Interaction structure constrains the emergence of conventions in group communication.

Real-world communication frequently requires language producers to address more than one comprehender at once, yet most psycholinguistic research focuses on one-on-one communication. As the audience size grows, interlocutors face new challenges that do not arise in dyads. They must consider multiple perspectives and weigh multiple sources of feedback to build shared understanding. Here, we ask which properties of the group's interaction structure facilitate successful communication. We used a repeated reference game paradigm in which directors instructed between one and five matchers to choose specific targets out of a set of abstract figures. Across 313 games (N = 1,319 participants), we manipulated several key constraints on the group's interaction, including the amount of feedback that matchers could give to directors and the availability of peer interaction between matchers. Across groups of different sizes and interaction constraints, describers produced increasingly efficient utterances and matchers made increasingly accurate selections. Critically, however, we found that smaller groups and groups with less-constrained interaction structures ("thick channels") showed stronger convergence to group-specific conventions than large groups with constrained interaction structures ("thin channels"), which struggled with convention formation. Overall, these results shed light on the core structural factors that enable communication to thrive in larger groups.

The role of translation equivalents in bilingual word learning.

Bilingual environments present an important context for word learning. One feature of bilingual environments is the existence of translation equivalents (TEs)-words in different languages that share similar meanings. Documenting TE learning over development may give us insight into the mechanisms underlying word learning in young bilingual children. Prior studies of TE learning have often been confounded by the fact that increases in overall vocabulary size with age lead to greater opportunities for learning TEs. To address this confound, we employed an item-level analysis, which controls for the age trajectory of each item independently. We used Communicative Development Inventory data from four bilingual datasets (two English-Spanish and two English-French; total N = 419) for modeling. Results indicated that knowing a word's TE increased the likelihood of knowing that word for younger children and for TEs that are more similar phonologically. These effects were consistent across datasets, but varied across lexical categories. Thus, TEs may allow bilingual children to bootstrap their early word learning in one language using their knowledge of the other language. RESEARCH HIGHLIGHTS: Bilingual children must learn words that share a common meaning across both languages, that is, translation equivalents, like dog in English and perro in Spanish. Item-level models explored how translation equivalents affect word learning, in addition to child-level (e.g., exposure) and item-level (e.g., phonological similarity) factors. Knowing a word increased the probability of knowing its corresponding translation equivalent, particularly for younger children and for more phonologically-similar translation equivalents. These findings suggest that young bilingual children use their word knowledge in one language to bootstrap their learning of words in the other language.

Limited evidence of test-retest reliability in infant-directed speech preference in a large preregistered infant experiment.

Test-retest reliability-establishing that measurements remain consistent across multiple testing sessions-is critical to measuring, understanding, and predicting individual differences in infant language development. However, previous attempts to establish measurement reliability in infant speech perception tasks are limited, and reliability of frequently used infant measures is largely unknown. The current study investigated the test-retest reliability of infants' preference for infant-directed speech over adult-directed speech in a large sample (N = 158) in the context of the ManyBabies1 collaborative research project. Labs were asked to bring in participating infants for a second appointment retesting infants on their preference for infant-directed speech. This approach allowed us to estimate test-retest reliability across three different methods used to investigate preferential listening in infancy: the head-turn preference procedure, central fixation, and eye-tracking. Overall, we found no consistent evidence of test-retest reliability in measures of infants' speech preference (overall r = 0.09, 95% CI [-0.06,0.25]). While increasing the number of trials that infants needed to contribute for inclusion in the analysis revealed a numeric growth in test-retest reliability, it also considerably reduced the study's effective sample size. Therefore, future research on infant development should take into account that not all experimental measures may be appropriate for assessing individual differences between infants. RESEARCH HIGHLIGHTS: We assessed test-retest reliability of infants' preference for infant-directed over adult-directed speech in a large pre-registered sample (N = 158). There was no consistent evidence of test-retest reliability in measures of infants' speech preference. Applying stricter criteria for the inclusion of participants may lead to higher test-retest reliability, but at the cost of substantial decreases in sample size. Developmental research relying on stable individual differences should consider the underlying reliability of its measures.

Unsupervised neural network models of the ventral visual stream.

Deep neural networks currently provide the best quantitative models of the response patterns of neurons throughout the primate ventral visual stream. However, such networks have remained implausible as a model of the development of the ventral stream, in part because they are trained with supervised methods requiring many more labels than are accessible to infants during development. Here, we report that recent rapid progress in unsupervised learning has largely closed this gap. We find that neural network models learned with deep unsupervised contrastive embedding methods achieve neural prediction accuracy in multiple ventral visual cortical areas that equals or exceeds that of models derived using today's best supervised methods and that the mapping of these neural network models' hidden layers is neuroanatomically consistent across the ventral stream. Strikingly, we find that these methods produce brain-like representations even when trained solely with real human child developmental data collected from head-mounted cameras, despite the fact that these datasets are noisy and limited. We also find that semisupervised deep contrastive embeddings can leverage small numbers of labeled examples to produce representations with substantially improved error-pattern consistency to human behavior. Taken together, these results illustrate a use of unsupervised learning to provide a quantitative model of a multiarea cortical brain system and present a strong candidate for a biologically plausible computational theory of primate sensory learning.

