morphosyntactic goals. In the control condition, SLTs used different techniques and programmes, in the experimental condition the game with additional tangible material was used. The game contains targets ranging from simple sentences with 4-6 constituents to complex sentences. Treatment fidelity was monitored with logbooks. Grammatical growth was measured using language sample analysis of story retelling tasks and CELF-5-NL Sentence Repetition task (CELF-5-NL-SR).

Results:

At group level, no difference between conditions were found for CELF-5-NL-SR and for LSA measures MLU, percentages mean constituents, and percentages complex syntax. However, percentage grammatically correct sentences did increase significantly after the experimental condition. Over the whole 16-week treatment period children told longer stories, but no growth in grammatical complexity was found. Furthermore, substantial individual differences between participants were observed. We will present data from the planned retention measurement and from the individual children. Both children, parents, and SLTs were highly motivated to use the game.

Conclusion:

Possibly, sixteen weekly sessions of 20-minutes grammatical therapy with homework is not enough to show growth in complex syntax at group level, in a study targeting diverse morphosyntactic goals.”

07/16/2024, from 10:30 AM to 12:30 PM , Room P018

Symposium: Input studies across languages and communities in Meso-South America

Speakers: Andrea Taverna; Lourdes de León; Cecilia Rojas; Celia Rosemberg; Susana Mendive

The study of children’s communicative environments has engendered considerable research over the years with findings indicating there are important cross-cultural differences across communities. The present symposium provides an overview of research in Meso-South America (e.g., Mexico, Argentina, and Chile) that portrays different disciplinary approaches and methods to study input and language development in the context of socio-economic, cultural, and linguistic diversity of this region.
The first presentation focuses on the developmental trajectory underlying the acquisition of verb and noun morphology in children learning Wichi, a polysynthetic language. It assesses the role of input from the environment in the early production of morphology in spontaneous speech data from a longitudinal study.
The second paper examines the role of interactional formats (IF’s) in child-directed-communication in Mayan Tsotsil. It identifies the pragmatic functions of IF’s and their association with different caregivers of multigenerational families. Results show that grandparents and siblings play key roles in providing IF’s to young children.
The third is a case study employing conversational analysis in a corpus of spontaneous interactions in Mexican Spanish, monolingual, educated middle-class families. The analyses focuses on the different types of conversational moves that caregivers deploy in response to toddlers’ messages when trying to achieve mutual understanding.
The fourth examines variations in quantitative and qualitative properties of child-directed and overheard speech in a longitudinal corpus of audio recordings in socioeconomically diverse households of Argentinian Spanish speaking toddlers.The analyses also considers SES and input effects on children’s vocabulary comprehension.
Finally, the last study examines the relationship between parents’ level of education, home literacy environment, and the use of language by adult and child during shared book reading at home in a sample of 114 low-SES Chilean children aged 2 and 3 years.

07/16/2024, from 02:00 PM to 04:00 PM , Room P104

Symposium: Unraveling the predictive role of gesture in children’s linguistic and cognitive development

Speakers: Ingrid Vilà-Giménez; Júlia Florit-Pons; Lucía Domingo-Moscardó; Begum Yilmaz; Ying Li

This symposium aims to provide interdisciplinary insights that unravel the predictive nature of gesture in shaping linguistic and cognitive development across different populations, learning environments and developmental stages.

Multimodal skills but not motor skills predict language skills in neurotypical and clinical children

Authors: Júlia Florit-Pons; Mariia Pronina, Alfonso Igualada, Pilar Prieto, & Courtenay Norbury

“Research has examined the relationship between either motor (i.e., non-communicative limb movements) or multimodal skills (i.e., communicative movements involving hand gestures, body movements, and facial expressions) and linguistic skills, highlighting the predictive role of multimodality (see Hübscher & Prieto, 2019), and with unclear findings for motor skills (Leonard & Hill, 2011). To our knowledge, no previous study has assessed altogether the relationship between both motor and multimodal abilities and narrative and pragmatic skills across neurotypical (NT) children and children with neurodevelopmental disorders (NDD), who might have impairments with multimodal and motor skills. For this, the current study aims to compare the predictive power of multimodal versus motor skills for narrative and pragmatic abilities in two groups of children (NT and NDD).

Participants of this study were 140 children (57 girls; 87 NT and 53 NDD) aged 3 to 7, who were administered different tasks measuring linguistic, narrative, pragmatic, multimodal, and motor skills.

A first correlational analysis showed positive correlations between multimodal skills and narrative and pragmatic skills in both populations, while motor skills were only found to positively correlate with narrative skills in neurotypicals. Additionally, two multiple regression analyses were run, with narrative and pragmatic abilities as dependent variables. The models indicated that multimodal, but not motor skills, were significant predictors of both narrative and pragmatic skills. Language, age, and group also contributed to the prediction of narrative and pragmatic competence. The interaction between group and multimodal abilities was not found to be significant, suggesting that the relationship between multimodal and narrative and pragmatic skills is not different depending on group.

The results of the study reveal a close relationship between multimodal, narrative and pragmatic abilities in NT and NDD populations, highlighting that communicative gesture skills (i.e., multimodality), as opposed to non-communicative motor skills, predict narrative and pragmatic competence.”

Multimodal narrative perceptive assessment serve to identify language needs of children with NDD

Authors: Lucía Domingo-Moscardó; Júlia Florit-Pons, Pilar Prieto, & Alfonso Igualada

“Multimodal narrative patterns have not been clearly established comparing children with typical development (TD) and neurodevelopmental disorders (NDD), although previous evidences of gesture’s role as a language compensation mechanism in NDD children with communication needs. Research has shown that professional judgment serves to identify language needs, but this methodology has not been implemented under a multimodal perspective. This study aims to investigate: a) how a novel multimodal narrative perceptive assessment outcomes converges with independent evidence from multimodal narrative analysis and b) how the two methodologies can predict language needs.

A total of 162 4-to-7-year-old children explained a wordless cartoon. Two assessment methodologies were applied. First, a multimodal narrative perceptive assessment served to rate four different multimodal narration profiles that measured children’s gesture use associated with oral information: compensatory (gestures replacing lack of oral information), transitional-learning (gestures reinforcing almost every oral production), discursive (gestures adding relevant information) and self-regulatory (gestures as an emotional regulator). Second, a multimodal narrative analysis measured microstructure (lexical, syntactic, discourse-markers and referents), macrostructure and gesture measures (rate, type, semantic function and self-adaptors).

Preliminary results with 10 TD and 10 NDD children revealed that both multimodal assessments converged evidences to identify children’s language needs. First, the transitional-learning profile was more frequent in the NDD group. Also, there was a positive correlation between compensation profile with sentences without verbs and different gesture measures. In the TD group, there was a negative correlation between discursive profile and self-adaptors use. Second, the multimodal narrative analysis showed that the NDD group had lower macrostructure and microstructure scores (lexical, discourse-markers and referents) and more extending function gestures than the TD group. These findings showed that a multimodal perceptive assessment is a valid and rapid measure of language abilities. Future regression analyses will investigate whether both multimodal methodologies can predict children’s language needs.”

The link between fluent gesture rate, vocabulary and narrative skills in 7- to 9-year-old children

Authors: Ingrid Vilà-Giménez; Pilar Prieto

“Previous research has highlighted the strong link between both receptive and expressive vocabulary skills and narrative performance (Uccelli & Páez, 2007). While longitudinal evidence indicates that gestures produced in narrative retellings can predict narrative performance (Demir et al., 2015), the concurrent association between narrative abilities and gesture rate remains unclear. Our study aims to assess the link between 7- to 9-year-old children’s narrative structure and fluency scores, and their vocabulary knowledge and production of fluent referential iconic and non-referential gestures.

Participants were 83 children (43 girls) aged 7 to 9 who completed a narrative retelling task. All narratives (n = 166) were coded for duration, narrative structure and fluency using standard scales, and for referential iconic (referring to semantic content in speech) and non-referential (lacking semantic content) gesture rates. Receptive vocabulary was assessed using the Peabody Picture Vocabulary Test-III (Dunn et al., 2010) adapted into Catalan.

A set of six GLMM analyses were conducted, with narrative structure and fluency as dependent variables. Two initial models showed a positive association between vocabulary and narrative structure scores (R2 = 60%) and a negative association between narrative duration and fluency scores (R2 = 74%). When referential iconic gesture rate (n = 212) was introduced in the two models, both vocabulary and referential iconics positively predicted narrative structure (R2 = 48%). Referential iconics were also positively associated with fluency, while narrative duration showed a negative association with fluency (R2 = 69%). Conversely, the addition of non-referential gesture rate (n = 236) did not account for additional variance in either narrative structure or fluency scores.

While our findings corroborate previous research in highlighting the importance of vocabulary as a good predictor of better-structured narratives, they crucially provide novel evidence for the predictive role of referential iconic gesture rate in both narrative structure and fluency scores.”

The Predictive Role of Gestures and Parental Input in Mathematical Development:An Intervention Study

Authors: Begum Yilmaz; Işıl Doğan, Emel Nur Kaya, Dilay Z. Karadöller, Ece Demir-Lira, & Tilbe Göksun

“Early mathematical skills are crucial for later academic achievement. One of the predictors of early variability in children’s proficiency in these skills is parental numerical input (Levine et al., 2010). Parents provide numerical input through different modalities, such as speech or gestures (Karadöller et al., under review). Previous work suggests that both math instructions accompanied by gestures (Singer & Goldin-Meadow, 2005) and children’s gesture use (Novack et al., 2014) help them solve math problems. However, research remains scarce on the relation between multimodal input, children’s gesture, and subsequent math proficiency. We examined the role of multimodal input on children’s math development through a parent-administered book-reading intervention study. We assessed how preschoolers’ gesturing and parental use of gestural math input interactively relate to children’s math development.

Sixty-three children (33 girls, M age=49.9 months, SD=3.68, Range:45-59) were pretested on their verbal counting and cardinal principle knowledge (i.e., “What is on the card?”) tasks, followed by a six-week book-reading intervention period. Intervention Group 1 received a book with numerical language prompting parents to use gestures; Intervention Group 2 received the same book and asked not to use gestures. The control group received a book without numerical language and gesture instructions. At the end of the intervention period, children were post-tested on mathematical proficiency. Posttest analyses are ongoing to investigate the group differences with varying parental input patterns. Preliminary results for the pretest showed that children’s gestures while labeling objects were related to their cardinal principle knowledge, regardless of the group (r=-.380, p<.01). Children who used more gestures had lower performance on “What is on the card?” task after controlling for age and verbal counting skills (p<.01). Contrary to previous work, our findings suggest that children might use gestures to compensate for their proficiency in mathematics, highlighting the differential functions of gestures in mathematical development.”

Neural synchrony during parent-child spatial problem-solving: Role of parent gesture strategy

Authors: Ying Li; Ece Demir-Lira

Synchronous interactions between parent and child are fundamental for children’s development. Prior work examining synchronous interactions focused on the relations to children’s social-emotional development (Nguyen et al., 2020), but such interactions play a vital role in children’s cognitive development as well (Casey et al., 2000; Mahy et al., 2014). Gestures constitute a pivotal element in parent-child communication. However, little is known about how and why gesture might relate to the neurocognitive basis of parent-child interactions. Here we leveraged fNIRS (functional near-infrared spectroscopy)-based hyperscanning, which measures neural data from all participants simultaneously using multiple devices, to assess the role of parental gesture in semi-natural parent-child interactions in the laboratory. We applied fNIRS-based hyperscanning in a cooperative spatial task (tangram puzzle). Two main brain areas from both hemispheres, the dorsal lateral prefrontal cortex (dlPFC) which is relevant to problem-solving, and the temporoparietal junction (TPJ) area which is relevant to social mentalization, were covered using 8 by 8 fiber distribution. During the task, parent-child dyads were asked to cooperate to solve the tangram puzzle together. All task-relevant parental gestures during the interaction were coded based on video recordings. Neural synchrony between the parent and child pairs was measured by WTC. The results revealed higher WTC scores in regular pairs than in random pairs, (F(1, 287) = 99.2, p < 0.001). Gesture played a role leading a higher overall WTC (t = 2.555, p < .05). Also, gesture negatively predicted slope of WTC during the interaction (t = -3.251, p < 0.01). Although WTC decreased during the interaction, presence of gesture diminished the decreasing slope WTC compared with no-gesture. WTC predicted dyad’s performance on task (number of puzzles solved). Overall, our study is the first study to pinpoint how gesture relates to dynamic changes in neural synchrony during parent-child interactions using fNIRS-based hyperscanning.

07/16/2024, from 02:00 PM to 04:00 PM , Room P018

Symposium: Discourse development: argumentative, explanatory and narrative skills

Speakers: Maia Migdalek; Alejandra Stein; Alejandra Meneses; Jarmila Bubikova-Moan; Martha Shiro

The purpose of this symposium is to present an integrating view of discourse development by reviewing recent studies on how children acquire argumentation, explanatory and narrative discourse skills.

A longitudinal study of the discourse units in mealtime family interactions

Authors: Alejandra Stein; Maia Julieta Migdalek, Celia R. Rosemberg

“The aim of the study is to contribute to the research on the discursive environment at home in which the development of narration, argumentation and explanation takes place during preschool years. The development of these discursive forms is relevant to the linguistic, cognitive and social dimensions of child development and to how their mastery constitutes a requirement to participate in classroom practices (Heller, 2014). Although the discursive units are intertwined in everyday interactions, previous studies have mainly analyzed each form in isolation without addressing how they are deployed together and how they configure the texture of the conversation.

In the present paper we analyze the structural and interactional organization of the narratives, explanations and arguments produced during mealtimes in Argentinian Spanish monolingual homes. The corpus consists of 22 mealtime situations audio-recorded in 11 middle-income households at two time points, when children were 4 and 5 years old (Author, 2013-2019). Following Quasthoff et al. (2017) we identified the discourse units as segments delimited from the surrounding turn-by-turn talk, each type serving a different communicative function.

Preliminary findings from the statistical analysis carried out on 181 discourse units showed that narratives were the most frequent and complex units (they included other embedded units) as compared to argumentations and explanations. Most narratives were intraconversational, while explanations and argumentations were mainly triggered by an object or an action salient in the situation. Children globally initiated 24% of the units. They contributed to most of the narrative accounts, but mainly listened to the argumentative and the explanatory units produced by adults. The findings of this study contribute to the understanding of home discursive practices, which are relevant to the requirements of classroom discourse, from a perspective that comprehensively contemplates different forms of discourse, as well as diverse interactional and structural aspects involved.”

Time reference in justifications as argumentative strategies used in disputes of children at play

Authors: Maia Migdalek; Florencia Alam, Martha Shiro, Celia R. Rosemberg

At an early age already, children at play participate in confrontational interactions using argumentative strategies. As they take a stance in a dispute, they justify their position in different ways. The purpose of this study is to determine the time framework in which the justifications are inserted and how they contribute to the child’s argumentative sequence. Our hypothesis is that justifications that refer to the past imply the reconstruction of the event from a specific perspective; reference to the future or hypothetical events serve as justifications of a course of action that can be perceived as acceptable or not; reference to the present brings evidence supported by some aspect of the situational context of the interaction; non-temporal justifications display generalizations that the child has derived from the social and natural environment. Focusing on how children choose the time reference in justifications helps us understand the type of evidence they consider relevant to defend their point of view and to what extent this choice contributes to the acceptance or not of their argument. Furthermore, we can ask ourselves whether the children’s socio-economic status influences the choice of time reference in the justifications. Thus, we analyzed 432 hours of recordings of 37 Argentinian 4 year-old children from a socioeconomically diverse population (Author, 2005-2012) to extract the justifications present in disputes during naturalistic play situations. Our preliminary findings suggest that most children at this age choose justifications entrenched in the present, whereby the evidence is based on the situational context. However, the mid SES children used an anticipatory framework more frequently than their low SES peers, indicating thus possible differences in the cultural practices of the two groups with respect to what constitutes a strong evidence in the argumentative interaction.

The development of scientific explanation: relations between genre and lexicogrammatical resources

Authors: Alejandra Meneses; Maximiliano Montenegro, Evelyn Hugo, Daniela Acevedo, Javiera Figueroa

This research explores language development during the school-age years, focusing on extended discourse, decontextualized and academic language (Grøver et al., 2019). Students have difficulties when they need to engage in this type of discourse genres at school because they require new linguistic forms and functions (e.g., Silvestri, 2002; Uccelli, 2019). One of the pivotal discourses learned during primary education is explanation, particularly relevant to science learning. Scientific explanations require students to construct causal mechanisms to display how scientific phenomena occur (affirmation), supported by empirical scientific data (evidence) (Fitts et al., 2020; McNeill & Krajcik, 2012). In this study, 117 Chilean boys and 152 Chilean girls in 4th grade participated in a language and science-integrated learning sequence. Each student wrote four explanations, two at the beginning and two at the end of each instructional unit. A total of 986 scientific explanations were coded by their use of genre (affirmation and evidence) and lexicogrammatical resources (scientific vocabulary, cross-disciplinary vocabulary, metadiscursive vocabulary, nominalizations, causal links, evidential markers, self-mentions, and contextualized language). The results indicated a significantly higher dominance of the final explanation over the initial explanation in affirmation construction, particularly in the identification of the relevant components (F(1, 487) = 334.55, p < .001, η2 = .442), as well as in producing evidence, especially in the specification of the data used (F(1, 487) = 295.12, p < .001, η2 = .429). Regression models revealed that scientific vocabulary, causal links, and a lower presence of metadiscursive vocabulary and contextualized language predict affirmation quality (R2=0.376). In comparison, evidence quality (R2=0.698) is predicted by evidential markers, metadiscursive vocabulary, and causal links. These findings emphasize the distinct roles that various resources play in constructing disciplinary genres.

Children’s argumentative competence: untangling the concept

Authors: Jarmila Bubikova-Moan;

Much like argument and argumentation, argumentative competence (AC) is a complex and contested term. It has, for example, been conceptualized as a threefold set of argumentation-relevant skills at play in a person’s argumentative performance: metacognitive (declarative), metastrategic (procedural) and epistemological (Rapanta, Garcia-Mila & Gilabert, 2013). The manifestation and further development of AC as a metaknowing competence has been studied especially in children in early and late adolescence, considered developmentally ready for argumentation. Referred to as the skills-view of AC, it has been questioned, and investigations of argumentative contributions in children as young as preschool have been offered. Providing a broad conceptual overview of advances in the field, a recent meta-synthesis has documented that studies of argumentative discourse in the youngest age groups have been on the rise especially in the last decade. Nonetheless, the review has also warned of the terminological and conceptual heterogeneity in the field and called for a careful scrutiny of the potential overlap between central terms and concepts in circulation. This paper aims at taking a closer look at the terminological and conceptual conundrum by reviewing some of the argumentation-related terms in use in early childhood education and care, such as sustained shared thinking, inferential thinking and exploratory talk, and by examining in more detail key empirical and theoretical literature, concerned with conceptualizing what AC in the early years may entail. Based on this, a tentative conceptualization of AC will be proposed that takes into account the youngest children’s argumentative contributions and accommodates them as competent arguers in their own right. Practical implications for future studies of children’s AC will also be considered.

07/16/2024, from 02:00 PM to 04:00 PM , Room P131

Symposium: Using multimodal corpora to study language development

Speakers: Gabriella Vigliocco; Ruthe Foushee; Ingrid Vilà-Giménez; Abdellah Fourtassi

Investigations of language and communicative development using corpora of multimodal language

The impact of caregiver’s multimodal behaviours on word learning: Insights from the ECOLANG corpus

Authors: Gabriella Vigliocco; Ed Donnellan, Yan Gu, Antonia Jordan de Barros, Beata Grzyb, Gwen Brekelmans, Margherita Murgiano

Studies have shown the importance of caregivers’ multimodal behaviours (e.g., prosody, gestures, gaze) on children’s word learning. However, most studies focus on only one specific behaviour (e.g., only prosody). Here, we investigate which multimodal caregiver behaviours best predict word learning and vocabulary growth. Using data from the ECOLANG corpus, we analysed caregiver (N = 36) behaviours in semi-naturalistic interactions with their child (3 to 4 years old) in which they talked about known and unknown toys. Measures of learning of the unknown words (using a picture-word matching task after the interaction) and of vocabulary (at the time of testing and one year later) were obtained. We analysed caregivers’ use of multimodal cues while labelling the objects, specifically their use of yes/no questions, pitch, representational gestures, pointing, object manipulations and gaze. We separately assessed the impact of the frequency/mean scores for each cue and difference scores (differences between caregiver behaviours when naming known and unknown toys). We used logistic mixed effect models to assess the effect of multimodal cues on immediate learning, and linear regression to assess vocabulary growth. Preliminary results show that only caregivers’ pitch, use of yes/no questions and pointing predicted children’s word learning. In particular, higher pitch when labelling unknown toys predicted immediate word learning while the pitch difference between known/unknown toy’s labels predicted both immediate learning and vocabulary growth. Furthermore, the difference in use of yes/no questions between known and unknown toys predicted immediate learning, while the frequency of yes/no questions when naming unknown toys predicted vocabulary growth. Lastly, caregiver pointing predicted immediate learning and vocabulary growth, but in the opposite direction from pitch: the more they pointed towards known toys, the better children’s learning of novel toy labels. Overall, these results provide evidence for the important role of multimodal caregiver behaviours on children’s lexical development.

Nonverbal communication between hearing caregivers and their deaf and hearing children

Authors: Ruthe Foushee; Michelle Madlansacay, Zena Levan, Susan Goldin-Meadow

“What is caregivers’ role in language acquisition? The literature often emphasizes the caregiver as a source of linguistic feedback and language data. Here, we turn our attention from caregivers’ provision of linguistic evidence to examine the nonlinguistic evidence that caregivers may provide children regarding what it is possible to communicate, and what they can expect from future interlocutors.

To do so, we capitalize on data from two longitudinal studies featuring naturalistic videorecordings of children and caregivers (Goldin-Meadow & Mylander, 1984; Goldin-Meadow et al., 2014). We focus on caregivers’ responses to early (often partial or imperfect) communicative attempts in: (1) deaf children born to hearing parents—particularly relevant because these children cannot easily process the verbal linguistic evidence that their parents provide—and (2) hearing children of hearing parents. To shed light on the role of nonverbal communication in the development of language, we match the children in these two groups on the basis of syntactic complexity (MLU in gestures and words, respectively), and compare the behavior of their hearing caregivers along two dimensions: contingency, and pragmatic appropriateness.

Method. For each child communicative attempt (an utterance and/or gesture), we code contingency based on whether the parent acknowledges or responds within 2s—and if so, in what modalities (e.g., touch, speech, gesture, action). Pragmatic appropriateness indexes caregivers’ attributions of communicative intent to their children, and their visible efforts to interpret them. To code it, two research assistants first independently ‘gloss’ the communicative attempt, then evaluate the relevance and sensitivity of the caregiver’s response relative to this interpretation.

We present descriptive results revealing both commonalities and differences in the multimodal feedback that hearing caregivers provide their deaf and hearing children, as well as statistical tests using these dimensions of caregiver nonverbal responsiveness to predict conversational, structural, productive, and semantic language outcomes in children.”

Children’s multimodal development: Insights from a corpus of Catalan narrative speech

Authors: Ingrid Vilà-Giménez; Júlia Florit-Pons, Patrick L. Rohrer, Sara Coego, Pilar Prieto

“Studying gesture-speech interactions at the discourse level is a necessary step for gaining a better understanding of how children acquire multimodal language. The aim of this presentation is twofold: (a) to introduce the features of the “”Audiovisual corpus of Catalan children’s narrative discourse development”” (Vilà-Giménez et al., 2023, https://osf.io/npz3w/); and (b) to delve into the specific contributions made using this corpus.

The audiovisual corpus consists of 332 narratives carried out by 83 children (43 girls) at two time points in development (at 5-6 years old and two years later). All narratives were annotated for prosody (Cat_ToBI), gesture referentiality (referentials and non-referentials, following the M3D labeling system), and Information Structure (IS; LISA guidelines). On the basis of these comprehensive annotations, five studies were carried out. The first study explored gesture-speech temporal alignment patterns, revealing that by ages 5-6 children already closely associate the stroke of both gesture referentiality types with pitch-accented syllables. The second and third studies investigated the relationship between IS, prosody and gesture referentiality at both time points, and found that while gestures are sensitive to IS categories, pitch accentuation is not a stable marker of IS. More refined analyses are still ongoing. The two final studies demonstrated that (a) children’s referential iconics produced in narratives at 5-6 years predict their narrative structure scores two years later; and that (b) referential iconics produced at 7-9 years are positively associated with both narrative structure and fluency scores at that age.

Altogether, these studies help refine our knowledge about multimodal development. In this presentation, we will delve into these findings and explore other possibilities that this longitudinal narrative corpus offers for gaining valuable insights into the multilevel analysis of discourse speech in the window spanning between 5 and 9 years of age.”

A developmental corpus of face-to-face natural conversations in middle childhood

Authors: Abdellah Fourtassi;

“Existing studies of naturally occurring talk-in-interaction have largely focused on the two ends of the developmental spectrum, i.e., early childhood and adulthood, leaving a gap in our knowledge about how development unfolds, especially across middle childhood.

The current work contributes to filling this gap by introducing a developmental corpus of child-caregiver conversations at home, involving groups of children aged 7, 9, and 11 years old (N=30). We capitalized on recent advances in mobile, lightweight eye-tracking and head motion detection technology to optimize the naturalness of the recordings, allowing us to obtain both precise and ecologically valid data. Further, we mitigated the challenges of manual annotation by making full use of large data processing tools in speech processing and computer vision.

In order to demonstrate the usefulness of this multimodal corpus in the study of face-to-face communicative development, we present preliminary results on children’s use of gaze to regulate conversational turn-taking. Methods. To quantify gaze aversion, we used computer vision tools to categorize continuous gaze data from the mobile eye-tracking device. First, the face of the interlocutor was detected from the egocentric video camera using a state-of-the-art computer vision algorithm for face detection. Second, the gaze coordinate data (overlaid on the pixel space of the egocentric video) were used to determine when gaze coordinates did or did not intersect with the interlocutor’s face. Results. Preliminary results show that older children (compared to younger ones) and adults (compared to children) spent more time gazing at interlocutors while listening compared to speaking, suggesting that children in middle childhood are still developing in terms of the systematic use of gaze in regulating conversational turn-taking.”

07/16/2024, from 02:00 PM to 04:00 PM , Room P217

Symposium: Applying the LUNA framework to analysing children’s personal narratives: Clinical implications

Speakers: Marleen Westerveld; Vani Gupta; Sara D. Ferman; Mateja Gabaj; Kristine Jensen de López

This symposium illustrates how the LUNA framework may be used when analysing personal narrative samples produced by children with and without identified language difficulties across pragmatics, macrostructure planning, propositional, and/or linguisti

A systematic review of methods for eliciting and analysing school-age children’s personal narratives

Authors: Vani Gupta; Marleen Westerveld, Stephanie Malone

“Talking about past experiences, also referred to as personal (event) narratives, is a key form of human interaction across age, culture, and community. To date, personal narrative development has been researched across different disciplines, using a range of terminology and definitions, making it difficult for speech therapy clinicians to translate the findings into practice. To address this gap in present knowledge we systematically reviewed the literature regarding personal narrative assessment and analysis methods, focusing on children ages 4 to 18 years with and without identified language difficulties or disorders.

The pre-registered systematic review of six databases yielded 17,531 articles and theses published in English. Only empirical studies in which children talked about a specific event that had previously happened to them were included (i.e. considered as ‘personal narratives’). Included studies were further required to analyse language structure (e.g., macrostructure), form (e.g., syntax, morphology), or content (e.g., topics, themes, vocabulary) within the personal narratives elicited. Studies in which adults (including caregivers) elicited, prompted or co-constructed personal narratives were included, but studies in which only the adult’s narrative was analysed (and not the child’s) were excluded.

After title and abstract screening, 412 articles met our selection criteria for full-text screening; After full-text screening 207 articles were eligible for inclusion in the systematic review. Data extraction and synthesis of the data includes detailing the assessment tasks as well as the analysis methods. Data analysis methods will be categorized into LUNA components (pragmatics, macrostructure planning, propositional, and linguistic). Further, the review will identify which measures were found to differentiate between children with identified language difficulties and their peers without language difficulties. The presentation will provide a detailed overview of the findings, outline which LUNA components have been studied, and identify potential gaps in the existing research. Clinical implications will be highlighted.”

Personal narratives of Hebrew and Arabic speakers in Israel: Investigating pragmatic and linguistic

Authors: Sara D. Ferman; Khaloob Kawar

“This study investigated how the discourse context (face-to-face [F2F] vs. tele-communication [TC]) affected the linguistic performance of two groups of 10-year-old children living in Israel (Hebrew-speaking Jewish children and Arabic-speaking children) and whether children’s language/culture influenced their performance. A total of 38 Arabic-speaking and 51 Hebrew-speaking children produced personal narratives in response to the six emotion-based prompts of the Global TALES protocol. Forty participants were assigned to a TC group (via Zoom) and 49 to a F2F group. Personal narratives were analyzed on the following linguistic measures: total number of words (TNW), total number of utterances (TNU), number of different words (NDW), mean length of utterance in words (MLU-W). At macrostructure level each narrative was categorized according to its topic.

Analysis revealed no significant main effect of discourse context (F2F vs TC) on any of the linguistic measures. However, discourse context influenced the chosen topics. Furthermore, the Hebrew-speaking children obtained significantly higher scores compared to their Arabic-speaking peers, and language/culture also influenced the chosen topics. Finally, a significant interaction effect between language/culture and discourse context was found, with a significant main effect for the Arabic speaking children only. The Arabic-speaking children produced more utterances in the F2F condition compared to TC.

These findings suggest that the discourse context (pragmatic component) may influence (1) the topics 10-year-old children choose to narrate (macrostructure), and (2) the linguistic performance of children from some cultures more than others. These results will be discussed in light of potential differences in the interpersonal quality of communication between TC and F2F discourse contexts, taking differences between the collectivist-Arab and individualistic-Jewish cultures into account. Altogether, the results support the notion of the LUNA framework that pragmatic factors (discourse context) and language/cultural factors should be considered when analyzing the linguistic features of children’s personal narrative discourse performance.”

Emotional valence of topic in the personal stories of children with developmental language disorder

Authors: Mateja Gabaj; Jelena Kuvač Kraljević, Marleen Westerveld

“When children are encouraged to tell personal event stories, they need to access their episodic memory to select a topic (either pleasant or unpleasant) and convey the semantic content of the event in a coherent and linguistically appropriate manner. Although it is known that children with developmental language disorder (DLD) have difficulty with linguistic aspects of discourse, this is the first study to examine both macrostructural planning and propositional components as per the LUNA framework. The macrostructure component investigates what children choose to talk about (topic or story content). The propositional component includes the semantic content needed to relate a past personal event about the chosen topic that can be categorised as positive or negative depending on emotional valence.

The aim of this study was to determine (1) the topics that children with DLD talk about in their personal narratives, (2) the frequency with which they recall (un)pleasant experiences with positive or negative outcomes, compared to children with typical language development (TLD). The study involved 32 children with DLD (M = 10;09 years) and 33 children with TLD (M = 10;01 years) whose personal narratives were elicited using six emotion-based prompts from the Global TALES protocol.

Both groups of children share similar topics, such as academic achievement or success in sports or competitions, especially in stories prompted by negative prompts or referring to problems and important events. In terms of the type of event and the emotional ending of the story, children showed the greatest variability when talking about important events. Overall, children with DLD talk more often about unpleasant events than pleasant ones, which, however, have a positive outcome in the majority of cases (53.33%). In contrast, children with TLD talk more often about pleasant events, with a higher proportion of positive endings (76.26%). Clinical implications are discussed.”

The personal narrative skills of school-age students on the autism spectrum

Authors: Marleen Westerveld;

“Sharing personal experiences through storytelling (personal narration) is a fundamental aspect of social communication. Despite the importance of personal narrative proficiency for participating in daily activities, and the knowledge that many autistic children show challenges in social communication, research into the personal narrative skills of this group of children is limited. The current study addressed this knowledge gap by analysing personal narrative samples produced by autistic children in response to the six emotion-based prompts of the Global TALES protocol.

Ten autistic children (8 male) participated (ages 7;11 – 12;05). All children scored within age expectations on tests of receptive vocabulary (PPVT) and nonverbal cognition. All personal narratives (stories; max 6 per child) were transcribed and coded for topic. Stories containing at least two past tense actions were analysed at macrostructure level (global coherence / theme), and propositional level (context, chronology), using the Narrative Coherence Coding Scheme (NCCS).

The autistic children produced 46 personal narratives (max 60 = 77%), compared to a 95% response rate in our sample of 44 ten-year-old Australian children. Further, 50% (30) of the children’s stories did not contain two actions (thus not scored on the NCCS). Global coherence (context) ranged between 0 (23%; difficult to identify topic), 1 (66%; identifiable topic with only limited evidence of causal connections or evaluations, and 2 (10%; substantial development of topic). At propositional level, 70% of stories contained no information about time or place; with time and place mentioned in two stories. Finally, 60% of the stories contained minimal/no information about temporal order, with 13% of the stories containing an identifiable timeline.

The discussion will highlight the importance of personal narrative coherence for supporting social communication. A case example will be provided to illustrate how analysing personal narratives at macrostructure planning and propositional levels may inform clinical assessment and intervention practice.”

Language and narrative skills of Danish school children with DLD: Investigating child profiles

Authors: Kristine Jensen de López; Hanne Søndergaard Knudsen

“Language and narrative skills are often investigated as separate domains and evaluated by comparing group results rather than within child profiles. This study provides a detailed analysis of individual children’s narrative skills across different genres (pragmatics) at linguistic, macrostructure planning, and propositional levels and compares their performance to standardized language results and parental reported skills. It also investigates topics addressed in personal narratives.

Five Danish monolingual children (2 girls) with Developmental Language Disorder (DLD), aged 7- 10, participated in a range of language tasks (CELF; Parental reported Language on the CCC and the 5-15; Narratives: Global TALES personal narratives, MAIN telling, BUS retelling). All language samples were scored and/or coded, transcribed and analyzed for linguistics (MLUw, number of different words [NDW], errors), macrostructure, topic, and global coherence (CUDP-A) where relevant.

Results from standardized language tests and parent-reported language measures showed substantial language difficulties of all children compared to typically developing peers. The preliminary results comparing different narrative genres show that for personal narratives children with DLD overall addressed topics like those identified cross-linguistically in previous studies with typically developing children. However, some responses reflected lack of propositional and semantic knowledge necessary to respond to the prompts (e.g., proud). Global coherence in the children’s personal narratives ranged between 48-73% and did not align with parental reported coherence on the CCC. Children with the lowest standard scores on CELF were most challenged in producing narratives, although this was not consistent.

Across the three narrative genres, despite individual differences, children generally showed higher MLUw in personal narratives compared to their fictional narrative re/tells. There were large variations in NDW produced within each of the narrative genres, but relatively consistent productions of error types. The presentation will provide case examples of language profiles of Danish children with DLD to illustrate these results.”

