Meg Cychosz
Research Subjects
Research Interests
Meg Cychosz has a total of 79 co-authors. Here are the top co-authors by number of shared publications:
Alejandrina Cristia
8 shared publicationsCamila Scaff
6 shared publications- Jan Edwards
Department of Hearing and Speech Sciences, University of Maryland, College Park
6 shared publications - Rochelle Suzanne Newman
University of Maryland
6 shared publications - Benjamin Munson
Department of Speech-Language-Hearing Sciences, University of Minnesota, Twin Cities, Minneapolis
4 shared publications Marisa Casillas
4 shared publications
Latest Publications
- Open Access
- 6 November 2026
- Published by Center for Open Science in PsyArXiv
Multiparty conversation predominates in children's language environments across diverse cultural settings
Little is known about the dynamics of children’s conversations outside of urban Western contexts, where research focuses on one-on-one interactions with parents. Our study addresses this gap, aggregating naturalistic speech data gathered in 10 countries across multiple labs, and using novel speech diarization algorithms to extract and classify different types of conversation. We analyzed 23,000 hours of child-centered audio recordings taken from 844 children (mean age 23 months; 389 girls) across urban, agricultural, and foraging societies. We found that multiparty conversations, involving at least three speakers, are more frequent than any other type of conversation. Across settings, children’s relative exposure to conversation differed based on their age, number of siblings, and mother’s education. We conclude that the kinds of conversation most often studied differ from those children typically encounter. Greater focus on children’s conversations in multiparty contexts may provide a more representative picture of their daily experience with language.
- 6 November 2026
- Published by Center for Open Science in PsyArXiv
Open AccessMultiparty conversation predominates in children's language environments across diverse cultural settings
Little is known about the dynamics of children’s conversations outside of urban Western contexts, where research focuses on one-on-one interactions with parents. Our study addresses this gap, aggregating naturalistic speech data gathered in 10 countries across multiple labs, and using novel speech diarization algorithms to extract and classify different types of conversation. We analyzed 23,000 hours of child-centered audio recordings taken from 844 children (mean age 23 months; 389 girls) across urban, agricultural, and foraging societies. We found that multiparty conversations, involving at least three speakers, are more frequent than any other type of conversation. Across settings, children’s relative exposure to conversation differed based on their age, number of siblings, and mother’s education. We conclude that the kinds of conversation most often studied differ from those children typically encounter. Greater focus on children’s conversations in multiparty contexts may provide a more representative picture of their daily experience with language.
- Open Access
- 1 October 2026
- Published by Center for Open Science in PsyArXiv
Beyond speed and beyond prediction: The multi-faceted nature of speech processing in preschoolers and how it relates to language development
Efficient language processing requires both prediction and flexible revision. Studies show young children can predict during online language processing, but less is known about their ability to quickly revise interpretations when initial predictions are infelicitous. We examined four- andfive-year-olds’ (N=60) phonological revision using an eye-tracking paradigm that captured children’s initial interpretation of a word and their subsequent revision and recovery. On average, word mispronunciations slowed children’s looks to the target object. But that average hid different responses: some children initially treated the mispronounced word as a new word before settling on the target one, while others treated it as the target word from the start. Standard analyses of children’s looking data often capture how quickly and completely looks converge on a target, not a trajectory that involves hypothesis and revision. We therefore characterized each child’s looking trajectory along numerous dimensions spanning mispronunciation sensitivity, rejection, recovery, and word-recognition efficiency. Children whose processing profiles showed the least rejection and revision had modestly lower vocabulary and phonological awareness scores, consistent with Bayesian inference and error-driven learning accounts in which the probabilistic prediction and resolution of mismatches may drive language development. Overall, preschoolers vary widely in how flexibly, not just how quickly, they interpret unexpected pronunciations, with implications for language development.