Comparing apples to manzanas and oranges to naranjas: A new measure of English-Spanish vocabulary for dual language learners.

The valid assessment of vocabulary development in dual-language-learning infants is critical to developmental science. We developed the Dual Language Learners English-Spanish (DLL-ES) Inventories to measure vocabularies of U.S. English-Spanish DLLs. The inventories provide translation equivalents for all Spanish and English items on Communicative Development Inventory (CDI) short forms; extended inventories based on CDI long forms; and Spanish language-variety options. Item-Response Theory analyses applied to Wordbank and Web-CDI data (n = 2603, 12-18 months; n = 6722, 16-36 months; half female; 1% Asian, 3% Black, 2% Hispanic, 30% White, 64% unknown) showed near-perfect associations between DLL-ES and CDI long-form scores. Interviews with 10 Hispanic mothers of 18- to 24-month-olds (2 White, 1 Black, 7 multi-racial; 6 female) provide a proof of concept for the value of the DLL-ES for assessing the vocabularies of DLLs.

Referential transparency of verbs in child-directed input by Japanese and American caregivers.

Children acquiring Japanese differ from those acquiring English with regard to the rate at which verbs are learned (Fernald & Morikawa, ). One possible explanation is that Japanese caregivers use verbs in referentially transparent contexts, which facilitate the form-meaning link. We examined this hypothesis by assessing differences in verb usage by Japanese and American caregivers during dyadic play with their infants (5-22 months). We annotated verb-containing utterances for elements associated with referential transparency and compared across groups. Contrary to our hypotheses, we found that Japanese caregivers used verbs in fewer referentially transparent contexts than American caregivers, or did not significantly differ from American caregivers, depending on the measure. These findings cast doubt on cross-cultural differences in referential transparency between Japanese and American child-directed input.

Testing the relationship between preferences for infant-directed speech and vocabulary development: A multi-lab study.

From early on, infants show a preference for infant-directed speech (IDS) over adult-directed speech (ADS), and exposure to IDS has been correlated with language outcome measures such as vocabulary. The present multi-laboratory study explores this issue by investigating whether there is a link between early preference for IDS and later vocabulary size. Infants' preference for IDS was tested as part of the ManyBabies 1 project, and follow-up CDI data were collected from a subsample of this dataset at 18 and 24 months. A total of 341 (18 months) and 327 (24 months) infants were tested across 21 laboratories. In neither preregistered analyses with North American and UK English, nor exploratory analyses with a larger sample did we find evidence for a relation between IDS preference and later vocabulary. We discuss implications of this finding in light of recent work suggesting that IDS preference measured in the laboratory has low test-retest reliability.

An individual differences perspective on pragmatic abilities in the preschool years.

Pragmatic abilities are fundamental to successful language use and learning. Individual differences studies contribute to understanding the psychological processes involved in pragmatic reasoning. Small sample sizes, insufficient measurement tools, and a lack of theoretical precision have hindered progress, however. Three studies addressed these challenges in three- to 5-year-old German-speaking children (N = 228, 121 female). Studies 1 and 2 assessed the psychometric properties of six pragmatics tasks. Study 3 investigated relations among pragmatics tasks and between pragmatics and other cognitive abilities. The tasks were found to measure stable variation between individuals. Via a computational cognitive model, individual differences were traced back to a latent pragmatics construct. This presents the basis for understanding the relations between pragmatics and other cognitive abilities. RESEARCH HIGHLIGHTS: Individual differences in pragmatic abilities are important to understanding variation in language development. Research in this domain lacks a precise theoretical framework and psychometrically high-quality measures. We present six tasks capturing a wide range of pragmatic abilities with excellent re-test reliability. We use a computational cognitive model to provide a substantive theory of individual differences in pragmatic abilities.

Registered Reports in Child Development: Introduction to the Special Section.

Registered Reports (RRs) are an emerging format for publishing empirical journal articles in which the decision to publish an article is based on sound conceptualization, methods, and planned analyses rather than the specific nature of the results. This article introduces the Special Section on Registered Reports in Child Development by describing what RRs are and why they are necessary, outlining the thought process that guided the Special Section, describing key thematic insights across the eight articles included in the collection, and providing recommendations for developmental researchers interested in publishing via the RR format. This article also serves as a formal announcement that RRs will be a standard publishing option at Child Development, effective immediately.