07/16/2024, from 02:00 PM to 04:00 PM , Room P300

Symposium: Developing materials on healthy language development with and for Indigenous communities

Speakers: Shanley Allen; Randeana Peter; Melvatha Chee; Rebecca Defina; Catherine Dench; Carmel O’Shannessy

This symposium focuses on the development of language-related materials in Indigenous communities around the world, in collaborations between community members and academics, to revitalize and strengthen the acquisition of those languages.

t’ut’a’thut ‘un thathun: Beatboxing exercises for children learning Hul’q’umi’num’ sounds

Authors: Randeana Peter;

“hakwush ‘un shqwultun ‘i’ thuythut ‘un sqwal. ‘e’ut wulh t’ut’a’thut ‘un thathun ‘i’ ‘un tuhwthulh ‘i’ ‘un shhw’uthqun. This is how we begin our Hul’q’umi’num’ beatboxing alphabet. I am a Kindergarten and Grade 1 Hul’q’umi’num’ teacher at the Quw’utsun Smuneem Elementary School in my home of Quw’ustun (Cowichan). The Hul’q’umi’num’ language spans from Snuw’nuw’us (Nanoose) to Me’luxutth’ (Malahat) on Vancouver Island in Western Canada, and is an Indigenous Coast Salish language. Our communities have been working on revitalizing our language for many years.

My research is on how children learn sounds of Hul’q’umi’num’, which uses glottalization, ejectives, and complex consonant clusters not found in English. Pronouncing Hul’q’umi’num’ involves learning new muscle movements and I remembered what people said about speaking it: “Our people, when they spoke, it was like they were singing to each other,” and “the muscles we are using [when speaking Hul’q’umi’num’] aren’t used all the time.” This had me thinking about when my kids would hold their throat because it was sore.

I discuss here how I teach beat boxing warm-ups to help us speak and keep our Hul’q’umi’num’vocal muscles in shape. In this presentation, I will share examples of nonsense words and sentences that keep us walking around the house practicing everyday, reinforcing the kinds of movements that help us produce the new sounds. For example, in the following passage, all the consonants sounds are used for making beats and the vowels are used for harmony or a transition:

p’uq’ (white) lhsuq’ (half) shewuq (carrot)

p’, q’, s, lh, q, sh, w – consonants

u, e – vowels

Our Hul’q’umi’num’ speaking at home and school has become stronger because of these beatboxing activities that prime our muscles to work in a Hul’q’umi’num’ way, and has made us – both children and adults – more confident speakers.”

Linguistic investigation at Saad K’idilye: Developing evidence-based pedagogy to support urban Diné

Authors: Melvatha Chee; Warlance Chee (Saad K’idilyé), Cheryl Yazzie (Saad K’idilyé), Alec Goldberg (University of New Mexico)

“In collaboration with Saad K’idilyé, the Indigenous Child Language Research Center at the University of New Mexico has been documenting caretaker-child speech interactions. At Saad K’idilyé, in Albuquerque, NM, USA, children up to eighteen months of age are immersed in Diné Bizaad, the Navajo language. This presentation is based on one year of documentation from which a cross-section of 6 hours of video recordings were closely analyzed.

Through this work we discovered a set of frequently-produced words in child-directed speech that could not be easily categorized as verb, noun, or particle. These words are unique as they express a complete thought and are used to convey several but intersecting concepts. We refer to this interesting set of words as Navajo holophrases.

Navajo holophrases cannot be inflected, nor do they alternate form. These words recur throughout each day and are among the most frequent words in child-directed speech, precisely because they have a simple form alongside a salient meaning.

For example, na’ ‘here, I give you this’ has been identified as the fourth most common word in child-directed speech at Saad K’idilyé. Na’ was also produced by children within several months of being enrolled. The use of na’ circumvents the necessity to classify an object as the Navajo classificatory verb system typically requires. Rather, na’ can function for the giving of any object.

Thus, these words may allow the language to be penetrated and the acquisition process to begin. This presentation will show how we take this information and use it to teach parents so that they can use these Navajo holophrases with their children in the home. Saad K’idilyé also used this set of words to create original resources for speaking with the children in their care.”

Pitjantjatjara scaffolding techniques for early childhood education

Authors: Rebecca Defina; Katrina Tjitayi (Pitjantjatjara Yankunytjatjara Education Committee)

“Of the hundreds of Indigenous languages spoken in Australia before colonization, only 12 are still being learnt by children as a first language (AIATSIS 2020). Pitjantjatjara is one of these few languages and is in a relatively strong position within this context. It is used as the main language of everyday life by around 3000 people across the vast desert region of the Anangu Pitjantjatjara Yankunytjatjara (APY) lands in central Australia. Across the APY lands, there is also a network of early childhood education centers with strong Anangu (preferred self-reference for Pitjantjatjara-speaking people) in educator and leadership positions. However, the training, planning, and design of these programs is often led by research-informed pedagogy from other cultures and languages, particularly from Europe.

In this collaboration, we combine research on Pitjantjatjara child language acquisition with experience in Anangu early childhood education in order to create strength-based, locally appropriate resources. We examined a 15-hour longitudinal corpus of Pitjantjatjara children aged 10 months to 5 years interacting with their peers and caregivers. We described the scaffolding techniques commonly used by adults to engage children in talk and facilitate language acquisition, such as questions, attention direction techniques, and prompting. We then utilized this documentation to build research-informed resources and tools for Anangu and non-Indigenous early childhood educators working with Pitjantjatjara-speaking families. These resources include video examples and discussions, as well as suggested activities for classrooms based on local scaffolding practices.

This work seeks to build on the strengths of the Pitjantjatjara-speaking community and the research resources currently available, to build culturally and linguistically appropriate research-informed early childhood education resources. A main aim is to combat the deficit-based framing that is common in Indigenous education contexts throughout Australia.”

A collaborative process to formalize an Inuktitut child language developmental sequence and support

Authors: Catherine Dench; Annie Novalinga & Maaji Putulik & Malaiya Weetaluktuk (Inuulitsivik Health Cent), Nunia Anoee & Rhoda Karetak (NU Board of Ed), Shanley Allen (U Kaiserslautern), Shirley Tagalik (Aqqiumavvik Society)

“Inuktitut, a polysynthetic language, is the second largest Indigenous language in Canada, spoken by nearly 40,000 Inuit. With a very young population, and concerns about language loss, there is a pressing need for culturally and linguistically appropriate resources for early assessment and intervention in Inuktitut. In this talk, we describe the process we used to develop a community-validated and accessible Inuktitut preschool language development chart, as well as culturally appropriate language stimulation ideas for parents and educators.

Over the last forty years, a wealth of information about Inuktitut child language development has been gathered. Much of this information comes from Inuit elders and child development support workers. For example, Inuit Elders in Nunavut documented traditional childrearing knowledge and advice over a series of meetings, including information about child language development. This information was later published as a series of pamphlets organized by age and incorporated into Inuit Qaujimajatuqangit – the unified system of beliefs and knowledge characteristic of the Inuit culture, which is translated as “that which Inuit have always known to be true.” Another set of information on Inuktitut language development comes from extensive research by Allen, Crago and colleagues in small communities in Nunavik in Quebec, which has been published in a series of academic journal articles (e.g., Allen, 2017).

In a process based on Pence and Ball’s (2006) Generative Curriculum Model, we engaged in the collaborative construction of a set of language development materials combining knowledge from both Inuit knowledge and academic research, including community validation of the materials. This approach emphasizes “how Eurowestern self-assertive thinking and values can exist in creative dialogue with the more integrative thinking and values that are characteristic of many Indigenous cultures, resulting in positive transformations for all individuals, institutions, and communities involved” (Ball, 2004).”

Development of a CDI for multilingual children speaking Indigenous languages in Central Australia

Authors: Carmel O’Shannessy; Vanessa Davis & Denise Foster (Tangentyere Research Hub), Jessie Bartlett & Alice Nelson (Red Dust Role Models)

“There is a focus in Australian national government policy on early childhood development, yet little is known about young children’s paths of language development in Indigenous languages. This includes the language children hear directed to them, and that they therefore learn. Contexts in which there is language change and where people speak in more than one language are even less well understood.

There are MacArthur Bates Communicative Development Inventories (CDI) for about 90 languages world-wide, but until now none for the languages spoken in Central Australia. A CDI is a list of the most common words that young children up to age 3 years are likely to know and say.

In the Central Australian town of Alice Springs Indigenous children grow up hearing and learning more than one language, with differing degrees of multilingualism, and with differing access to literacy practices in their home languages. This context presents a specific challenge for describing children’s paths of language development and for developing a tool such as a CDI.

A team of Indigenous and non-Indigenous researchers worked with 22 families for data collection. The sessions with families had three parts: a) an interview where caregivers were asked which words their young children know and say, b) caregivers talking with the children about textless picture books, and c) free play and talk between children and families. The sessions were transcribed, and translated where they were not in English.

From this, an online multilingual ‘spoken’ CDI for five of the languages spoken by young children in Alice Springs has been developed, along with other resources in preparation. In this talk we outline the methods of the study and present the CDI and work in progress.

https://little-kids-learning-languages.net/

https://mywordlist.app/app/little-kids-word-list“

07/16/2024, from 02:00 PM to 04:00 PM , Room P301

Symposium: Cross-linguistic comparison in studies of morphosyntactic acquisition

Speakers: Virve Vihman; Gordana Hrzica; Adele Vaks; Caroline Rowland; Elena Tribushinina; Evan Kidd

This symposium focuses on studies comparing the acquisition of morphosyntax across unrelated Slavic, Finnic and Germanic languages, using a matched design. The studies discuss method, results and challenges arising in the cross-linguistic design.

Overgeneralisation and overabundance in the production of 5-year-olds: a crosslinguistic study

Authors: Gordana Hrzica; Mari Aigro, Sara Košutar, Tomislava Bošnjak Botica, Virve Vihman

“Children are known to produce variable forms for a single target. Some of these derive from the overgeneralization of patterns in the input (‘goed’ pro ‘went’). It has been proposed that children eventually retreat from overgeneralization errors via entrenchment and a probabilistic process of construction competition (Ambridge et al. 2015). A related issue is that the target language itself may contain parallel forms. Overabundance refers to a situation where more than one lexical form is available in a morphological paradigm cell, e.g. ‘puddle’ in Croatian (lokva-nom.sg, lokava ~ lokva ~ lokvi-genitive.plural) and Estonian (loik-nom.sg, loikusid ~ loike-partitive.plural). The acquisition of overabundance in a language appears to be in conflict with the simultaneous process of retreat from overgeneralisation.

This talk investigates how children acquiring Croatian or Estonian navigate nouns exhibiting overabundance in the target languages, and whether overgeneralisation follows the same patterns as overabundant nouns. We tested 139 four- to six-year-olds on noun forms with an elicited production task. The same design was used in both languages, using both overabundant noun stimuli, with at least two forms (genitive plural in Croatian, partitive plural in Estonian), and those with only one form.

The results showed significant main effects of age (improved accuracy with age in Estonian), differences between item types (lower accuracy for overabundant nouns) and language (lower overall accuracy in Croatian). Croatian children produced overgeneralised forms of overabundant nouns, using the most frequent genitive plural ending. Errors made by Estonian children were less consistent. Thus, effects of the linguistic system and properties of the input were found. We discuss overgeneralisation errors across age groups, and what overgeneralisation in the presence of overabundance in the target system reveals about entrenchment and form competition in cross-linguistic acquisition. We also discuss methodological challenges of cross-linguistic experimental paradigms and theoretical implications.”

Crosslinguistic influence in bilingual acquisition: the production of Estonian unreal conditionals

Authors: Adele Vaks; Virve Vihman

“Cross-linguistic influence (CLI) is one of the many factors contributing to the heterogeneity of bilingual experience. The direction and form of CLI depend on several factors like language dominance and typological differences. In a study with Estonian-Norwegian (NOR, N=24) and Russian-Estonian (RUS, N=20) bilinguals aged 5;0 to 7;10, we compared the results of a Sentence Repetition Task (SRT), focusing on structures that might reveal different effects of CLI from their respective languages. Estonian is spoken as heritage language by NOR and as majority language by RUS. However, we did not find differences in the two groups’ overall SRT and vocabulary scores.

In this talk, we focus on unreal conditionals. The structure proved most challenging for both groups, with mean accuracy scores at 43% for NOR and 38% for RUS. The conditional mood is distinctively morphologically marked in Estonian, while both Norwegian and Russian employ past tense forms, combined with a subjunctive particle (Russian) or modal verbs (Norwegian). Both groups frequently omitted the conditional suffix (36% of targeted verbs in NOR, 52% in RUS) and often substituted it with the past indicative (14% of targeted verbs in NOR, 26% in RUS). We discuss possible sources of variation in accuracy rates, considering the different language situations of the two groups. We also take a closer look at the strategies used by children, demonstrating that cross-linguistic transfer can be a resourceful way to deal with a complex structure when the child understands the semantic content, but lacks the grammatical resources to express it in a target-like manner. To better tease out what stems from CLI, we compare the bilinguals’ responses to each other and to those of their monolingual Estonian peers.”

The relationship between lexicon, morphology and syntax in English and Estonian

Authors: Caroline Rowland; Seamus Donnelly, Adele Vaks, Ada Urm, Piia Taremaa, Izabela Jordanoska, Tiia Tulviste, Virve Vihman

A long-standing question in language development is the nature of the relationship between early lexical and grammatical knowledge. The very strong correlation between the two has led some to argue that lexical and syntactic knowledge may be inseparable, consistent with usage-based theories that eschew a distinction between the two systems. However, little research has explicitly examined whether early lexical and syntactic knowledge are statistically separable and, if so, whether morphology patterns with the lexicon or syntax. Moreover, there are two underappreciated methodological challenges in such research. First, the relationship between the lexical, morphological and grammatical knowledge may change during development. Second, non-linear mappings between true and observed scores on the language measurement scales we use could lead to spurious multidimensionality. Here we overcome these challenges by using data from several time points and a statistical method robust to such non-linear mappings. In Study 1, we examined item-level vocabulary and syntax data on American English from the Wordbank database (data collected using Communicative Development Inventories, CDIs). In Study 2, we used two longitudinal corpora of American and British English, which used the same CDI scale at 18/19, 21, 24 and 30 months. In both studies, we found clear evidence that while there is clearly a very strong relationship between vocabulary and syntax knowledge in early language development, the two are clearly separable from about 18 months. In Study 3, we are extending the analysis to morphology, using data from Estonian, a language with a much more extensive and complex morphological system than English. The goal is to determine (a) whether the same patterns hold for morphology when we compare cross-linguistic data, and (b) whether morphology aligns with lexical or syntactic development. The project has implications for our understanding of the mechanisms underlying lexical, morphological and syntactic acquisition, which will be discussed.

The acquisition of pronominal gender in the two languages of Russian-Dutch bilinguals

Authors: Elena Tribushinina; Pim Mak

“The acquisition of grammatical gender has been extensively studied across languages and populations. Much less is known about how children acquire pronominal gender. This talk reports a study testing production of pronoun gender in the narratives of Russian-Dutch bilinguals in comparison to Russian and Dutch monolinguals. This language combination is theoretically interesting because Russian has a rich and transparent gender system, whereas the Dutch gender system is morphologically sparse and opaque. In addition, Russian pronominalizes based on grammatical gender, whereas pronoun gender in Dutch is largely determined by semantic principles. Hence, we expect both positive and negative transfer in children simultaneously acquiring Dutch and Russian.

The participants were 93 simultaneous Russian-Dutch bilinguals growing up in the Netherlands, 97 Dutch-speaking monolinguals and 93 Russian-speaking monolinguals (age range 4;1-7;11). In addition, we included 83 age-matched Russian monolinguals with DLD to compare the effects of reduced exposure in bilingualism with the effects of reduced intake in DLD. Narratives were elicited using two sets of pictures: the Fox Story and the Cat Story. The bilinguals produced one narrative in each language; the monolinguals were randomly assigned to one of the narratives. Each pronoun was coded as either correct or incorrect based on the gender of its antecedent. Data were analysed with a logistic regression.

The results reveal cross-linguistic differences in the acquisition pace. Russian monolinguals with and without DLD performed at ceiling from age 4 onwards, whereas Dutch-speaking monolinguals still made errors at age 7. Russian-Dutch bilinguals acquired Russian pronoun gender more slowly, and reached the monolingual level only by age 7. In Dutch, the bilingual children outperformed their monolingual peers from age 5 onwards, which suggests positive transfer from Russian. However, we also found transfer errors, where bilinguals used a pronoun compatible with the gender of the antecedent in Russian.”

07/16/2024, from 02:00 PM to 04:00 PM , Room P429

Symposium: Dynamic Systems in Early Language Development: A Multifaceted Exploration

Speakers: Elitzur Dattner; Dorit Ravid; Paul Ibbotson; Mira Calin Gatenyo; Ayhan Akcu Koc

The transition into syntax: dynamic relations in syntactic development using network analysis. Corpus studies of Hebrew and English child-caregiver dyads, focusing on the emergence of definite marking, formulaticity, and syntactic constructions.

The transition into syntax: Early syntax in a dynamic perspective

Authors: Elitzur Dattner; Dorit Ravid

In the pivotal transition between ages two and three, children undergo a remarkable linguistic transformation, moving from isolated words to complex syntax. This talk introduces a comprehensive framework that aims to capture this intricate process, focusing on the role of syntactic cognitive representation within a dynamic-systems oriented Usage-Based, Construction Grammar approach. At the core of this framework is the idea that syntactic cognitive representation involves an implicit understanding of which linguistic units are likely to follow others given a particular pragmatic/discursive/communicative needs. This understanding shapes the linear array of words that best convey specific semantic relations in diverse communicative contexts. Connection weights between these units are dynamic, changing in relation to their frequency of co-occurrence and the contextual, communicative information they carry. To illustrate this, the framework employs a network-based model where words act as nodes and word combinations serve as links. This network-based representation allows for the dynamic, non-linear nature of language development to be statistically captured and analyzed. Child Directed Speech (CDS) is posited as a crucial filter in this framework, highlighting the most frequent and meaningful lexical and grammatical contrasts. The talk will elaborate on how this framework can be applied to longitudinal speech data, offering insights into the evolving relationship between lexical and grammatical components over time. By introducing this framework, the talk aims to provide a nuanced understanding of the dynamic systems that underlie the complex process of syntactic development and CS-CDS relations. It emphasizes the role of cognitive representation in this evolution, offering a new lens through which to view and study the journey of early language acquisition.

Emergence of Hebrew definite marking in early mother-toddler interaction

Authors: Dorit Ravid; Naama Feldman, Orit Ashkenazi and Elitzur Dattner

“Definiteness (involving semantic-pragmatic notions such as uniqueness and familiarity) is a fundamental property expressed in many languages either lexically of morphologically, referring to a clearly identifiable entity in a given context. According to the developmental literature, children start producing a small number of definite articles with a restricted set of nouns towards their second birthday, while six months later definiteness is already used in most obligatory contexts. This rapid acquisition process requires young children to experience and learn the contexts of definite marking, with special focus on common ground, deixis and anaphoricity, as these pragmatic cues highlight the language-specific structures associated with the definite article in the ambient language.

The current study investigates the emergence of Hebrew definite marking by a toddler aged 1;9-2;2 and definite article usages in the CDS, in a 20,028-utterance corpus of mother-child interaction across six months of development. Structural coding involved identifying and classifying all occurrences of the definite prefix ha-, including morpho-phonological variations, on nouns in a range of syntactic roles and on adjectives in agreement with nouns. Functional / pragmatic coding examined the ways in which common ground (CG) was presented to the child (and used by her), including deictic, lexical, syntactic and discourse cues and strategies. Results indicated a clear association of child’s age with frequency and diversity of definite marking, with definite usage categories (both structural and pragmatic) becoming increasingly related to parental input with child’s age, as well as correlated with each other. The child started marking definiteness productively on nouns around age 2 years, with definite overmarking as the most common error. CG contexts and anaphors were most frequent in both CS and CDS, however earlier CS had many usages of deixis. This study demonstrates how parental input guides the acquisition of a complex pragmatic category in mother-child interaction.”

Emerging Formulaticity in Dynamic Networks of Speech: English and Hebrew comparisons.

Authors: Paul Ibbotson; Orit Ashkenazi and Elitzur Dattner

In the realm of language development, understanding the emergence and evolution of formulaic language structures is crucial. This study introduces a novel measure to quantify formulaiticity in dynamic networks of naturalistic child speech and child directed speech. The Clustering coefficient captures how interconnected the language network is, specifically the ratio of connected triangles to triplets in a directed network. Each time a new node (word) is added to the network, the clustering coefficient is recalculated, providing a measure of formulaticity that is sensitive to cumulative linguistic experience. The results are then rescaled to fall between -1 and +1, offering a nuanced view of how formulaic structures evolve over time. Using this framework, we analyze language corpora from two Hebrew-speaking and two English-speaking child-caregiver dyads, as well as an adult Hebrew dataset from the Israeli Parliament. Our findings reveal that the actual clustering coefficient navigates between the theoretical maximum and minimum formulaic values. As children’s multi-word uses accumulate, the network becomes increasingly complex, and interconnected, and the formulaticity, as measured by the clustering coefficient, drops. This dynamic measure allows us to compare formulaticity across developmental stages and between languages with very different morphosyntax, contributing to a more comprehensive understanding of language acquisition and use. This study highlights the importance of dynamic, context-sensitive measures in capturing the complexities of language development and cross-linguistic analysis.

Emerging syntax as a dynamically developing network: CS-CDS relations in early Hebrew

Authors: Mira Calin Gatenyo; Elitzur Dattner, Orit Ashkenazi and Dorit Ravid

Toddlers globally undergo a captivating linguistic journey, transitioning from single words to creative expressions. Usage-based syntactic theories propose that, in the course of meaningful interactions, toddlers grasp concrete structures through imitation and gradually form abstract syntax. In this study, we propose to conceptualize the cognitive knowledge of “syntax” as a dynamic network of links between words. We present a case study to explain syntactic development through dynamic systems analysis tools, characterizing the syntactic network evolution between a toddler and their mother from early word combination emergence to basic sentence formation (ages 1;8-2;2 years). 49 recordings from the Ashkenazi corpus, comprising natural parent-child interactions, were transcribed and encoded, culminating in 184,961 words based on syntactic categories and morphology. Toddler and mother transcriptions were separated, and each recording yielded an independent ‘syntactic network’ where each node represented a lexical/syntactic category (e.g., Determiner, Preposition, Noun.Fm, Verb.3sg.Fm.Past) and each link a connection between them. Measures of dynamic networks were analyzed (node count, network size, mean network distances, node eigenvector centrality, link betweenness centrality, importance hubs and network density) alongside the mutual influence of network measures between the toddler and their caregiver. Results indicated the expansion of the toddler’s network and an increase in the variety of syntactic categories and specific connections, leading to consistent production of syntactically-linked simple sentences. The caregiver’s utterances show stable syntactic richness relative to their toddler’s. Mutual alignment was found in terms of node centrality and link betweenness measures and alignment in the density of the Child-Directed Speech network relative to the Child Speech network as a function of the child’s age. These findings strengthen the notion of a dynamic syntactic network development within a communicative context.

07/16/2024, from 02:00 PM to 04:00 PM , Room P200

Symposium: Theory of Mind, Language and Cognition in children with Developmental Disorders

Speakers: Theo Marinis; Dafni V. Bagioka; Stephanie Durrleman; Eleni Peristeri; Franziska Baumeister; Mikhail Kissine

This symposium brings together studies addressing the relationship between ToM, language, and cognition in monolingual and bilingual children with DLD and ASD as well as studies using non-verbal and low-verbal ToM tasks.

Testing first- and second-order Theory of Mind through novel verbal and non-verbal False Belief task

Authors: Dafni V. Bagioka; Maria Andreou, Anna Czypionka, Angelika Golegos, Theodoros Marinis, Eleni Peristeri, Arhonto Terzi

“ToM-abilities are often assessed using False-Belief (FB) tasks that test first-order ToM (monitoring another person’s beliefs) and second-order ToM (monitoring person-A’s beliefs about person-B’s beliefs). Although language is a powerful predictor of ToM, most widely used ToM tasks involve complex language. Therefore, low scores in ToM tasks, especially in low-verbal autistic children and children with Developmental Language Disorder (DLD), could result from low language skills that make it difficult to understand and/ or solve ToM tasks. To date there is a scarcity of non-verbal first-order FB tasks and a lack of non-verbal second-order FB tasks.

To assess ToM independently from language we developed a toolkit that comprises first- and second-order FB tasks. The toolkit is available in verbal (Greek, German) and non-verbal versions and consists of video clips. We present data from six neurotypical (NT) groups and one group of children with DLD:

(1) NT Adults (n(Greek)=50, n(German)=50): better performance in first-order-FB than second-order-FB, better in verbal than non-verbal tasks.

(2) NT children, 1-2-graders (n(Greek)=50, n(German)=22): Ongoing data collection to reach 50 per group; better performance in true-belief (TB) than FB; better in non-verbal than verbal tasks.

(3) NT-children, 3-4-graders (n(Greek)=30, n(German)=22): Ongoing data collection to reach 50 per group; better performance in TB than FB; similar in non-verbal and verbal tasks.

(4) Children with DLD, 3-4-graders (n(Greek): Ongoing data collection to reach 30 children.

Neurotypical adults seem to use language to mediate ToM performance. NT 1-2-graders do not benefit from language in ToM performance, suggesting that language supports adults but adds processing cost for young schoolchildren during FB-tasks. Data from NT 3-4-graders show development of ToM-abilities throughout primary-school. We predict that children with DLD will perform similarly to age-matched controls in non-verbal FB tasks but will show deficits in verbal FB-tasks and will perform similarly to NT 1-2 graders.”

Bilingualism effects in children with Developmental Language Disorder (DLD)

Authors: Stephanie Durrleman; Clémence Gordon-Dana, Anamaria Bentea

“A subgroup of children diagnosed with DLD exhibits difficulties in their ToM. Maintaining a balanced level of proficiency in two languages has been proposed to possibly confer certain cognitive advantages, although this hypothesis remains unexplored for ToM, which according to preliminary work might be enhanced at the simpler, first-order level, in DLD (Peristeri et al 2019).

This study sought to answer, for the first time, the following two research questions: (1) Can higher scores be observed in bilingual children with DLD compared to their monolingual peers for both first-order and second-order ToM? (2) Can a more balanced level of bilingualism maximize ToM benefits? We administered ToM tests inspired by Durrleman et al. (2022) designed to minimize the influence of language, thus ensuring a clear assessment of ToM, to two groups of 14 children with DLD, aged 5 to 10 years, matched for age (average age = 7) and language abilities on standardized tasks. We assessed their bilingual experiences by means of the PaBiQ parental questionnaire. Our analysis revealed that bilingual children with DLD performed significantly better in understanding first-order false beliefs compared to their monolingual peers, but no differences were observed in second-order ToM tasks. Moreover, having a more balanced exposure to the two languages did not correlate with improved ToM.

The higher performance of bilingual children on first-order ToM tasks compared to their monolingual peers suggests a potential socio-cognitive advantage of bilingualism in the context of DLD. While previous studies have reported associations between balanced bilingualism in typically developing children, this relationship does not emerge for ToM in our study of children with DLD. These findings suggest that bilingual experiences can have positive effects on at least one aspect of ToM development in children with DLD.”

Bilingualism & Theory of Mind (ToM) development in autistic children over time

Authors: Eleni Peristeri; Margreet Vogelzang, Ianthi Tsimpli, Stephanie Durrleman

“Autistic children often display ToM difficulties, which can be alleviated by better vocabulary, executive functions (EF), IQ and the linguistic experience of bilingualism. Research on bilingualism effects on ToM has focused on cross-sectional comparisons of first-order ToM. This is the first longitudinal study focused on second-order ToM in monolingual and bilingual autistic children and aims to shed light on the roles played by linguistic and cognitive factors on ToM trajectories.

The study included 21 autistic bilingual children (ASDbi), and 21 age-, socioeconomic status-, and verbal IQ-matched autistic monolingual children (ASDmono). Both cohorts were followed for three timepoints, specifically at ages 6, 9 and 12. Measures of second-order ToM, expressive vocabulary, EF, and Performance IQ (PIQ) were taken.

Both groups’ ToM performance improved significantly with age. ASDmono children outperformed ASDbi in expressive vocabulary at all timepoints, but ASDbi had higher EF and PIQ (except for PIQ at timepoint 1, which did not reach significance. ASDbi children scored higher on second-order ToM at timepoints 2 and 3 and this was associated with PIQ and expressive vocabulary scores.

The results reveal positive relationships between vocabulary and PIQ, and trajectories of second-order ToM in ASDbi children, most noticeably at 9 and 12 years. ASDbi children showed higher EF across all three timepoints, however, this did not seem to boost their ToM. In contrast, vocabulary, which was lower in ASDbi children, and PIQ, which was higher for this group at timepoints 2 and 3, positively impacted ToM, suggesting that bilingual children capitalize on cognitive (PIQ) and linguistic (vocabulary) resources effectively to yield beneficial cascading effects on ToM development.”

Impact of cumulative linguistic exposure on Theory of Mind in children with and without autism

Authors: Franziska Baumeister; Pauline Wolfer, Stephanie Durrleman

“Bilingualism reportedly confers advantages in ToM in neurotypical (NT) children, and in one preliminary study involving autistic children (Peristeri et al. 2021). However, the “False-Belief” tasks used in these studies often involve confounding verbal abilities; moreover, the binary nature of the tasks’ response choices increases the likelihood of chance performance. Finally, studies comparing monolinguals and bilinguals typically hypothesize rather than operationalize the underlying mechanisms explaining the potential impact of the bilingual experience on ToM: bilinguals would show a “privileged language-knowledge-person connection” thanks to an increased exposure to people speaking different languages, which should consequently lead to more frequent reasoning about interlocutors’ languages, and by extension, their mental states.

In this study, we administer a new low-verbal ToM task providing 3 possible response choices, thus steering away from chance-level, and we incorporate justification questions to yield additional insight into ToM reasoning. We furthermore quantify language experience through a “cumulative language exposure entropy score” based on the Q-BEx questionnaire. These tools allow us to address the questions (1) Does higher cumulative language exposure positively impact ToM in both NT and ASD children? (2) How do NT and ASD children justify their perspectives?

Preliminary analysis involving 114 TD children (3;3-11;7) and 31 ASD children (3;10-11;7) using ordinal regression analyses suggests that for both NT and ASD groups only age (for both groups) and IQ (only for NT children) are identified as significant predictors but not language exposure. To provide a more robust assessment of these findings for question (1) and to address question (2), an additional 40 NT and 70 ASD children will be tested by June 2024 and included in the dataset. Bayesian analyses will be used with this larger dataset to provide a deeper understanding of the observed results and to address the research questions in a robust way.”

07/17/2024, from 02:15 PM to 04:15 PM , Room P200

Symposium: Let’s talk about feelings: Multilinguals’ Language-related Emotions and Well-Being

Speakers: Graziela Dekeyser; Nathalie Topaj; Jessica Willard; Meagan Driver; Annick De Houwer

How do multilingual children, adolescents, and their teachers feel about speaking, learning and teaching their languages? This symposium gives their feelings a platform by drawing on rich qualitative and large-scale quantitative data.

“I am scared Miss.” Multilingual Children’s Emotional Experiences Associated with Languages

Authors: Graziela Dekeyser;

“Topic: In Flanders, the Dutch-speaking region of Belgium, an increasing number of children grow up multilingual. Prior research has shown how language learning is a highly emotional process. For example, a significant amount of studies investigated how foreign language learning can be associated with feelings of anxiety as well as enjoyment (Dewaele & Li, 2020). However, several gaps in the literature remain. Relatively few studies focused on the emotions experienced in everyday situations beyond the context of the formal foreign language class. Next, most research has been restricted to late(r) learners of a L2 (either late adolescents or (young) adults), while early child multilinguals have received less attention. Especially for the latter group, more research on this topic is important as emotions surrounding a language may have a crucial impact on the further development of that language (Sevinç and Backus 2017). Moreover, in Flanders, multilingualism is generally problematized due to the region’s strong Dutch monolingual ideology (Pulinx et al., 2017). Consequently, emotions associated with the heritage language as well as the majority language may be heightened among Flemish multilingual children. However, to our knowledge, no prior studies on this topic among this research group exist.

Research questions: RQ1: Which emotions do multilingual children associate with their heritage language and Dutch? RQ2: What are the sources of these reported emotions according to the children?

Method: 10 single-sex focus groups were conducted with multilingual children living in the city of Antwerp, Belgium. In total, 50 children were included. Thematic analysis was performed.

Results: The results show how growing up multilingual in Flanders is a very ambivalent emotional experience. Children reported a wide variety of positive ànd negative emotions associated with both languages. Interpretation of the results points to the importance of identity motives in producing emotional experiences related to languages.”

Well-being and Language Skills in Heritage Language Classes

Authors: Nathalie Topaj; Natalia Gagarina

“Berlin is a multilingual and multicultural city. Community interest, changes in policies, and the newly established legal framework, giving school students an opportunity to attend heritage language (HL) classes, are contributing to the visibility of multilingualism and the expansion of HL instruction in schools, provided by the Senate of Berlin. The offer for Turkish and Arabic, which are among the most spoken heritage languages in Berlin, has been considerably increased in the last years. More than 2200 Turkish- and over 1400 Arabic-speaking children are enrolled in HL classes.

Within the framework of the evaluation of HL instruction at Berlin primary schools (Evaluation ESU), over 200 Turkish-German and Arabic-German-speaking children were tested in both languages at the end of grade 3 and 4. The test battery included tests on reading, writing, and narrative skills, and a questionnaire on children’s language environment, attitudes towards languages, self-evaluation of language skills, experience with HL classes etc. Among other factors, children’s well-being might play a considerable role in the overall success in learning family languages at school.

For the symposium, we will explore questions on children’s well-being (Do you feel comfortable in first language classes and enjoy going there? Do you enjoy learning Turkish/Arabic at school?) in relation to attitudes towards their languages (Which language do you prefer to speak?), self-evaluation (Which language do you speak best? How well can you understand/speak/read/write German and Turkish/Arabic?), improvement (Have you improved in understanding/speaking/ reading/writing in Turkish/Arabic through first language classes?), and to language skills based on the performed tests. The data is currently being processed. While many children feel comfortable and enjoy learning their family language in HL classes, others give different answers to these questions. The attitude, language preferences, self-evaluation and improvement also differ within subgroups. The results will be presented at the conference.”

