12 months of data collection, 317 lab sessions, 22,958 phrases, 54,594 vowels, and a gazillion supplementary tables. I'm proud to say that our longitudinal study on the acoustics of infant-directed speech is now out in Royal Society Open Science! 👶🧵 1/

Consonants, like vowels, appear to be less distinct in IDS than ADS, thus reinforcing the interpretation that IDS may serve an attentional and/or affective aim, rather than a didactic purpose. /end.

We did not find any association between features of VOTs and infants' consonant production and/or babbling as reported by parents. This held whether VOTs were operationalised as the IDS input itself or parents' adjustment/effort in IDS from their ADS. 4/.

We found that, compared to adult-directed speech, voiceless stops in IDS had longer VOTs, while voiced stops had shorter, leading to overall less distinct consonant contrasts (/b-p/, /d-t/, /g-k/) in IDS than ADS. From 6 to 12 months, VOTs in IDS became more similar to ADS. 3/.

We examined voice onset time (VOT)—the interval between a consonant's release and vocal fold vibration, which distinguishes voiced and voiceless stops (e.g. /b/-/p/)—in speech recorded during shared reading interactions in Norwegian parent-infant and parent-experimenter dyads. 2/.

New paper out! Following from our work on vowels and prosody in infant-directed speech, we turned our attention to features of consonants and their role in language development. Great collab w/ @julienmayor, Nina Varjola and Natalia Kartushina 🧵 1/.

Data, code and materials to the study is of course openly available at OSF. Watch this space for whether and how these acoustic properties might relate to infants’ language skills! /end.

Our results show that IDS is both dynamic and static, and we suggest that future studies should investigate how the acoustic constants and variations influence the proposed attentional, socio-emotional, and linguistic functions of speech to infants. 7/.

We also found that IDS was characterized by increased vowel variability and decreased vowel distinction, and more so in mothers and fathers, with no change with infant age. In other words, this highlights that IDS is not across-the-board ‘clear’ input. 6/.

Analyses of vowel space expansion (vowel hyperarticulation) told a more complicated story; While fathers’ overall expanded their vowel space in IDS, mothers did not, perhaps as they appeared to shift from vowel space reduction to expansion with infant age. 5/.

What did we find? Compared to their ADS, both mothers and fathers’ IDS had: Higher pitch, wider pitch range, slower articulation rate, and longer vowels. With infant age, pitch range widened and vowels shortened, while pitch and articulation rate remained relatively stable. 4/.

We followed 69 Norwegian families for one year and five lab sessions, from infants were 6–18 months. Parents’ speech were recorded while reading a picture book to their infant (IDS) and an experimenter (ADS), ensuring no distortion from different linguistic content of speech. 3/.

First things first, this was only possible because of the great team of @NataliaKartush1, @julien__mayor, @CristiaAlex, Arun Singh and Roger Mundry, as well as fantastic RAs and helpful reviewers! 2/.

RT @MPI_NL: Testing the relationship between preferences for infant-directed speech and vocabulary development: A multi-lab study. New pap….

We consider several linguistic and methodological explanations for these findings, incl. the acoustic subtlety of the contrasts, and conclude that our results challenge the strict "universal phoneticians" account while echoing calls for diversity in the literature on PC.🌍🔍 /end.

However, we did find a negative correlation between infants' native and non-native discrimination scores, suggesting that these babies could be on the cusp of reorganizing their speech perception toward their native language.📉📈 4/.

Surprisingly, on a group level, infants' looking behavior were not indicative of discrimination of either contrast—neither in our preregistered pipeline nor from several variations of it that were tested in a multiverse analysis.❌ 3/.

N=67 infants were tested on their discrimination of the native 🇳🇴 vowels /y-i/ and the non-native 🇬🇧 vowels /ʌ-æ/, with an automatised, gaze-contingent habituation procedure. As perceptual narrowing supp. occurs >6mo, we predicted successful discrimination of both contrasts.✅ 2/.

Can't get enough of infant habituation designs!?💥 In our new paper, @julien__mayor, @CristiaAlex, @NataliaKartush1, and I tested native and non-native vowel discrimination in 6-month-old Norwegian infants using eye-tracking.👶👀 1/.

Of course, our study is correlational, and other factors might be moderating the associations. Still, our results support current recommendations to engage in shared reading while limiting screen time early in life. 📚💪 Data, code and materials is of course openly available! 5/5.