Infrequent Child-Directed Speech Is Bursty and May Draw Infant Vocalizations
Authors: Margaret Cychosz, Adriana Weisleder
Organizations: Dept. of Linguistics Stanford University Stanford, USA · Dept. of Communication Sciences and Disorders Northwestern University Evanston, USA
Children in many parts of the world hear relatively little speech directed to them, yet still reach major language development milestones. What differs about the speech input that infants learn from when directed input is rare? Using longform, infant-centered audio recordings taken in rural Bolivia and the urban U.S., we examined temporal patterns of infants' speech input and their pre-linguistic vocal behavior. We find that child-directed speech in Bolivia, though less frequent, was just as temporally clustered as speech input in the U.S, arriving in concentrated bursts rather than spread across the day. In both communities, infants were most likely to produce speech-like vocalizations during periods of speech directed to them, with the probability of infants' speech-like vocalizations during target child-directed speech nearly double that during silence. In Bolivia, infants' speech-like vocalizations were also more likely to occur during bouts of directed speech from older children than from adults. Together, these findings suggest that the developmental impact of child-directed speech may depend not only on quantity, but on temporal concentration and source, with older children serving as an important source of input in some communities, including where adult speech to infants is less frequent.
Figures & tables
Fig. 1: Daily speech register patterns for two representative infants in Bolivia (left) and the U.S. (right). Y-axis=the proportion of a 5-minute rolling window containing a given register. Dashed line indicates the mean proportion of each register or vocalization rate over the entire recording period.
Community
Register
M (SD)
Range
Bolivia
TCDS
0.27 (0.16)
0.09–0.46
CDS incl.
0.22 (0.10)
0.13–0.38
CDS excl.
0.34 (0.13)
0.14–0.47
ADS
0.37 (0.17)
0.13–0.59
Overheard
0.21 (0.12)
0.07–0.31
U.S.
TCDS
0.15 (0.11)
0.04–0.31
TABLE I: Burstiness ( B ) prior to infant-level baseline correction, by register and community.
Fig. 2: Temporal distribution of speech registers and target infant vocalizations for two representative infants in Bolivia (left) and the U.S. (right). Each vertical slice represents the presence or absence of the register/vocalization within a single 30-second annotation bin. White space indicates silence (no speech present in the segment). Green tick marks above the register bar indicate bins containing a speech-like vocalization from the target infant.
Fig. 3: Conditional probability of target infant vocalization given each speech register, by community. Points represent individual infants. Vocalizations are more likely to co-occur with TCDS than most other registers, in both communities (see Table 2 for detail). TCDS = speech directed to the target infant and no other child. CDS incl. = child-directed speech to the target child and other children. CDS excl. = child-directed speech to other children only. ADS = adult-directed speech.
Bolivia
U.S.
P(voc ∣ register)
M (SD)
M (SD)
Target child-directed speech
0.74 (0.13)
0.69 (0.11)
Child-directed speech incl.
0.44 (0.15)
0.64 (0.16)
Child-directed speech excl.
0.34 (0.16)
0.42 (0.24)
Adult-directed speech
0.36 (0.14)
0.41 (0.28)
TABLE II: P(vocalization ∣ register) and one-tailed paired t -tests comparing TCDS to each register, by community. p -values are Bonferroni-corrected (4 comparisons per community). Bold indicates significantly higher probability of vocalization in TCDS than the listed register, after correction.
Automatically distinguishing child-directed speech from adult-directed speech in long-form recordings is key to scalable analyses of children's language environments. Existing approaches process utterances in isolation and have been evaluated primarily on English. We address these gaps along three dimensions. First, we fine-tune and evaluate six-self supervised models on a multilingual dataset of 182 children, showing that in-domain pre-training on child-centered recordings substantially outperforms models trained on adult speech. Second, we demonstrate that incorporating surrounding context substantially improves classification, with an absolute gain of 13.8% in average F1-score. Third, we evaluate our model in a realistic end-to-end pipeline, from adult speech detection to addressee classification, showing that performance drops under automatic segmentation but still consistently outperforms a rule-based baseline.
Théo Charlot, Tarek Kunze, Kaveri K. Sheth +2
LSCP, DEC, ENS, EHESS, CNRS, PSL University, France · Laboratoire d’Informatique et Syst`emes, Universit´e Aix-Marseille, CNRS, France
Recent studies suggest that child-directed speech is not conducive to language learning in BabyLMs. However, current evaluations focus predominantly on comprehension and not production, which is central to usage-based theories of language acquisition which argue how CDS facilitates early language use through constructional ''frames'' (frequent lexical patterns with open slots). We introduce a novel generation-based evaluation inspired by such theories in form of a frame-completion task, and compare Llama models trained with CDS, the BabyLM corpus, and web-crawl data (FineWeb-edu) on comprehension benchmarks and our novel framework. Our results reveal a clear dissociation between models' comprehension and production capabilities: while FineWeb-trained models excel at minimal pairs, CDS-trained models produce grammatical completions substantially earlier in training and concentrate probability mass on appropriate slot-fillers. These findings show that comprehension benchmarks underestimate what CDS affords to BabyLMs.
Bastian Bunzeck, Sina Zarrieß
Computational Linguistics, Department of Linguistics Bielefeld University, Germany
Structured phonological representations provide an interpretable alternative to generic speech embeddings, but existing models are largely trained on adult speech. We adapt PhonoQ-2.0 to child speech using CHILDES-Aligned data and compare three alignment-supervision conditions (Adult, Adult+Child, and Child-only) across two initialization strategies (Adult PhonoQ and scratch). Generalization is evaluated against manual child-speech annotations. On 1,352 consonant targets from 58 typically developing children, child-speech adaptation improves voicing recognition across all supervision conditions, from 0.922 macro-F1 for Adult PhonoQ to 0.972--0.987 after adaptation. Manner is more sensitive to alignment supervision: Adult+Child MFA reaches 0.804 and 0.796, compared to approximately 0.70 under Adult MFA supervision. Place remains comparatively strong across systems (0.871--0.902), although per-class performance varies substantially. The velar-fronting contrast is preserved across all seven model variants. Longitudinal UltraPhonix analysis further reveals speaker-specific velar and post-alveolar changes that are largely preserved across models and broadly consistent with reported clinical progress.
Abner Hernandez, Tomás Arias Vergara, Andreas Maier +1
Pattern Recognition Lab, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany · GITA Lab. Facultad de Ingenier´ıa. Universidad de Antioquia UdeA, Medell´ın, Colombia · Chair for AI in Healthcare and Medicine, Technical University of Munich, Munich, Germany