Linking Multilingual Adolescents’ Language-related Emotions to Individual and Contextual Correlates

Authors: Jessica Willard; Beyhan Ertanir, Mualla Kaya, Linda Kololli, Larissa Trösch

First findings from North America and the Netherlands have demonstrated that being multilingual and speaking not only the majority but also a heritage language can be a source of positive and negative emotions. Individual and contextual factors may play a role in determining the valence of emotions (Driver, 2020; Jean & Geva, 2012; Sevinç & Dewaele). For a large sample of multilingual adolescents in Germany, we examined: (1) whether positive and negative emotions with regard to using a heritage and majority language are reported, and (2) how individual (e.g., generational status, self-perceived language proficiency, ethnic and national identity), peer group (e.g., language use), family and school factors relate to positive and negative emotions. N= 818 multilingual adolescents in Germany with numerous different heritage languages participated in this cross-sectional online survey. Positive and negative emotions in relation to using the heritage and majority language were assessed with 28 adapted self-report items (Driver, 2020). Descriptively, adolescents reported experiencing both positive and negative emotions with regard to the heritage and majority language. A set of multiple regressions revealed that many individual-level, family, and school factors were connected to positive and negative emotions: for example, stronger self-rated proficiency in a language was related to more positive and fewer negative emotions in that language. Stronger ethnic and national identity commitment were mostly related to more positive and fewer negative emotions. Finally, teachers’ negative and positive attitudes towards adolescents’ heritage languages were connected to positive and negative emotions for both the heritage and majority language. These findings suggest no single factor such as self-perceived language proficiency is the sole determinant of feeling good or bad about using a language. Instead, intricate processes involving individual factors and several levels of contextual factors appear to be involved in emotions related to using a heritage and majority language.

From Linguistic Insecurity to Pride: The Emotional Rollercoaster of the Heritage Language Teacher

Authors: Meagan Driver;

“As the number of heritage speakers (HSs) in the U.S. continues to rise, scholars across a range of disciplines have set out to explore HS needs both in and outside the classroom setting. Encouragingly, increased interest on affective variables and wellbeing initiatives in second language acquisition (SLA) have also motivated many scholars to highlight the need to empower heritage language learners and stakeholders (e.g., Ortega, 2020) and explore emotional aspects within the heritage language (HL) classroom. Specifically, linguistic insecurity has appeared as an emotion particularly relevant to the HL context (e.g., Driver, 2023; Tseng, 2021). Still, few studies explore the experiences of HL educators who also identify as members of a HS community (e.g., Cho, 2014), which leaves an essential perspective and valuable expertise absent from HL education discussions. This study aims to explore the emotional experiences of HL educators from HL backgrounds in order to uncover not only challenges but also coping mechanisms in an effort to create spaces that support HSs along the HL teacher pipeline.

Qualitative data for this study include survey and interview responses from five middle (N = 2) and high school (N = 3) HL teachers in the U.S. from three different HL backgrounds (Arabic, Chinese, Spanish). Findings shed light on the emotional benefits, challenges, and barriers associated with entering the HL teacher pipeline for those from language minoritized backgrounds. Specifically, results highlight mixed feelings of linguistic insecurity, shame, and frustration, but also pride and confidence for HL educators in the HL classroom. Findings also suggest a relationship between common language ideologies (e.g., native speaker ideology) and manifestations of linguistic insecurity for educators from HL communities. Pedagogical implications for language teacher training programs and HL classrooms are provided, as well as suggestions for future research and professional initiatives that may foster HL teacher wellbeing.”

07/17/2024, from 02:15 PM to 04:15 PM , Room P301

Symposium: Extending the clinical value of tense omissions into complex syntax and grammaticality judgments

Speakers: Sean Redmond; Melanie Schuele; Mabel Rice; Alyson Abel

Converging evidence for persistent tense-marking deficits among English-speaking children with language impairment is presented across four research projects. Effects of clausal contexts, co-occurring conditions, predictors of growth are discussed

Tense marking in complex syntax by children with specific language impairment

Authors: Melanie Schuele;

Tense/finiteness marking has been studied most often in the production of simple clauses. Interventions to increase accuracy on tense marking appear to have targeted primarily, if not exclusively, simple clauses. To meet academic and social language demands, children must produce a variety of dependent clauses in spoken and written language. Tense marking in dependent clauses (e.g., spoken as mono-clausal utterances) and multi-clauses utterances (i.e., dependent + main clauses) may present greater challenges to children. Language samples from English-speaking children with specific language impairment (SLI) between 5 and 7 years of age (n = 25) and MLU-matched typical language (TL) learners (n = 20) were analyzed for obligatory tense marking. TL children demonstrated no differences in tense-marking when comparing simple clauses, dependent clauses, and main clauses in complex syntax utterances across three structures: regular past tense (jump/ed), 3s (jump/3s), and auxiliary/copula BE. However, SLI children demonstrated greater difficulty marking obligatory tense in regular past tense (jump/ed) and 3s (jump/3s) but not auxiliary/copula BE. Ongoing analyses are exploring whether tense marking is particularly vulnerable for children with SLI by complex syntax type (e.g., relative clauses, sentential complements) or a general vulnerability in complex syntax is apparent. The details of finite marking deficits are critical to understand in order to design the most efficacious language interventions.

English-speaking SLI children’s persistent acceptance of tense errors in complex sentence: 5-18 yrs

Authors: Mabel Rice; Kathleen Kelsey Earnest

Language impairments can be difficult to detect in otherwise typically developing healthy children, a condition known as Specific Language Impairment (SLI), sometimes referred to as the broader category, Developmental Language Disorders. The difficulties in identification persist as the children’s language impairments lag those of age peers throughout childhood. Identification methods for young children focus on the acquisition of simple sentences and vocabulary growth, methods that can lack sensitivity for later grammar skills. Previous studies identified a valid clinical grammar marker featuring tense/finiteness marking in simple clauses, although limited by ceiling effects around the age of 8 years. The focus of this presentation is a study of a new, more grammatically challenging complex sentence task in children affected or unaffected with SLI in longitudinal data encompassing the ages 5-18 years. There were 483 monolingual English-speaking children (213 unaffected, 270 SLI affected), tested annually over ages 5-18 years. The new experimental task was a grammaticality judgment (GJ) task designed to be interpretable as an extension of the simple clause formats of tasks designed for young children. The task evaluated detection of errors of tense/finiteness in a multi-clause complex of this format: When did Mary say she paint(ed) a door? The outcomes were consistent with the outcomes of earlier studies of younger children with a simple sentence, such as she paint(ed) a door. As in these studies, growth models for the SLI group were consistently lower than the unaffected group, although the growth trajectories across groups did not differ. The types of error were similar across groups, i.e., acceptance of omissions and better identification of non-allowed insertion of tense/finiteness markers. Covariates of child nonverbal IQ, mother’s education, and child sex did not significantly moderate these effects.

Grammaticality judgments of tense omissions by children with and without ADHD

Authors: Sean Redmond;

Academic underachievement, disrupted peer relations, and difficulties following directions could be the result of underlying attention deficits, language impairments, or both. Robust psycholinguistic protocols are needed to sift through the relative contributors to these generic clinical signs and develop appropriate treatment plans to address them. In this presentation, the extent to which grammaticality judgments (GJ) could be used to asses the linguistic abilities of children with attention-deficit/hyperactivity disorder (ADHD) is evaluated. ADHD provides an interesting stress test for GJ tasks. GJ tasks have been characterized by some as drawing heavily upon children’s metalinguistic skills requiring sustained attention, concentration, and working memory. For these reasons, we would expect GJ tasks to be challenging for children with ADHD and prone to generate underestimations of their linguistic abilities. Others have characterized GJs as largely reflexive decisions drawing primarily upon children’s underlying syntactic representations. To the extent this is true, GJ task performances should track with the linguistic abilities of children with ADHD in the same way it does for unaffected children. A 4-group design (n=78) consisting of students with neurotypical development (TD), attention-deficit/ hyperactivity disorder (ADHD), specific language impairment (SLI), and comorbid ADHD and language impairment (ADHD+LI) was used to test these contrasting predictions with judgments involving finite errors in simple declarative and interrogative sentences. Groups were matched for age, gender, race/ethnicity, and maternal education levels. All participants (age range: 6-8 years) were monolingual speakers of General American English and had nonverbal IQ scores > 80. ANOVAs revealed the following significant pairwise comparisons: TD = ADHD > SLI = ADHD+LI. Correlations between children’s ADHD symptoms and their GJ performances were weak and nonsignificant (rs < -.15). These outcomes align best with the contention that GJs tap into children’s linguistic representations regardless of their ADHD status.

Judgments of tense marking as an English proficiency measure in mono/bilingual preadolescents

Authors: Alyson Abel;

With heightened public awareness of developmental language disorders (DLD), including specific language impairment (SLI), comes increased efforts to identify clinical markers of the disorder. Two widely-accepted clinical markers of SLI are deficits in tense marking and phonological working memory. Tasks assessing grammaticality judgment (GJ) of tense marking accuracy in two simple question forms, wh- and yes/no questions, capture the persistent tense marking deficits in monolingual children with SLI through age 15 and correlate with figurative language abilities in 16-year-olds. Measures of phonological working memory, particularly nonword repetition (NWR) tasks, are sensitive to SLI in both monolingual and bilingual children through at least age 9. Older children with SLI often go misdiagnosed or unnoticed, especially if they’re bilingual, due to the lack of other deficits and relatively low demands on language through the middle-school years. Accurate clinical markers for pre-adolescent children from variable language backgrounds are needed to improve identification. Here, 9-12 year old children (n=33, 18 monolingual, 15 bilingual) completed an English language omnibus assessment (CELF), NWR task, GJ wh-question task (GJWH), and GJ yes/no question task (GJYN). Participants were identified as bilingual if they received Spanish input at or before age 5. Spoken Spanish proficiency was variable across the bilingual group. Participants were not clinically selected; CELF scores ranged from 56-132 (M=102.5). Groups did not differ on any measure. CELF scores correlated with all other measures for the bilingual group but only correlated with NWR and GJWH performance for the monolingual group. A series of linear regression analyses further identified that NWR did not predict CELF score for either group. GJWH predicted CELF score for both groups and GJYN predicted CELF score for the bilingual group only. These outcomes support GJ measures of tense marking as being sensitive to language ability in pre-adolescent monolingual and bilingual children.

07/17/2024, from 02:15 PM to 04:15 PM , Room P104

Symposium: Linking child speech perception and production

Speakers: Fleur Vissers; Iris Berent; Tamar Keren-Portnoy; Irene Lorenzini; Henny Yeung; Natasja Delbar

Researching the link between children’s speech perception and production can be challenging. In five talks, a variety of methods are presented showing different ways to study this link in children of different ages – from newborn to the age of six.

Phonology isn’t’ about talking!

Authors: Iris Berent;

The links between phonology and the articulatory motor system are controversial. One view asserts that phonology is governed by abstract principles (Halle & Reiss, 2008); another states that phonological intuitions in perception arise from tacit motor simulation, i.e., talking (Hayes et al., 2004). The “talking” position is supported by a large literature, showing that motor simulation guides phonetic categorization (e.g., ba/da?) (Berent et al., 2020; Bruderer et al., 2015; D’Ausilo et al., 2009; Möttonen et al., 2009; Murakami et al., 2018; Smalle et al., 2014). Moreover, many phonological principles are articulatorily “sensible”; for example, syllables like bla are easier to articulate than lba (Mattingly, 1981). Of interest, then, is whether phonological intuitions in perception (e.g., bla>lba) arise directly from motor pressures that occur in production. Here, I address this question by evidence from the sonority hierarchy (e.g., bla>bna>bda>lba). I first show that young children are sensitive to the sonority hierarchy (Berent et al., 2011). I next evaluate the likely origins of these intuitions. Building on transcranial magnetic stimulation work from adults, I show that sensitivity to the sonority hierarchy does not require motor simulation, as it emerges even when the brain motor system is disrupted by transcranial magnetic simulation (Berent et al., 2015; Berent et al., 2023), or mechanically (Berent & Platt, 2022; Zhao & Berent, 2018). Furthermore, newborns show sensitivity to the hierarchy well before they can talk (Gómez et al., 2014). The dissociation of phonological constraints operating in perception from the articulatory motor system is particularly striking, given that motor simulation demonstrably constrains phonetic perception. As such, these results suggest that phonological intuitions are distinct from the phonetic pressures, and they are irreducible to the constraints on talking.

The contribution of production practice to perception, memory and lexical development

Authors: Tamar Keren-Portnoy; Rory DePaolis (James Madison University), Helena Daffern (University of York), & Mona Kanaan (University of York)

Research on the connections between perception and production in infants’ language development tends to focus on the ways in which infant’s perception of the input language affects their production. This study focuses on the reverse relationship, asking how infants’ own productions, and in particular babble or prelinguistic consonant production, affect the ways in which they listen, remember and learn language. For example, infants ability to move their articulators affects their ability to distinguish between different sounds (Choi et al., 2019); that sounds familiar to infants from their own production capture their attention more than other sounds (DePaolis et al., 2011); that novel words which contain sound that are within a toddler’s repertoire are easier to repeat in a nonword repetition task (Keren-Portnoy et al., 2010) and that early mastery of sound production through babble predicts age at production of first words (McCune & Vihman, 2001; McGillion et al., 2017). I will then report on an ongoing experiment with infants from low-mid socioeconomic status (n=96), in which one group of infants aged 0;7 received an app that we developed (Daffern, 2020; Keren-Portnoy et al., 2021), to encourage them to vocalise and babble more, while another group received an app that encourages reaching for a moving shape. We measured the infants’ vocabulary size using the UK-CDI (Alcock et al., 2020) monthly until age 1;0 and then again at age 1;6. This study was pre-registered, and we can therefore not report on its results midway. The final data will be collected in November 2023. We will be able to report on the study’s results by July 2024.

More on the action-perception link in language development: speech-related sensorimotor development

Authors: Irene Lorenzini; Henny Yeung (Simon Fraser University), & Thierry Nazzi (Université Paris Cité)

Speakers of a large variety of languages tend to rely more on consonants than on vowels for lexical processing (Cutler & Nazzi, 2019). In French, this ‘Consonant-bias (C-bias)’ appears between 6 and 8 months, preceded by a ‘Vowel-bias (V-bias)’ (Nishibayashi & Nazzi, 2016). Linguistic experience might explain this switch: contrary to consonants, vowels are perceivable even prenatally, and young infants might first focus on such familiar sounds and later switch to consonants. However, auditory experience might not be the only factor at play. Indeed, the onset of consonant production is synchronous with the switch to the C-bias. On these bases, we asked whether a perception-production link could be identified in this domain. Using the Headturn Preference Paradigm, 32 French-learning 7-month-olds (when the switch to the C-bias is observed; M=7m; 14d) completed a word segmentation task in which they were tested on C- and V-mispronunciations of the segmented words, to determine the presence of V-/C-biases (extending Nishibayashi & Nazzi, 2016). In parallel, we measured infants’ babbling (human-annotated LENA home recordings) and general oro-motor activity (Infant Behavior Questionnaire). A One-way ANOVA on the difference in looking times to C-/V-mispronunciations revealed a main effect of IBQ (p = 0.006), no significant effect of babbling (p = 0.236) and a significant interaction IBQ x babbling (p = 0.012): oro-motor activity is linked to the development of the C-bias, but only in participants who produce more complex babbling. Work in progress is investigating links between babbling development, conversational turn-taking, quantity of input words uttered in the environment (as calculated by the LENA) and general oro-motor development (evaluated by chewing and eating skills, Lemarchand et al., 2020). These data will allow us to draw detailed individual profiles of perception/production skills in our infant group.

Visual speech enhances children’s looking but does not speed naming: A priming study

Authors: Henny Yeung; Theresa Rabideau (University of Ottawa), Margarethe McDonald (The University of Kansas), & Tania Zamuner (University of Ottawa)

How multimodal information is integrated in children’s speech remains an active area of research. Following predictions from adult neuroimaging models, we ask here whether visual speech processing has distinct connections to speech production versus perception tasks. We previously measured 2-8 year-olds’ looking to a target object (e.g., ball) over a distractor object (e.g., coat) after children either saw a visual prime (a face silently articulating ball), heard an auditory prime (the word ball), or were presented with an audiovisual prime (a face saying the word ball). Results suggested increasing target-object looking across age in V, A, and AV modalities, and although visual speech had comparably weaker effects, it reliably increased in strength from the youngest to oldest ages. The current study investigates whether children were able to use visual speech to prime spoken word production in an almost identical procedure. Instead of two target images, however, only one image was shown (e.g., ball), which was preceded by the same V, A, or AV primes, which presented either the target word (ball), or an unrelated word (coat). Naming latencies for the target object were also recorded instead of looking times. Data collection is ongoing but results to date from 48 children in the 4-6 year range (M = 5;4 years) show strong priming effects in AV and A conditions, but surprisingly, no priming effect or age-related improvement in the V condition. Overall, we replicate the finding that visual speech is less effective than audio speech at activating (or inhibiting) lexical representations in children. Critically, we also show that visual speech is far more effective at enhancing looking (a perceptual skill) than speeding word naming (a production skill). This suggests a rich avenue of future research that investigates why visual speech may interface differently with children’s perceptual versus productive lexical development.

Toddlers’ speech monitoring of their own and others’ (deviant) productions

Authors: Natasja Delbar; Fleur M.H.G. Vissers (Radboud University), Imme Lammertink (Radboud University), Paula Fikkert (Radboud University), & Clara C. Levelt (Leiden University)

Typically developing children around the age of two often do not (yet) produce all words correctly, for example, they may produce the Dutch word bloem ‘flower’ as *[bum] (Beers, 2003). What underlies these deviations in the children’s productions remains unclear (e.g., articulatory difficulties, encoding difficulties, and/or deviant lexical representations). To locate the source of these deviant productions we need to study both speech production and perception. We will present pilot data from two experiments with typically developing two-year-olds who participated in both a production and a perception experiment. In the production experiment, the children were prompted (by the experimenter saying ‘hmm?’) to repair deviant productions of target onset consonant clusters during a picture-naming game. This triggers self-monitoring, the mechanism of perceiving one’s own productions and using observed mismatches with the target to adjust subsequent productions (see also Levelt et al., 2023). By analyzing both phonologically and acoustically how deviant productions are repaired, this experiment disentangles word-form encoding errors from segmental representation errors. In the perception experiment the children listened to stories containing target words pronounced correctly (e.g., bloem ‘flower’) or with phonological deviation. Two types of deviant productions were included; (1) the pronunciation the children used themselves (e.g., *[bum]) and (2) random deviations (e.g., *[dum]). This experiment provides information about the level of detail in children’s lexical representations of words they know either actively or only passively, and words they produce either correctly or with deviation. We will end by discussing the pros and cons of these methods and the implications for theoretical models of (toddler) language processing.

07/17/2024, from 02:15 PM to 04:15 PM , Room P429

Symposium: Parent-child interaction in deafness and language development: methods, data and applications

Speakers: Gary Morgan; Martina Curtin; Evelien Dirks; Mario Figueroa; Ciara Kelly

Parent–child interaction in the early years plays a central role in a deaf child’s language development. The symposium addressees why this is challenging and how it can be changed.

Everyday routines for deaf and hard of hearing children: A systematic review assessing PCI

Authors: Martina Curtin; Evelien Dirks and Amy Szarkowski

“Parent-child interaction (PCI) is known to influence deaf children’s developmental outcomes. Even with advances in early identification of hearing differences and earlier exposure to sign language and/or earlier introduction of the hearing technology (when families chose to use them and when they are likely to benefit a child), developmental outcomes of deaf children can be altered, informed at least in part by their early experiences with caregivers. Ninety percent of deaf children are born to hearing parents who have little experience of deafness, so families need to rapidly acquire knowledge and skills in how best to communicate with and foster the development of their deaf children. There is a need to better understand the relationships and bi-directional interaction effects between caregivers’ and their deaf children, particularly as they pertain to developmental outcomes. Further, the majority of PCI research for deaf infants centres around play and book reading, which could be seen as a luxury activities for some families. This current systematic review will focus on parent child interaction within the context of everyday routines and investigate a wide range of developmental outcomes.

Aims: 1) What are the range of parent behaviours assessed or described in PCI among dyads in which the child is deaf?

2. How are PCI behaviours assessed?
3. Which daily routine contexts involving deaf children have been explored?
4. How are parent behaviours, child behaviours, parent-child interactions, and parent-child routines related to child development outcomes?

Methods Used: A systematic review (SR) with narrative synthesis was conducted on Covidence software, following guidance from the Cochrane Handbook for SRs. Each included study was independently reviewed by two authors. The main themes of the review will be presented.”

Look at me! Interactions between deaf babies and their parents

Authors: Evelien Dirks;

The early parent-child interactions are critical for children’s overall development. They shape children’s language, social-emotional and cognitive development. Interactions between deaf children and their typical hearing parents can be more challenging. Most typical hearing parents have no prior experiences with deafness and they may experience feelings of insecurity when raising their child. They are faced with challenges in interaction and communication. Previous research showed more difficulties in establishing and maintaining joint attention and more parent directive behaviour. In the present study we examine the early interactions of deaf babies and their parents in comparison to typical hearing babies. This is one of the first studies that focuses on a very young age group of deaf children. In the study 30 deaf and 30 typical hearing babies between 2 and 6 months of age and their caregivers participated. Video recordings of a free play activity and a daily routine activity (diaper changing) of the babies and their parents were made in the home environment. The interactions were coded for parental sensitivity, intrusiveness and joint attention. Results suggest systematic differences in the timing and quality of early interaction patterns across groups. This study contributes to our knowledge on the early interactions of babies and their parents. Based on the findings recommendations for early intervention will be discussed.

Linguistic input to deaf toddlers with hearing parents

Authors: Mario Figueroa; Gary Morgan

Children’s language may be affected by the linguistic-communicative environment, especially during the first years of life. In the case of deaf children, previous studies have shown the importance of good stimulation for later language and academic development. However, fewer studies have focused on very early stages of development and therefore on infants with little language and auditory experience. Fifty-one toddlers with and without deafness participated in this study. All were living in Catalonia and ranged in age from 10 – 40 months. Deaf toddlers wore hearing aids or cochlear implants and were exposed to a spoken language. Toddlers and their parents participated in the Gallery Art Task. In this task there are 5 pictures distributed around the room and the adult must capture the child’s attention and explain and discuss each of the pictures for 5 minutes. Parent-child interactions were recorded, transcribed, and coded for analysis.The results showed significant differences between the communicative exchanges of parents with deaf children and parents of hearing children. Linguistic stimulation received by the infants shows a great variability, both in quality and quantity of language provided by their parents. Linguistic input differs from group to group and is therefore conditioned by deafness. Speech therapy must take these aspects into account in order to provide optimal language learning environments.

Arriving at a consensus on the assessment of parent-child interaction with deaf and hard of hearing

Authors: Martina Curtin; Madeline Cruice, Ros Herman and Gary Morgan

“Most deaf babies are born to hearing parents who need to learn how to adapt their communication with their deaf child. This is important because the quality of parent-child interaction (PCI) predicts how well a deaf child develops language. Teachers of the Deaf and Speech and Language Therapists support families with communication in the home. At present, there are not any deaf-specific assessments that can appraise PCI. Preliminary work has uncovered which parent behaviours and approaches are used in PCI assessments across international research and practice. The next step is to arrive at agreement on the core content and the best practices of a new PCI tool for deaf infants and their families.

Methods. An international sample of expert academics and practitioners (n=83) were recruited to take part in this two-round electronic Delphi study. Participants were presented with 69 statements focusing on (i) which parent behaviours were important to assess and (ii) which approaches should be used in a PCI assessment. Participants rated the extent to which they agreed or disagreed with each statement on a five-point Likert scale and gave comments to support their response. Consensus was defined as >80% of participants rating the statement as ‘highly important’ or ‘essential’. If consensus was not reached, participant comments were used to generate new statements which were rated in the next round.

Results. Expert participants achieved consensus on 52 statements. Consensus on statements ranged from 80-99%. A further six statements (achieving consensus of 75-79%) were also subsequently included following a review of the participants’ comments, of how ratings of these six items had changed between rounds, and feedback from our patient and public involvement group. A syntheses of the main themes to come out across these statements will be discussed in the presentation. The e-Delphi has enhanced our understanding of which parent behaviours and approaches to assessment should be included in a PCI assessment.”

Family-centred intervention to promote parent communication strategies with deaf infants and increas

Authors: Ciara Kelly; Danielle Mathews, Gary Morgan

Infant–parent interaction forms the foundation for language learning. For the majority of deaf infants, hearing loss can impact access to, and the quality of, communicative interactions, placing language development at risk. Support for families to meet the challenges faced during interaction is highly variable in the United Kingdom. Further, hearing aid use is at its lowest and most variable in the early years, reducing access to communicative interactions and subsequently undermining the benefits of family-centred support to provide a rich, language-learning environment. We discuss a series of studies that take steps towards more standardised but tailorable family support. In study 1, we co-produced an instructional, video-based intervention, testing for feasibility in terms of behaviour change in seven communicative strategies and the acceptability of the strategies and video delivery. Parents of deaf infants (n=9) increased their use of the majority of behaviors and found content and delivery acceptable. However, further development was required to: (a) support use of semantically contingent talk and attention getting strategies to elicit infant attention, and (b) ensure the information was provided in a bite-size format that could be tailored to individual families. In study 2, the intervention was refined based on findings from study 1 and assessed for seven dimensions of acceptability with 9 parents and 17 professionals, who reported similar high acceptability scores. In the final study, we discuss the development of a co-produced intervention grounded in the principles of behaviour change theory, that aims to increase hearing aid use during the first years of life critical for spoken language development. Future research needs to test the effectiveness of such interventions and explore the best means of laying a strong social-communicative foundation for later language in deaf infants.

07/17/2024, from 02:15 PM to 04:15 PM , Room P018

Symposium: Screen time and children’s emerging language and literacy skills, and parental attitudes

Speakers: Suvi Stolt; Anett Sundqvist; Monika Abels; Peixin Nie; Riikka Mustonen

This symposium focuses on the associations between screen time use, parental attitudes on screen time, and emergence of language ability in typically developing children aged between 1 month – 5 years, and in those with hereditary risk for dyslexia.

Media and home literacy environment and the emergent literacy development in five-year-olds

Authors: Anett Sundqvist; Felix Koch, University of Linköping; Rachel Barr, Georgetown University

“Background and aim: The child’s emergent literacy development is of utmost importance when starting formal schooling. The present study has examined factors in the home environment (parental teaching of letters, reading, colors, sound, shapes), with home screen media environment and parental attitudes and the relation to the child’s emergent development.

Methods: The study involved 88 participants, 5-year-olds (55% boys) with their parents. Emergent literacy development was conceptually defined as a composite measure including vocabulary knowledge, letter recognition, phonemic awareness, and rhyme understanding. The home media environment was assessed with the Media Assessment Questionnaire-2 (Barr et al., 2020). Parental engagement in the home environment was measured using the STIMQ-2, including aspects such as book reading habits (e.g., number of books and dialogical reading), time spent reading, and educational activities at home, such as teaching colors, letters, and shapes. Kendall rank correlation was used to analyze the relationships between the variables.

Results: The findings revealed that the amount of time spent reading books and engagement in educational activities at home positively correlated with emergent literacy. However, watching television content and playing digital games showed a negative correlation with emergent literacy. It is not uncommon for children who watch content on YouTube to watch content not in Swedish, which was negatively correlated to the child’s emergent literacy.

There was a positive correlation between the child’s emergent literacy and the parents’ active limitation of certain content and screen time. However, there was no correlation to whether the parents said they let the child use screens to learn educational content. A detailed content analysis will be conducted.

Conclusions: The child’s emergent literacy development is positively associated with parental home activities such as teaching letters and numbers and the amount of time the parent reads to the child, and negatively associated with time spent with screens.”

Parental phone use and children’s communication attempts

Authors: Monika Abels;

“Background and aims: Previous work has shown that caregivers are distracted from their children’s communication attempts when caregivers are using their phone (Abels et al., 2018; Van den Abeele et al., 2020). The current study is an attempt to replicate these results in Norway. Caregiver-child interactions will be analyzed in terms of communication that takes place while caregivers are engaged with only their child or other activities, particularly their mobile phones. Additional characteristics of the interactions are explored, such as: do children adapt their communication attempts depending on the length of the caregiver’s involvement with the phone and is there adjustment to the modality of the caregivers’ distraction.

Methods: For this study, 73 caregiver -child (1-66 moths of age) dyads were observed for approximately 10 minutes in different public places in northern Norway. The observations were done with a time sampling procedure focusing on caregivers’ and children’s behaviors. Caregivers’ behaviors included child-directed behaviors and distractions through other activities, including using mobile phones. Children’s behaviors included different types of bids for attention (e.g. by gaze, touch, using objects).

Results and conclusions: Approximately 55% of the caregivers used their phones during the observation period. In contrast to previous studies, Norwegian caregivers used their phones more frequently but overall not longer than caregivers in previous studies in the Netherlands. In line with the previous results, Norwegian caregivers showed fewer, less timely and weaker responses to their children when they are using their phones. Preliminary results indicate, that children adapted to the caregivers’ distraction by increasing movements and decreasing auditory signals. Exploratory analyses also showed that caregivers were less likely to respond, the longer the phone usage last, although the children did not change their bid intensity. The results are discussed in terms of potential effects on caregiver-infant communication in general these results imply.”

Screen time from infancy to pre-school age – associations with language outcomes and dyslexia risk

Authors: Peixin Nie; Paula Virtala and Teija Kujala, University of Helsinki

“Background and aims: Excessive amounts of screen time can have negative effects on language and literacy skills in early childhood. These environmental factors are essential to acknowledge in children who are already at inherited risk for language development delays due to, for example, parental dyslexia. The present study aimed to investigate 1) the trajectory of screen time in early childhood and whether it was affected by parental dyslexia; 2) whether the amount of screen time at early ages could predict the later verbal and cognitive test performance.

Methods: In the DyslexiaBaby longitudinal study, we have followed the language development of over 200 children from birth to school age — 155 children in the dyslexia-risk group and 48 children in the control group. The amount of screen time of various contents, e.g., screen time on child programs, adult programs, games, or others, has been collected with parental questionnaires at 6 months, 1, 2, 3, 4, and 5 years. The children’s language and cognitive skills have been evaluated with standardized neuropsychological tests at 2 years and 4-5 years.

Results and Conclusions: Our preliminary results show increases in the amount of screen time with age during early childhood with no marked effects of familial dyslexia risk on the amounts. Higher verbal IQ scores, for example, in word knowledge and word reasoning subtests, were associated with less amount of screen time. These potential relationships can contribute to providing recommendations to parents and early childhood educators on the amount and content of screen time for young children, especially those at increased risk for language difficulties.”

Parental attitudes, children’s screen time and lexical development in five European countries

Authors: Riikka Mustonen; R Torppa R, J Kuvac Kraljevic, A Matic Skoric, A Schults, K Muszyńska, P Karp, G Krajewski, H Gram Simonsen, N Gram Garmann, T Tulviste, S Stolt

“Background and aim: Previous studies suggest an inverse association between children’s screen time and lexical ability. However, the association may not be straightforward, since factors like parental attitudes toward screen time may influence the time children spend on screen, and, in turn, on children’s lexical development. This study examined how parental attitudes toward screen time relate to both children’s screen time and lexical skills in five European countries.

Methods: Participants were 36- to 42-month-old children and their parent (mothers 95.6%) from five countries (N=319): Finland (n=73), Estonia (n=99), Croatia (n=49), Norway (n=12; recruitment ongoing), and Poland (n=86). Children’s lexical skills were assessed with the Communicative Development Inventory III (vocabulary section, n=100 words). Data on children’s screen time spent alone in minutes, and the parental attitude toward screen time in terms of talking ability (screen time is useful/harmful to talking) were collected using the Screen Time Questionnaire (Schults & Tulviste, 2019). Statistical analysis involved Mann-Whitney U-test and linear regression (dependent variable: lexical skills; independent variables: screen time alone; screen time attitude (0=useful [very+rather useful], n=170; 1=harmful [very+rather harmful], n=60; category could not say was excluded, n=87); children’s age; maternal education (0=below Bachelor’s degree; 1=Bachelor’s degree or higher).

Results: Of the parents, 53% considered screen time useful to talking ability. However, children had better lexical skills (U=6008.5; p=.03) and they spent less screen time alone (U=3592.0; p=.03) when parents considered screen time harmful to talking ability. When children’s age and maternal education were controlled in the regression analysis, the independent variables children’s screen time alone (β=-.06; p=.12) and the attitude toward screen time (β=5.74; p=.06) were approaching significance when explaining lexical skills.

Conclusions: Parental attitudes are important for lexical skills and can affect the time children have screen time alone. Recognizing parental beliefs about screen time is essential for practitioners.”

07/17/2024, from 02:15 PM to 04:15 PM , Room P300

Symposium: The future of research in language acquisition is constructivist

Speakers: Caroline Rowland; Padraic Monaghan; Anna Theakston; Elena Lieven

We describe a constructivist framework for research on language acquisition, focusing on five concrete principles which, we argue, characterise the process by which the learning and processing mechanisms in the brain build mature linguistic knowledge

Introducing a constructivist framework for language acquisition research

Authors: Caroline Rowland; Gert Westermann, Anna L. Theakston, Julian M Pine, Padraic Monaghan, Elena V. M. Lieven

In this first presentation we introduce the five principles of the constructivist framework. We describe our framework as constructivist because at its heart is the idea that each child uses a toolkit of learning and processing mechanisms to construct their language using information sampled from their multimodal environment (principle 1). However, the framework also incorporates key ideas from other traditions in developmental science, combining and extending them, informed by empirical findings and theoretical debates. As specified by traditional nativist approaches, it recognises that children are born with sophisticated learning mechanisms that are innately specialised to “organise and incorporate experience” (Greenough et al., 1987: 539; (principle 2). As specified by traditional empiricist approaches, it acknowledges that the role of the environment is substantial and multifaceted, and that learning mechanisms combine multimodal information in the environment at different levels (e.g., statistical, conceptual and socio-cultural) such that the knowledge that they construct integrates across these levels (principle 3). In line with traditional constructivist approaches, the framework proposes that children are not passive receivers of information, but selectively interact with their environment to navigate the large number of potential cues available to them (principle 4). And finally, the approach is developmental, in line with dynamic systems theories (Thelen & Smith, 1994): the learning process is constrained at each point in development by children’s current knowledge state and the processing limitations under which they operate, which together determine what gets learned (principle 5). Note that individually none of these ideas is new. Thus, our intention is not to claim that we are presenting, in each principle, a radically new concept. Instead, we are attempting to shift the foundations of the debate, by showing how bringing these ideas together in a constructivist framework yields a more tractable, more plausible, approach to explaining the process of language acquisition.

Simultaneously learning words and grammar via simple associative learning

Authors: Padraic Monaghan; Rebecca L. A., Frost, Kirsty Dunn, Patrick Rebuschat

Different levels of language structures, such as vocabulary and grammar, are often assumed to be underwritten by different cognitive processes (Marchetto & Bonatti, 2014; Peña et al., 2002), operating at distinct stages of children’s development (Gleitman et al., 2005), such that basic vocabulary tends to be acquired before grammar. Evidence for the differentiation has typically emerged from controlled settings where multiple potentially confounding cues are not available to support the learner. However, these multiple environmental cues in naturalistic language learning situations may constitute a vital role in how the child is learning, and from what information (as per principle 2). In a series of studies, we have investigated how learners may acquire both vocabulary and grammar from cross-situational statistics between sentences in a novel language describing dynamic scenes. From a single pairing between a sentence and a scene, it is not possible to determine which word refers to which object, action, or property of the objects, and nor is it possible to determine what constraints between words in the sentence define the grammar. However, over multiple trials, the associations between certain words occurring in different grammatical roles in the sentence can be acquired. We have found that when the multiple sources of information in the environment interact with the language, both vocabulary and grammar are acquired, simultaneously, but some aspects of the language are learned less easily than others, which appears to depend on the variability present in the language structure. For instance, when the language varied in word order, morphological information about subject and object roles of was learned less quickly than information about word order constraints and verb and noun vocabulary. Word order constraints and verb vocabulary were learned quickly and robustly from a mechanism sensitive to the simple associations available in the combination of world and language (see principle 3). Though there remains the possibility that vocabulary and grammar are learned in very different ways (e.g., Ullman, 2016), our work highlights for which language learning phenomena such distinctions are not yet necessary in order to explain behaviour.