- 1 October 2026
- Published by Center for Open Science in PsyArXiv
Open AccessBeyond speed and beyond prediction: The multi-faceted nature of speech processing in preschoolers and how it relates to language development
Efficient language processing requires both prediction and flexible revision. Studies show young children can predict during online language processing, but less is known about their ability to quickly revise interpretations when initial predictions are infelicitous. We examined four- andfive-year-olds’ (N=60) phonological revision using an eye-tracking paradigm that captured children’s initial interpretation of a word and their subsequent revision and recovery. On average, word mispronunciations slowed children’s looks to the target object. But that average hid different responses: some children initially treated the mispronounced word as a new word before settling on the target one, while others treated it as the target word from the start. Standard analyses of children’s looking data often capture how quickly and completely looks converge on a target, not a trajectory that involves hypothesis and revision. We therefore characterized each child’s looking trajectory along numerous dimensions spanning mispronunciation sensitivity, rejection, recovery, and word-recognition efficiency. Children whose processing profiles showed the least rejection and revision had modestly lower vocabulary and phonological awareness scores, consistent with Bayesian inference and error-driven learning accounts in which the probabilistic prediction and resolution of mismatches may drive language development. Overall, preschoolers vary widely in how flexibly, not just how quickly, they interpret unexpected pronunciations, with implications for language development.
- Open Access
- 13 September 2026
- Published by Center for Open Science in PsyArXiv
Speech Maturity Dataset: A cross-cultural corpus of naturalistic child and adult vocalizations
Over the first years of life, children’s vocalizations become increasingly adult-like, and laythe foundation for later phonetic and phonological development. Yet research in this area hasbeen limited to a narrow set of languages and communities, mainly Indo-European languagesfrom Western(ized) speaker communities, and focused on a narrow age range (0-24mo). Wepresent a new publicly-available dataset, the Speech Maturity Dataset, consisting of over 1million labels, provided by over 20,000 unique annotators, of ~250,000 clips from long-formrecordings of over 400 children (aged 1mo-6yrs) from 10 communities (ranging from richindustrialized societies to farmer-forager communities) speaking one or more of 41 languages.The clips include both (i) key child vocalizations labeled with their vocalization type (canonicalvocalization, containing an adjacent consonant and vowel; non-canonical vocalization, with noadjacent consonant/vowel; laughing; or crying) as well as (ii) vocalizations from other speakersin the environment which are additionally labeled with speaker information (baby, child,female/male adolescent, female/male adult). This dataset, which includes child-level metadata(sex, age, monolingual status, diagnoses, etc.), can be used to study vocal development in anunprecedented way across a wide variety of communities. It can also be used to trainvocalization-type classifiers in an effort to make software dedicated to the study of child languageacquisition free, open-source, and reproducible. In sum, our dataset represents an ongoing andcollaborative effort between field researchers, psycholinguists, and citizen scientists, andpromises to allow scientists to significantly expand their knowledge of early vocal development.
- 13 September 2026
- Published by Center for Open Science in PsyArXiv
Open AccessSpeech Maturity Dataset: A cross-cultural corpus of naturalistic child and adult vocalizations
Over the first years of life, children’s vocalizations become increasingly adult-like, and laythe foundation for later phonetic and phonological development. Yet research in this area hasbeen limited to a narrow set of languages and communities, mainly Indo-European languagesfrom Western(ized) speaker communities, and focused on a narrow age range (0-24mo). Wepresent a new publicly-available dataset, the Speech Maturity Dataset, consisting of over 1million labels, provided by over 20,000 unique annotators, of ~250,000 clips from long-formrecordings of over 400 children (aged 1mo-6yrs) from 10 communities (ranging from richindustrialized societies to farmer-forager communities) speaking one or more of 41 languages.The clips include both (i) key child vocalizations labeled with their vocalization type (canonicalvocalization, containing an adjacent consonant and vowel; non-canonical vocalization, with noadjacent consonant/vowel; laughing; or crying) as well as (ii) vocalizations from other speakersin the environment which are additionally labeled with speaker information (baby, child,female/male adolescent, female/male adult). This dataset, which includes child-level metadata(sex, age, monolingual status, diagnoses, etc.), can be used to study vocal development in anunprecedented way across a wide variety of communities. It can also be used to trainvocalization-type classifiers in an effort to make software dedicated to the study of child languageacquisition free, open-source, and reproducible. In sum, our dataset represents an ongoing andcollaborative effort between field researchers, psycholinguists, and citizen scientists, andpromises to allow scientists to significantly expand their knowledge of early vocal development.
