Publications

Abstract

The rate at which young children are directly spoken to varies due to many factors, including (a) caregiver ideas about children as conversational partners and (b) the organization of everyday life. Prior work suggests cross-cultural variation in rates of child-directed speech is due to the former factor, but has been fraught with confounds in comparing postindustrial and subsistence farming communities. We investigate the daylong language environments of children (0;0–3;0) on Rossel Island, Papua New Guinea, a small-scale traditional community where prior ethnographic study demonstrated contingency-seeking child interaction styles. In fact, children were infrequently directly addressed and linguistic input rate was primarily affected by situational factors, though children’s vocalization maturity showed no developmental delay. We compare the input characteristics between this community and a Tseltal Mayan one in which near-parallel methods produced comparable results, then briefly discuss the models and mechanisms for learning best supported by our findings.

Abstract

This study evaluates whether early vocalizations develop in similar ways in children across diverse cultural contexts. We analyze data from daylong audio recordings of 49 children (1–36 months) from five different language/cultural backgrounds. Citizen scientists annotated these recordings to determine if child vocalizations contained canonical transitions or not (e.g., “ba” vs. “ee”). Results revealed that the proportion of clips reported to contain canonical transitions increased with age. Furthermore, this proportion exceeded 0.15 by around 7 months, replicating and extending previous findings on canonical vocalization development but using data from the natural environments of a culturally and linguistically diverse sample. This work explores how crowdsourcing can be used to annotate corpora, helping establish developmental milestones relevant to multiple languages and cultures. Lower inter‐annotator reliability on the crowdsourcing platform, relative to more traditional in‐lab expert annotators, means that a larger number of unique annotators and/or annotations are required, and that crowdsourcing may not be a suitable method for more fine‐grained annotation decisions. Audio clips used for this project are compiled into a large‐scale infant vocalization corpus that is available for other researchers to use in future work.

Abstract

Language acquisition is complex. However, one thing that has been suggested to help learning is the way that information is distributed throughout language; co-occurrences among particular items (e.g., syllables and words) have been shown to help learners discover the words that a language contains and figure out how those words are used. Humans’ ability to draw on this information—“statistical learning”—has been demonstrated across a broad range of studies. However, evidence from non-WEIRD (Western, Educated, Industrialized, Rich, and Democratic) societies is critically lacking, which limits theorizing on the universality of this skill. We extended work on statistical language learning to a new, non-WEIRD linguistic population: speakers of Yélî Dnye, who live on a remote island off mainland Papua New Guinea (Rossel Island). We performed a replication of an existing statistical learning study, training adults on an artificial language with statistically defined words, then examining what they had learnt using a two-alternative forced-choice test. Crucially, we implemented several key amendments to the original study to ensure the replication was suitable for remote field-site testing with speakers of Yélî Dnye. We made critical changes to the stimuli and materials (to test speakers of Yélî Dnye, rather than English), the instructions (we re-worked these significantly, and added practice tasks to optimize participants’ understanding), and the study format (shifting from a lab-based to a portable tablet-based setup). We discuss the requirement for acute sensitivity to linguistic, cultural, and environmental factors when adapting studies to test new populations.

Abstract

Recordings captured by wearable microphones are a standard method for investigating young children’s language environments. A key measure to quantify from such data is the amount of speech present in children’s home environments. To this end, the LENA recorder and software—a popular system for measuring linguistic input—estimates the number of adult words that children may hear over the course of a recording. However, word count estimation is challenging to do in a language-independent manner; the relationship between observable acoustic patterns and language-specific lexical entities is far from uniform across human languages. In this paper, we ask whether some alternative linguistic units, namely phone(me)s or syllables, could be measured instead of, or in parallel with, words in order to achieve improved cross-linguistic applicability and comparability of an automated system for measuring child language input. We discuss the advantages and disadvantages of measuring different units from theoretical and technical points of view. We also investigate the practical applicability of measuring such units using a novel system called Automatic LInguistic unit Count Estimator (ALICE) together with audio from seven child-centered daylong audio corpora from diverse cultural and linguistic environments. We show that language-independent measurement of phoneme counts is somewhat more accurate than syllables or words, but all three are highly correlated with human annotations on the same data. We share an open-source implementation of ALICE for use by the language research community, allowing automatic phoneme, syllable, and word count estimation from child-centered audio recordings.