A multiple cue approach to the acquisition of complex adverbial sentences

Authors: Anna Theakston;

Complex sentences such as multi-clause adverbial sentences (e.g., ‘before she submitted the paper, she had a coffee’) present a puzzle in acquisition because their use and interpretation depends on the integration of multiple sources of information. For example, the child must establish a rich lexical representation of specific connectives (e.g., the temporal meanings of before and after, the causal/conditional meanings of because and if) as well as knowledge of the syntax governing correct use. In addition, adult-like use requires an understanding of specific form-meaning mappings. For instance, in languages like English, the child must learn when to place an adverbial clause sentence-initially or sentence-finally (e.g. Tom painted the fence before he washed the car vs. Before Tom washed the car, he painted the fence), a process that likely relies to some extent on both principles of iconicity (the order of events in the real world maps onto the order of events in the sentence) and an understanding of discourse pragmatics. Sentences involving causal (e.g., because) and conditional (e.g., if) connectives can additionally require integration of information about the real world and/or from the speaker’s knowledge and beliefs (e.g., I think she’s a queen, because she is wearing a crown). Furthermore, many complex sentence types convey a range of different pragmatic relations (e.g. Content, Speech-Act, Epistemic) and perform a number of illocutionary acts (e.g. Ask, Agree, Command, State/Claim). In this talk we will present empirical data from a number of corpus and experimental studies of adverbial clause acquisition that demonstrate how children are sensitive to a range of different sources of information, and how their knowledge changes over development. We will argue that constructivist accounts based on the five principles outlined in this symposium (especially principles 3, 4 and 5) provide a fruitful avenue to understanding these data.

Sensitivity to low-level surface features in input explains the acquisition of complex inflection

Authors: Elena Lieven; F Engelmann, B. Ambridge, S. Granlund, J Kolak, J Pine, M Szreder, A. Theakston, V-M Vihman

Many languages show complex patterns of inflectional morphology. Because children seem to master these types of complex system early, many theories of acquisition posit innate symbolic rules or a combination of symbolic rules with some kind of surface-form analogy and/or stored exceptions. We report two studies of nominal and verbal inflectional marking. In the study of noun case-marking in Polish, Finnish and Estonian, 3-to-5-year-old children were tested on their production of different cases when presented with nominative singular forms. Overall error rates were low, but with pockets of high error in lower-frequency parts of the system. Age, token frequency (Polish and Estonian) and phonological neighbourhood density (PND; Polish, Finnish and Estonian) were significant predictors of correct production, with the effect of PND greater for less frequent forms. Thus, the errors patterned as if learning was from surface-level distributional regularities in linguistic input and analogizing across forms that share phonological similarities. The errors usually involved replacing low-frequency targets with higher-frequency lexical forms of the same verb or higher-frequency markers from the correct target case. A second study reports the learning of Polish and Finnish verbal inflections. Using a novel elicitation method, all six person-number forms were elicited from 3-to-5-year-old children. The results were: Low overall error rates, but higher error rates for low-frequency forms; Substitution of more frequent forms of the same verb; Near misses (e.g., correct person but incorrect number) or conjugation class errors (correct person/number but suffix from wrong class). Thus, error patterns were predicted by token frequency and the process of analogy across phonologically similar forms, as demonstrated by an effect of PND (see principles 3 and 4). Both studies show learning based on surface forms which cannot be explained by approaches that posit abstract symbolic rules. We conclude by considering how these findings can be applied to other typologically complex inflectional systems.

07/17/2024, from 02:15 PM to 04:15 PM , Room P131

Symposium: Timing in language acquisition: Why are some phenomena acquired late?

Speakers: Petra Schulz; Angeliek van Hout; Esther Ruigendijk; Esther Rinke; Ianthi Tsimpli

Overall, children are very good language learners. We will discuss why certain syntactic and semantic phenomena are acquired late, after age 5, and how consideration of different approaches can help us understand timing in language acquisition.

The late acquisition of (certain) postverbal subjects by Italian heritage children

Authors: Jacopo Torregrossa; Andrea Listanti

Heritage language (HL) speakers seem to diverge from monolingual speakers in the acquisition of syntax–discourse interface phenomena. However, most of the studies reporting this finding do not make any distinction between different types of syntax–discourse interface structures. Therefore, it is an open question whether these structures are difficult for HL speakers across the board or whether different types of structures are associated with different acquisition outcomes. We investigate whether the timing of the acquisition of syntax–discourse interface structures among monolingual children affects their acquisition by HL children. We consider the acquisition of the alternation between preverbal and postverbal subjects with unaccusative, unergative and transitive verbs in Italian. This alternation involves the syntax–discourse interface across all verb types. However, mastery of this alternation has a different timing in monolingual acquisition depending on the verb type: it emerges earlier in association with unaccusative verbs than unergative and transitive ones. We elicit narratives in Italian from 42 Italian heritage children living in Germany and ranging in age from 7 to 14 years. The children show a good mastery of the alternation between preverbal and postverbal subjects with unaccusative verbs, whereas they exhibit difficulties with the same alternation in association with transitive and unergative verbs. The results of the study suggest a gradient interpretation of heritage speakers’ difficulty with syntax–discourse interface structures, with timing in monolingual acquisition being a relevant factor that modulates the degree of this difficulty.

Why some complex sentences containing temporal connectives are very late: Evidence from child Greek

Authors: Petra Schulz; Christos Makrodimitris (Goethe University)

“Children have difficulty comprehending complex sentences with temporal connectives until late childhood, but the reasons for this difficulty remain controversial. We tested six- to twelve-year-old children to assess how the iconicity of event-language mapping, type of connective, and clause order mediate the comprehension of temporal sentences. Sixty monolingual Greek-speaking children and 15 adult controls completed a picture-sequence selection task in which they judged after- and before-sentences in iconic and non-iconic order (e.g., He read a letter, after/before he ate an apple; After/Before he ate an apple, he read a letter).

Up to the age of twelve, children did not reach adult-like comprehension of the connectives; performance in non-iconic AFTER-sentences was significantly lower than in the other three conditions across ages. This pattern suggests that violation of iconicity negatively affected comprehension of AFTER but not of BEFORE.

We conclude that neither iconicity, connective, nor clause order can fully explain these findings and propose an account based on the interaction of iconicity and clause order: in languages with clause-initial connectives like English or Greek, non-iconic, sentence-medial AFTER requires revision of the initial event representation, resulting in an event-semantic kindergarten-path that children find difficult to overcome. In this case, children may maintain their initial incorrect representation. Non-iconic BEFORE-sentences are not harder than their iconic variant, because sentence-initial BEFORE serves as an early cue of the non-iconic order, so no reanalysis of the event-representation is needed.

The event-semantic kindergarten-path effect predicts that children should master non-iconic BEFORE earlier than non-iconic AFTER. This was borne out in our results; 23 children had mastered non-iconic BEFORE but not non-iconic AFTER, whereas no child had mastered non-iconic AFTER but not non-iconic BEFORE.

The second prediction is cross-linguistic in nature: if cues for the event-semantic (re)analysis appear non-clause-initially (e.g. Tamil, Mandarin), the acquisition path may be different.”

Pronoun interpretation in acquisition: early in German, and late in Dutch. Bilingual evidence

Authors: Esther Ruigendijk; Petra Hendriks (University of Groningen)

“Dutch-speaking children, like English-speaking children, make mistakes in the interpretation of pronouns until age 7 (the so-called Pronoun Interpretation Problem, PIP), whereas children speaking German or a Romance language already interpret pronouns correctly from age 4. The question is why the interpretation of pronouns is early in some and late in other languages. Explanations are sought in differences in the pronominal systems of the languages, or in external task related factors. To understand whether language internal or external factors are relevant for the lateness of this phenomenon, we address the question: what happens if a bilingual child acquires a language with late (Dutch) and one early acquisition (German) of pronoun interpretation? There are four possible outcomes: Dutch influences German: a PIP in both languages; German influences Dutch: no PIP in either language; bidirectional influence: smaller PIP in Dutch, increased PIP in German; no influence: a PIP in Dutch, no PIP in German.

We tested 21 Dutch-German bilingual children, age 3;8 to 6;11 with a picture selection task that included transitive sentences with a reflexive or a personal pronoun. Each child was tested in both languages, in separate sessions, with at least 1 week in between. Whereas the children performed very well on the interpretation of both pronouns and reflexives in German, the same children made more errors on pronouns than on reflexives in Dutch.

These results indicate that there is no cross-linguistic influence in pronoun interpretation. We find a PIP in Dutch, but not in German. The results also show that the pronoun interpretation problem is not a task effect or a language-independent effect of processing as has been argued before. Rather, it seems to originate in the grammatical system of the specific language: the observed cross-linguistic difference may arise from a stronger ambiguity of Dutch pronouns, which causes their late acquisition.”

Are late acquired structures challenging for heritage speakers?

Authors: Esther Rinke; Cristina Flores (University of Braga)

“Heritage speakers (HSs) acquire their Heritage language (HL) mainly in the family and in the context of a dominant environmental language. Their linguistic competence may show certain particularities which are often similar across different HL and independent of the environmental language. In this talk, I will put forward the hypothesis that late acquired structures in monolingual language acquisition are particularly challenging for bilingual speakers. I will present results of two studies on European Portuguese (EP) as HL: one on the interpretation of null/overt subjects and one on the production of null/clitic objects.

Based on an offline sentence interpretation task, we investigated subject interpretation in German/EP and Spanish/EP children/teenagers (aged 9-16) and compared them to age-matched monolingually raised Portuguese children and adult controls. The results show no effect of the contact language in the bilingual groups. Differences between monolingual and bilingual children arise in the null subject condition with respect to syntactic context (intersentential/cataphoric vs. anaphoric- context). However, in the overt pronoun conditions, monolingual children show the same non-adult-like behaviour like the bilinguals. This indicates that the adult-like interpretation of overt pronominal subjects is challenging for bilingual speakers and also acquired very late in monolingual acquisition.

Based on an elicited production experiment, we investigated object realization in monolingual EP and bilingual German/EP children (age 7-10). The bilingual children were sensitive to the different pragmatic conditions (not immediately accessible vs. immediately accessible referents), but they produced higher amounts of object omissions in comparison to age-matched monolinguals. However, younger monolinguals (age 5-6) produce the same omission rates as the bilinguals, showing that this area also develops late in monolingual language acquisition.

Taken together, the results of both studies indicate that structures which are acquired late in monolingual acquisition are particularly challenging for heritage speakers.”

07/17/2024, from 02:45 PM to 04:15 PM , Room P270

Symposium: Validating the Q-BEx questionnaire on language experience in multilingual children

Speakers: Sharon Unsworth; Ludovica Serratrice; Cecile DeCat; Laurie Tuller; Draško Kašćelan

The goal of this symposium is to present the results of an international validation study whose goal was to test and optimize the design and use of questionnaire on language experience in multilingual children..

Predicting language outcomes using the Q-BEx questionnaire

Authors: Ludovica Serratrice; Philippe Prévost

Typically, the more opportunities children have to hear and use a language, the better the outcomes, although distinct domains may be differentially affected. The association between outcomes and amount of exposure has been shown to be generally stronger for vocabulary breadth than for morphosyntax, while there is less information for vocabulary depth. The extent to which more qualitative aspects of bilingual children’s language experience predict language outcomes is less clear. The Q-BEx questionnaire offers a detailed profile of both quantity and quality of language experience. In this presentation we investigate the extent to which the language background and experience variables derived from Q-BEx predict children’s SL outcomes for vocabulary breadth and depth (Study 1) and morphosyntax (Study 2). These studies involved 233 multilingual children who were administered the PPVT, the Word classes subtest of the CELF-5, and the LITMUS sentence repetition task to assess vocabulary breadth, vocabulary depth and morphosyntax, respectively. In both studies we included the same set of Q-BEx predictors and added measures of short-term memory (STM), working memory (WM), phonology, and non-verbal IQ (NVIQ). Significant predictors were i) for vocabulary breadth: age, NVIQ, STM, homework in the SL, highest caregiver education, and number of older siblings; ii) for vocabulary depth: language concerns before 4, age, STM, and highest caregiver education overall and in the SL, and iii) for morphosyntax: age of first words, age, STM and WM, phonology, vocabulary breadth and depth, frequency of reading and tech activities in the SL, and current exposure to the SL. Decision tree analyses furthermore allowed us to tap into individual profiles and better understand the non-linear relationships and complex interactions between children’s language background and experience, their cognitive skills, and language outcomes. The theoretical and applied implications of these findings will be discussed.

How detailed do measures of language experience need to be? A cost-benefit analysis using the Q-BEx

Authors: Cecile DeCat;

Parental questionnaires vary considerably in how they quantify language exposure and use. Some require a detailed description of children’s daily routines, whereas others restrict themselves to more general questions about children’s language experience. The rationale behind more detailed questionnaires is that more information has the benefit of greater accuracy, though it costs more time to complete. The Language & Exposure module in the Q-BEx questionnaire offers users the options of collecting information at varying levels of detail. For example, exposure over time can be estimated either by using a single question about age of first exposure or as a cumulative measure weighted for different time periods. In this presentation, we use data gathered using these different questions to investigate which level of detail is (i) optimal to effectively balance “cost” and “benefit”, and (ii) minimally required to reliably inform research and practice. To do this, we determined the extent to which using different measures of language exposure were better able to predict bilingual children’s outcomes in the societal language (SL) (scores on morphosyntax, vocabulary breadth and depth) and the heritage language (HL) (parental estimates of oral proficiency). After first generating an optimal model containing all other relevant variables (e.g., age, working memory), we subsequently re-fitted the model each time using a different measure of exposure. Subsequently, adopting an information-theoretic approach to model fit comparison, we were able to identify the best-fitting model. For SL outcomes, this contained age of onset, with the models containing more detailed measures of exposure also receiving support for vocabulary. For HL outcomes, the best-fitting model contained the most detailed exposure measure, though the models with less detailed measures were likely good enough (depending on the goal and population). We will use these findings to make suggestions for the use of Q-BEx in research and practice.

Using the Q-BEx questionnaire to identify risk for language impairment

Authors: Laurie Tuller;

Inclusive population sampling entails that participant samples will necessarily subsume children who have a language disorder. This is especially important in studies on bilingual children, who are frequently subject to either over- or under-diagnosis for language impairment because of their bilingualism. Researchers therefore need to be able to identify children likely to have a language disorder. There is likewise a need, in educational settings, to understand the nature of language difficulties experienced by all students. And, of course, in clinical settings, there is an obvious need for a rich empirical base for optimal detection of language disorders in bilingual children. Identifying risk for language impairment was incorporated into Q-BEx in a dedicated module (Risk Factors), containing just three short questions: age of first word (in any of the child’s languages), age of first sentence (in any of the child’s languages), and any parental concerns about early language development (excluding languages new to the child). In this presentation, we investigate whether parents’ answers to these questions can be used to derive a score which can flag children who are likely to be at risk for language impairment. Our results suggest that this is the case. Scores using Q-BEx flagged children independently known to be at risk and furthermore predicted children’s performance on the quasi-universal version of LITMUS-nonword repetition, as well as sentence repetition in the SL. Equally important to researchers, educators, and clinicians is obtaining information about a child’s skills in the HL. It will be shown that parental assessment of children’s skills in the HL, taking into account their assessment of the SL as well, may provide another angle for flagging a child at risk for language impairment. In other words, this “quick and easy” module delivers, yielding a red flag index useful for researchers, educators, and clinicians.

Putting the Q-BEx questionnaire to a quality test

Authors: Draško Kašćelan;

Language background questionnaires are commonly completed by parents without help from researchers. While this approach is practical when an interview with the parent cannot be arranged, self-administration may lead respondents to accidentally provide inaccurate information. Researchers using such questionnaires do not generally report on respondents’ errors of this kind, even though they may be informative about the reliability of certain questions and furthermore provide useful information for questionnaire design and use more generally. As part of our validation study, and in the interests of full transparency, this presentation puts the Q-BEx questionnaire to a quality test by cross-validating parents’ answers internally (i.e., with other questions) and externally (i.e., with children’s scores on language tasks). Most parents (n=246) completed Q-BEx independently, but for others (n=54) assistance was given. The analyses focus on: (i) data loss, (ii) inconsistencies, and (ii) implausible answers. Our findings demonstrate minimal data loss caused by the questionnaire design. Where inconsistencies were found, these involved parents’ estimations of age of first exposure, and their ability to complete the most detailed version of the Exposure & Use module. For example, for about 1 in 10 children, the reported age of first exposure to each language was inaccurate when cross-checked against answers concerning amount of exposure in the early years. Similarly, for about 1 in 5 children, weighted estimates of exposure/use sometimes contained implausible answers (e.g., the child did not spend any time at school), though this dropped to 1 in 10 in France, where assisted administration was comparatively frequent. Finally, when parents were asked to estimate children’s proficiency in various ways (i.e., in general terms, in comparison to other multilinguals, in terms of their own satisfaction), their answers were highly consistent, but also different enough to suggest they tapped into different aspects. Implications for Q-BEx users are discussed.

Using the Q-BEx questionnaire in schools

Authors: Sharon Unsworth;

Teachers in mainstream schools are often unaware of the language experience and background of bilingual children in their classroom. For many (but unfortunately not all), this is not because of a lack of interest, but more often due to a lack of time to obtain such information and in many cases the result of a language barrier between them and the children’s parents. The Q-BEx questionnaire is available in many languages and thus offers teachers access to information about children’s language experience and background in both the societal and heritage languages which may otherwise remain unavailable. Information about prior experience with the societal language is essential as this has direct consequences for children’s level of proficiency in that language at school entry as well as their rate of acquisition throughout the early years of schooling. This can serve as a starting point for differentiated teaching strategies. Information about the heritage language is equally important as this allows teachers to better understand (and hopefully make use of) children’s full linguistic repertoire. As part of the output available to users, Q-BEx provides this information in a compact report containing key findings alongside an explanation of key variables written in accessible language. In short, then, the Q-BEx questionnaire has considerable potential for use in schools. In this presentation, we investigate (i) to what extent teachers also see this potential, (ii) how they assess its useability and the compact report, and (iii) whether there are factors which would prevent them from using Q-BEx in the future. To this end, we are conducting a focus group study with teachers (intended n=20) from a range of schools in the Netherlands. Teachers will report on their thoughts and experiences by completing an online questionnaire and/or during an interview. Data collection is ongoing.

07/18/2024, from 11:30 AM to 01:30 PM , Room P131

Symposium: A direct approach to global vocabulary assessment

Speakers: Margaret Friend; Celia Rosemberg; Florencia Alam; Naomi Havron; Michael Frank

Introducing four new adaptations of the Computerized Comprehension Task (CCT; Friend et al., 2012) presenters will discuss applications to bilingualism and understudied languages.

Scaling up: Increasing the age range and accessibility of early vocabulary assessment

Authors: Margaret Friend; Matthew McArthur, Diego Leon, and Melisa Gonzalez

“Vocabulary is a strong predictor of children’s cognitive, academic, and social outcomes and an important measure of proficiency in the languages of bilingual children. However, direct measures often require lab- or clinic-based administration, limiting equitable and representative access. Recent research focuses on measures easily administered in community settings. For example, Bleses et al., (2021) extended a lab-based measure of receptive vocabulary for children up to 24 months of age in English, Spanish, and French (the Computerized Comprehension Task or CCT; Friend, et al., 2012) to 35-month-old children learning Danish. The task was implemented on a tablet and administered by preschool teachers. Similarly, Lo, et al. (2021) used a similar, tablet-based, task with Norwegian children administered by parents in the home. This report concerns the development of the Web-CCT in English and Spanish, bringing continuous vocabulary assessment from 18 to 60 months of age to scale.

We present data on a preliminary sample of 52 (27F) children (M=35.9; range=19 to 64 months): English monolingual (n=43), Spanish-English bilingual (n=7), and Spanish monolingual (n=2). The psychometric properties of the Web-CCT are strong. The English Web-CCT correlates well with age (r42=.866, p<.001) and converges with the MCDI:WS (Marchman et al., 2023; r17=.721, p<.001) and the ROWPVT-4 (Martin & Brownell, 2011; r16=.750, p<.001). Internal consistency is excellent (=.982) and test-retest reliability is satisfactory (r13=.679, p=.008). Web-CCT Spanish-English conceptual vocabulary correlates well with age (r7=.916, p<.001) and converges with the ROWPVT-SBE (Brownell, 2012; r7=.775, p=.041). Internal consistency is excellent (=.974) and test-retest reliability is promising (r5=.822, p=.088). These measures expand the opportunity for early vocabulary assessment to families without ready access to a lab or clinic and can be completed on a desktop, laptop, tablet, or cellphone anywhere with Wi-Fi access. We anticipate presenting data on 100 participants. Applications of this extended measure will be discussed.”

Documentation of vocabulary acquisition in Bilingual Qom-Spanish children in Argentina

Authors: Celia Rosemberg; Florencia Alam, Gladys Ojea, Leandro Garber, Alejandra Stein, Carla De Benedictis, and Margaret Friend

This study documents vocabulary acquisition in bilingual Qom – Spanish indigenous children. The Qom ethnic group is one of the largest in Argentina (INDEC, 2010). Qom l’aqtaqa (a Guaycuruan language) is characterized by the morphological complexity of nouns and verbs, tending towards polysynthesis and agglutination. Only 44.3% of the Qom population over 5 years of age speak Qom l’aqtaqa at home (ECPI, 2004-2005). In rural communities language attrition has increased, due to the spread of social media and schooling in traditional Spanish monolingual contexts (Hetch, 2017). High quality bilingual education in the early years of schooling is crucial to fostering bilingual acquisition, intergenerational transmission, and cultural identity. We assessed Qom and Spanish vocabulary from preschool through the early school years to inform educational interventions and extend the literature on bilingual development. We developed adaptations of the Computerized Comprehension Task (Author, 2012) in Qom and Chaco Spanish. Item selection was based on word frequency in recordings of household interactions (corpus: Author, 2011- 2019) and interviews with native speakers. 70 bilingual children (x= 5.5 years; range=3.0 to 7.0) were assessed by monolingual Spanish and bilingual Qom teachers. Regression analyses revealed a positive trajectory in Spanish: the association between age and accuracy was positive (from 3 to 5 ⏀= .751, p= .00 and from 5 to 7: ⏀= .289, p= .048) and between age and latency was negative (from 3 to 5: ⏀= -.661, p= .00 and from 5 to 7: ⏀= -.222, p=.057). Qom showed a positive association with vocabulary from 3 to 5 (accuracy ⏀= .506, p= .037 and latency ⏀= -.780 p= .00), but the association with accuracy was negative from 5 to 7 years (⏀=.-.58 p=.00). This points to potential Indigenous language attrition relative to the majority language in the early school years.

Socioeconomic and cultural differences in lexical comprehension

Authors: Florencia Alam; Celia R. Rosemberg

The majority of the studies that have documented the effects of social inequality on lexical comprehension were built on a relatively small range of SES variation, since caregiver’s education generally ranges between 12 and 16 years of schooling in the samples analyzed. Furthermore, almost all the studies have been carried out in the USA or in Western Europe. This paper delves into the impact of SES on the comprehension of different types of content words in Spanish-speaking Argentinian children, a population in which socio-economic differences are more striking than in the previously studied populations. Previous findings have shown group variability, particularly in low SES groups (Author, 2021), which need to be considered in vocabulary development research. Therefore, in this study we examine children’s vocabulary comprehension in three socio cultural groups, distinguished by maternal education, place of residence and Indigenous heritage: semi-urban low-income Indigenous communities (IndLow), urban low-income (UrbLow) and urban middle-income (UrbMid). Using a performance based forced choice lexical recognition task implemented on a tablet (an adaptation of the Computerized Comprehension Task -CCT- Friend and Keplinger 2003), recognition accuracy and haptic response time to nouns, verbs and adjectives were assessed. 143 toddlers were tested. Regression analysis was carried out. Preliminary results that considered only data from UrbMidd and UrbLow showed overall SES effects on recognition accuracy (⏀=1.1, p=.00) but not on response time. Further analyses that considered lexical category revealed an interaction between SES and lexical category (⏀=-0.23, p=.00): UrbMidd children were only more accurate on the recognition of nouns and adjectives but not verbs. We expect to find differences between the three groups. The discussion links socio-cultural differences in children’s performance to previous evidence regarding the characteristics of the three groups of children’s linguistic experiences.

Adapting the Computerized Comprehension Task to Sub-dialects of Palestinian Arabic

Authors: Naomi Havron; Maali Jammal-Agbaria, Jawana Zoubi, Rawan Abu-Baker Watad, 

“The CCT is validated on resource-rich languages such as English and Spanish. It is important to adapt it to understudied languages. Arabic is such a language: There are few studies on language acquisition, despite its large number of speakers (e.g., the first CDI-WG parental-report questionnaire in Arabic was developed in 2021, but the data are not openly available, Abdelwahab et al., 2021). Thus, there is a need for assessment tools, like the CCT, for Arabic-learning children; however, developing them is challenging. Spoken Arabic has many dialects, varying greatly from one another – each dialect might need its own CCT. Second, the lack of resources and validated tests in Arabic make it hard to choose items, and to validate the CCT.

In this project, we are developing a CCT for two sub-dialects of Palestinian Arabic: the Northern-Triangle-Area and the Northern-Rural dialects. We based our word list on two sources: words from a Palestinian-Arabic parental-report tool (PA-CDI) for 18- to-36-month-olds (validated only for production, Hashoul Essa, 2018); and culturally-appropriate words from the English CCT. We created a preliminary list of 177 words, 22 of them were manipulation checks (easy words, e.g., “mother”, and hard words, e.g., “pier”). Pediatric speech-language pathologists rated the words on their difficulty level. The chosen 100 words constitute the items for the CCT. They are matched for syntactic and semantic categories, and difficulty level.

We will test 150 infants aged 18-22 months and should have results by July 2024. Validity will be assessed in comparison to a parental-report of comprehension of the same 100 words as the CCT items, and by examining well-known effects from the language-development literature, such as an advantage for girls over boys. Reliability will be assessed by repeating 12 of the word pairs a second time at the end of the test."	<p>

07/18/2024, from 11:30 AM to 01:30 PM , Room P200

Symposium: Bilingual Language Development in Varying Contexts: Insights from Longitudinal Studies

Speakers: Tamara Lautenschlaeger; Jessica Willard; Julie Smith; Johanne Paradis

How do young bilinguals develop in their two languages and why do they develop the way they do? This symposium presents findings from studies tracing bilingual development over several years for preschool- and school-age children in three countries.

Between-language Associations in Vocabulary Development in Bilingual Preschool Children

Authors: Tamara Lautenschlaeger; Alla Sawatzky, Jens Kaiser-Kratzmann, Steffi Sachse

“Theoretical background: Possible interdependencies between the languages of bilingual children regarding performance level and developmental trajectories are often discussed. However, to date only few studies have simultaneously investigated vocabulary development in both the majority and the heritage language.
Research question: We aim to investigate the vocabulary development in bilingual preschool children regarding the relationship between the majority language German and the heritage language Turkish.
Method: Data were taken from a larger longitudinal study, which investigates the language development in bilingual children in Germany (n = 423) over the course of the preschool age. Children in the subsample for the present analysis attended early childcare facilities in Germany and were exposed to Turkish as a heritage language through family members. Receptive and expressive vocabulary in German and Turkish were assessed annually. Only children with complete data in both languages for at least two measurement points were included in the present analysis (n = 40).
Results: Cross-sectionally vocabulary performance in the two languages was positively correlated for receptive vocabulary, while a negative correlation was found for expressive vocabulary. Vocabulary growth in the two languages was positively correlated for receptive vocabulary, but there was no correlation between languages regarding growth in expressive vocabulary.
Discussion and conclusion: The results suggest that the relationship between majority and heritage language differs between receptive and expressive vocabulary. These differences could indicate that receptive and expressive vocabulary are influenced by different factors. While receptive vocabulary might be more influenced by factors that affect both languages, expressive vocabulary may be more influenced by factors that are independent or in competition with each other.”

How do Turkish– and Russian–German Speaking Preschoolers Develop Strong Bilingual Vocabulary?

Authors: Jessica Willard; Nathalie Topaj, Natalia Gagarina

“Understanding how strong bilingual language skills develop requires examining both heritage and majority language outcomes over time to identify periods of growth, slowing growth or even attrition, and how internal and external factors relate to such developmental patterns. Child-internal factors have been theorized to support development in both languages, while certain input-related external characteristics could support only one of the two languages (Hoff et al., 2021). However, such findings could also be specific to children’s cultural context, specific language domains or analysis strategies.

Drawing on the BIVEM-study conducted within the Berlin Interdisciplinary Network for Multilingualism, we examine bilingual expressive vocabulary development from age 3-5 (n = 161) with two analytic approaches asking: 1) Is there growth in heritage and majority language vocabulary? 2) What developmental patterns co-occur? 3) Are there shared internal or external supports that predict high skills in both languages? Picture naming tests assessed expressive Turkish or Russian heritage and expressive German majority language vocabulary (SRUK, Gagarina, 2010; PDSS, Kauschke & Siegmüller, 2010), the SON-R nonverbal reasoning (Tellegen et al., 2007).

Variable-oriented growth curve models show much heterogeneity, but overall consistent growth in each language and no shared supports: Across the study period, girls have an advantage in heritage language vocabulary and children with stronger nonverbal reasoning in German vocabulary. Both an early start in preschool and exclusive use of the heritage language in the family support vocabulary in one language while posing a “risk” to the other. Preliminary person-oriented latent profile analyses suggest that various patterns of growth exist. A pattern of comparably strong vocabulary in two languages is supported by female gender and nonverbal reasoning, and can occur when exclusively the heritage language is spoken in the family. We discuss limitations (e.g., aggregating over heritage languages), and the specificity of findings (e.g., German context, research design).”

Examining Variability in the Oral Language and Early Literacy Trajectories of U.S. DLLs

Authors: Julie Smith; Jessica A. Willard, Carol Scheffner Hammer

“Children need a robust foundation in oral language and early literacy skills to become strong readers. Understanding differences in oral language and early literacy trajectories during early elementary school is crucial for supporting reading outcomes in diverse learners. It is particularly important to understand bilingual development, as the population of dual language learners (DLL) in the United States continues to grow. This study focuses on Spanish-English DLLs, who constitute the majority of school-age DLLs in the United States.

Academic expectations for oral language and early literacy development in DLLs are based on research suggesting that children can establish academic proficiency after five years of schooling in English (Cummins, 1981). However, DLLs’ time to academic proficiency may vary depending on their timing of exposure to English (Hammer et al., 2008) and the language of instruction (e.g., heritage language, majority language).

This study investigated differences in the time to Spanish and English academic proficiency between DLLs who were sequential language learners (SEL) and those who were simultaneous language learners (SIL). Spanish and English oral language and early literacy trajectories during kindergarten through second grade were examined in 83 children who had attended two years of English-only preschool and elementary school. By the end of second grade, differences between SEL and SIL groups were maintained in oral language skills but diminished in early literacy skills. In English, both groups approximated academic proficiency in early literacy skills, but only children in the SIL group reached academic oral language proficiency. Children in both groups demonstrated a decline in Spanish oral language and early literacy skills. Findings provide new insights on variability in DLLs’ time to academic proficiency. Implications for reading development and academic instruction will be discussed.”

Age of Arrival and Input Factors Modulate Arabic Heritage Language Maintenance in Child Refugees

Authors: Johanne Paradis; Adriana Soto-Corominas, Evangelia Daskalaki, Redab Al Janaideh, Xi Chen, Alexandra Gottardo

“Most studies on heritage language/HL development are cross-sectional and include second-generation children. By contrast, the recent resettlement of 44,620 Syrian refugees in Canada offered the opportunity to examine – in vivo – HL maintenance/attrition in first-generation children who vary in age of arrival/AOA, but all began acquiring the societal language/SL at the same time. It is possible that first-generation children, especially older arrivals, might have different profiles of HL maintenance/attrition and dominance shift to the SL than second-generation children. Accordingly, we examined Arabic-English lexical and morphosyntactic development in children during their first 4.5 years in Canada to address the following questions: (1) Is dominant language shift taking place during this time and is it modulated by AOA? (2) How do AOA and HL input factors (concurrent and longitudinal) predict individual differences in the HL after 4.5 years?

Data collection took place at three time periods (T1-T3). At T1, 133 children (mean age = 9;5(2;0); mean AOA = 7;7) participated from three Canadian cities. Children were administered sentence repetition and receptive vocabulary tasks in English and Arabic, and information about AOA and language input and experience was obtained via parent report. Linear modelling revealed that, across time periods, older AOA children had superior Arabic vocabulary and morphosyntax than younger AOA children, but growth in English was stronger than in Arabic for both. The younger AOA children showed a shift to English dominance for morphosyntax by T3. Concurrent and longitudinal modelling revealed that speaking more Arabic with friends, having more Arabic schooling pre-migration, and engaging in language-rich activities in Arabic contributed positively to HL proficiency at T3; using more English with siblings was a negative contributor. We conclude that the HL is vulnerable even in first-generation migrant children, but older AOA and quantity and quality of HL input can boost HL maintenance.”

07/18/2024, from 11:30 AM to 01:30 PM , Room P301

Symposium: TALK (Tackling Acquisition of Language in Kids): a focused US NIH initiative on late talkers

Speakers: Bonnie Lau; Helen Tager-Flusberg; Nan Bernstein Ratner; Laura Justice; Mary Alt

Five research teams report on outcomes of research funded by a US NIH focused initiative to understand late talking: its roots, outcomes and effective interventions.

Early brain and behavioral predictors of late language emergence

Authors: Bonnie Lau;

There is converging evidence that starting intervention programs earlier, such as in the first year of life, leads to more optimal language and learning outcomes. However, the identification of infants at higher likelihood of language delay that would benefit from intervention remains challenging. Current language screening tools are often administered between 18 to 24 months of age, or even later. In this project, we obtain brain measures of speech, music, and binaural stimuli using magneto-/electroencephalography in combination with a behavioral assay of cross-domain development including language abilities, auditory skills, vocabulary, and adaptive behaviors at 3-, 6-, and 11-months in infants with and without hearing loss. As part of the NIH TALK supplement, these children will return to the laboratory for an additional 30-month visit or 5-year-old visit which includes a standardized language assessment, a natural language sample, measures of intellectual ability, in addition to parent-report measures of early language, adaptive behaviors, and language experience – to identify the late talkers in our sample. As we have longitudinal brain measures in combination with a clinical battery including both parent-report and direct assessments administered in the laboratory, we are able to compare objective brain measures to clinically relevant language measures longitudinally from as young as 3 months of age. We hypothesize that as language development is complex, effective infant language screening will require a combination of brain and behavioral measures interpreted in the context of demographic and social determinant of health variables. This presentation will summarize the preliminary findings from this project.