- Open Access
- 30 June 2026
- Published by Center for Open Science in PsyArXiv
Syllables, not words, minutes, or utterances, to accurately characterize verbal input to children
The verbal input that children hear from caregivers around them can explain some of the well-known variability in trajectories of early child language development. Verbal input has traditionally been quantified by counting words, minutes, or utterances of child-directed speech from an adult interlocutor. This paper argues that exclusive reliance on these units can result in inaccurate estimation of verbal input, especially when comparisons are made between different languages, sociocultural settings, and speakers. Using spontaneous speech data from children and adults, we illustrate some confounds that these units create. Though illustrated with children’s verbal input, these confounds apply to any quantification of language structure across speakers, registers, or languages. We show that words, which vary in internal grammatical complexity between and even within speakers and individual languages, result in cross-linguistic and speaker age confounds. Utterances are likewise confounded with typological differences and speaker age. Minutes, which appear to be language-neutral, are in fact confounded with speaking rate which similarly varies within and between speakers, languages, and registers, especially since children speak slower than adults. We put forth measures based instead in information theory that employ the syllable as the base unit to resolve many of these issues: total information transmitted (syllable count x information density per syllable) and information rate (information density x syllable rate). These results demonstrate that accurate characterization of children’s verbal input requires paying close attention to linguistic structure and speaker attributes, and that information-theoretic measures can provide more reliable estimates across languages, speakers, and registers.
- 30 June 2026
- Published by Center for Open Science in PsyArXiv
Open AccessSyllables, not words, minutes, or utterances, to accurately characterize verbal input to children
The verbal input that children hear from caregivers around them can explain some of the well-known variability in trajectories of early child language development. Verbal input has traditionally been quantified by counting words, minutes, or utterances of child-directed speech from an adult interlocutor. This paper argues that exclusive reliance on these units can result in inaccurate estimation of verbal input, especially when comparisons are made between different languages, sociocultural settings, and speakers. Using spontaneous speech data from children and adults, we illustrate some confounds that these units create. Though illustrated with children’s verbal input, these confounds apply to any quantification of language structure across speakers, registers, or languages. We show that words, which vary in internal grammatical complexity between and even within speakers and individual languages, result in cross-linguistic and speaker age confounds. Utterances are likewise confounded with typological differences and speaker age. Minutes, which appear to be language-neutral, are in fact confounded with speaking rate which similarly varies within and between speakers, languages, and registers, especially since children speak slower than adults. We put forth measures based instead in information theory that employ the syllable as the base unit to resolve many of these issues: total information transmitted (syllable count x information density per syllable) and information rate (information density x syllable rate). These results demonstrate that accurate characterization of children’s verbal input requires paying close attention to linguistic structure and speaker attributes, and that information-theoretic measures can provide more reliable estimates across languages, speakers, and registers.
- Open Access
- 30 November 2025
- Published by Center for Open Science in PsyArXiv
Child-directed speech and infant vocal development in rural, highland Bolivia and immigrant families in urban United States
Decades of research have established links between speech input and children’s vocabulary growth. However, it is unclear if input also facilitates phonological development. Phonology has a strong biological component—–implicating fine motor development—–so language socialization factors like speech input may matter less. Further complicating matters, children in many cultures are exposed to vastly different amounts of speech input, and yet these children still reach many major language development milestones on age-appropriate timelines. How does speech input relate to infant phonological development in the first years of life? We estimated infants’ (1) speech input and (2) vocal maturity using daylong audio recordings taken in an Indigenous Quechua- and Spanish-speaking community in Bolivia (n=10, M age=12 months, 5 females, 5 males) and an immigrant Spanish- and English-speaking community in the U.S. (n=10, M age=9 months, 4 females, 6 males; all Hispanic or Latino). Although we found no differences in the overall amount of total speech input between communities, infants in the U.S. were 2.5x more likely to hear speech directed to them than the infants in Bolivia. When child-directed speech was instead characterized to include all speech directed to any child within the infants’ vicinity, there were no differences between communities. When employing these different definitions of child-directed speech, we found positive relationships between quantity of speech input and different metrics of the Bolivian infants’ vocal maturity. These results paint a nuanced picture showing that directedspeech input, even when less common, is related to early precursors of phonological development, and that by expanding the definition of speech input to accommodate diverse cultural settings we can understand how infants’ language development is resilient to differences in speech input.