The BabyView camera: Designing a new head-mounted camera to capture children's early social and visual environments.

Head-mounted cameras have been used in developmental psychology research for more than a decade to provide a rich and comprehensive view of what infants see during their everyday experiences. However, variation between these devices has limited the field's ability to compare results across studies and across labs. Further, the video data captured by these cameras to date has been relatively low-resolution, limiting how well machine learning algorithms can operate over these rich video data. Here, we provide a well-tested and easily constructed design for a head-mounted camera assembly-the BabyView-developed in collaboration with Daylight Design, LLC., a professional product design firm. The BabyView collects high-resolution video, accelerometer, and gyroscope data from children approximately 6-30 months of age via a GoPro camera custom mounted on a soft child-safety helmet. The BabyView also captures a large, portrait-oriented vertical field-of-view that encompasses both children's interactions with objects and with their social partners. We detail our protocols for video data management and for handling sensitive data from home environments. We also provide customizable materials for onboarding families with the BabyView. We hope that these materials will encourage the wide adoption of the BabyView, allowing the field to collect high-resolution data that can link children's everyday environments with their learning outcomes.

Peekbank: An open, large-scale repository for developmental eye-tracking data of children's word recognition.

The ability to rapidly recognize words and link them to referents is central to children's early language development. This ability, often called word recognition in the developmental literature, is typically studied in the looking-while-listening paradigm, which measures infants' fixation on a target object (vs. a distractor) after hearing a target label. We present a large-scale, open database of infant and toddler eye-tracking data from looking-while-listening tasks. The goal of this effort is to address theoretical and methodological challenges in measuring vocabulary development. We first present how we created the database, its features and structure, and associated tools for processing and accessing infant eye-tracking datasets. Using these tools, we then work through two illustrative examples to show how researchers can use Peekbank to interrogate theoretical and methodological questions about children's developing word recognition ability.

Automated detections reveal the social information in the changing infant view.

How do postural developments affect infants' access to social information? We recorded egocentric and third-person video while infants and their caregivers (N = 36, 8- to 16-month-olds, N = 19 females) participated in naturalistic play sessions. We then validated the use of a neural network pose detection model to detect faces and hands in the infant view. We used this automated method to analyze our data and a prior egocentric video dataset (N = 17, 12-month-olds). Infants' average posture and orientation with respect to their caregiver changed dramatically across this age range; both posture and orientation modulated access to social information. Together, these results confirm that infant's ability to move and act on the world plays a significant role in shaping the social information in their view.

Improving the generalizability of infant psychological research: The ManyBabies model.

Yarkoni's analysis clearly articulates a number of concerns limiting the generalizability and explanatory power of psychological findings, many of which are compounded in infancy research. ManyBabies addresses these concerns via a radically collaborative, large-scale and open approach to research that is grounded in theory-building, committed to diversification, and focused on understanding sources of variation.

Context-Dependent Learning of Linguistic Disjunction.

What are the constraints, cues, and mechanisms that help learners create successful word-meaning mappings? This study takes up linguistic disjunction and looks at cues and mechanisms that can help children learn the meaning of or. We first used a large corpus of parent-child interactions to collect statistics on or uses. Children started producing or between 18-30 months and by 42 months, their rate of production reached a plateau. Second, we annotated for the interpretation of disjunction in child-directed speech. Parents used or mostly as exclusive disjunction, typically accompanied by rise-fall intonation and logically inconsistent disjuncts. But when these two cues were absent, disjunction was generally not exclusive. Our computational modeling suggests that an ideal learner could successfully interpret an English disjunction (as exclusive or not) by mapping forms to meanings after partitioning the input according to the intonational and logical cues available in child-directed speech.

Continuous developmental change explains discontinuities in word learning.

Cognitive development is often characterized in terms of discontinuities, but these discontinuities can sometimes be apparent rather than actual and can arise from continuous developmental change. To explore this idea, we use as a case study the finding by Stager and Werker (1997) that children's early ability to distinguish similar sounds does not automatically translate into word learning skills. Early explanations proposed that children may not be able to encode subtle phonetic contrasts when learning novel word meanings, thus suggesting a discontinuous/stage-like pattern of development. However, later work has revealed (e.g., through using more precise testing methods) that children do encode such contrasts, thus favoring a continuous pattern of development. Here, we propose a probabilistic model that represents word knowledge in a graded fashion and characterizes developmental change as improvement in the precision of this graded knowledge. Our model explained previous findings in the literature and provided a new prediction - the referents' visual similarity modulates word learning accuracy. The models' predictions were corroborated by human data collected from both preschool children and adults. The broader impact of this work is to show that computational models, such as ours, can help us explore the extent to which episodes of cognitive development that are typically thought of as discontinuities may emerge from simpler, continuous mechanisms.