Exploring predictors of late talking in a diverse, low-resource population-based urban sample

Authors: Helen Tager-Flusberg; Charles A. Nelson

“Using a prospective study design in a racially, ethnically, and socioeconomically diverse primary care population, we will identify the EEG features measured during the first year of life at 4, 9 and 12 months that are associated with late talkers identified at 12, 18 or 24 months. Building on our past work we will explore several cross-sectional and longitudinal approaches to analyzing EEG data including a) data-driven model selection using EEG power spectra; and b) complex signal processing of nonlinear EEG measures.

We are currently carrying out a large-scale (projected N=720) longitudinal study of infants enrolled in early infancy exploring neural (resting EEG markers), behavioral, demographic, and environmental factors that predict autism and related neurodevelopmental outcomes. Our participants are recruited from a clinic at an urban hospital serving a racially, ethnically, and socioeconomically diverse primary care population. At 4, 9 and 12 months of age a five-minute resting EEG sample is collected in the clinic while the children sit on their parents’ lap watching a screen-saver. Additional measures collected at each time point (in English or Spanish) include behavioral development (ASQ-3), demographic information, prenatal, perinatal and medical history (both child and mother), maternal depression, perceived stress, knowledge of child development and life events. The CSBS at 12 months, and the MCDI at 18 and 24 months are used to identify late-talking children (lowest 20%) and parent-child interactions collected at 24 and 30 months provide rich information about the children’s expressive language and parental input. To date, over 250 infants have been enrolled in the study. We will present preliminary data on EEG features (e.g., periodic and aperiodic power) and demographic/environmental

factors that are associated with late talking and discuss plans for future explorations of this unique dataset.”

Using phonology and conversational turn taking profiles to predict outcomes of late talking

Authors: Nan Bernstein Ratner; Carly Rosvold, Brian MacWhinney

“An enduring challenge is predicting which LTs will basically “catch up” to their peers (~ two-thirds or more), while others continue to demonstrate lags in language skill development. This project takes two related phenomena (child phonological skill) and parental response to communicative gambits (e.g., recasting, requesting clarification) to assess whether they can differentiate outcomes in two large longitudinal corpora which tracked families from 2-6 years of age (total N=86 LT, as well as a comparison corpus of age- and sex-matched TD peers).

In our first hypothesis, we propose that some children’s basic limitation in achieving early language milestones may reflect phonological immaturity, rather than underlying deficits in word learning or grammatical development; such children are more likely to catch up once they attain more intelligible speech. To evaluate this hypothesis, we are exhaustively evaluating phonological profiles (using PHON to code for accuracy, inventories, phonological processes) of the late-talking children engaged in conversational play with parents.

Further, we believe that individual differences in parental efforts to recast or elaborate child language efforts, together with individual child responses to such efforts, may partially explain differential outcomes in cohorts of late talking children. To this end, we are concurrently examining parental responses to child turn attempts, including recasting and expansion behaviors, and requests for child clarification. We also believe that some children are more attentive to such adult behaviors than others and predict that child re-attempts may be predictive of stronger language profiles at later time points. Results of both analytical aims will be presented at the meeting in July.”

Testing the Family Stress Model for Modeling the Relations between Early Childhood Poverty and LLE

Authors: Laura Justice; Hui Jiang, Britt Singletary

“Heightened susceptibility of children in poverty for developmental language disorder (DLD) suggests that ecological factors likely contribute to DLD (Norbury et al., 2016). However, there is limited understanding of the mechanisms through which poverty may disrupt early language development and lead to DLD.

The Family Stress Model presents an overarching model for detailing how, in low-income families, economic hardship leads to economic pressure, which in turn contributes to inter-relationship conflict, disrupting parenting, and parental distress in the family unit. These serve as the pathway through which poverty negatively affects child and youth outcomes.

In this study, part of the larger NIH-supported Small Talk project, we evaluate the applicability of the Family Stress Model for understanding how poverty may disrupt early language trajectories and contribute to late language emergence (LLE; AKA late talking) for children born into poor homes. For the present work, 353 mothers (all low-income) and 6-month-old children participated in data-collection sessions every 3 to 6 months to the present (underway). Measures of the home context were collected at time-point 1 to represent the contextual aspects of the Family Stress Model: economic hardship, economic pressure, inter-relationship conflict, disrupted parenting, and parental distress. Measures of child language skill were captured every six months.

Analyses are underway in advance of the conference. We shall present findings pursuant to three specific aims utilizing the time-point 1 and 2 data: (1) to identify the prevalence of LLE and the theorized domains in the Family Stress Model that differentiate children with LLE from those without LLE; (2) to examine early language trajectories from ~6 months to ~27 months for those with LLE and those without LLE; and (3) to identify the pathways through which the ecology of poverty disrupts early language trajectories and contributes to LLE.”

Describing the Skills and Lived Experiences of Late Talkers Following Toddlerhood

Authors: Mary Alt;

“Our team has provided vocabulary intervention to more than 60 late-talking toddlers, determining which treatment parameters lead to the best outcomes. However, our outcomes currently only extend to 4-6 weeks post treatment-and focus solely on vocabulary. We will follow up with children who will range in age from 30-months to 9-years of age, based on when they initially participated in our treatment protocol. These children will all be verified as having been late talkers.

We plan share the follow-up data we collect from the children who completed the Vocabulary Acquisition and Usage for Late Talkers (VAULT) treatment protocol to determine: (1) How many children continue to show signs of risk/impairment in language, speech, phonological awareness/reading, cognitive skills, or educational performance and (2) If there are early indicators (e.g., demographic information, response to treatment, SES) that predict later outcomes.

We will also collect qualitative data from families of late talkers and former late talkers themselves to better understand the functional and social effects of late talking and to understand the lived experience of being a late talker. We will work with families to co-create the interview protocol, and involve families in member checking, to ensure that the themes that emerge from our interviews accurately represent their points of view. For the children’s interviews, we will modify McCormack et al.’s (2022) Drawing Talking protocol, as an age-appropriate manner to elicit information.

Our qualitative data will add unique information to the literature about late talkers. In addition to understanding how late talking affects family dynamics, children’s lived experiences, and how families perceive their children’s communication, academic, and social skills, we will be able to shed light on families’ impressions of treatment, their goals for their children, and their views and experiences on the supports available to them.”

07/18/2024, from 11:30 AM to 01:30 PM , Room P018

Symposium: Language acquisition in children who are deaf or hard of hearing from preschool to adolescence

Speakers: Elizabeth Walker Walker; Elizabeth Walker; Krystal Werfel; Elizabeth Heinrichs-Graham; Ryan McCreery; Tina Grieco-Calub

This symposium presents research from studies focusing on speech perception and language development in children who are deaf or hard of hearing compared to hearing peers, as well as the underlying processes that drive these trajectories. We will als

Growth Trajectories in Vocabulary Depth and Breadth for Adolescents with Hearing Loss

Authors: Elizabeth Walker; Jacob J. Oleson, Ryan W. McCreery

Children with hearing loss (CHL) show vocabulary delays relative to typical-hearing peers, which can lead to cascading negative effects in reading achievement. Prior research on vocabulary knowledge in CHL has focused primarily on their breadth of knowledge (how many words one knows). While the breadth of one’s receptive vocabulary is important to measure, it does not fully capture their knowledge of their lexicon. Therefore, it is important to also assess vocabulary depth (how much one knows about a word). The primary aim of the current presentation is to examine trajectories of vocabulary breadth and depth from age 7 to age 18 years in CHL (n = 114) and age-matched children with typical hearing (n = 58). A secondary aim is to determine the factors that influence variation in vocabulary breadth and depth for CHL. We used a linear mixed model to test for group differences while accounting for age and correlated observations. Group, age, and a group by age interaction were fixed effects, with a random intercept to account for within subject correlation. Preliminary results suggest that children with typical hearing had significantly higher scores for vocabulary breadth compared to CHL at younger ages, but the effect size diminished over time. In contrast, differences between groups in vocabulary depth remained constant with age out to age 9. These results suggest that CHL face persistent deficits in their depth of vocabulary knowledge. Given the emerging research on the importance of both vocabulary breadth and depth on reading comprehension, these data suggest that the focus of vocabulary intervention may need to shift from quantity to quality.

Spoken Language Morphosyntax by Children who are Deaf or Hard of Hearing Across the Preschool Years

Authors: Krystal Werfel; Emily A. Lund, Lisa Fitton

The purpose of this presentation is to describe morphosyntax development in children who are deaf or hard of hearing (DHH) over the preschool years. Recently, we reported that children who are DHH and use spoken language lag behind their peers with typical hearing on performance on an elicited morphosyntax task throughout preschool (Werfel et al., 2022). In this presentation, we will extend this finding and report on longitudinal morphosyntax acquisition measured via naturalistic language samples as well as elicited tasks for children who are DHH (n = 100) and children with typical hearing (n = 70). All children primarily use spoken English. Children participated in a 12-minute language sample following the Hadley (1998) protocol, as well as completed the Test of Early Grammatical Impairment (Rice & Wexler, 2001), which contains two subtests that elicit production of (a) third person singular markers and (b) regular and irregular past tense markers. Each child was assessed initially at age 4 and then at 6-month intervals until they turned 6. Preliminary findings indicate that children make gains in morphosyntax production across the preschool years in elicited as well as naturalistic language tasks. For elicited tasks, an interaction of group and time is observed, such that children with typical hearing display ceiling effects while children who are DHH continue to exhibit growth. This interaction is not observed for language sample variables; all groups exhibit growth but children who are DHH consistently lag behind children with typical hearing across all language sample morphosyntax variables. Additionally, for children who are DHH, morphosyntax performance at age 5;0 accounts for 60% of the variance in performance on an omnibus norm-referenced measure of spoken language at age 6;0. Elicited and naturalistic tasks contribute similar unique variance. Theoretical implications of morphosyntax acquisition in children who are DHH will be discussed.

Uncovering the Brain Dynamics Serving Grammaticality Judgement in Children who are Hard of Hearing

Authors: Elizabeth Heinrichs-Graham;

Children who are hard-of-hearing (CHH) are at a heightened risk of a myriad of language delays through development. Among the most prominent delays is that of grammar processing. Moreover, grammar ability has been shown to mediate relationships between hearing status and narrative storytelling, reading, and other important developmental language milestones. Unfortunately, it is unclear what the underlying mechanisms that lead to these alterations in grammar ability in CHH are, which makes therapeutic decision-making more difficult for these youth. This study sought to determine the effects of age and hearing status on the neural dynamics of grammaticality judgement in a large cohort of children aged 7-15 years old with normal hearing, as well as a matched cohort of children who are hard-of-hearing. Magnetoencephalographic (MEG) data was collected while participants performed a grammaticality judgement task, where they listened to a 6-7 syllable sentence and were instructed to determine whether the last (target) word was grammatically correct. Artifact-free trials were decomposed into the time-frequency domain, significant spectro-temporal neural responses were imaged using beamforming, and these images were subject to whole-brain statistical analysis. We found significant age-related increases in beta desynchronous activity throughout the left language network, including inferior frontal, superior temporal, and parietal regions, as well as the anterior cingulate and left dorsolateral prefrontal cortex. We also found significant age-by-hearing status interactions in the beta band within this language network, such that these age-related increases were diminished in CHH. Neural dynamics correlated with behavioral performance. These data provide new neurophysiological evidence of altered grammar processing in CHH through development and may hold promise in guiding individualized therapeutic approaches in the future.

Effects of language and executive function on speech recognition in children with hearing loss

Authors: Ryan McCreery;

Children with mild to severe hearing loss exhibit delays in speech recognition compared to peers with typical hearing levels, even when auditory access is improved with hearing aids. These persistent deficits are even greater in degraded conditions, which creates challenges for listening and learning in everyday conditions like classrooms. Development of speech recognition in degraded conditions is thought to be supported by parallel maturation of language and cognitive abilities, but previous cross-sectional analyses showed varied patterns of association between speech recognition, language abilities, and executive function. We analyzed longitudinal development of speech recognition in a large cohort (n=125) of children with mild to severe hearing loss between 18 months-old and 9 years-old and a group of children with typical hearing matched for age, socioeconomic status and nonverbal intelligence. Early word recognition was associated with receptive vocabulary. Word and sentence recognition in noise at school age was associated with vocabulary and working memory skills, as well as factors related to cumulative auditory experience. These results suggest that the relationships between speech recognition and language and cognitive skill depend on the child’s age and complexity of the speech recognition task. These results have implications for clinical assessment of speech recognition in children with mild to severe hearing loss who use hearing aids.

Using gated word recognition to test the effect of visual speech cues on lexical access

Authors: Tina Grieco-Calub;

Listeners process words incrementally to facilitate lexical access and speed of understanding. Children who are deaf and who use cochlear implants (CIs) have altered auditory and linguistic experience early in life and are typically slower at processing speech in real time. This may be due, in part, to disrupted access to the acoustics of speech related to their hearing loss or technical limitations of their hearing devices. The purpose of this presentation is to discuss the extent to which visual speech cues (i.e., lipreading) facilitate word recognition in children when speech is degraded. Forty-seven children between ages 6 and 10 years who used CIs (n = 14) or who have typical hearing ability (TH, n=33) participated in a gated word recognition task. In this task, children are presented with partial phonological information of target words and asked to verbally produce the identity of the word. Children are provided with additional “gates” of phonological content of the target word on subsequent trials. Children are assigned to a condition where speech stimuli have high acoustic fidelity (HF) or are spectrally degraded by an 8-channel noiseband vocoder (D). Within a condition, stimuli are presented in an auditory-only or audiovisual modality. Performance is quantified by proportion correct per gate (1-5), modality (auditory-only, audiovisual), and group (CI-HF, TH-HF, TH-D). Preliminary results suggest that CI-C and TH-D groups have reduced performance compared to the TH-HF group in the auditory-only modality across all gates. Additionally, visual speech cues improve performance in the CI-C and TH-D groups. These data support the idea that although speech degradation disrupts incremental speech processing, children can utilize visual speech cues to promote lexical access in these situations. The discussion will include implications for future research and clinical practice.

07/18/2024, from 11:30 AM to 01:30 PM , Room P300

Symposium: Links between language and cognitive development: insights from childhood deafness

Speakers: Gary Morgan; Judit Gervain; Claire monroy; Mario Figuroa; Laura Bosch; Beatriz de Diego-Lázaro

How does wider cognitive development influence language. This symposium addresses this question with studies of children born deaf

Rhythmic discrimination of languages in deaf infants

Authors: Judit Gervain; Gaia Lucarini, Caroline Nallet, Davide Brotto, Alessandro Martini, Patrizia Trevis

“At birth, hearing newborns show sensitivity to the prosody, i.e. melody and rhythm, of their native language, i.e. the language they heard prenatally (Peña et al., 2003, May et al., 2018). For instance, they can discriminate rhythmically different languages (Ramus et al. 2000) and show a heightened, left-lateralized brain response to the language heard prenatally (Peña et al. 2003). A current hypothesis (Gervain, 2018; Nallet & Gervain, 2021) suggests that this prenatally heard prosody provides the basis of early speech perception and helps infants discover other linguistic units after birth, when the full-spectrum speech signal is available. Prenatal experience is thus hypothesized to be relevant for language learning. But what happens when prenatal experience is different?

To investigate this, we are testing the ability of 0-10-month-old deaf infants (DI) with varying hearing thresholds (40-90dB) to discriminate their native language (Italian) from a rhythmically different unfamiliar language (English). Sentences in both languages are presented forward and backward. Backward speech, with perturbed temporal features, is a standardly used non-linguistic control (Peña et al., 2003). A group of age-matched normal hearing (NH) infants is also tested. Infants’ brain responses are recorded using functional near-infrared spectroscopy (fNIRS) recorded in twenty channels covering the frontal, temporal and parietal regions, bilaterally.

Data collection in deaf and hearing infants (n=20 D, n=20 H). Results suggest that both deaf and hearing infants respond differentially to Italian and English, although the exact pattern of responses is different. If confirmed, these results suggest that some basic, broad acoustic features of language are perceived, even prenatally, by deaf infants with varying levels of auditory thresholds, with a possible contribution from vibrotactile experience. Such abilities need to be considered in the language acquisition process and may be built on during the developmental path of deaf children.”

Cognitive markers in deaf infants and associations with language

Authors: Claire monroy; Derek Houston

“Children with hearing loss demonstrate atypical performance across many general cognitive skills, such as visual working memory and visual statistical learning. However, the evidence for these deficits has been challenged, with mixed findings emerging in recent years, and is limited to studies involving school-age children. In our work we are investigating cognitive skills in young babies, and associations with later language, to investigate the effects of hearing loss on cognition early in development. In two studies, we compared visual habituation (study 1) and visual statistical learning (VSL; study 2) between deaf and hearing babies.

Study 1 implemented a visual habituation-oddity paradigm to deaf and hearing infants. Study 2 used mobile eye-tracking to compare VSL between groups of deaf and hearing infants, prior to cochlear implantation (pre-CI), and toddlers 12 months post-implantation (post-CI). Study 1 revealed that deaf infants were slower to habituate to a visual stimulus than hearing infants. For deaf pre-CI infants, habituation measures correlated with language scores on a standardized assessment. Study 2 similarly revealed a significant difference between deaf pre-CI vs. hearing infants, with evidence for learning only in the hearing infants. However, there were no differences between deaf vs. hearing toddlers post-CI, with both groups demonstrating learning. VSL performance was also positively correlated with language scores for the deaf toddlers. These findings provide the first evidence that differences in essential cognitive abilities between deaf and hearing children emerge in infancy. They also suggest that cognitive differences between deaf and hearing infants can explain some of the variability in the neurocognitive, social, and language outcomes in children with hearing loss. We will discuss the potential mechanisms that drive the effects of hearing loss on general cognitive development and offer some ideas for future experiments that could shed light on these mechanisms.”

Social cognition in processing overheard communication: A comparison of deaf and hearing toddlers.

Authors: Mario Figuroa; Gary Morgan

“The early years of life are considered a critical period for the development of theory of mind (ToM). The foundations of this construct can be acquired through direct communication, where the child consciously seeks interaction from an adult interlocutor. Input to early ToM can also come from overhearing/seeing the conversations of others. This avenue for receiving ToM input is more difficult to process for deaf infants. This study will investigate how deaf children use their social-cognitive skills in order to understand intentions in situations which require overhearing.

Fifty-one toddlers with and without deafness participated in this study (deaf N=26, hearing N=25). Participants ranged in age from 10 – 40 months. Deaf toddlers wore hearing aids or cochlear implants and were exposed to a spoken language. All of them were evaluated with a social-cognition task adapted from Behne et al. (2005) with 4 incrementally difficult conditions. In the non-verbal conditions, an experimenter addressed an assistant and indicated the whereabouts of a hidden toy’s location by (1) pointing or (2) gazing ostensively. In the case of the verbal conditions, the experimenter indicated the location by (3) uttering a verbal interjection or (4) having a conversation which that included the spoken deictic ´there´. The results showed that deaf toddlers understood the intentions of others when they were expressed through pointing and less so by gaze. They did not process overheard verbal intentions. Hearing children showed an understanding of both non-verbal and partly the verbal conditions. The present study thus shows that deaf toddlers are able to monitor and comprehend some aspects of third-party interactions in order to process early aspects of ToM but these are delayed compared with normally hearing peers.”

Fast mapping skills in congenitally deaf children two years after implantation

Authors: Laura Bosch;

“Fast mapping, the ability to establish a rapid audio-visual label-referent association from minimal exposure, is connected to word learning and can enhance lexical growth around 2 years of age. For congenitally deaf infants, late access to auditory information can hinder the deployment of fast mapping skills, overall constraining progress in lexical learning. Although early implanted deaf children seem to catch up with their normal-hearing peers, variability is high and successful learning post-implantation remains difficult to predict. Here, fast mapping data from a behavioral task, involving 4 novel words, was obtained in two groups of participants at age 2 years (chronological age and age after implant activation, for n=24 normal hearing -NH-, and n=22 implanted children -CI-, respectively). Cognitive and language development measures (BSID-III), and expressive language levels (MacArthur CDI), were obtained.

After a training trial, the task involved a referent selection phase, followed by two recall phases, immediate and delayed recall. Each phase included 8 trials, always presenting 3 objects, 2 novel and 1 familiar in both recall phases. Each novel object was requested twice. This multiple label-object task is demanding at age 2 years and suitable to explore variability in performance. Although a lower performance could be expected in the CI group, older (chronological) age at test and normal cognitive capacity would lead to a more similar between-group performance.”

Executive Function and Language in Children who are Deaf or Hard of Hearing

Authors: Beatriz de Diego-Lázaro;

“Auditory and linguistic deprivation have a widespread impact on language and executive function (EF) abilities in children. While this has been extensively shown in children who are deaf, it is unclear whether children who are hard of hearing (HH) show EF difficulties and what factors are related to EF in children who are deaf or hard of hearing (DHH). The purpose of this study was to (1) assess differences in EF in children who were DHH and (2) identify auditory and linguistic factors related to EF in this group. Twenty-six children with normal hearing (NH), 15 children who were deaf, and 16 children who were HH between 4 and 8 years completed hearing, vocabulary, word learning, and direct and reported EF measures. Children completed the Color Stroop subtest from the Leiter Scale and caregivers completed the Behavior Rating Inventory of Executive Function (BRIEF).

ANCOVAs controlling for age showed a significant main effect of hearing group (NH, HH, and deaf) on selective attention measured by the Color Stroop (F [2,53] = 3.15, p <.05). Children who were deaf (p = .03) and HH (p = .04) showed poorer selective attention than children with NH, but they did not differ from each other (p = .90). We did not observe any significant differences by hearing group in the reported EF (BRIEF). Correlations revealed that age at first expressive word, speech perception, vocabulary, and word learning correlated positively with direct and reported EF measures in children who were DHH. We observed minimal correlations between direct and reported EF measures. Both children who were deaf and HH showed EF difficulties compared to NH peers, suggesting that even partial auditory and linguistic deprivation might impact EF abilities. Hearing and language abilities contribute to enhanced EF in children who are DHH.”

07/18/2024, from 11:30 AM to 01:30 PM , Room P104

Symposium: Expanding Diversity in Language Development Research through the Acquisition Sketch Project

Speakers: Rebecca Defina; abdellah Elouatiq; Dorjderem Byambasuren; Lucy Davidson; Birgit Hellwig

“The field of child language acquisition has a proud history of crosslinguistic research, highlighting the importance of taking into account the many ways that languages vary when building theoretical models of the acquisition process. However, Kidd and Garcia (2022) recently reported that while research on less-commonly studied languages is increasing, our evidential base is still severely skewed: we only had papers on around 103 of the world’s current 7,000 languages published in the last 40 years in the four main journals in the field. Over the last few years, we have developed a model of data collection that aims at rapidly increasing language coverage in the field, with an eye to also giving back to communities that contribute data: the Acquisition Sketch Project. Drawing inspiration from the field of language documentation, we propose that researchers write an “acquisition sketch” based on a minimum of 5 hours of naturalistic data from children aged 2 – 4 years. A sketch is a broad overview of properties of the input and children’s productions, covering core topics in acquisition.
The presentations in this symposium highlight work across different language groups and topics, following the guidelines of the Sketch Acquisition Manual. Each talk highlights one topic of particular interest within one language. The symposium thus presents a showcase of research in underdescribed languages: featuring five unrelated languages from Australia, Mongolia, Morocco, and Papua New Guinea with topics including child and child-directed language, phonology, lexicon, and morphology, as well as feedback signaling and gesture. The aim is to illustrate the potential of the sketch format by demonstrating its application in a variety of contexts.”

07/18/2024, from 02:30 PM to 04:30 PM , Room P429

Symposium: Nonword Repetition – The role of administration criteria and nonword characteristics

Speakers: Maren Eikerling; Anna-Lena Scherger; Maria Luisa Lorusso; Juliane Hinnerichs; Kathrin Heeg

Nonword repetition tasks (NWRT) are widespread in child language assessment. Tests vary in nonword characteristics and administration criteria. Their influence on children’s repetition performance is high-lighted and discussed within the symposium.

Nonword repetition in bilingual German children by age 4 to 8

Authors: Anna-Lena Scherger; Angela Grimm

“Nonword repetition tasks vary with respect to the nature of nonwords, the focus on linguistic phenom-ena or working memory and the language-specificity. The present talk presents a nonword repetition task that has been developed within the COST Action IS0408 (2009-2013) and has been further adapted and evaluated since: the LITMUS quasi universal nonword repetition task (LITMUS-QU-NWRT) by Grimm (2022).

In the development of the LITMUS-QU-NWRT, the focus was placed on a bilingualism-sensitive con-struction of the items in the way that typologically well-attested vowels and consonants were used. Further, the task aims to tap into children’s phonological knowledge rather than to rely on multisyllablic nonwords that measure working memory load. Therefore, only nonwords with one to three syllables are included. The task consists of a language-dependent and a language-independent part and is prere-corded in a set of powerpoint slides in order to control for speech rate and word stress.

In the present talk, we will present data of several hundred children within the age range between 4 and 8 years of age that has been collected in various research projects. We will highlight data from children with very low exposure to German (>12 months; N = 85) where results show amongst others difficulties with consonant clusters. Further, we will present the state of the art of a research project running from 2023-2026 which aims at collecting norms for the LITMUS-QU-NWRT for children growing up with more than one language in their input.

Results will be discussed with respect to the validity of this tool for assessing children with DLD in a bilingual population and will address strengths and limitations.”

Influence of nonword characteristics as rated by adults on bilingual children’s repetition

Authors: Maria Luisa Lorusso; Maren Eikerling; Maria Luisa Lorusso

“Nonword repetition tasks are considered to be a suitable clinical marker for Developmental Language Disorders (DLD), both in mono- and multilingual children. However, the characteristics concerning pro-nounceability and language-specificity of nonwords seem to play a major role for children’s repetition performance. Thus, specific linguistic characteristics of nonwords should be taken into account in the construction, administration and scoring of such tasks. Moreover, the interaction of such effects with the degree of exposure to the language of assessment should be considered, to avoid penalizing chil-dren with little exposure.

To this end, we first identified the pronounceability and language specificity (i.e., language allocation) of a collection of nonwords, based on the online ratings of 68 adult native speakers of the two lan-guages. A selection of German- and Italian-language-specific as well as language-non-specific non-words yielding the best ratings according to pre-specified parameters and cut-offs was administered to 28 Italian-speaking children attending kindergarten in Germany using an automated screening platform. Some of the children were typically developing, whereas others had a diagnosis or suspect of DLD.

The results indicate that language specificity has different effects depending on DLD risk status. It is further shown that language-specific stimuli appear to be particularly sensitive indicators, possibly due to reduced sensitivity to frequent, familiar characteristics of the linguistic stimuli in children with DLD. The findings further confirm the role of language specificity (in terms of target language alikeness) but also of other characteristics related to word structure complexity. Furthermore, the study shows that it is possible to administer pre-recorded language-specific nonwords of a heritage language in an auto-mated way and to have the children repeat them, which simplifies the application of repetition tasks with language-specific nonwords in different languages.”

Using the LITMUS-CL-NWRT in Germany to differentiate between bilingual children with and with-out DL

Authors: Juliane Hinnerichs;

“Many German SLTs use Nonword Repetition Tasks (NWRT) to measure phonological short-term memory (Eikerling et al. 2023). The often-highlighted benefit of this task is that the items are less de-pendent on a specific language than real words – making it a promising tool for the diagnosis of De-velomental Language Disorders (DLD) in bilingual children.

However, it is impossible to construct nonwords which are equally compatible with all the worlds’ lan-guages. Within an international research network the LITMUS-CL-NWRT (Language Impairment Testing in Multilingual Settings – Crosslinguistic NWRT; bi-sli.org) was designed in order to minimize language specific features with the aim of creating a universally applicable task (Chiat 2015).

The LITMUS-CL-NWRT was conducted with 100 children divided into four groups (BiliSAT-Project: Chil-la/Hamann DFG CH-1112/4-1). Two groups with monolingual and bilingual typically developing children and correspondingly two groups with DLD. The main research questions were:

Is the LITMUS-CL-NWRT a suitable tool to differentiate between children with and without DLD, with-out disadvantaging bilingual children?

Does the LITMUS-CL-NWRT show better values of diagnostic validity compared to an already well-established German NWRT?

The outcome shows better diagnostic validity for the LITMUS-CL-NWRT compared to the German NWRT. This is in line with the findings of other NWRT studies that crosslinguistically constructed items are more appropriate to test bilingual children on DLD (Ortiz 2021). Further analyses of the data show that the results are also influenced by the language profile of children with DLD. This finding should be kept in mind for future NWRT studies as it may shed light on the role phonological short-term memory plays as an underlying cause for DLD.”

Factors of item presentation influencing children’s performance in nonword repetition tasks

Authors: Kathrin Heeg; Nathalie Frey, Theresa Bloder, Maren Eikerling, Anja Starke, Carina Lüke

“Nonword repetition tasks (NWRT) help to identify children with developmental language disorders (DLD). For these tasks several factors, e.g., structure of nonwords, have been examined to determine their potential impact on children’s performance. However, only a limited amount of discussion has been focusing on the way the items are presented to the children. Given that NWRT often lack a stand-ardized administration, it is relevant to consider how certain presentation aspects might affect the in-terpretation of test results. We therefore investigated the influence of pace (slow vs. fast) and mode (prerecorded vs. life) on children’s performance in a common German NWRT.

For both conditions the test was employed in an AB-BA design. Pace: We presented the children (n = 42, 13% multilingual, M = 6;2 years, SD = 9.09 months) with prerecorded audio items, half of them at a rate of one syllables (“slow”) the other half with two syllables (“fast”) per second. Mode: The items for this group (n = 54 children, 46% multilingual, M = 5;5 years, SD = 6.88 months) were partly presented live by the examiner, while the other half was presented using prerecorded audios.

Our results show that items presented at a faster rate yielded in more nonwords being repeated cor-rectly in comparison to the slower presentation (M = 2.52, t(41) = 8.12, p < .001, d = .833). Likewise, children showed better results for the items that were presented live compared to the prerecorded items (M = -1.741, t(53) = -4.813, p < .001, d = .688).

Disparities in the presentation impact both children’s performance in repeating them and the diagnostic outcomes. In order to attain valid results, it is imperative to work towards establishing consistent in-structions for NWRT administration.”

07/18/2024, from 02:30 PM to 04:30 PM , Room P217

Symposium: Language acquisition in Indigenous North and Central American contexts

Speakers: Henny Yeung; Martina Joe; Pedro Mateo Pedro; Melvatha Chee; Shanley Allen; Titia Benders

Here we explore how children acquire a variety of linguistic levels (phonology, morphology, syntax) in the languages of several Indigenous communities in North and Central America (Canada, US, and Guatemala).

Child pronunciation in a language revitalization context: Evidence from Hul’q’umi’num’

Authors: Martina Joe; Boey Kwan, Elise McClay, Henny Yeung, Sonya Bird

Hul’q’umi’num’ is an Indigenous language spoken on southeastern Vancouver Island (British Columbia, Canada). Through intense grassroots activism, many communities have a remarkable increase in the number of young children learning Hul’q’umi’num’ through early home- and school-based language programs. Hul’q’umi’num’ has a rich consonant inventory, including many place and manner contrasts not used in English, as well as complex consonant clusters, and children most commonly learn Hul’q’umi’num’ from their teachers and parents, who are themselves adult L2 speakers. In our research, we ask at what point children master specific sounds, and how L2 adult input affects children’s pronunciation of these sounds. As community- and university-based linguists, including an adult learner of Hul’q’umi’num’ who is a parent of young children enrolled in this language programming, we have been examining children’s recitations of word and phrase lists that are representative of the language’s sound system. In our talk, we present an overview of a corpus of 173 transcribed words from 8 children, and we track the acquisition of individual consonants based on children’s age and adult input patterns. Results show that (1) the more often children hear a particular sound, the more often they faithfully reproduce it, (2) consonants present in both English and Hul’q’umi’num’ are more easily produced at earlier ages, whereas consonants unique to Hul’q’umi’num’ are produced more accurately at older ages, and (3) some select non-English sounds are nevertheless produced accurately from an early age (e.g., uvular /q/). These patterns are beginning to build a full picture of the developmental pathway of pronunciation acquisition among young Hul’q’umi’num’-learning children. Our research furthers our understanding of phonological acquisition in languages with more complex consonant inventories than those currently well-studied in the literature. Results will also provide benchmarks that can be used by parents, teachers, and clinicians in supporting Hul’q’umi’num’ speaking children.

Acquiring ejectives in L1 and L2 learners of Mayan languages

Authors: Pedro Mateo Pedro;

In this presentation I will provide an overview of child and L2 acquisition of Mayan languages and focus specifically on the acquisition of ejectives and related patterns of sound substitution, with special focus on Q’anjob’al and Chuj. Based on studies of child data from naturalistic settings, which consisted of a child’s interactions with relatives from home and with research assistants, children’s production of ejectives often results in either the production of a plain or the glottal stop. Additionally, when children are acquiring the sound system of a Mayan language, they undergo a process of substitution. For example, instead of producing a retroflex as in the word [ t͡ʂitam] ‘pig’, they would produce an affricate as in [ t͡ʃitam]. I ask here how this data can predict L2 learner performance in a context of language revitalization, and whether this resembles what is reported in first language acquisition. For this section of the talk, I will discuss original data from the revitalization project of Itzaj in Guatemala that involves adults and teenagers who are L2 learners of this language. The prediction would be that L2 learners of Itzaj will undergo a similar pattern of sound reduction and the modified production of ejectives as observed in other studies of L1 child language acquisition in other Mayan languages.

Learning to navigate the Navajo verbal prefix system

Authors: Melvatha Chee;

Navajo has a verb-based lexicon. The polysynthetic Navajo verb construction is morphologically rich, often compared to an entire English predicate. It expresses verbal semantics plus agreement, argument structure, and adverbial information. The Navajo verb construction comprises a concatenation of lexical and inflectional morphemes, all of which are bound. Their individual meanings are dependent on the verb construction as a whole. While the verb structure is notoriously complex in all polysynthetic languages, identifying the specific morphemes in the Navajo verb is particularly difficult due to morphophonological interactions that blur morpheme boundaries. Data for this study was collected from four Navajo speaking children in the form of audio recorded child-caretaker conversations. From this data, 1600 verb tokens, 400 per child ranging in age from 4;07 to 11;02, were analyzed. This talk focuses on the acquisition of agreement and mode making in the verb construction to answer the question how do children learn the elements that compose the base verbal prefix string? The analysis of child-produced Navajo verbs, specifically verbal prefixes, shows that children first focus their attention on adding semantically salient verbal prefixes. Although all verbal prefixes have semantic content, children especially pay attention to those that have a one-to-one form-meaning pair. Verbal prefixes that children pay less attention to and therefore do not use as much are homophonous prefixes. Homophonous prefixes present a challenge because they cannot be easily distinguished due to a single form containing numerous meanings. The meaning of homophonous prefixes tends to come out when they are used with neighboring prefixes, usually resulting in syllables. Navajo children must learn to navigate this verbal prefix system to begin to produce adult-like verb constructions. Findings will contribute to the study of child language development and to the revitalization of Navajo where a shift to English is prevalent.