- 30 November 2025
- Published by Center for Open Science in PsyArXiv
Open AccessChild-directed speech and infant vocal development in rural, highland Bolivia and immigrant families in urban United States
Decades of research have established links between speech input and children’s vocabulary growth. However, it is unclear if input also facilitates phonological development. Phonology has a strong biological component—–implicating fine motor development—–so language socialization factors like speech input may matter less. Further complicating matters, children in many cultures are exposed to vastly different amounts of speech input, and yet these children still reach many major language development milestones on age-appropriate timelines. How does speech input relate to infant phonological development in the first years of life? We estimated infants’ (1) speech input and (2) vocal maturity using daylong audio recordings taken in an Indigenous Quechua- and Spanish-speaking community in Bolivia (n=10, M age=12 months, 5 females, 5 males) and an immigrant Spanish- and English-speaking community in the U.S. (n=10, M age=9 months, 4 females, 6 males; all Hispanic or Latino). Although we found no differences in the overall amount of total speech input between communities, infants in the U.S. were 2.5x more likely to hear speech directed to them than the infants in Bolivia. When child-directed speech was instead characterized to include all speech directed to any child within the infants’ vicinity, there were no differences between communities. When employing these different definitions of child-directed speech, we found positive relationships between quantity of speech input and different metrics of the Bolivian infants’ vocal maturity. These results paint a nuanced picture showing that directedspeech input, even when less common, is related to early precursors of phonological development, and that by expanding the definition of speech input to accommodate diverse cultural settings we can understand how infants’ language development is resilient to differences in speech input.
- Open Access
- 19 November 2025
- Published by Center for Open Science in PsyArXiv
Connecting preschoolers’ spontaneous speech patterns to future language skills: A three-year concurrent and longitudinal cohort study of canonical proportion as a developmental index
Purpose: Understanding how children’s spontaneous language behavior relates to standardizedmetrics of speech and language remains a crucial challenge in studies of language acquisition andclinical assessment. Traditional lab- and clinic-based paradigms are time- and resource-intensive,and may not fully capture a child’s underlying language competencies. This study aimed toinvestigate whether automated analysis of naturalistic, child-centered audio recordings can indexthe developmental trajectory of speech-language abilities from ages 3 to 5 years.Methods: A longitudinal design was employed with N=155 preschoolers, who were followedfrom age 3 to 5 years. Deep learning speech classification methods were used to computecanonical proportion—the proportion of a child’s speech that is produced as canonical syllables(well-formed syllables with a consonant–vowel transition, such as “ba”), a key marker of speechmotor control development and phonological representation building—from child-centered audiorecordings collected at ages 3 and 4. Standardized lab-based assessments of speech and languagedevelopment were administered one year later at ages 4 and 5.Results: Canonical proportion measures significantly predicted multiple dimensions of laterspeech-language development. The strongest longitudinal associations were observed forconsonant articulation skill and vocabulary size. Weaker, though still significant, relationshipswere found for meta-phonological skills, including phonological awareness and phonologicalworking memory.Conclusion: Findings suggest that children’s spontaneous, everyday speech productionpatterns provide moderate indices of subsequent language development across multiple domains.Automated measures such as canonical proportion hold promise for expanding and diversifyingapproaches to studying language acquisition and for complementing traditional lab-basedassessments.