Abstract

Rapid response latencies in conversation suggest that responders start planning before the ongoing turn is finished. Indeed, an earlier EEG study suggests that listeners start planning their responses to questions as soon as they can (Bögels, S., Magyari, L., & Levinson, S. C. (2015). Neural signatures of response planning occur midway through an incoming question in conversation. Scientific Reports, 5, 12881). The present study aimed to (1) replicate this early planning effect and (2) investigate whether such early response planning incurs a cost on participants’ concurrent comprehension of the ongoing turn. During the experiment participants answered questions from a confederate partner. To address aim (1), the questions were designed such that response planning could start either early or late in the turn. Our results largely replicate Bögels et al. (2015) showing a large positive ERP effect and an oscillatory alpha/beta reduction right after participants could have first started planning their verbal response, again suggesting an early start of response planning. To address aim (2), the confederate's questions also contained either an expected word or an unexpected one to elicit a differential N400 effect, either before or after the start of response planning. We hypothesized an attenuated N400 effect after response planning had started. In contrast, the N400 effects before and after planning did not differ. There was, however, a positive correlation between participants' response time and their N400 effect size after planning had started; quick responders showed a smaller N400 effect, suggesting reduced attention to comprehension and possibly reduced anticipatory processing. We conclude that early response planning can indeed impact comprehension processing.

Abstract

Speaker diarization (answering 'who spoke when') is a widely researched subject within speech technology. Numerous experiments have been run on datasets built from broadcast news, meeting data, and call centers—the task sometimes appears close to being solved. Much less work has begun to tackle the hardest diarization task of all: spontaneous conversations in real-world settings. Such diarization would be particularly useful for studies of language acquisition, where researchers investigate the speech children produce and hear in their daily lives. In this paper, we study audio gathered with a recorder worn by small children as they went about their normal days. As a result, each child was exposed to different acoustic environments with a multitude of background noises and a varying number of adults and peers. The inconsistency of speech and noise within and across samples poses a challenging task for speaker diarization systems, which we tackled via retraining and data augmentation techniques. We further studied sources of structured variation across raw audio files, including the impact of speaker type distribution, proportion of speech from children, and child age on diarization performance. We discuss the extent to which these findings might generalize to other samples of speech in the wild.

Abstract

Word count estimation (WCE) from audio recordings has a number of applications, including quantifying the amount of speech that language-learning infants hear in their natural environments, as captured by daylong recordings made with devices worn by infants. To be applicable in a wide range of scenarios and also low-resource domains, WCE tools should be extremely robust against varying signal conditions and require minimal access to labeled training data in the target domain. For this purpose, earlier work has used automatic syllabification of speech, followed by a least-squares-mapping of syllables to word counts. This paper compares a number of previously proposed syllabifiers in the WCE task, including a supervised bi-directional long short-term memory (BLSTM) network that is trained on a language for which high quality syllable annotations are available (a “high resource language”), and reports how the alternative methods compare on different languages and signal conditions. We also explore additive noise and varying-channel data augmentation strategies for BLSTM training, and show how they improve performance in both matching and mismatching languages. Intriguingly, we also find that even though the BLSTM works on languages beyond its training data, the unsupervised algorithms can still outperform it in challenging signal conditions on novel languages.

Abstract

Conversational turns often proceed with very brief pauses between speakers. In order to maintain “no gap, no overlap” turntaking, we must be able to anticipate when an ongoing utterance will end, tracking the current speaker for upcoming points of potential floor exchange. The precise set of cues that listeners use for turn-end boundary anticipation is not yet established. We used an eyetracking paradigm to measure adults’ and children’s online turn processing as they watched videos of conversations in their native language (English) and a range of other languages they did not speak. Both adults and children anticipated speaker transitions effectively. In addition, we observed evidence of turn-boundary anticipation for questions even in languages that were unknown to participants, suggesting that listeners’ success in turn-end anticipation does not rely solely on lexical information.