Early development of morphological and syntactic structure in Inuktitut

Authors: Shanley Allen; Hannah Lee

Inuktitut (Inuit) is spoken by some 100,000 speakers in eastern Canada. It is an agglutinative polysynthetic language which allows up to 10 morphemes per word. Thus, the challenge of language learning is very different from that in English because much of syntax occurs within the word rather than across words. Studies with eight children aged 1;4-3;4 show a gradual increase of mean length of utterance in morphemes with age (from MLUm 1.18 to MLUm 3.25). However, most of this growth comes from more morphemes per word rather than from more words per utterance. Verb and noun roots at the earliest ages typically appear with no inflection, but verbal and nominal inflections appear in most obligatory contexts by age 2. By 2;10 almost a fifth of words contain four or more morphemes, but even by 3;4 no utterances are longer than three words and most utterances contain only one word. Children show early use of morphology with syntactic functions including passives, causatives, and complement-taking verbs. They also show increasing mastery of polysynthesis including devices that change word class from noun to verb or verb to noun up to four times within a given word. As is typical in the target language, argument omission is very common and few if any utterances are produced using both a subject and an object. In multi-word utterances, they are initially strict in using the default SV and OV structures, and only start to vary word order for pragmatic purposes after age 3. These findings expand on data from a new “acquisition sketch” of Inuktitut following the recently published Sketch Acquisition Manual – a set of guidelines to facilitate acquisition research in understudied languages, which we will also briefly introduce.

07/18/2024, from 02:30 PM to 04:30 PM , Room P018

Symposium: Evolving professional roles of SLPs to support students with language and literacy needs in school

Speakers: Marie-Catherine St-Pierre; Wenonah Campbell; Pamela Filiatrault-Veilleux; Edith EL KOUBA; Victoria Joffe; Chantal Desmarais

SLP professional roles and school-based practices are expanding to contribute to a more inclusive education for all students. Research findings related to this paradigm shift for students with oral and written communication needs will be shared.

Use of universal design for learning in tiered school-based speech language pathology practice

Authors: Wenonah Campbell;

“Tiered service delivery models create equitable access to school-based speech language services. Instead of focusing solely on supports for individual children, tiered models target the learning needs and well-being of all students. Services typically consist of three tiers, with the first Tier, Universal Services, offering a promising strategy for promoting equity and inclusion, through an emphasis on Universal Design for Learning (UDL).

UDL is an educational framework to support the full participation of students with diverse abilities in education settings by removing unnecessary barriers to learning. It guides one type of universal service provided within a tiered service delivery model and involves speech language pathologists collaborating with educators to provide multiple means of engaging children, representing information, and expressing learning. Use of UDL principles leads to the development of educationally relevant and accessible strategies, tools, and environments for all.

This presentation will define UDL and describe this framework; share evidence on the roles of speech language pathologists in implementing UDL, including to support language and literacy; and identify specific UDL strategies and tools that speech language pathologists can use when collaborating with educators. Resources from a free health professional development program, FIRST, available at, first.machealth.ca also will be briefly introduced to enable further learning about tiered models, universal services, and UDL.

Given the increased use of tiered service models in education settings, speech language pathologists need to be aware of UDL and its role in the provision of universal services. This session will support speech-language pathologists to better advocate for their role in implementing UDL and collaborating with educators to support the participation of all students, including students with oral and written communication disorders.”

School-based SLPs’ perceptions on their involvement in Tier 1 interactive reading intervention

Authors: Pamela Filiatrault-Veilleux; Paméla McMahon-Morin (Université de Montréal); Claire Croteau (Université de Montréal); Wenonah Campbell (McMaster University)

Background: With the advent of tiered models of service delivery, school-based SLPs’ roles are evolving in ways that promote inclusive education. Of note, interactive reading programs are designed to support language and literacy skills of children with diverse profiles in inclusive settings. Nevertheless, SLPs identified barriers to implementing such practices, including a lack of collaboration time and an increased workload. To our knowledge, no study has yet examined the benefits and levers for change perceived by SLPs from their involvement in Tier 1 interactive reading intervention in francophone contexts. Objective: This qualitative study examines school-based SLPs’ perceptions regarding their involvement in Tier 1 interactive reading intervention to determine the perceived benefits and levers for change in their practice in two French-language educational contexts in Canada. Methods: Seven SLPs who had received training to deliver an interactive reading intervention program in kindergarten classrooms from either minority or majority francophone context were recruited. Individual semi-structured interviews were carried out followed by two focus group meetings, within which the SLPs from both contexts were distributed. A thematic analysis is underway. Results: Preliminary themes identified as levers for change include: (1) the time allocated for planning and implementing the program within a coordinated collaborative service delivery model; (2) the adaptability of the program according to the linguistic, cultural and socio-demographic context of each classroom; and (3) the importance of the teacher-SLP partnership. The SLPs reported substantial benefits to their own practice, including a broader impact within their school district resulting in a better understanding of their roles and providing earlier support to more children with various support needs including those with language and communication disorders. Conclusions. Study findings offer school community members several insights about how to facilitate SLPs’ involvement and crucial contribution in implementing inclusive educational practices.

Supporting language in preschoolers: using SOLEM, a collaborative approach between SLPs and teachers

Authors: Edith EL KOUBA; Valentine Levaux (Université de Liège); Christelle Maillart (Université de Liège)

Objective: The present study aims to determine the effects of a preventative, preschool-based child language intervention program, SOLEM (Soutenir et Observer le Language en Maternelle), on the use of language support strategies by preschool teachers. SOLEM features a collaborative coaching model involving Speech and Language Pathologists (SLPs) alongside preschool teachers, in French-speaking Belgian preschools. Methods: A single-case experimental design with repeated measures was applied within tailored on-site interventions. Specific language support strategies were targeted after being selected based on the contextual needs of each teacher and her classroom. Teachers’ use of language practices was documented through video sequences collected before, during and after the intervention. Videos were analyzed using a measurement grid that allowed for the identification of language practices that have been documented in the literature. Results: Based on the repeated measures, the results indicated a significant decrease in directive language strategies. They also revealed a significant increase in the use of high-level decontextualized language within classroom activities. The data also allowed for the identification of the co-occurrence of different language support practices within the same interaction sequence. The use of strategies extended to different classroom activities (rituals, storytelling, free play, etc.). In particular, the findings highlight the variability in teacher’s proficiency in using targeted language practices, which reinforces the importance of tailoring language support goals based on the contextual needs of the teacher and her classroom. Clinical implications: Implementing a collaborative program to support language development in preschoolers may have positive effects on teachers’ use of positive language practices. The study findings support the importance of embedding prevention programs in real-world contexts, through a collaborative coaching process that actively involves the speech and language pathologists and teachers.

Exploring Collaborative Practice working with Speech, Language and Communication Needs in Schools

Authors: Victoria Joffe;

“Educational services for children and young people with Speech, Language and Communication Needs (SLCN) in the UK, and in other parts of the world are provided at three levels: universal, targeted, and specialist, with universal provision available to all children in the classroom through whole class teaching, targeted to those at risk for or with mild to moderate SLCN, delivered individually or in small groups, and specialist support focusing on the students with the most severe and complex difficulties, typically delivered by the Speech and Language Therapist (SLT). Frequently, the contribution from the SLT will be at its maximum when working at the specialist level of provision, but it can be argued that best practice in the management of children and young people with SLCN in schools is embedded in partnership working, incorporating input from teaching and support staff and the SLT across all three levels of provision.

This presentation provides an overview of the evidence base in working with children and young people with SLCN across the three levels of service provision, targeting secondary school aged students and vocabulary and narrative outcomes. The nature, type, and duration of the intervention at each level is explored, as well as the description and severity of the language disorder. The mechanisms underlying the delivery of the intervention in the education context, drawing on the expertise of the education and speech and language therapy staff is explored, and the best ingredients for effective collaboration are identified. Views from service users, teaching and support staff and SLTs are gathered to inform how best to meet the needs of children and young people with SLCN in schools. Facilitators and obstacles to collaborative practice are discussed and a framework for enhancing language and communication for all students in schools, through partnership working, is shared.”

Speech language pathologists supporting teachers in special education classrooms in high school.

Authors: Chantal Desmarais;

Background: A recent codesign and implementation study led to the publication of the intervention programme ESCALADE aimed at young adolescents with Developmental language disorder (DLD). ESCALADE has three main goals, (i) improve self-knowledge including characteristics of one’s communication abilities and challenges, (ii) increase autonomy in selecting and using communication strategies, and (iii) support the use of these communication strategies to improve social interaction. The 17 activities of the program were implemented over the course of one school year. This was preceded by training to teachers to reach a common understanding of DLD and of the needs of adolescents with DLD. Teachers were supported via Community of practice (CoP) meetings led by a speech-language pathologist (SLP). The aim of this study is to examine the strategies used by the SLP in this support role. Methods: 19 teachers or school personnel participated in the CoP meetings that were audio recorded and transcribed. A qualitative thematic analysis procedure was used to highlight categories of support strategies used by the SLP. Results: Four main support categories were used by the SLP to enrich the implementation of ESCALADE: (i) seeking the opinions of teachers, (ii) verifying their needs, (iii) sharing expert knowledge about DLD, (iv) problem-solving with teachers to find solutions to problems for implementing activities aimed at improving the youth’s communication. When surveyed, teachers confirmed the positive effect of the SLP support in putting in place strategies for language and literacy. Conclusion: Results suggest that SLPs can play a positive role, even if it is indirect, in supporting adolescents with DLD in high school and that this role should be encouraged. This will be discussed in view of recommendations on knowledge brokering and interdisciplinary support that stem from recent research.

07/18/2024, from 02:30 PM to 04:30 PM , Room P301

Symposium: Orchestrating Action, Gesture, Speech and Sign in Family dinners

Speakers: Aliyah Morgenstern; Christophe Parisse; Sophie de Pontonx; Corrado Bellifemine; Alice Brunet; Lourdes de León; Gary Morgan

This panel highlights the semiotic differences between adults and children using one or several spoken languages, French, English, Russian, and a sign language, LSF as well as the orchestration of language and dining practices.

Interlanguaging and dining in signing and speaking family meals

Authors: Christophe Parisse; Stéphanie Caët, Léa Chevrefils, Aliyah Morgenstern

“In family dinners, language practices can be analyzed as they occur in real life and real time in the framework of multiparty interactions and multiactivity (Haddington et al. 2014) to capture the multiple deployments of the embodied behaviors of speakers and signers.

In this study, we analyzed four French signing and four French speaking family’s coordination of interlanguaging and eating around the dinner table. The families composed of two parents and two to three children aged 3 to 10 were filmed with three cameras (including a 360° camera) to capture all family members’ behaviors. The three videos per dinner were synchronized and coded on ELAN. We annotated all participants’ acting, languaging, and gaze. We focused on adults’ expertise and on children’s socialization to the finely-tuned coordination and in situ organization of the joint activities of conversing and dining that fully engage the same body components according to the language used and their age.

Our quantitative analyses show how family members collaboratively manage multiple streams of activity through the embodied performances of dining and interacting. The hearing participants use the affordances of the visual and vocal channels to maintain the simultaneity of the two activities. The deaf participants also skillfully manage to alternate smoothly between dining and interacting, but within a single channel and differently than the hearing participants. Deaf and hearing younger children manifest how they are in the process of developing these skills and manage multi-activity as well as adults, with some differences between the developmental path of the deaf and hearing children. Our qualitative analyses focus on the ecology of visual-gestural (LSF) and audio-vocal-visuo-gestural (French) languaging in the context of co-activity according to language and participant. We open new perspectives on the management of action, gesture, gaze, speech and sign in multimodal interlanguaging in multiparty interaction.”

Building participation frameworks: the role of gaze in signing and speaking family dinners.

Authors: Sophie de Pontonx; Pauline Beaupoil-Hourdel, Diane Bedoin, Loulou Kosmala, Aliyah Morgenstern, Claire Danet, Stéphanie Caët.

“Goffman’s critique of the speaker-hearer dyadic model (1974) and cross-cultural studies suggest that children experience and learn language not just in dyadic but also in situated multiparty interactions (Ochs and Schieffelin, 1984). Family dinners involve several activities, among which eating and languaging are central. In this multi-activity setting (Haddington et al., 2014), children and adults play different participant roles (Goffman, 1981). Children can be integrated in collaborative talk and broaden their experience of various interactional practices. They are socialized to conduct multiparty interactions and are dynamically involved in various participation frameworks. This study focuses on the role and management of gaze and how it may indicate (dis)engagement throughout the conversation in coordination with dining activities. We compared signing and speaking family members according to their language, status and age.

We analyzed dinners in four families with two adults and two children, two using French and two using French Sign Language (LSF). The data was systematically coded for 1) audible or visible languaging or acting produced by each participants, 2) which participants the acting or languaging was directed at, 3) participants’ gaze orientation towards people or objects. Quantitative results from our coding are complemented with qualitative analyses of chosen extracts to provide a fuller picture of the identified features of family dinners.

Our analyses underline differences in gaze and turn distribution patterns according to the varying participation roles, age, and language. Both speaking and signing children in our data develop the skill to participate in the interlanguaging while dining and are socialized to the specific crucial functions of gaze according to the affordances of the language they use. Adults adapt to and scaffold their children’s management of gaze in order to orient their visual attention when and where it is most needed during the dining and languaging activities of family dinners.”

Focus and emphasis marking through multimodal prosodic cues in spoken and signed family discourse

Authors: Corrado Bellifemine; Christelle Dodane, Karine Martel, Fanny Catteau, Marion Blondel

“Prosody can be studied with a multimodal approach that integrates vocal and gestural cues in spoken and sign languages (Ferré, 2014; Lombart, 2022). Focalization is an important discursive feature that can be expressed, among other semiotic means, by prosodic cues. It is defined as the process of emphasizing certain elements of speech (syllables, words, phrases) within an utterance, in order to draw the interlocutor’s attention (Groussier & Rivière, 1996 : 84).

The goal of this presentation is to compare how speaking and signing family members establish focalization through multimodal prosody. Thus, we compared dinner data collected in two French-speaking and two LSF-signing families. We centered our study on the way adults and children emphasized elements of their utterances through different multimodal prosodic markers based on the type of language: pitch accent, intonation, sign amplitude, hand-gestures, gaze, head movements. These indissociable facets of prosody were integrated in our analyses of the pragmatic value and informational status of the focalized elements (information and comments, questions and answers, directives…) based on the participants’ dialogic moves. We coded gestures, signs and the informational status of the emphasized element in ELAN, and we made acoustic measurements using Praat to analyze the vocal cues associated with focalization.

Results show that, regardless of their profile or of their role in the family (adult vs child), both signing and speaking family members used different bodily components (especially hand-gestures and head movements) and specific prosodic patterns (i.e. scansion) to convey focalization. At the pragmatic level, comparisons between participants reveal that focalization was mostly associated with informative and directive moves in adults, and answers and comments in children.

Prosody can thus be considered as a multidimensional process that allows participants to establish joint attention, express meaning adequately, but also take the floor and manage the dynamics of turn-taking in multiparty interaction.”

The role of gesture in multimodal bilingual family dinners

Authors: Alice Brunet; Yana Kut, Aliyah Morgenstern

“In multilingual environments where languages meet and intermingle, it is challenging for children to carve out their own bilingual interactional space. Often shared cross-culturally, gestures might represent more stable forms since words may vary in each code for the same function. The study of bilingual first language acquisition allows the researcher to investigate whether asymmetrical bilingual development impacts the quantity and functions of gesture production on the communication flow. Multimodal and multilingual communication could reinforce bilingual children’s language socialization into their bicultural community.

To study bilingual children’s use of symbolic gestures as their multimodal skills blossom, we collected video-recorded data from 4 bilingual families (English/French and English/Russian). We transcribed multilingually and coded gesture production both when the children used their dominant and non-dominant language.

We show that gestures play an important role in bilingual children’s management of communication. Our results suggest that children do not produce more gestures when using their non-dominant language, but rather that they used more gestures during instances of multilanguaging. This would mean that gestures are not used primarily to compensate for access to more complex verbalizations in their weak language. Detailed contextual analyses of the children’s productions indicate that gestures produced by bilingual children in multilingual interactions with adults pave the way for the interlocutors’ access to meaning in each language and in their own weaker language; multimodality and multilingualism interact constantly in the meaning-making process.

Bilingual children seem to be particularly skilled at resorting to gesture as well as multilanguaging to enhance inter-fluency via all the semiotic resources at their disposal. In their search for the best way to package their message, the bilingual children in our dataset create successive transitory multimodal and multilingual systems.”

07/18/2024, from 02:30 PM to 04:30 PM , Room P104

Symposium: Mandarin-speaking children’s language acquisition and literacy experience in early childhood

Speakers: Cs Cs; Kanyu Yeh; Wun Tsong Chaou; Pao chuan Torng

The papers in this symposium examine Mandarin-speaking children’s linguistic abilities and literacy experiences in early childhood, addressing crucial issues on gesture-speech combinations, referential choice, and shared-book reading interactions.

Gesture-speech combinations and language development of Mandarin-speaking young children

Authors: Kanyu Yeh; Chiung-chih Huang Affiliation: National Chengchi University

“Children’s gestures provide them with a tool to extend their communicative repertoire to express meaning they cannot yet express through speech, and reveal their current cognitive ability and readiness to learn (Goldin-Meadow, 2007). Their gesture-speech combination may convey similar information about an object (e.g., pointing to a dog + “dog”) or supplement each other to convey sentence-like meaning (e.g., pointing to a dog + “sleeping” to mean “the dog is sleeping”). Children’s early uses of various types of gesture-speech combinations were related to different aspects of their later linguistic development, such as lexical and syntactic abilities (e.g., Fasolo & D’Odorico, 2012). However, earlier research focused mainly on children from Western cultures, less is known about the role of gesture in Mandarin-speaking children’s development.

The present study investigated Mandarin-speaking children’s development of multimodal communicative ability during early childhood. Longitudinal data of two Mandarin-speaking children’s daily conversations with their mothers were collected at ages 1;9, 2;0, and 2;3. The children’s gesture-speech combinations were coded for gesture types (deictic gestures, emblems, iconic gestures, and other gestures) and utterance types (vocal, single argument, multiple arguments), as well as the relationship between gesture and speech (complementary, supplementary) in each cross-modal combination. The results showed that the children frequently combined deictic gestures and emblems with speech to co-express meaning. Utterances with multiple arguments increased in the children’s production during their development, while utterances with single argument remained the majority type in their cross-modal combinations. The children’s gestures complemented speech at the younger age, and gradually provided supplementary information for their speech. Qualitative analysis further revealed increasing sophistication in the children’s multimodal communicative ability. This study contributed to a limited number of studies on Mandarin-speaking children’s communicative ability in speech and gesture and provided cross-linguistic evidence for children’s multimodal language development.”

Referential choice in oral narratives of Mandarin-speaking children: A developmental study

Authors: Cs Cs;

“A speaker’s use of referring expressions is associated with the degree to which a referent is linguistically retrievable or cognitively accessible to the listener. For instance, indefinite noun phrases (NPs) are frequently used for introducing new referents, whereas definite NPs are opted for subsequent mentions of referents. Unlike languages such as English which differentiates given from new referents in terms of definite and indefinite articles, Mandarin Chinese has no fully grammaticalized definite article, and usually conveys the given-new distinction through a combination of nouns with demonstratives, numerals, or classifiers. Given this, this study examined how Mandarin-speaking children use referring expressions to mark given vs new referents in oral narratives.

Thirty Mandarin-speaking 5-year-olds, thirty 9-year-olds, and thirty adults participated in this study. The narrative data were based on Frog, where are you? Arguments were coded in terms of referential form, function, and adequacy. Linguistic forms were classified into three major categories: nominals, pronominals, and null form; nominals were further divided into bare nouns, indefinite NPs (yī běn shū ‘a book’), and definite NPs (zhè běn shū ‘this book’). Three referential functions were differentiated: (1) introduction, (2) maintenance, or (3) reintroduction of a story character.

The results showed that children and adults primarily used nominals to introduce and reintroduce referents, whereas null forms were most preferred for reference maintenance. Regarding nominal subtypes, 9-year-olds and adults reserved indefinite NPs for introducing referents and definite NPs for subsequent mentions; 5-year-olds, however, chose between bare nouns and indefinite NPs for introducing referents, and opted for bare nouns for subsequent mentions, suggesting that 5-year-olds were still developing their skills in choosing appropriate nominal subtypes to mark given vs new referents. Additionally, our data showed less adequate referencing in children as compared with adults, and higher adequacy level for introduction than the other functions across age groups.”

Shared book reading with Mandarin-speaking preschool children from rural areas in Taiwan

Authors: Wun Tsong Chaou; Yu-Hsuan Cheng & Chi-Chi Yang, Puli Christian Hospital; Pao chuan Torng, National Taipei University of Nursing and Health Sciences

“Previous research has suggested that shared book reading (SBR) be a promising intervention approach for promoting and facilitating child language development. However, we still lack knowledge about the feasibility and effectiveness of SBR for young children from rural areas with limited resources. Therefore, the focus of this study was on SBR experiences and emergent literacy of Mandarin-speaking preschool children from rural areas in Taiwan.

195 preschool children from 10 kindergartens across Nantou, a mountainous county in central Taiwan, were recruited to take part in a nine-month SBR program, with assistance of speech-language pathologists. A questionnaire was developed to examine the quantity and quality of reading interactions at kindergartens and at homes, such as children’s motivations and interests in reading, reading habits parents have with their children, family time spent with reading per week, reading resources, demographic factors, and self-evaluation through the nine-month program. At the end of the program, children and their parents were asked to fill out the questionnaire.

Our results showed that all children were actively engaged during SBR interactions and a majority of the parents were aware of the benefits of SBR on facilitating children’s language skills, such as enhancing print and word awareness, and improving verbal fluency. Parents were positive about SBR interactions at homes, but reading resources were extremely lacking in the rural areas. The impact of SBR on young children’s emergent literacy, and difficulties of using SBR with rural children were analyzed. With the inclusion of equitable quality education for all in the 2030 Sustainable Development Goals, speech pathologists as well as early educators need to provide more supports for families of rural children. The paper concluded with suggestions for future research on rural children’s emergent literacy and discussion of implications for clinical and educational practices.”

The application of dialogic reading during shared book reading activity at home

Authors: Pao chuan Torng; Ching-Hsien Chang, National Taipei University of Nursing and Health Sciences

“Past studies in western societies have reported positive gains of using dialogic reading (DR), the interactive reading approach (Whitehurst et al., 1988), such as enriching emerging literacy skills and stimulating early language development. In Taiwan, though researchers have recognized the important of early reading experiences and tried to promote involving DR in shared book reading (SBR), research on the feasibility of this strategy with Mandarin-speaking children is sparse. This study aimed to investigate the application of DR during parent-child SBR in Chinese families.

This study focused on the implementation of DR with Mandarin-speaking preschoolers aged 3-5 years. Ten parent-child dyads were recruited to participate in a SBR program over 20 weeks. The researchers led the SBR sessions once a month, with each session lasting one and a half hours, training parents to read using the DR strategy. The parents were taught to use the PEER sequence, namely to prompt, evaluate, expand, and repeat a child’s responses, and were demonstrated how to begin the PEER sequence with the CROWD techniques including completion, recall, open-ended, wh-question, and distancing. At the end of each SBR session, parents were asked to fill out a questionnaire regarding SBR frequency at home, gains/problems with the DR techniques, attitudes towards DR, number and types of books read, and evaluation of the child’s performance.

Results showed that all participants were actively engaged in SRB activities led by the researchers, and that the parents did try to employ DR when doing SBR at home. Generally, though parents encountered difficulties in implementing PEER/CROWD, their abilities in using the techniques gradually improved after the monthly SBR session instructed and guided by the researchers. Findings on the effects and difficulties of implementing the DR strategy in Chinese families are discussed. This study has implications for future research and clinical practices.”

07/18/2024, from 02:30 PM to 04:30 PM , Room P200

Symposium: Children’s Complex Language and Theory of Mind Development

Speakers: Jill de Villiers; Stephanie Durrleman; Isabelle Charnavel; Hristo Kyuchukov; Jenna Croteau; Silke Brandt

The relationship of language acquisition to later Theory of Mind (ToM), such as false belief reasoning, remains debated: are the effects of language general or specific? If specific, which aspects matter?

Complements and FB reasoning: an ecological training study in children with and without autism

Authors: Stephanie Durrleman; Stéphanie Durrleman

“Sentential complement structures play a role in children’s explicit reasoning about false beliefs, and training complements enhances false belief (FB) reasoning in typically developing (TD) children and those with autism and delays in language and ToM. However, some studies suggest that acquiring relative clauses may also help with FB reasoning. A training study with a crossover design tested whether complements showed an advantage over relative clauses, with parents reading specially designed children’s books targeting syntax. The books each included 12 examples of the critical structure in a natural story format, and were closely matched in complexity.

Parents of 22 French-speaking children, 12 TD and 10 with autism, were recruited for an eight-week dyadic reading program at home. Children had to be above the 3-year-old level on standardized language tests, but to fail pretests of complements, relative clauses and FB reasoning. Two experimental groups received a different order of exposure to the target (complement) versus control (relative clause) books. Children were tested again with different variants of the pretests at the intermediate step when they changed books, and at the posttest. Parents reported their reading frequency, which varied widely across participants but did not differ across books.

The target (complement) book had a specific impact, regardless of order (Wilcoxon tests: p<.005), not only on complement understanding but on FB reasoning (Wilcoxon tests: p<.02). The training effects of the complement book did not statistically differ for TD and ASD children. There was no effect of the control book.

The results confirm the impact of training sentential complementation to enhance FB reasoning, which was not true for relative clauses. The effect occurred for children with and without autism, but for the first time we show that these benefits can be obtained via an ecological training of this structure, by parents at home.”

Taking Perspective in Motives and Purposes: The Case of Adjunct Clauses and Theory of Mind

Authors: Isabelle Charnavel; Stéphanie Durrleman, Jill de Villiers, Amber Liu

“The paper aims at investigating how the acquisition of adjunct clauses (such as causal – because – or purpose – in order to/so that – clauses) relates to the acquisition of Theory of Mind (ToM). Previous studies examining the relation between syntax and ToM usually focus on complement clauses (such as think that P), which are closely linked to the understanding of false belief. Adjunct clauses, however, have been neglected in this literature although they can involve an interaction of points of view which should be uniquely informative about ToM. For example, the interpretation of the causal relation in the sentence Dora left the party because the cupcakes were yucky requires adopting Dora’s point of view (like the interpretation of attitude clauses) without involving false belief (unlike the interpretation of attitude clauses).

In our study, we investigated whether the mastery of various types of adjunct clauses (causal/purpose, perspectival/non-perspectival) correlates with ToM performances. In an online experiment, (N=38) 3-8-year-old children’s understanding was tested using a set of nine specially designed stories, after each of which Memoji puppets attempted a summary using adjunct clauses (4 per story). The child had to judge whether the puppet retained the truth expressed in the story as they rephrased it. In the same session, the child’s ToM understanding was assessed based on classic false belief tasks (8). The results reveal a clear correlation between ToM and the understanding of perspectival adjunct clauses. The strong correlation (r (34) = .51, p<002) remains when controlling for the child’s chronological age and their performance judging the truth of purely factual adjunct clauses. Therefore, the child’s performance on false belief tasks is uniquely connected to their understanding of adjuncts entailing points of view, opening new avenues of exploration in the relationship between complex language and ToM.”

Retelling Narratives Entailing Second Order False Belief in Lyuli and Uzbek Bilingual Children.

Authors: Hristo Kyuchukov; Jill de Villiers

“The first known study of the socio-cognitive development of Lyuli children, a Roma-type group living in Bukhara in Uzbekistan was conducted in schools in Bukhara serving both Lyuli children and Uzbek children. Each group is multi-lingual in a different variety of Tajik (L1) with Uzbek (L2), but their sociolinguistic circumstances and educational preparation are different. There has been less cross-cultural work on later stages of Theory of Mind development, in which children have to make inferences about the mental states of characters in a complex narrative.

Two well-known second-order narratives were used with small cultural adaptations: the bake sale and the birthday story. The children from both groups (N=38, ages 7-8 years) did remarkably well on these tasks, and their multilingualism is hypothesized to be a possible source of their success relative to other children studied with similar narratives across cultures. The current report analyzes the children’s attempts to retell the second order false belief story told/questioned in one language (e.g., Tajik), into their other language (e.g., Uzbek) and vice versa. Do the children successfully capture the essential points of the story, and is that a function of how well they did on answering the belief questions?

The two groups were alike in two major respects. They both did better at retelling the story into their second language that they first heard in their native Tajik, rather than the reverse. Furthermore, both groups retained essential points of the complex stories very well (average overall 4.36 on a 5-point scale). The Lyuli children did better on retelling the more they understood the story in their native language, whereas the Uzbek children showed a correlation across retellings in their two languages. The small differences between the groups in story quality (p<.04), are attributed to the extra kindergarten year the Uzbek children received.”

When and How do Children Understand Referential Opacity?

Authors: Jenna Croteau; Jill de Villiers

“In our study children (N=79, ages 3-7) answered questions about characters with dual designations (e.g., Uncle Albert/an astronaut) concerning protagonists who either knew (ProtKNOW) or did not know (ProtIGNORANT) the character under both designations. Philosophy of language has long recognized that there are restrictions on substituting terms within the scope of propositional attitude verbs, i.e., “referential opacity”. But only intensional readings of NPs are restricted, not referential/extensional NPs. In developmental psychology, the focus has been on when children recognize that intensional readings are conditioned by the knowledge state of the agent. Rackoczy claims that referential opacity is mastered with 1st order false belief, but the Mental Files approach (Perner) argues for a protracted development.

In this study children could easily answer questions like:

a) “Does ProtIGNORANT know that Uncle Albert is an astronaut?”

yet struggled until age 7 with questions like:

b) “Does ProtIGNORANT know he took an astronaut to his room?”

We also found that false belief understanding correlated with (a) but not (b). We propose that “an astronaut” in an identity predicate like (a) is easily understood as intensional and thus performance reflects ToM. However, in other predicates like (b) it is unclear whether “an astronaut” is meant extensionally or intensionally without sophisticated comprehension of discourse dynamics. Both speaker and hearer have established a common ground for the dual identity, so is the question about the ProtIGNORANT‘s knowledge of the designation or of the event? Definite and indefinite articles in English do not uniquely constrain the status of the NP as extensional or not. We hypothesize that children initially default to a de re reading in which they treat “an astronaut” as extensional, giving the truth of the event priority. The phenomenon of referential opacity is thus best considered to entail semantic/pragmatic developments, not just conceptual change.”

07/18/2024, from 02:30 PM to 04:30 PM , Room P131

Symposium: Acquisition of complex predicates in diverse languages

Speakers: Jekaterina Mažara; Dorjderem Byambasuren; Alice Johnson; Dagmar Jung; Sabine Stoll

What challenges do complex predicates pose for L1 learners and what and how can we compare them cross-linguistically? The talks cover a broad range of language families and topics surrounding complex predicates like polysynthetic verbs and converbs.

Acquisition of converbs in Mongolian-speaking rural and urban children

Authors: Dorjderem Byambasuren; Shanley Allen

“Converbs are a complex morphosyntactic structure used for clause chaining in many languages but are not well-understood in language acquisition in general. Mongolian is one language in which converbs are prevalent. In Mongolian, converbs can be used to chain either coordinate or subordinate clauses together – usually two but potentially more. Additionally, subordinating converbs often appear with participles to express temporal relations. The last verb in the chain is the finite verb, carrying tense and other information.

Previous studies on converbs have mainly focused on analyzing these constructions in individual languages from a theoretical perspective. Unfortunately, little is known about how children’s converbs are acquired cross-linguistically and empirically. The present study investigates how Mongolian-speaking children use converbs in rural and urban communities, whether the use of converbs differs across the communities, and whether children’s age correlates with converb use within each community. A total of 56 typically-developing monolingual children (36 rural and 20 urban) participated in this study. To facilitate comparison across the children, elicited narrative data was collected using the “frog story” (Meyer, 1969).

We found that rural children outperformed urban children on the frequency of converbs per utterance and diversity of types of converbs. Further, we found a correlation between age and converbs per utterance in the urban group, but not in the rural group. Finally, children in both groups predominantly focused on using coordinating converbs rather than subordinating converbs. These findings imply that the urban and rural children showed a considerable difference in how they use converbs, and thus the developmental trajectory of converbs for Mongolian-speaking children may vary depending on their immediate socio-cultural context. These results contribute to expanding the diversity of the field of language acquisition and provide insights not only into Mongolian acquisition but also into the acquisition of this structure in agglutinative languages.”

Morphological simplification of complex structures in Inuktitut child-directed speech

Authors: Olga A. Johnson; Shanley Allen

“Caregivers typically use a simplified mode of the language – child-directed speech (CDS) – when addressing young children. The most studied domains of CDS are lexicon, phonetics, and prosody (Snow, 1995), while the morphological aspects of this speech mode have received much less attention. When it comes to the morphological properties of CDS, polysynthetic languages constitute a great source for investigation. In this study, we investigate the use of complex morphological structures with a word class change within a single word in Inuktitut CDS. Inuktitut is a polysynthetic agglutinative language of the Inuit-Yupik-Unangan language family spoken in arctic Quebec, which allows more than ten morphemes per word and in which the meaning of an entire sentence can be expressed in one word (1).

(1) Illujaraalummuulaursimannginamalittauq.

  illu-juaq-aluk-mut-uq-lauq-sima-nngit-gama-li-ttauq

  house-big-EMPH-ALL.SG-go-PAST-PERF-NEG-CSV.1sS-but-also

  “But also, because I never went to the really big house.” (Dorais, 2011)

Clearly, such a complex morphological system presents special challenges for young children, which raises a question of whether caregivers shape their CDS in ways that facilitate acquisition. Using the data from mothers addressing eight Inuktitut-speaking children aged 0;11 to 3;6, we investigated whether the frequency and complexity of polysynthetic structures in CDS are dependent on the stage of the children’s linguistic development. The results demonstrate that the number and morphological complexity of the structures with a word class change increased as the children developed linguistically. The variety of nominalizers and verbalizers – the key components of such structures – also increased through the stages and were used in variation sets, which help children acquire morphological items by providing examples of use of the same morpheme in morphologically contrasting environments. These results show the presence of morphological simplification in Inuktitut CDS and demonstrate that such simplification is fine-tuned, i.e., that mothers are sensitive to their children’s level of linguistic development.”