Children's interpretation of ambiguous pronouns based on prior discourse.

In conversation, individual utterances are almost always ambiguous, with this ambiguity resolved by context and discourse history (common ground). One important cue for disambiguation is the topic under discussion with a particular partner (e.g., "want to pick?" means something different in a conversation with a bluegrass musician vs. with a book club partner). Here, we investigated 2- to 5-year-old American English-speaking children's (N = 131) reliance on conversational topics with specific partners to interpret ambiguous or novel words. In a tablet-based game, children heard a speaker consistently refer to objects from a category without mentioning the category itself. In Study 1, 3- and 4-year-olds interpreted the ambiguous pronoun "it" as referring to another member of the same category. In Study 2, only 4-year-olds interpreted the pronoun as referring to the implied category when talking to the same speaker but not when talking to a new speaker. Thus, children's conception of what constitutes common ground in discourse develops substantially between ages 2 and 5.

Children's Social Information Seeking is Sensitive to Referential Ambiguity.

We examined children's spontaneous information seeking in response to referential ambiguity. Children ages 2-5 (n = 160) identified the referents of familiar and novel labels. We manipulated ambiguity by changing the number of objects present and their familiarity (Experiments 1 and 2), and the availability of referential gaze (Experiment 2). In both experiments, children looked to the face of the experimenter more often while responding, specifically when the referent was ambiguous. In Experiment 2, 3- to 4-year olds also demonstrated sensitivity to graded referential evidence. These results suggest that social information seeking is an active learning behavior that could contribute to language acquisition in early childhood.

Building a collaborative Psychological Science: Lessons learned from ManyBabies 1.

The field of infancy research faces a difficult challenge: some questions require samples that are simply too large for any one lab to recruit and test. ManyBabies aims to address this problem by forming large-scale collaborations on key theoretical questions in developmental science, while promoting the uptake of Open Science practices. Here, we look back on the first project completed under the ManyBabies umbrella - ManyBabies 1 - which tested the development of infant-directed speech preference. Our goal is to share the lessons learned over the course of the project and to articulate our vision for the role of large-scale collaborations in the field. First, we consider the decisions made in scaling up experimental research for a collaboration involving 100+ researchers and 70+ labs. Next, we discuss successes and challenges over the course of the project, including: protocol design and implementation, data analysis, organizational structures and collaborative workflows, securing funding, and encouraging broad participation in the project. Finally, we discuss the benefits we see both in ongoing ManyBabies projects and in future large-scale collaborations in general, with a particular eye towards developing best practices and increasing growth and diversity in infancy research and psychological science in general. Throughout the paper, we include first-hand narrative experiences, in order to illustrate the perspectives of researchers playing different roles within the project. While this project focused on the unique challenges of infant research, many of the insights we gained can be applied to large-scale collaborations across the broader field of psychology.

The role of salience in young children's processing of ad hoc implicatures.

Language comprehension often requires making implicatures. For example, inferring that "I ate some of the cookies" implicates that the speaker ate some but not all (scalar implicatures), and "I ate the chocolate chip cookies" where there are both chocolate chip cookies and raisin cookies in the context implicates that the speaker ate the chocolate chip cookies but not both the chocolate chip and raisin cookies (ad hoc implicatures). Children's ability to make scalar implicatures develops at around 5 years of age, with ad hoc implicatures emerging somewhat earlier. In the current work, using a time-sensitive tablet paradigm, we examined developmental gains in children's ad hoc implicature processing and found evidence for successful pragmatic inferences by children as young as 3 years in a supportive context and substantial developmental gains in inference computation from 2 to 5 years. We also tested whether one cause of younger children's (2-year-olds) consistent failure to make pragmatic inferences is their difficulty in inhibiting an alternative interpretation that is more salient than the target meaning (the salience hypothesis). Our findings support this hypothesis; younger children's failures with pragmatic inferences were related to effects of the salience mismatch between possible interpretations.

childes-db: A flexible and reproducible interface to the child language data exchange system.

The Child Language Data Exchange System (CHILDES) has played a critical role in research on child language development, particularly in characterizing the early language learning environment. Access to these data can be both complex for novices and difficult to automate for advanced users, however. To address these issues, we introduce childes-db, a database-formatted mirror of CHILDES that improves data accessibility and usability by offering novel interfaces, including browsable web applications and an R application programming interface (API). Along with versioned infrastructure that facilitates reproducibility of past analyses, these interfaces lower barriers to analyzing naturalistic parent-child language, allowing for a wider range of researchers in language and cognitive development to easily leverage CHILDES in their work.