- 19 November 2025
- Published by Center for Open Science in PsyArXiv
Open AccessConnecting preschoolers’ spontaneous speech patterns to future language skills: A three-year concurrent and longitudinal cohort study of canonical proportion as a developmental index
Purpose: Understanding how children’s spontaneous language behavior relates to standardizedmetrics of speech and language remains a crucial challenge in studies of language acquisition andclinical assessment. Traditional lab- and clinic-based paradigms are time- and resource-intensive,and may not fully capture a child’s underlying language competencies. This study aimed toinvestigate whether automated analysis of naturalistic, child-centered audio recordings can indexthe developmental trajectory of speech-language abilities from ages 3 to 5 years.Methods: A longitudinal design was employed with N=155 preschoolers, who were followedfrom age 3 to 5 years. Deep learning speech classification methods were used to computecanonical proportion—the proportion of a child’s speech that is produced as canonical syllables(well-formed syllables with a consonant–vowel transition, such as “ba”), a key marker of speechmotor control development and phonological representation building—from child-centered audiorecordings collected at ages 3 and 4. Standardized lab-based assessments of speech and languagedevelopment were administered one year later at ages 4 and 5.Results: Canonical proportion measures significantly predicted multiple dimensions of laterspeech-language development. The strongest longitudinal associations were observed forconsonant articulation skill and vocabulary size. Weaker, though still significant, relationshipswere found for meta-phonological skills, including phonological awareness and phonologicalworking memory.Conclusion: Findings suggest that children’s spontaneous, everyday speech productionpatterns provide moderate indices of subsequent language development across multiple domains.Automated measures such as canonical proportion hold promise for expanding and diversifyingapproaches to studying language acquisition and for complementing traditional lab-basedassessments.
- Open Access
- 15 April 2025
- Published by Center for Open Science in OSF Preprints
Canonical Babbling Development in Infancy: A Systematic Review and Meta-Analysis of Methodological and Ambient Language Influences
Canonical babbling, or the production of adult-like consonant-vowel syllables in infancy,represents a critical milestone in pre-linguistic vocal development and predicts later speech andlanguage outcomes. This systematic review and meta-analysis synthesized findings from 43studies and 1,291 infants aged 5-24 months across 16 language environments to examine howmethodological and contextual factors influence the most common measure of canonical babbling:the canonical babbling ratio (CBR). Results confirmed a robust, linear increase in CBR with age,reinforcing its role as a consistent developmental marker. Different CBR measures yieldedcomparable developmental trajectories. Sampling method significantly affected CBR values, withinteractive free play sessions eliciting higher CBRs than naturalistic home recordings, particularlyin older infants. In contrast, the location of data collection had no effect. Ambient languagecomplexity also shaped CBR: infants acquiring languages with more complex syllable structures(e.g., English, Dutch) initially exhibited lower CBRs compared to infants acquiring languageswith less complex syllable structures (e.g., Mandarin, Spanish). Despite these initial differences,babbling trajectories were predicted to converge by approximately 21 months as infants exposedto languages with more complex syllables demonstrated accelerated CBR growth. Publicationbias was detected, with smaller samples more likely to report inflated CBRs. To address thisissue, simulation-based analyses are reported to estimate sample size recommendations forimproved precision in future research. Together, these results support CBR as a meaningfuldevelopmental marker while offering practical guidance for future research directions andcontinued clinical applications.
- 15 April 2025
- Published by Center for Open Science in OSF Preprints
Open AccessCanonical Babbling Development in Infancy: A Systematic Review and Meta-Analysis of Methodological and Ambient Language Influences
Canonical babbling, or the production of adult-like consonant-vowel syllables in infancy,represents a critical milestone in pre-linguistic vocal development and predicts later speech andlanguage outcomes. This systematic review and meta-analysis synthesized findings from 43studies and 1,291 infants aged 5-24 months across 16 language environments to examine howmethodological and contextual factors influence the most common measure of canonical babbling:the canonical babbling ratio (CBR). Results confirmed a robust, linear increase in CBR with age,reinforcing its role as a consistent developmental marker. Different CBR measures yieldedcomparable developmental trajectories. Sampling method significantly affected CBR values, withinteractive free play sessions eliciting higher CBRs than naturalistic home recordings, particularlyin older infants. In contrast, the location of data collection had no effect. Ambient languagecomplexity also shaped CBR: infants acquiring languages with more complex syllable structures(e.g., English, Dutch) initially exhibited lower CBRs compared to infants acquiring languageswith less complex syllable structures (e.g., Mandarin, Spanish). Despite these initial differences,babbling trajectories were predicted to converge by approximately 21 months as infants exposedto languages with more complex syllables demonstrated accelerated CBR growth. Publicationbias was detected, with smaller samples more likely to report inflated CBRs. To address thisissue, simulation-based analyses are reported to estimate sample size recommendations forimproved precision in future research. Together, these results support CBR as a meaningfuldevelopmental marker while offering practical guidance for future research directions andcontinued clinical applications.