Mixed verbs in Dene language acquisition

Authors: Dagmar Jung; Olga Lovick, Allison Lemaigre, Jekaterina Mazara

“Dene Sųłıné (Athabaskan) is the primary language spoken in Clearwater River and La Loche, Saskatchewan. Children grow up in a bilingual setting with English dominating social media, children’s programming, Youtube, etc. Speakers older than 15 usually prefer speaking Dene. In this study, we examine the use of English-Dene mixed verb forms (uninflected English stem plus inflected Dene light verb) in the input and production of children aged 2;0-5;0 in a longitudinal naturalistic corpus.

Dene languages challenge learners with their long, morphophonologically opaque prefix strings, combined with intransparent verb stems that inflect for aspect. In child-directed speech (CDS) Dene forms alternate with mixed verbs in variation sets (Küntay&Slobin 1996). The most common mixed construction in CDS involves the Dene verb ‘do, make’.

(1) wë́ k’a sënǫ́-dhër

            over.there 	by	play.2SG-play.IPF

             wë́	        k’a 	play 	hán-lë

            over.there by play 	thus.2SG.IPV-do.IPV

           ’Play over there’

Some constructions with English stems and Dene light verbs are conventionalized, e.g. ‘I call/phone’ phone hë́sle, or ‘you send her a facebook message’ facebook message hę̈́leo, and used by speakers of all ages. Others represent alternating choices; adults rarely switch to fully English utterances, but can replace Dene verbs by complex English-Dene mixed forms. In this study, we analyze the most frequent mixed predicate constructions in CDS and children’s use. We compare the developmental stages as well as the amount of mixing. Children start to use this type of complex predicate from the age of around 3;6, building on simpler uses of ‘to do’. After generalizing the use for the expression of most transitive predicates some verb types are switched to Dene-only again (play hánlë -> sënǫ́dhër ’you-sg. play’). We show the developmental trajectory of these predicates and its correlation with the acquisition of the Dene verb prefix system.”

Disentangling statistical properties that drive morphological acquisition in diverse languages

Authors: Sabine Stoll; Giuachin Kreiliger, Jekaterina Mazara, Balthasar Bickel

“Children’s ability to leverage statistical distributions to segment and learn linguistic elements has been explored in a large body of literature. However, we know little about which specific statistical properties shape the process of acquisition in real-life data. Here, we assess the effect of five factors on the acquisition of verbs in a sample of morphologically diverse languages. We analyze predicates produced in the input as well as during morphological development of 20 children in naturalistic longitudinal corpora of five languages: Chintang (Sino-Tibetan); English (Indo-European); Japanese (Japonic); Turkish (Turkic); and Yucatec (Mayan).

To evaluate children’s development, we computed the entropy of their use of verb stems and morphology and compared it to the measures in surrounding adult speakers within the same recording context. This allows us to examine the development in children’s flexibility of predicate use. We ran a hierarchical nonlinear distributional model to fit the development, using the exponential function to describe the learning curves to evaluate the speed of acquisition. We fitted models with the following five factors: i) difference between heads (main stems) and dependents; ii) language; iii) log-number of heads and dependents; iv) deviation of the empirical input distributions from the theoretical Zipf distribution; v) entropy of heads and dependents.

Model comparison shows that convex deviations from the theoretical Zipfian distribution, i.e. more higher-frequency types, predicts the development best. The number of items has a weaker but still appreciable effect. All other factors explain the data less well, i.e., differences between languages, item types, and system entropies have no impact on the acquisition dynamic. This suggests that the acquisition process is shaped by the availability of a particularly large number of high-frequency items as well as by the number of elements that must be acquired as part of the system.”

07/18/2024, from 02:30 PM to 04:30 PM , Room P300

Symposium: Influences on young bilinguals’ lexical trajectories across languages

Speakers: Katrin Skoruppa; Jessica C. Weiner-Bühler; Leila Schächinger Tenés; Robin Segerer; Letizia Volpin; Ludovica Serratrice

Four contributions will present longitudinal trajectories of lexical acquisition across languages in German- and French-learning bilingual infants and preschool children, and link them to their phonological, pragmatic and metacognitive skills.

Phonological Memory and Vocabulary Skill Development in Young Dual Language Learners

Authors: Jessica C. Weiner-Bühler; Skoruppa, K., Schächinger Tenés, L. T., Segerer, R. K., & Grob, A.

“Children’s vocabularies expand tremendously during the preschool years. Nevertheless, considerable variation exists concerning children’s learning speed and skill development. The capacity for learning new words and expanding vocabulary knowledge is assumed to be related to children’s phonological memory skills, indicated by their ability to repeat novel words (i.e., nonword-repetition). This idea is supported by several cross-sectional studies examining monolingual preschoolers and school-aged children. However, research is still missing that investigates specific interrelations in children growing up speaking more than one language, i.e., dual language learners (DLLs). We hypothesized reciprocal booster effects between vocabulary skills and later phonological memory skills and vice versa, with similar effects for both of DLL-children’s languages.

We studied 58 DLL-preschoolers longitudinally (age-range at t1 = 36-63 months; 55% female). At three measurement time-point (t1-t3), we assessed phonological memory skills using a cross-linguistic nonword-repetition task and evaluated receptive-level vocabulary knowledge via standardized testing-battery in both of the children’s languages (i.e., Italian or Turkish as the heritage language; German or French as the societal language). Random-intercept cross-lagged panel modeling (RI-CLPM) explored the interrelations between DLL-preschoolers’ phonological memory and vocabulary skills over time. Specifically, RI-CLPM evaluated whether a deviation from a DLL-child’s phonological memory trait-level prospectively affected his/her deviation from the vocabulary skills trait-level, and vice versa.

Results revealed a trend-like effect where t1 societal language vocabulary skills predicted better t2 phonological memory abilities (z = 1.95; p = .052), while t1 heritage language vocabulary skills did not (z = 0.60; p = .552). Subsequent (t2-t3) cross-lagged effects were nonsignificant (z < 1.54; p = .123). These findings suggest that the relationship between phonological memory and vocabulary learning may differ slightly for DLL-preschoolers when learning their societal and heritage language, but more importantly challenge assumptions about the impact of phonological memory development on receptive-level vocabulary learning in children.”

How code-switching behavior predicts language attrition in dual language learning children

Authors: Leila Schächinger Tenés; Weiner-Bühler, J. C., Grob, A., Skoruppa, K., & Segerer, R. K.

“Dual language learning children often experience difficulties acquiring high competence levels in both languages. Particularly with the entrance into the educational system, societal language abilities increase, while heritage language skills stagnate or even decrease. However, an early prognosis of such a heritage language attrition is difficult. Beside linguistic factors, language attrition processes also seem to be related to cognitive and acculturative factors. These very factors are also likely to influence dual language learners’ code-switching (i.e., switching between languages) within (i.e., intrasentential) or between (i.e., intersentential) utterances. Code-switching could therefore serve as an early predictor for dual language learning children’s future heritage language attrition. We anticipate a heritage language attrition in dual language learning preschoolers, and expect early code-switching behavior into the societal language as well as intra- and intersentential code-switching to be of predictive value.

We examined 93 dual language learners (age-range = 37-74 months; 58% female) growing up with Italian or Turkish as the heritage language together with German or French as the societal language. At baseline, parents reported via a questionnaire their child’s intra- and intersentential code-switching from the heritage to the societal language and vice versa. Children’s expressive vocabulary in both languages at baseline and follow-up (after ~11 months) were assessed based on standardized testing batteries.

According to our expectations, growth analyses showed that dual language learners’ heritage language development stagnated (b = 0.02; p = .489). For children even experiencing heritage language attrition, their intra- (b = -0.03; p = .023) and intersentential (b = -0.05; p = .007) code-switching was of incremental value, beyond further linguistic and cognitive characteristics. However, code-switching direction was not predictive.

Our findings suggest code-switching to represent an easily observable predictor for heritage language attrition, potentially enabling parents and childcare professionals to shape and maintain dual language learners’ heritage language competencies.”

Disentangling the Bilingual Mind: Vocabulary Overlap in Dual Language Learning Preschoolers

Authors: Robin Segerer; Schächinger Tenés, L. T., Skoruppa, K., Grob, A., & Weiner-Bühler, J. C.

“Studies on preschool-aged dual language learners (DLL) often exhibit limited vocabulary overlap between language systems, with each language assigned to specific semantic domains. The phenomenon of non-overlapping vocabularies is thought to be caused by contextual factors in families and child-care facilities. However, in previous studies this conclusion was based primarily on parental ratings, but not on objective test data, and moreover, mainly focused on contextual factors without taking possible cognitive determinants into account. This research addresses these gaps by using linguistically parallelized standardized tests for heritage and societal languages. It also explores the assumed positive links between vocabulary overlap and cognitive abilities like intelligence, inhibition, and verbal memory.

A sample of 201 preschool-aged DLL (M(SD) age = 48.08 (8.50) months) was examined, with German (69%) or French (31%) as their societal language and Italian (52%) or Turkish (48%) as their heritage language. Linguistically parallelized standardized tests assessed receptive and productive vocabulary. The mean phi-coefficients between item responses in the heritage and the societal language were in the low positive range for both receptive (M(SD) = .10(.24)) and productive vocabulary (M(SD) = .20(.25)), indicating a stronger tendency towards overlapping than distributed vocabularies. The overlap of societal and heritage language vocabulary does not seem to be caused by stable contextual factors, as indicated by absent correlations between receptive and productive vocabulary overlap as well as with social background factors. Surprisingly, cross-sectionally, cognitive variables did not emerge as significant predictors of overlap. Moreover, preliminary longitudinal findings (retest after ~12 months) suggest that vocabulary overlap is more likely a predictor than a consequence of cognitive factors especially for inhibition and verbal memory. These findings challenge the assumption of distributed vocabulary systems in young DLL and suggest that DLL-vocabulary acquisition is primarily driven by the acquisition of concepts, rather than by contextual or cognitive factors.”

Phonological and pragmatic precursors facilitate the development of the early multilingual lexicon

Authors: Letizia Volpin; Schwob, S., Ballestraz, A. & Skoruppa, K.

“The emergence of the lexicon is facilitated by precursor skills in other linguistic domains, such as babbling and subsequent phonological development, but also early pragmatic skills, such as turn-taking and joint attention, as well as gesture development. However, these well-known links have mostly been established for monolingual, often English-learning, children, and rarely been studied together. Moreover, specifically bilingual phenomena, such as the influence of cognate status (phonological similarity across languages, e.g. tiger in English – tigre in French), have rarely been studied on an individual level at such a young age.

As part of a larger longitudinal study, we will analyze French-speaking mono-, bi- and trilingual children’s lexical development (n=72) from 8-9 to 24 months, using data collected via cross-linguistic parental questionnaires and live tasks.

Preliminary analyses up to the age of 12 months confirm previous literature showing that bilingual and monolingual children’s conceptual lexicons (that is, the number of concepts they know a word for in any language) are similar (p’s>.1), whereas monolinguals understand significantly more French words than bilinguals (p=.025). We found a significant correlation between gesture repertoire at 8-9 months and receptive and productive vocabulary in all children (p’s<.01). Finally, for bilingual children, we were also able to show that phonological similarity across languages significantly facilitates the emergence of a word in both receptive lexica of a given child (p<.001).

By the time of the conference, we will finish data collection at 18, 21 and 24 months. We will incorporate these data in longitudinal mixed effect models, which we will construct and compare stepwise in order to account for the large number of variables of interest and possible co-variates (e.g. parental education), and discuss our results in light of contemporary theories of bilingual language development.”

07/19/2024, from 08:30 AM to 10:30 AM , Room P300

Symposium: Associations and dissociations between phonological and semantic aspects of word learning

Speakers: Ron Pomper; Shelley Gray; Natalie Munro; Lisa Goffman; Sara Benham

This symposium investigates phono-semantic interactions in word learning for populations of children that are traditionally under-represented in research: bilinguals with typical development and monolinguals with speech or language disorders.

Structural Models of Spanish-English Bilingual Word Learning in Young Children

Authors: Shelley Gray; Mary Alt, Roy Levy, Tiffany Hogan & Nelson Cowan

In an earlier study, we established a comprehensive structural equation model of word learning that fit the data of typically-developing, English-speaking second graders (Gray et al., 2020). Our goal for the current study was to determine if that structural equation model was invariant for typically-developing Spanish-English-speaking bilingual 2nd graders from the same community. Understanding models of word learning is an important first step to understanding the word learning and vocabulary development of bilingual children. Vocabulary is one area that is notoriously difficult to assess for bilingual children, as their learning is distributed across their languages. As such, bilingual children’s vocabulary and word knowledge may be routinely underestimated. We tested 80 Spanish-English speaking second graders in a series of word learning tasks. The tasks were delivered via computer and required the children to learn a series of nonword nouns and verbs. The word learning tasks served as indicators for a range of potential models that tapped children’s (1) expressive and receptive knowledge and (2) ability to create and store phonological and semantic representations, to link those representations, and to retrieve, recreate and produce information about the nonwords. We tested whether the best-fitting model for English-speaking children from Gray et al., (2020) that included both phonological and semantic factors, was invariant for the Spanish-English speaking group. The model from Gray et al., 2020 did not fit the bilingual group well and we could not establish even the first level of invariance – configural invariance. This means that we cannot assume that the Gray et al. (2020) word learning model measures word learning constructs in the same way for monolingual English and bilingual Spanish-English-speaking children. We discuss the differences in the data and the implication for understanding word learning in bilingual children.

The word learning abilities of children with and without phonological impairment

Authors: Natalie Munro; Stephanie Hearnshaw, Elise Baker

“Children with phonological impairment have difficulty learning the sound system of their language, but what about their vocabulary abilities? Historically it has been assumed that children with phonological impairment are at a higher risk of difficulties with language acquisition, including morphosyntax and vocabulary. However, children with phonological impairment are heterogeneous. In this study, we focus on children with phonological impairment who are lexically precocious. We specifically examine their word learning abilities as a way of offering insight into these children who present as a theoretical conundrum—their phonological and lexical systems seem to be developing with an unusual degree of independence.

Forty-nine 4- to 5-year-old Australian-English speaking children—18 with phonological impairment and 31 with typically developing speech; 21 with lexically precocious vocabulary and 28 with average vocabulary—consented to participate in this study. Four novel non-words were taught through stories, and word learning was assessed at one week post-initial exposure using two measures: confrontation naming and story retell naming.

There was no significant difference in the word learning ability of children with versus without phonological impairment. Regardless of phonological ability, children with above average vocabularies presented with significantly better word learning abilities than children with average vocabularies. Of the 49 participants, 31 named more target non-words correctly in the story retell naming task than in the confrontation naming task. This suggests there may be value in adding semantic context to support new word learning.

Children with phonological impairment and lexical precocity challenge the theoretical understanding about the interdependence of phonological and lexical systems. Irrespective of phonology, children who present with above average vocabularies are better word learners. These findings offer insights for clinicians in understanding the individual strengths and needs of children with phonological impairment, the need to assess their vocabulary, and therefore tailor intervention accordingly.”

Enhancing the Word Learning of Children with Developmental Language Disorder

Authors: Ron Pomper; Karla McGregor

Individuals with Developmental Language Disorder (DLD) take longer than peers to learn new words. In two studies, we manipulated the teaching environment to boost word learning. In Study 1, we taught children (Mean age 7;3) novel words via indirect exposure or direct instruction. In the indirect condition, we presented an unfamiliar and familiar referent (e.g., giraffe-necked weevil and ant) and asked children a question (e.g., “are the antenna of the blavid up?”). In the direct condition, we presented an unfamiliar referent while naming it and asked children to remember the name. Using alternative-forced-choice tests, we measured children’s accuracy in identifying the correct referent, semantic category, and form (e.g., “did you learn blagid, blavid, or blazid?”) for each word. For children with typical development (TD, n=45), direct instruction improved all three aspects of word learning. For children with DLD (n=36), direct instruction only improved their accuracy in linking words to referents. In Study 2, we tested the hypothesis that the simultaneous presentation of new visual and auditory information (in Study 1) was too cognitively demanding for children with DLD who, as a group, had low scores on independent measures of working memory. We taught a subset of children from Study 1 (Mean age = 10;3; DLD N=27; TLD N=40) novel words and manipulated the presentation of the word and referent to occur simultaneously or sequentially. For both groups, sequential exposure improved children’s accuracy in identifying the correct referent and form of each word. These outcomes have clinical implications: A simple, no-cost change in teaching (naming the referent just prior to showing the referent) enhances the early stages of word learning. Taken together, these studies demonstrate associations and dissociations between phonology and semantics during word learning for children with DLD that are sometimes similar and sometimes dissimilar to their peers with TLD.

Interactivity across semantic, phonological, and articulatory levels as children learn words

Authors: Lisa Goffman; Sara Benham

In classic accounts of word learning, meaning is mapped to phonological and then to articulatory levels. Children with developmental language disorder (DLD) may show deficits at all levels, but particularly in word form learning. We will present a series of studies showing that there is significant interactivity across processing levels in children with DLD and with typical development (TD). We use a paradigm in which children are asked to produce novel two syllable phonologically complex (e.g., “puvgub”) nonwords or words. Children initially produce these sequences as nonwords, after which varying degrees of semantic content are incorporated. When 4- and 5-year-old DLD and TD children learn novel word forms, the inclusion of a semantic referent results in decreased variability in the implementation of speech motor templates, suggesting interactivity across semantic and articulatory levels. In another study, we show that, for children with DLD but not TD, syllable sequences become more stable when a semantic referent is included; semantic information provides a bootstrap for word form learning by inducing increased organization of syllable sequences. Thus, in the initial mapping of word forms to referents, both phonological and articulatory production levels are enhanced. We also asked about more protracted learning in which children return for additional sessions. Children with DLD continue to show deficits in word form, as indicated by decreased accuracy and increased intra-word variability in comparison with typical peers. However, counter to our predictions, the inclusion of increasingly complex semantic information leads to increased errors in children with DLD. In summary, word learning is a complex and interactive process, with semantic cues affecting how novel word forms are produced over time. Word form learning deficits are hallmarks of DLD, and the inclusion or exclusion of semantic cues influences the malleability of learning.

Phonological and semantic interactions in typical and atypical language development

Authors: Sara Benham; Lisa Goffman

When children learn new words, they must map a sequence of sounds to a meaningful referent. For many children, this process unfolds relatively seamlessly over time. However, for preschoolers with developmental language disorder (DLD), integrating novel sound sequences with a semantic referent leads to difficulties in sound accuracy. In this work, we examine the relationship between form production and a semantic referent by analyzing not only phonetic accuracy, but other aspects of phonology such as the organization of phonetic features and syllable sequences. We have applied tools from network science that are rooted in a graph theoretic approach to determine developmental change in phonological organization as toddlers and preschoolers map word forms to referents. We will present findings from a series of studies in which children between the ages of 2 and 8 years with typical and atypical language development produce novel disyllabic words, some that are paired with a consistent visual referent, and some that are not. Using a combination of novel network science analyses and standard phonological measures, we show that, for preschoolers with DLD, the incorporation of a referent with a novel word form induces the production of stable syllable sequences, but does not affect segmental or phonetic feature accuracy. Typically developing children who are 2 years old show a different pattern of results: sound feature accuracy is disrupted by a referent, but not syllable organization. These findings elucidate how sounds and referents interact within the mental lexicon, and also point to new directions in our understanding of the phonological factors underpinning word learning. We show how the study of phonology across key developmental periods can be enhanced by tools of network science, providing new insights into the shifting organization of sounds and words.

07/19/2024, from 08:30 AM to 10:30 AM , Room P301

Symposium: Language Acquisition in Real-World Contexts: Head-Mounted Eye Tracking in Diverse Populations

Speakers: Jennifer Sander; Yayun Zhang; Chen Yu; Hana D’Souza; Caroline Rowland

This symposium is aimed at understanding how children’s real-time attentional abilities constrain language development across different populations, communication modalities, and learning contexts using head-mounted eye-tracking.

How does shared book reading support language development?

Authors: Yayun Zhang; Thalassia Kontino, Utrecht University; Caroline Rowland, Max Planck Institute for Psycholinguistics; Chen Yu, The University of Texas, Austin

“Shared book reading positively affects language development by providing children with types of complex language that might be otherwise rare. However, because children’s earliest input is likely too full of unknown words to attract and maintain enough attention necessary for learning, the causal pathways of this positive relationship are not well understood. In this study, we examined how children process diverse linguistic input from a book-reading context by analyzing real-time information processing and how parent and child jointly create interactional routines to achieve joint focus and maintain the attention necessary for learning.

We conducted a dual head-mounted eye-tracking study to investigate how toddlers learn the correct word-object mappings by attending to the right object at the moment of parent naming in shared book reading. Specifically, we investigated (1) linguistic constructions in object naming utterances generated by parents; (2) whether gestures from both parent and child influence the linguistic constructions of parent speech and (3) how gestures play a role in directing the child’s attention to the named referent.

We collected 45 book reading sessions from 16 parent-child dyads (M= 19.03 m.o.) and found that parents used a mix of simple and complex grammatical constructions, with a preference for subject-predicate forms (67%). Gestures, whether from parents or children, did not significantly affect the grammatical constructions used in parent speech. However, gestures successfully directed the child’s attention to the named objects.

Our findings suggest that book reading offers toddlers a linguistically diverse environment for language development. This diversity is advantageous as it provides children with more opportunities to gradually build the multi-layered structure of language. Gestures, in particular, offer a helpful pathway for learners to connect words to objects, facilitating real-time language comprehension. This study sheds light on how dyads coordinate various information sources during shared book reading to support language learning.”

Quantifying the input of early word learning to children with and without ASD

Authors: Chen Yu; Julia Yurkovic-Harding, University of South Carolina; Daniel Kennedy, Indiana University

“The success of early word learning relies on children’s ability to associate the words they hear with the objects or events they see. When hearing an object label in parent speech, looking at the right referent at the right moment allows children to build correct word-referent mappings. In the current research, we aim to quantify and compare the differences among children with and without ASD in directing their visual attention to select the correct referents.

We recruited 60 young children (M=23.3 mo, SD=5.8; 17 with ASD and 43 without ASD) and their parents who were asked to play with a set of 24 unfamiliar toys in a home-like environment. We used head-mounted eye tracking to record child gaze direction when parents spontaneously named toy objects. In a 3-second window beginning at the onset of a parent naming event, the proportion of time that child gaze was directed to the parent-intended referent was measured.

There were 1246 naming events from children without ASD and 423 from children with ASD. For children without ASD, their attention to named objects across all naming events forms a bimodal distribution. For much of the 3-second window after a naming event, children without ASD either looked to the intended referent or, equally often, looked to some other object. For children with ASD, their attention across all of the naming events forms a skewed distribution. Among 60% of naming events, children with ASD didn’t attend to the target object at all. Only within 20% of naming events, they spent over 50% of time attending the target object. The marked differences among children with and without ASD suggest that differences in visual attention while hearing parent speech may play a critical role in vocabulary development, via selecting and altering the word-referent statistics that children perceive in everyday contexts.”

Sensorimotor patterns of parent-child interaction during word learning in Down syndrome

Authors: Hana D’Souza; Kate Mee, Cardiff University; Catalina Suarez-Rivera, New York University & UCL; Chen Yu, The University of Texas at Austin

“A lot of our knowledge about how young neurodiverse children develop language currently comes from screen-based tasks, standardized tests, or parental reports. Less is known about children’s everyday learning experiences, where parent-child interaction is at the core. During this free-flowing activity, parents direct or react to their child, while their child—who is often surrounded by interesting objects and in pursuit of their own goals—directs, reacts to, or ignores the activity of their parent. What sensorimotor patterns (e.g., looking, object handling, speech) define this rich context? In order to understand this, we need to step into children’s shoes and experience the world as they do. Recent technological advancements allow us to do this by enabling us to transition from traditional screen-based eye-trackers to head-mounted eye-trackers/cameras. This technology has already provided us with insights into the dynamic interplay between various components of parent-child interaction in typically developing children, challenging fundamental assumptions about the sensorimotor properties of language development (e.g., how much the child looks at faces). However, much less is known about neurodiverse children. This is the first study to use the same technology with young children with Down syndrome.

Fifteen children with Down syndrome aged 3-5 years, and 15 typically developing children matched on ability level, took part in this head-mounted eye-tracking study of parent-child interaction with novel objects and their labels. The interaction data was coded frame-by-frame and analysed for looking behaviours, object handling, and parental speech. The sensorimotor properties of these interactions will be discussed in the context of the strengths and difficulties young children with Down syndrome experience, as well as large individual differences we observe. We hope this research will inform our theories of language development as well as provide insights for parents and practitioners into how to better support young children with Down syndrome.”

Through the Eyes of the Learner: Attentional Dynamics in Signed and Spoken Parent-Child Interactions

Authors: Jennifer Sander; Dilys Eikelboom, MPI for Psycholinguistics; Yayun Zhang, MPI for Psycholinguistics, Caroline Rowland, MPI for Psycholinguistics

“Joint attention (JA) plays an important role in spoken language acquisition. However, signed language input looks very different to spoken language input and only very few studies have examined the effect this has (e.g. how signing children focus and sustain their attention during naturalistic interactions, how this affects language development).

Among the few sign language studies examining real-time attentional behaviors, it has been found that signing children and caregivers are more sensitive to gaze cues of their communication partner and spent more time in mutual gaze than speaking dyads, with a higher frequency in gaze shifting (Lieberman et al., 2011 & 2014). However, these studies used videos of naturalistic interactions recorded from an observer’s perspective. Less is known about the perspective from first-person view, which is an important perspective because it uniquely accounts for the visual perspective of each interaction partner.

To provide such a fine-grained measure of naturalistic interactions we plan to investigate gaze behaviours in signing as well as speaking dyads with children between the age of 1 and 5 years using head-mounted eye tracking in a mobile lab. This allows us to test a larger number of participants in a three-folded comparison: 1) cross-modal, 2) cross-developmental, 3) cross-contextual.

We predict that signing dyad interactions will have different characteristics than those of speaking dyads to accommodate the visual, as opposed to auditory, modality (e.g. have more frequent gaze shifting and shorter periods of JA and mutual gaze). However, in both speaking and signing dyads, we expect JA success to increase with children’s age, and we expect children which navigate the interaction most efficiently to display the biggest vocabulary growth. Our study will contribute new insights into what characterises a language learning facilitating environment, and which different interaction strategies can lead to successful language acquisition.”

07/19/2024, from 08:30 AM to 10:30 AM , Room P104

Symposium: Why does social contingency matter for language learning?

Speakers: Elena Luchkina; Morgane Jourdain; Gideon Salter; Elizabeth Che

This symposium focuses on a crucial component of social interaction – contingency – that facilitates language development explores the underlying mechanisms of this effect across languages, SES, and typical and atypical populations.

Mechanisms underlying the effects of social contingency on word learning in the first year of life

Authors: Elena Luchkina; Fei Xu

“We investigated the role of socially contingent interactions between typically developing infants and their parents in early vocabulary development. Convergent evidence from prior research shows that infants of parents are more likely to engage in socially contingent interactions with their infants tend to have larger vocabularies. An open question is how social contingency facilitates vocabulary growth.

One possibility is that parents who speak in response to their infants more often produce larger amount of language input, which could accelerate vocabulary growth. Another possibility is that the properties of socially contingent language input are uniquely suitable to support early word learning. Yet another possibility is that the frequency of parents’ contingent responses helps infants build a link between their own words or vocalizations and others’ behaviors, leading to further language advances, including vocabulary growth.

To distinguish between these hypotheses, we analyzed relations between parent-infant interactions when infants were 9 and 12 months and their vocabulary size at 12 months. Our findings suggest that the frequency of both verbal and non-verbal responses to infants’ vocalizations facilitate vocabulary development longitudinally. Our findings also suggest that this development is unlikely to be due to the amount of language input or the properties of language within socially contingent interactions. Follow-up work explores whether and how contingent responses indeed help infants infer the communicative nature of language and how this inference facilitates vocabulary growth.”

Simplification in contingent child-directed speech is the result of responsive alignment

Authors: Morgane Jourdain; Sabine Stoll

“Child-directed-speech (CDS), particularly interactive exchanges, crucially predict language learning. Emerging evidence suggests that contingent CDS, responsive to a child’s utterance, is simpler compared to CDS in general. The underlying mechanisms of this complexity matching remain unclear. We ask: is this adaptation limited to CDS, or does it reflect a broader mechanism inherent to human communication?

To test this, we compare contingent, non-contingent CDS, and contingent and non-contingent adult-directed speech (ADS) in naturalistic corpora of three languages with complex verb morphology: Chintang (Sino-Tibetan, Nepal), Qaqet (Baining, New Guinea) and Turkish (Turkic, Turkey). We measure verbs’ complexity both in length, in number of morphemes, and in diversity of form, measured with the Gini index that determines the skewness of the distribution of verb forms.

Study 1 shows that in all three languages verbs are shorter and have fewer forms in contingent than in non-contingent CDS, but that verbs are not simpler in non-contingent CDS than ADS in all languages. This highlights the specific role of contingency on simplification in CDS. Study 2 further reveals a correlation between verbs and utterance length in contingent CDS and the child’s prior utterance. This could entail that simplification in contingent CDS is the result of the adults aligning to the complexity of child speech. Child speech production serves as a gauge that adults use to base the complexity of their CDS. Study 3 demonstrates a similar effect of contingency on complexity matching in ADS.

Because ADS shows the same processes as CDS, we propose that responsive alignment is a general property of language interaction. For the language learning child it has a beneficial effect because adults utilize it to adapt their own speech, aligning with the child’s linguistic abilities. This adaptation facilitates effective communication and, in turn, enhances learning outcomes.”

Effects of promoting caregiver responsiveness on early communication and language: An RCT

Authors: Gideon Salter; Colin Bannard, Silke Fricke, Penny Levickis, Julian Pine, Kiera Solaiman, Emma Thornton, Danielle Matthews

“This study tested whether it is possible to mitigate the risk of language delay associated with social disadvantage by promoting caregiver linguistic responsiveness. In doing so it assessed a) the causal role of social contingency in language learning and b) the potential value and acceptability of low-intensity digital interventions.

435 families with infants aged 4-10 months at baseline participated from their homes across the four nations of the UK (postcodes in the lowest 5 deciles of the national Index of Multiple Deprivation). Families were randomly assigned to either a language intervention promoting responsive linguistic interaction or a matched physical health intervention, both of which involved receiving three text messages a month with links to short parenting-focussed videos. Analyses were pre-registered.

The language intervention led to a significant increase in overall caregiver linguistic responsiveness as measured by the PaRRiS at 12, 17 and 24 months. It also led to increased infant communication at 12 months and increased caregiver responses to infant communication. It did not lead to a significant increase in caregiver-reported expressive vocabulary at 17 or 24 months (primary outcome measure, N = 399, 400). Blind coding of home videos suggested a significant increase in expressive vocabulary at 17 and 24 months (N = 123, 154). Questionnaires and focus groups suggested the intervention was acceptable to parents (although one requested more support for children with developmental delay).

Overall, the findings indicate that relatively low-intensity interventions can promote caregiver responsiveness and this appears to have a causal effect on infant prelinguistic communication. More support would be needed to promote vocabulary development. More detailed coding of changes in specific developmentally-attuned response types and of focus groups suggest a promising avenue would be to provide one-to-one support for families at higher cumulative risk. We are currently exploring the feasibility and acceptability of such an option.”

Using CHIP to Explore Effects of Overlapping Word Usage on Early Language Development

Authors: Elizabeth Che; Patricia Brooks

fast mapping, the ability to establish a rapid audio-visual label-referent association from minimal exposure, is connected to word learning and can enhance lexical growth around 2 years of age. For congenitally deaf infants, late access to auditory inform

07/19/2024, from 08:30 AM to 10:30 AM , Room P018

Symposium: Interventions that Improve Vocabulary and Narrative Skills in Ethnically and Linguistically Diverse

Speakers: Diana Leyva; Si Chen; Sarah Surrain; Vibeke Grover; Catherine Snow

This symposium discusses four innovative interventions designed to improve preschoolers’ vocabulary and narrative skills, while targeting ethnically and linguistically diverse communities in both the U.S. and Norway.

Empowering ethnically diverse parents to support preschoolers’ language skills while cooking

Authors: Diana Leyva; Sofina Shekhar, Eden Galan, and Shante Antrom

Decontextualized talk (DT) is talk that goes beyond the here and now (e.g., talking about past events; Snow, 1983). Promoting DT in preschool is important because it predicts children’s academic achievement in middle school (Uccelli et al., 2019). Literacy Eats is a culturally responsive intervention program designed to empower low-income and ethnically diverse parents in using DT at home while cooking with children. Literacy Eats was co-developed in partnership with a community organization, where parents learned to use DT while cooking simple recipes and watched video clips of real families using DT while cooking. This study evaluated the feasibility of Literacy Eats using a pre-post-test design. Participants were 37 low-income, ethnically diverse parents (32% Black/African American, 27% Latino/Hispanic, 27% Asian, 5% White and 9% other; 46% spoke a language other than English at home) and their preschoolers (M age = 54.7 months; 60% girls) living in a northeastern U.S. city. We examined associations between program attendance frequency and improvements in: (a) children’s vocabulary and narrative skills and (b) the frequency of parent-child participation in food-related activities (e.g., cooking and eating together). At both pre-test and post-test, we assessed children’s vocabulary using the IDELA expressive vocabulary items, their narrative skills through a personal narrative task, in which the child told two stories (which were transcribed and coded for elaboration and lexical complexity), and the frequency of parent-child participation in food-related activities using a parent questionnaire. Parents who attended more Literacy Eats meetings had children with greater improvements in vocabulary and narrative skills from pre-test to post-test. However, increased attendance in the program did not necessarily relate to an increase in parent-child participation in food-related activities from pretest- to post-test. Literacy Eats is a promising culturally responsive program that supports low-income and ethnically diverse parents in fostering preschoolers’ language and literacy skills.

Enhancing narrative skills of young children by sustaining preschool impacts

Authors: Si Chen; Olivia Horne, and Catherine E. Snow

“The Expanding Children’s Early Learning (ExCEL) project is a collaboration between MDRC, Boston Public Schools, the University of Michigan, and Harvard Graduate School of Education that aims to examine several approaches to sustaining children’s preschool achievement. In Boston Public Schools, ExCEL has an excellent opportunity to evaluate whether a preschool program combining evidence-based curriculum with coaching and professional development for teachers, as well as alignment of instruction across elementary schools, is likely to result in sustained improvements in student outcomes. ExCEL seeks to generate rigorous evidence to guide the development and delivery of early childhood education on a large scale by harnessing the commitment to expand early childhood education.

The current study included 284 3-to-4-year-old children in the Boston Public School Prekindergarten program (29.6% Black, 25.3% Hispanic, 19.7% White, 14% Asian, 5.6% multiracial, and 5.6% other; 57.8% monolingual (English) and 42.3% bilingual (English and other language). Half of the classrooms were randomly assigned to ExCEL and the other half were randomly assigned to a control (business-as-usual) condition. We investigated whether ExCEL improved children’s narrative skills at post-test. Children’s narrative skills were assessed at post-test using the Renfrew Bus Story narrative retelling task (Renfrew, 1969). In this task, children heard a story and were prompted to retell it using illustrations as support. Children stories were scored for inclusion of information, sentence length, and the amount of prompting required, using an adapted version of the examiner’s manual (Cowley & Glasgow, 1994). We found that ExCEL had moderate but significant positive impacts on children’s narrative skills at post-test (effect size = 0.17). We discuss the differences in effect sizes among different racial and linguistic groups.”

Effects of Play/Learning Strategies for Spanish-dominant Latine Families in Reading and Free Play

Authors: Sarah Surrain; Susan Landry, Tricia Zucker, and Yoonkyung Oh

The way parents respond to their child’s initiations and guide their learning in informal settings is associated with children’s subsequent language development. However, less is known about the efficacy of responsive parent interventions for families from diverse linguistic and ethnic backgrounds. This study is a secondary analysis of a randomized controlled trial of the Play and Learning Strategies (PALS) intervention, which facilitates parents’ learning of responsive parenting strategies via online modules, 1-on-1 coaching, and video reflections. Parents of preschool-aged children (M age = 52.8 months) were randomized to PALS or a control condition and observed interacting with their child in two settings – book reading and free play – before and after the intervention period. The current study examines the effect of PALS on targeted parent behaviors (e.g., Responsiveness, Language Building Strategies) and child behavioral outcomes (e.g., Language Use, Engagement) for three subgroups: Spanish-dominant Latine families (n = 142), English-dominant Latine families (n = 112), and English-dominant non-Latine families (n = 137). Spanish-dominant Latine parents received PALS in Spanish, while the other two groups received PALS in English. There were significant, moderate-to-large, main effects of PALS on parent and child behaviors during the book reading for all three subgroups. However, PALS effects varied by group for parent and child behaviors observed during free play. The Spanish-dominant Latine parents experienced larger gains in parent and child outcomes when observed in the free play setting, compared to the other two groups. These differential effects were not explained by the number of completed sessions or coach-rated engagement but may be related to differences in family composition and/or cultural orientations toward book reading vs. free play activities. Our results highlight the importance of observing multiple contexts in research with families from minoritized backgrounds and inform the development of new parent interventions that build on families’ strengths.

Dual-language learners demonstrate longer-term narrative effects of a shared-reading intervention

Authors: Vibeke Grover; Jan-Eric Gustafsson, Veslemøy Rydland, and Catherine Snow

“In this presentation we examine the longer-term effects of taking part in a shared-reading intervention in preschool and at home. The study included 464 dual-language learners, 3-5 years old, who lived in Norway and spoke several different first languages at home. The best represented first languages were Urdu, Somali, Polish and Arabic. The 123 classrooms that the children attended were randomly assigned to either an experimental condition, in which the children received a shared-reading intervention in preschool and at home or a control (business-as-usual) condition. We assessed children pre-intervention, post-intervention and seven months post-intervention. We asked whether there were longer-term effects of receiving the intervention on second-language vocabulary and narrative skills and used second-order latent growth modelling to answer the question.

During the intervention year, children shared several books with their teachers who all spoke Norwegian. The teachers were asked to discuss targeted words, invite child reasoning, and explore ideas through questions. Additionally, parents at home shared some of the books read in preschool, using their preferred language. These books were mostly wordless, depicting a narrative through pictures. We found immediate effects of the intervention on children’s second-language vocabulary skills, but not on their narrative skills. Seven months later we identified a remaining, but somewhat fading effect on vocabulary skills and an emerging effect on narrative skills, a skill that may take time to develop.

Children shared the same books with different interlocutors, in different settings and in both of their languages, requiring them to take into consideration the perspectives of their reading partners and various interpretations of the books. The discussion will focus on how these specific features of the intervention may have contributed to the longer-term emerging effects on narrative skills.”

07/19/2024, from 08:30 AM to 10:30 AM , Room P200

Symposium: Individual differences in the heritage language development of school-age children

Speakers: Johanne Paradis; Silvina Montrul; Jacopo Torregrossa; Evangelia Daskalaki; Adriana Soto-Corominas

Bilingual children are at risk for HL attrition after school entry. Research on factors that determine successes and challenges in the development and maintenance of HLs is presented from North American and European contexts.

L1 Text Experience Contributes to Individual Differences in the Acquisition of Spanish Passives

Authors: Silvina Montrul; Andrew Armstrong

“Most research on the development of heritage Spanish in children is grounded in comparisons between heritage speakers (HSs) and age-matched Spanish monolinguals. However, this comparison overlooks variability in Spanish outcomes among HSs themselves, and fails to consider input and other individual factors that contribute to these differences. To understand the sources of this variability, we investigated Spanish textual input as one contributor to individual differences in the acquisition of complex syntax in 8-12-year-old children’s heritage Spanish in the United States by testing their production of verbal passives and other complex structures on a sentence repetition (Srep) task. 25 HS children attending English-only schools and 25 HS children attending dual-immersion schools completed tests for Spanish literacy (5), rapid automatized naming, and working memory. The SRep task targeted (im)plausible verbal passives, a structure frequent in written language, and actives, in addition to other morphosyntactically simple or complex structures.

Dual-immersion school HSs scored higher than English-only school HSs on all literacy and working memory measures (p < .001). On the Srep task, structural accuracy was significantly more likely for HSs in dual-immersion schools and with higher reading vocabularies. Descriptively, the decrease in accuracy from actives to passives was smaller for the dual-immersion school group (98% to 84%) than the English-only school group (82% to 38%). Comparing all simple and complex structures found a significant school × complexity interaction (p < .05) such that the negative impact of syntactic complexity on repetition accuracy was greater for English-only school HSs.

The results indicate that school-age HSs with higher levels of Spanish literacy have stronger morphosyntactic representations which results in more accurate production of complex structures. This provides evidence of text experience as an important factor that leads to individual differences in the oral production of school-age child heritage speakers.”

The role of literacy for the acquisition of difficult structures in Portuguese heritage language

Authors: Jacopo Torregrossa; Cristina Flores, Esther Rinke

“Most research on heritage language (HL) acquisition have examined either early stages (Meisel, 2019) or ultimate attainment (Benmamoun et al., 2013; Polinsky et al., 2019). By contrast, very few studies have analyzed HL acquisition among school-age children (Torregrossa et al., 2023). This age is crucial because HL children are usually exposed to literacy in the societal language and not the HL. We investigate which linguistic phenomena in the HL pose particular difficulties to school-age children and how far language- and literacy-exposure variables affect their acquisition.

We tested 180 HL children/adolescents (age range: 8;06-16; M: 11;07) speaking Portuguese as a HL and French, German and Italian as societal languages. We used parental questionnaires to collect information on their language and literacy exposure. The children were tested in Portuguese by using a cloze-test targeting linguistic structures of different levels of complexity and a written narrative task (Schneider et al., 2006). We identified the linguistic structures that posed more difficulties to the children by considering response accuracy in the cloze-test. Then, we examined whether these structures were produced accurately in the written narratives. We interpreted the results through the lens of language- and literacy-exposure variables.

Based on a cluster analysis of the cloze-test data, we identified 15 (out of 40) complex structures, including third-person clitics, relative pronouns and concessive connectors. Children’s response accuracy with these structures was affected by literacy-exposure variables. Preliminary results of a cluster analysis conducted on the narratives showed that the narratives exhibiting the most complex structures were produced by older children and children with a greater amount of formal instruction in the HL. Quantity of exposure to the HL did not affect the results in either case.

This study points to the importance of literacy exposure in the school-age period for acquiring and maintaining complex phenomena in the HL.”

The role of attitudes, home language use and schooling in Mandarin heritage language development

Authors: Evangelia Daskalaki; Johanne Paradis, Adriana Soto-Corominas

“A growing number of studies show that the amount of HL use at home as well as the amount and type of HL schooling are predictive of children’s HL performance. In the present study, we extend this line of research by asking (i) how these factors are shaped by maternal attitudes towards HL transmission (distal factors, in Paradis 2023, terms); (ii) how they affect vocabulary and syntax, and (iii) how they interact with each other.

To this end, we tested 46 Mandarin-English bilingual children in Canada (Mean Age=10.5; Range=6.8-16.2). The children were divided into three groups based on the school they attended: children (N=10) in English-only school, children (N=15) in English-Mandarin bilingual school, and children (N=21) in English-only school who attended heritage language classes on weekends. Children were administered a picture-naming task targeting Mandarin vocabulary (LITMUS-CLT) and an experimental elicitation task targeting Mandarin interrogative sentences, which is an early acquired structure. Parents were administered a questionnaire about home language environment, attitudes towards Mandarin transmission, and education program choices.

First, regression analyses showed that positive maternal attitudes were associated with more Mandarin use at home. Second, more Mandarin use at home was associated with larger vocabularies and more accurate production of interrogatives. By contrast, school type was only associated with vocabulary: children attending a bilingual school performed better than the other groups. Third, an interaction between school type and home language use showed that the positive role of Mandarin use at home was less pronounced for children attending bilingual schools than for those in English-only schooling.

Taken together, these results show how attitudinal factors may influence input factors, which in turn, may differentially affect the acquisition of vocabulary versus early-acquired syntactic structures. Results also suggest that schooling in the HL is protective against variation in HL use at home”

Minority languages under the microscope: Influences on the development of Catalan object pronouns

Authors: Adriana Soto-Corominas; Silvia Perpinan

“Most studies investigating variability in heritage language (HL) development focus on immigrant contexts, where the quantity and quality of the HL input is limited. This study investigates the development of direct object pronouns in Catalan, a minority language in Catalonia (Spain) in contact with Spanish. Despite being a non-immigrant HL subject to protective policies, such as being the language of schooling, Catalan is learned in intense contact with the majority language, and adult speakers have been shown to not converge on the same end-state grammar (Perpiñán, 2017). However, little is known about the sources of variation in the development of Catalan and national minority languages more broadly.

Catalan-Spanish bilingual children ages 4-8 (N=333) completed an oral production task eliciting three direct object pronouns in Catalan: accusative masculine l, neuter/unspecific ho, and partitive n. Whereas Spanish has a pronoun to carry out the functions of l and ho, partitive n is equivalent to morphological ellipsis (i.e., null pro). Parents completed a questionnaire that yielded information on participant/family demographics and their linguistic environment.

We used regression to investigate the effect of pronoun (l, ho, or n), age at testing, Catalan age of onset, home language use, Catalan nativeness of input providers, and frequency of Catalan-rich activities on the target production of the pronouns. Four predictors were associated with the outcome: age, language use, nativeness of input providers, and the target pronoun itself.

Results showed that the development of Catalan direct object pronouns is subject to similar processes that influence acquisition in immigrant HLs: bilinguals acquiring the minority language under reduced input and low-quality conditions show protracted development, especially with respect to the only pronoun without an overt counterpart in Spanish. Importantly, results reveal that factoring in the quality of the input is crucial in studies of national minority languages.”

07/19/2024, from 11:00 AM to 01:00 PM , Room P131

Symposium: Collaborating with parents to support the language and communication of children with Down syndrome

Speakers: Susan Foster-Cohen; Tamar Keren-Portnoy; Sue Buckley; Vesna Stojanovik; Kelly Burgoyne

The five studies in this symposium explore the value of learning from parents as observers of their children’s language development, and as collaborators in the development and implementation of interventions from infancy to school age.

Using the Babbleplay App to promote vocalising in babies with Down syndrome

Authors: Tamar Keren-Portnoy; Sue Buckley, Kelly Burgoyne, Helena Daffern, Mona Kanaan, Laura Boundy, Sab Arshad

Infants with Down Syndrome tend to vocalise less than typically developing infants (Parikh & Mastergorge, 2018). Given that earlier babbling correlates with earlier word production in typically developing infants (McGillion et al., 2016), it is predicted that actively encouraging infants with Down syndrome to babble more would have positive outcomes for early lexical development. Moreover, given the dose effects of intervention for children with Down syndrome found by Yoder et al. (2014), it is predicted that approaches that can deliver intensive doses of babble encouragement would be particularly effective. To achieve high dose input, we piloted a novel parent-led intervention that uses an app, BabblePlay, developed by our team (Daffern et al., 2020). BabblePlay responds to children’s vocalisations with moving shapes on a screen that are designed to attract and hold the child’s attention and provide encouragement to vocalise. The app records their vocalisations, measures the timing of their vocalisations, and counts their frequency. We will report on a proof-of-concept pilot with a group of infants with Down syndrome (n = 29) aged 0;7-1;3. The aim was to test whether the infants engaged with the app and whether parents found it easy to use. Each infant played with a mirror 5 times one week and with BabblePlay 5 times the following week, for 5 minutes each time. Parents video recorded the final session of each week. Assessment of engagement with BabblePlay was sought through parental questionnaires as well as analysis of the videos, coding for number and duration of looks to the screen. We also compared the amount of vocalisation when playing with BabblePlay versus playing with the mirror. Interim findings suggest a positive impact of BabblePlay in terms of quantity of vocalisation indicative of its potential as an intervention to promote vocalisations and lexical development. We will end by reporting on plans for a feasibility Randomised Controlled Trial of BabblePlay with infants with Down syndrome.

Babble Boot Camp for infants with Down syndrome – a proactive, parent led approach

Authors: Sue Buckley; Beate Peter, Jennifer Davis, Laurel Bruce, Linda Eng, Nancy Potter, Mark VanDam, Lauren Thompson, Lizbeth Finestack, Susan Loveall

Children with Down syndrome are at high risk of significant speech and language delays and the need for intensive intervention from birth is indicated since the foundations for both speech and language are being laid from the first weeks of life. Babble Boot Camp (BBC) is a programme of activities and routines designed to boost the speech and language skills of any child at risk of delays when challenges are predictable at birth. A paediatric speech-language pathologist (SLP) implements the intervention via parent training, using a telehealth platform. In weekly meetings, the SLP coaches parents in techniques designed to boost children’s communicative skills beginning in earliest infancy. The intervention typically spans ages <6 months to 24 months. It has been evaluated in an implementation with children with classic galactosemia and outcomes include more advanced babble skills and higher speech and language skills at follow-up, compared to standard care. Given that babies with Down syndrome are identified at or before birth, this programme offers the opportunity to intervene proactively and is predicted to be of value to their development. A pilot evaluation of BBC with 10 families with infants with Down syndrome (4 to 16 months at start) has been underway since September 2022 and was at the midpoint at time of abstract submission. Families are receiving 10 months of weekly intervention sessions and then monthly sessions to 24 months. Outcome measures include implementation fidelity, detailed measures of phonological and language development using LENA recordings, IPA transcriptions, and the MacArthur-Bates CDI, as well as parent feedback on their experience with the intervention. Child results and parent feedback mid-intervention indicate positive outcomes on all measures and results of the full implementation, currently being completed, will be presented. Plans for a full RCT study will be described.

Evaluating an early social communication intervention for young children with Down syndrome

Authors: Vesna Stojanovik; Emma Pagnamenta, Sarah Sampson, Rachel Sutton, Benjamin Jones, Victoria Joffe, Kate Harvey, Elena Pizzo, Sarah Rae

While aspect of social skills are often more preserved in Down syndrome than other aspects of language development (Fidler, 2005), capitalising on that base as a support for language development requires interventions specifically designed to promote the links from social skill to language skill (Mattie & Fanta, 2023). This presentation reports the results of a feasibility trial of a parent-delivered social communication intervention for young children with Down syndrome (‘ASCEND’) focused on developing children’s early social communication skills through responding to shared attention. The study used a two-arm feasibility randomised controlled trial (RCT), with 1:1 randomisation stratified by trial site, comparing the ASCEND intervention plus standard care and provision of standard care only. Nineteen children with Down syndrome aged 0;11 – 2;11 months took part (n=10 in intervention group, n= 9 in control group). Pre and post intervention and 6-month follow-up assessments included language, social communication skills, adaptive behaviour, quality of life (parents and children), parental anxiety and depression. Data were collected on recruitment and retention, standard care, treatment fidelity, acceptability of the intervention by the parents and speech and language therapists, feasibility of collecting health economic measures and suitability of the primary outcome measure. The intervention (manual, support, materials) was positively received by the participating parents. The children in the intervention group were reported to have significantly higher vocabulary measured on the Reading Communicative Development Inventory (Hamilton et al., 2000) and also had a higher communication subdomain score on the Vineland Adaptive Behaviour Scale (Sparrow et al., 2016). Speech and language therapists also evaluated the acceptability of the intervention positively. Treatment fidelity and retention were acceptable. The preliminary health economic data suggest that this intervention will be low cost.

Development and Evaluation of a Parent-Delivered Early Intervention for Children with Down Syndrome

Authors: Kelly Burgoyne; Kirstie Hartwell, Rebecca Baxter, Vesna Stojanovik , Emma Pagnamenta

Children with Down syndrome have significant language learning difficulties that require support from an early age to reach their full potential. While parents are well-placed to provide this support, they often need guidance, resources, and support to do so. This presentation describes the development and evaluation of an early language intervention for 3-6 year old children with Down syndrome, designed for parents to deliver to their child at home. The programme is based on Parents and Children Together (PACT), an evidence-based language intervention, originally developed for typically developing children at risk of language delays, which leads to gains in language and literacy skills (Burgoyne et al., 2018). In Stage 1 of the project, we worked closely with six families with children with Down syndrome to pilot and adapt the intervention using a mixed methods approach. Families completed 15-25 sessions of the programme over 5 weeks. All families reported enjoyment of the programme and children’s active involvement; observations of delivery further highlighted positive features as well as adaptations to enhance delivery and engagement, including ways to support non-verbal participation. Following adaptations, families completed 13-20 sessions of the adapted programme. Parents reported children were more engaged and actively participating and that delivery was easier and more enjoyable than in the original version. Stage 2 of the project is a feasibility randomized controlled trial (September 2023-December 2024) that will determine the feasibility of a definitive trial of the adapted intervention programme. The results from the Stage 1 pilot will be presented as well as progress to date on the Stage 2 RCT.

Language and language use in children with Down syndrome: Individual and group trajectories

Authors: Susan Foster-Cohen; Jayne Newbury, Toby Macrae

Children with Down syndrome find the acquisition of some aspects of communicative development more difficult than others. For example, aspects of pragmatics may be easier while aspects of linguistic resource development (e.g. morphology) are reported to be more difficult (e.g. Roberts et al 2007). Less is known, however, about how the development of pragmatics and linguistic resources pattern in relation to each other over time: whether there are moments where the development of linguistic resources outpaces the capacity to use them or whether there are aspects of pragmatics that might be delayed due to a lack of resources for expression. To explore this question in preschool children, this study used parent observational data from the Communicative Developmental Inventory (Fenson et al 2007) and the Language Use Inventory (O’Neill, 2009) collected by the mothers of 34 children at six-monthly intervals from 30 to 66 months. Using mixed effects regression modelling in R (Bates et al. 2015), we modelled the relationships between language resources (vocabulary and morphosyntax) and language use development over time with a focus on both group trends and individual differences. While language use development appeared to be ahead of linguistic resource development at most time points, we also found important individual differences as well as an early prevalence of language resources over language use at the earliest and the latest stages of development, independent of chronological age. Comparison of the children with Down syndrome with a group of 53 children with other causes of language delay and a group of 25 developmentally matched typically developing children showed similarities in the patterns of relationships suggestive of a delay rather than a disorder interpretation of the data. The results are important for clinical goal-setting as they can help determine when and where intervention energies might best be spent at different time-points.

07/19/2024, from 11:00 AM to 01:00 PM , Room P104

Symposium: Romanian in the bilingual world

Speakers: Anamaria Bentea; Elena Soare; Bianca E. Babei-Popa; Mihaela Pirvulescu; Silvina Montrul

This symposium addresses questions related to variation and change in heritage languages, with special emphasis on the acquisition and development of various grammatical phenomena in heritage Romanian in contact with different societal languages.

Differential object marking in child heritage Romanian: on the role of linguistic proximity

Authors: Elena Soare; Alexandru Mardale, Larisa Avram

“In this study we investigate differential object marking (DOM) in child heritage Romanian in contact with a language which has and one which does not have morphological DOM: French and Spanish, with a view to shedding light on the role of linguistic proximity in heritage language acquisition.

We used an acceptability judgment task with 16 test sentences across 2 conditions balanced for animacy: DOM with proper names (PN) (obligatory with animate, incorrect with inanimate PNs) and DOM with definite common nouns (CN) (optional with animate, incorrect with inanimate CNs).

45 child heritage speakers of Romanian in contact with French and 45 in contact with Spanish (7- and 9-year-olds) took part in the study. Their responses were compared to those of 45 age-matched monolinguals living in the homeland. All participants accepted DOM with animate objects at high rates (between 75% and 93.3%) but also extended DOM to inanimates. The acceptability ratings are higher with inanimate PNs (between 40% and 63.3%) than with inanimate CNs (between 18.3% and 40%). No significant difference between the children living in France and those living in Spain was attested, with any age group (no significant combined effect of condition x group for either the 7-year-olds – F(3,84) = 2.439, p > .05, or for the 9-year-olds – F(3,84) = 1.449, p > .05). The monolinguals also accepted DOM with inanimate PNs (45%) but only at age 7.

These findings indicate that: there is no effect of linguistic proximity on the acquisition of DOM in heritage Romanian and full acquisition is slower under conditions of language contact. We argue that this extension does not reflect a deteriorated DOM system. It reflects an incipient change in Romanian, which favours expansion of DOM to inanimate objects. Retraction of this input divergent use is slower under conditions of reduced input.”

DOM in heritage Romanian: Is language change accelerated under conditions of language contact?

Authors: Bianca E. Babei-Popa;

“Language acquisition can drive language change. Incipient changes tend to get amplified to rates higher than those in the input. According to some authors, such amplification is accelerated under conditions of language contact. Other authors argue that not every type of language change is amplified under conditions of language contact. In order to evaluate these views, I investigate differential object marking (DOM) in child heritage Romanian in Italy.

Romanian has two DOM markers: the functional preposition pe and clitic doubling (CD). The system is undergoing a diachronic change, from a system with two markers to one which uses exclusively CD. Following the view that language change is accelerated under conditions of language contact, child heritage speakers (CHSs) are predicted to opt for the innovative system and use CD as DOM at a level beyond the one in the input.
In order to test this prediction, I examined DOM use in 52 ‘frog story’ narratives told by Romanian CHSs living in Italy (age range 4;7–14;8). They are all simultaneous bilinguals. The data were compared to DOM use in the narratives told by 15 Romanian first-generation immigrants, 52 age-matched monolinguals and 20 Romanian adults living in the homeland.

Results indicate that CHSs use CD as DOM at a rate lower than 35% out of the total number of marked objects, in all age groups. This is significantly lower than the one with first-generation immigrants (χ2 = 31.79, p = .000) and with child monolinguals (χ2 = 34.31, p = .000). These results support the view according to which not every language change is amplified under conditions of language contact. I account for the lower rates of CD in child heritage Romanian in terms of the interface properties of this DOM marker, which signals D-linking, a property inherited from the clitic.”

Heritage Romanian in the multilingual context: the acquisition of object clitic pronouns

Authors: Mihaela Pirvulescu; Virginia Hill

“This paper focuses on the acquisition of object clitic pronouns by 31 children (ages 8-17) who grow up in Toronto and are exposed to Romanian since birth (L1) in their home; this is the heritage Romanian (HR). These children are in contact with both English (the societal language) and French (the immersion schooling system). The investigation focuses on the use of single clitic pronouns, as in “L-am citit” ‘it.CL-have.1 read’/’I read it’. Considering that the parametric setting for clitic pronouns (vs. lack of clitic pronouns) is well set in dominant Romanian (DR), the aim of the paper is to establish if the same parameter is also well set in HR.

Both comprehension and production were measured: results were compared between a Clitic Elicitations Task and a Comprehension Task (picture choice). The findings are the following: (i) The clitic parameter is well set in HR, i.e. the syntax of object clitic pronouns is well acquired. (ii) Both comprehension and production show morphosyntactic divergences with respect to the gender of the pronominal clitic and clitic omission; however, there is variation across the participants, with children who perform on target in both tasks, some children who only perform on target only on one task, and those who give divergent responses in both tasks. (iii) The variable performance of the participants ties with the external variables investigated, language experience and use predicting correct clitic comprehension while literacy predicting correct clitic production. Working memory and age are negatively correlated with clitic omissions.

Overall, the study highlights the strong similarity between the heritage and the dominant language in the domain of pronominal object clitics while, at the same time, confirming that in heritage language acquisition, target mastery of the morphosyntax is dependent on heritage language maintenance at home and in the heritage speakers community.”

What’s in a question? Comprehension and production of which-questions in child heritage Romanian

Authors: Anamaria Bentea; Theodoros Marinis

“This study compares the comprehension and production of which-questions in the Heritage Language/HL of early bilinguals and aims to understand whether online interpretation in both monolingually-raised and HL children is guided by the presence of morphosyntactic information like case-marking and/or number agreement and how this compares to production. Wh-dependencies in Romanian present morphosyntactic properties that lend themselves well to assessing the use of grammatical cues in comprehension and production: a Differential Object Marker/DOM, pe, precedes the wh-phrase in object questions and indicates that this element should be interpreted as object; number agreement on the verb disambiguates between a subject and an object interpretation. Thirty-one child HSs of Romanian with L2 German (5;6-10;0) and thirty monolingually-raised children (6;4-10;4) took part in a visual-world eye-tracking task and an elicited production task.

In the eye-tracking task, children saw pairs of pictures while listening to a which-question and had to choose the picture that matched the question. Eye-movements to 32 experimental trials and offline responses were recorded. In the elicited production task, children had to produce 32 subject and object which-questions. Our findings indicate that DOM on the wh-pronoun in Romanian does not eliminate the subject-object asymmetry found with which-questions in children. At a group level, HL children have more difficulties with the comprehension and production of object which-questions. Although number mismatch does not impact offline comprehension in either group, the online data suggest that this mismatch guides monolingual children’s online processing of object-which questions more than DOM-marking on its own. The fact that the heritage children do not seem to recruit the number information in the online processing of a short syntactic dependency could indicate more protracted processing, i.e., the number information is processed after the end of the sentence because of slower processing speed.”

07/19/2024, from 11:00 AM to 01:00 PM , Room P300

Symposium: Where do new words come from?

Speakers: Kristen Gilyard; Kennedy Casey; Samantha Durrant; Yuchen Jin; Elika Bergelson

Children’s vocabulary size and content is sensitive to their input environments—the frequency of words in directed, interactive input is a strong predictor of age of acquisition. But children’s lexical knowledge is expanded and supported by a wide variety of other mechanisms beyond linguistic input frequency. In four talks on two languages (English, Tseltal) from three linguistic regions (Eastern US, Central UK, Southern Mexico), we explore how early lexical development is guided by everyday routines (bookreading), embodied experience (object handling), inferential meaning-mapping (novel word learning), and overhearable (i.e., non-child-directed) input. Presentation 1 describes a novel noun learning study in which young US English-learning toddlers (14–22mo) are exposed to new words via home bookreading. Both age and book familiarity play a role in their engagement and word learning—by 22 months toddlers repeat most novel words and, as they become more familiar with the books, vocally engage more. Presentation 2 uses a unique daylong photo stream corpus and parent-reported productive vocabulary data from Tseltal (Mayan; Mexico) children, showing that children’s at-home manual activity reliably predicts the age at which children are typically reported to produce a word. Presentation 3 dives into real-time processing, testing how and when UK English-learning toddlers (18–30mo) identify, retain, and generalize object-label mappings discovered on the basis of disambiguation. Disambiguation ability was present already at 18 months (improving with age) and predicted both concurrent and future parent-reported productive vocabulary size. Presentation 4 combines daylong transcription data, age of acquisition estimates, and parent questionnaires (US English-learning children, 6–42mo) to examine which household object names are distinctly associated with overhearable input, opening opportunities for future work to test word recognition for primarily-directed versus -overheard words when controlling for frequency and interest. Together, these studies illustrate how new words enter children’s worlds and how real-time, multimodal processing supports word learning.

07/19/2024, from 11:00 AM to 01:00 PM , Room P200

Symposium: Children’s multimodal development across approaches, methods and languages

Speakers: Aliyah Morgenstern; Pauline Beaupoil-Hourdel; Amanda Bateman; Sara Coego; Maria Graziano; Pilar Prieto

The panel presents how research in our field has developed theories and methods to analyze multimodal productions, using both naturalistic and experimental data. All participants discuss the necessity to use a variety of approaches.

A developmental perspective on multimodal negation in French speaking children

Authors: Pauline Beaupoil-Hourdel; Christelle Dodane, Aliyah Morgenstern

“First negative constructions take over from early forms of rejection and avoidance. From the end of their first year on, children can express negation with headshakes, palm-ups and index waves. Prosody and gestures are combined to convey refusals, protests, epistemic negations or powerlessness, sometimes before the emergence of first verbal negation markers. The expression of negation in longitudinal adult-child data is thus an excellent source to study multimodal language development. But tracing the transitions between modalities is quite complex. It is therefore crucial to analyze gesture and speech with an integrative approach.

We used ecological spontaneous data and mixed methods to analyze the longitudinal development of four French monolingual children recorded monthly at home with a parent for one hour between 1;0 and 4;0. We focused on the productions containing the word “non” (no). Prosodic properties were coded with PRAAT, actional and gestural behavior with ELAN, functional analyses as well as directional and temporal synchronization patterns were integrated in EXCEL.

Results of quantitative and qualitative analyses show clear evolutions in 3 phases. During phase 1, vocal productions of “non” were amplified at the prosodic level. Body movements were synchronized with intonation contours. During phase 2 “non” was mainly produced with rise-fall intonation contours and children used upper-body movements in close parallel with their prosodic contours. During phase 3, “non” was produced with flat or falling intonation contours and reduced syllabic duration. As their mastery of speech developed, children gradually resorted to more subtle multifunctional gesture-speech orchestration.

Gestures, body movements and prosody thus provide powerful resources for children to make their multimodal entry into language. They use each modality more and more skillfully thanks to adults’ scaffolding in everyday life interactions. This study gives us insights on how children become experts in face-to-face social interaction, which is multimodal in nature.”

Multimodal recruitments and offers of help in infant early childhood education mealtimes

Authors: Amanda Bateman; Amelia Church

“This presentation explores children’s multimodal development from the perspective of ethnomethodology and conversation analysis (EMCA), which allows us to see how children use their bodies and voices to achieve particular social actions. Building on prior work exploring the social and linguistic practices in infant eating interactions (e.g. Wiggins 2023), we explore how mealtime interactions are managed as a collaborative, co-constructed activity between infants and early childhood teachers.

To illustrate how children achieve social actions – such as requests for assistance – we analyzed video data recently collected in an early childhood English-speaking setting in Mid-Wales. Our findings show that infant embodied displays of trouble communicate recruitment for assistance from the teacher, who responds with offers of assistance through gesture and by verbalising the infant’s actions, therefore modelling language use. Importantly the children’s actions are contingent on the educator responding in alignment with the child’s intention.

Micro-analyses of the video data illustrate how infants recruit help through embodied ‘showing’ of an item (Kidwell & Zimmerman, 2007), which sets up a problem (as a first pair part) that needs a response from the teacher (as a second pair part), establishing joint attention. The teacher offers assistance in the shape of ‘would you like me to [x]’, and uses accompanying embodied gestures that uphold infant children’s rights to choose, at their pace, when/if they want to receive help.

This presentation focuses on the interactional function of children’s multimodality, how they use their bodies, gaze and gesture to enact agency and achieve particular social actions, such as requests. It also highlights teachers’ multimodal scaffolding and reformulations. Implications for early childhood teacher practice and training, and the usefulness of an EMCA approach for analysing teacher-infant multimodal communication for future research will conclude the presentation.”

Prosody and gesture to express focus types: a cross-sectional study using task-based measures

Authors: Sara Coego; Nuria Esteve-Gibert, Pilar Prieto

In adult speech, prosodic prominence and co-speech gestures are relevant cues to distinguish focus types (information, contrastive, corrective), and they are frequently combined into multimodal ensembles. Despite the contribution of both modalities to mark focus type, developmental research has mainly studied children’s use of prosody independently from their gesture production. Esteve-Gibert and colleagues (2021) considered the use of both cues in distinguishing focus types and point towards a precursor role of head gestures in French 4- to 5-year-olds’ speech. The present study has two main goals. First, to determine whether the seemingly precursor role of gestures in the marking of focus types can be replicated in earlier stages of acquisition by looking at a larger developmental window than in Esteve-Gibert and colleagues’ study. Second, to investigate how the use of prosodic prominence and gestures interact in the creation of such complex pragmatic meanings as contrast and correction across age groups. A total of 120 Catalan-speaking children belonging to three age groups (3-4, 4-5, and 5-6) were video-recorded during an interaction-based semi-controlled task designed to elicit utterances in three focus contexts (information, contrast, and correction). Target productions were coded for perceived prosodic prominence using an adapted version of the DIMA system (Kügler et al., 2015). Visual data was coded for gesture presence. Preliminary results with 27 children showed, first, that both prosodic prominent patterns and gestures were used to differentiate focus types already at ages 3-4, showing no evidence of the precursor role of gesture. Words in contrastive and corrective focus contexts were prosodically more prominent and they overlapped more frequently with a gesture at all age groups. Secondly, a tendency to combine prosodically prominent words with gestures was observed across all ages and focus contexts. This study highlights how gesture and prosody must be analyzed together in child language development.

On the development of reference tracking in speech and gesture in Italian children’s narratives

Authors: Maria Graziano;

“Reference tracking (who does what to whom) is a multimodal phenomenon – speakers can use both verbal referential expressions and gestures to track entities in discourse.

Studies have shown that the development of referential expressions can be influenced by several factors, such as language specificity and contextual constraints (De Weck, 1991; Hickmann 2003). Similarly, the ability to use gestures with different functions is linked to the development of discourse and pragmatic skills (Colletta, 2004; Graziano, 2009). The few studies looking at children’s use of anaphoric gestures have produced divergent results, arguing either for a disambiguating function (Cristilli et al., 2010), or a more general late acquisition (Alamillo et al., 2010).

This study investigates the development of reference tracking in a pro-drop language (Italian) to observe whether and how gestures contribute to achieve referential cohesion in a narrative discourse. Analyses are conducted on elicited narratives produced by four-, six-, and nine-year-olds (11 per group). We examine how referents are introduced, maintained and re-introduced, and which type of anaphoric expression is used (pronouns, nouns and zero anaphora). All gestures identified are coded for function (referential vs. pragmatic; Kendon, 2004) and examined for whether they co-occur with an anaphoric expression.

Preliminary observations indicate that, in speech, four-year-olds tend to use more ellipsis than older children, where six- and nine-year-olds use more nouns phrases. The use of gestures with anaphoric functions appears in 6-years olds, in particular with abstract deictic gestures. Those results are in line with previous studies, indicating a late mastery of anaphoric devices (both in speech and in gesture). The present multimodal study of gesture alignment with different anaphoric expressions will shed light on whether and how the use of anaphoric gestures can depend on language-specific aspects, such as the possibility of dropping the subject which is a characteristic of Italian.”