- Open Access
- 3 July 2024
- Published by Center for Open Science in PsyArXiv
Establishing the reliability of metrics extracted from long-form recordings using LENA and the ACLEW pipeline
Long-form audio recordings are increasingly used to study individual variation, group differences, and many other topics in theoretical and applied fields of developmental science, particularly for the description of children's language input (typically speech from adults) and children’s language output (ranging from babble to sentences). The proprietary LENA software has been available for over a decade, and with it, users have come to rely on derived metrics like adult word count (AWC) and child vocalization counts (CVC), which have also more recently been derived using an open-source alternative, the ACLEW pipeline. Yet, there is relatively little work assessing the reliability of long-form metrics in terms of the stability of individual differences across time. Filling this gap, we analyzed eight spoken-language datasets: four from North American English-learning infants, and one each from British English-, French-, American English-Spanish, and Quechua-Spanish-learning infants. The audio data were analyzed using two types of processing software: LENA and the ACLEW open-source pipeline. When all corpora were included, we found relatively low to moderate reliability (across multiple recordings, intraclass correlation coefficient attributed to the child identity (Child ICC), was <50% for most metrics). There were few differences between the two pipelines. Exploratory analyses suggested some differences as a function of child age and corpora. These findings suggest that, while reliability is likely sufficient for various group-level analyses, caution is needed when using either LENA or ACLEW tools to study individual variation. We also encourage improvement of extant tools, specifically targeting accurate measurement of individual variation.
- 3 July 2024
- Published by Center for Open Science in PsyArXiv
Open AccessEstablishing the reliability of metrics extracted from long-form recordings using LENA and the ACLEW pipeline
Long-form audio recordings are increasingly used to study individual variation, group differences, and many other topics in theoretical and applied fields of developmental science, particularly for the description of children's language input (typically speech from adults) and children’s language output (ranging from babble to sentences). The proprietary LENA software has been available for over a decade, and with it, users have come to rely on derived metrics like adult word count (AWC) and child vocalization counts (CVC), which have also more recently been derived using an open-source alternative, the ACLEW pipeline. Yet, there is relatively little work assessing the reliability of long-form metrics in terms of the stability of individual differences across time. Filling this gap, we analyzed eight spoken-language datasets: four from North American English-learning infants, and one each from British English-, French-, American English-Spanish, and Quechua-Spanish-learning infants. The audio data were analyzed using two types of processing software: LENA and the ACLEW open-source pipeline. When all corpora were included, we found relatively low to moderate reliability (across multiple recordings, intraclass correlation coefficient attributed to the child identity (Child ICC), was <50% for most metrics). There were few differences between the two pipelines. Exploratory analyses suggested some differences as a function of child age and corpora. These findings suggest that, while reliability is likely sufficient for various group-level analyses, caution is needed when using either LENA or ACLEW tools to study individual variation. We also encourage improvement of extant tools, specifically targeting accurate measurement of individual variation.
- Open Access
- 19 June 2024
- Published by Center for Open Science in OSF Preprints
Bilingual language input to infants in Bolivia and the United States
Characterizing dual language input in children’s environments is critical to understand how early language experiences influence bilingual language development. However, little is known about how dual language exposure is distributed across factors known to influence the kinds of input children receive. This study examined how infants’ exposure to each of their languages is distributed across different speakers (adults versus other children) and speech registers (child- versus adult-directed speech). We examined daylong audio recordings of infants’ language environments in two bilingual communities: an indigenous Quechua- and Spanish-speaking community in Bolivia (n=10, M age = 9 months, 5 females, 5 males) and an immigrant Spanish- and English-speaking community in the United States (n=10, M age = 9 months, 4 females, 6 males). Infants in both communities were more likely to hear the societal language from older children than from adult caregivers. Infants were also more likely to hear the societal language in child-directed speech, and more of the minoritized language in adult-directed speech, by a factor of more than 4 to 1. Thus, infants in both cultures hear each of their languages in contexts that do, and do not, facilitate minoritized language transmission. These findings have implications for bilingual language development and maintenance, as distinct social contexts afford different opportunities for engagement and interaction over the course of learning two languages in infancy.
- 19 June 2024
- Published by Center for Open Science in OSF Preprints
Open AccessBilingual language input to infants in Bolivia and the United States
Characterizing dual language input in children’s environments is critical to understand how early language experiences influence bilingual language development. However, little is known about how dual language exposure is distributed across factors known to influence the kinds of input children receive. This study examined how infants’ exposure to each of their languages is distributed across different speakers (adults versus other children) and speech registers (child- versus adult-directed speech). We examined daylong audio recordings of infants’ language environments in two bilingual communities: an indigenous Quechua- and Spanish-speaking community in Bolivia (n=10, M age = 9 months, 5 females, 5 males) and an immigrant Spanish- and English-speaking community in the United States (n=10, M age = 9 months, 4 females, 6 males). Infants in both communities were more likely to hear the societal language from older children than from adult caregivers. Infants were also more likely to hear the societal language in child-directed speech, and more of the minoritized language in adult-directed speech, by a factor of more than 4 to 1. Thus, infants in both cultures hear each of their languages in contexts that do, and do not, facilitate minoritized language transmission. These findings have implications for bilingual language development and maintenance, as distinct social contexts afford different opportunities for engagement and interaction over the course of learning two languages in infancy.
- Open Access
- 2 February 2024
- Published by Center for Open Science in PsyArXiv
Bursty, irregular speech input to children predicts vocabulary size
Children learn language by listening to speech from caregivers around them. However, the type and quantity of speech input that children are exposed to changes throughout early childhood in ways that are poorly understood due to the small samples (few participants, limited hours of observation) typically available in developmental psychology. Here we used child-centered audio recorders to unobtrusively measure speech input in the home to 292 children, aged 2-7 years, over 555 distinct days (approximately 8,600 total hours of observation, or 29.62 hours/child). These large timescales allowed us to compare how different dimensions of child-directed speech input (quantity, burstiness) varied throughout early childhood. We then evaluated the relationship between each dimension of input and children's concurrent receptive vocabulary size. We found that the burstiness of speech input (spikes of words) was a stronger correlate with age than the quantity of speech input. Input burstiness was also a stronger predictor than input quantity for children's vocabulary size: children who heard spiky, more intense bouts of input had larger vocabularies. Overall, these results reaffirm the importance of speech input in the home for children's language development, and support exposure-consolidation models of early language development.
- 2 February 2024
- Published by Center for Open Science in PsyArXiv
Open AccessBursty, irregular speech input to children predicts vocabulary size
Children learn language by listening to speech from caregivers around them. However, the type and quantity of speech input that children are exposed to changes throughout early childhood in ways that are poorly understood due to the small samples (few participants, limited hours of observation) typically available in developmental psychology. Here we used child-centered audio recorders to unobtrusively measure speech input in the home to 292 children, aged 2-7 years, over 555 distinct days (approximately 8,600 total hours of observation, or 29.62 hours/child). These large timescales allowed us to compare how different dimensions of child-directed speech input (quantity, burstiness) varied throughout early childhood. We then evaluated the relationship between each dimension of input and children's concurrent receptive vocabulary size. We found that the burstiness of speech input (spikes of words) was a stronger correlate with age than the quantity of speech input. Input burstiness was also a stronger predictor than input quantity for children's vocabulary size: children who heard spiky, more intense bouts of input had larger vocabularies. Overall, these results reaffirm the importance of speech input in the home for children's language development, and support exposure-consolidation models of early language development.