Growing out of infant-directed speech
The study in a sentence
Caregivers often use infant-directed speech (IDS) when speaking to their babies, which is thought to reflect their accommodation to the infant's immature language knowledge. Infants respond preferentially to IDS over adult-directed speech (ADS).
But what happens as the infant develops increasingly sophisticated language knowledge?
Does the caregiver's use of IDS adapt to the child's developing skills?
This study uses a meta-analysis to show that some properties of IDS change with infant age. The authors propose that this reflects differences in the form and function of IDS: caregivers adapt their speech to their infant's developing language knowledge and attentional preferences.
This study performs a meta-analysis of 88 different studies to find out:
Which features of IDS occur across different languages? That is, are any features of IDS 'universal', and are any more specific to certain languages?
Which features of IDS change as the child gets older? If features change over time, what does it tell us about the form and function of IDS?
Roughly half of the infants in the study were hearing IDS in English: in Australian English, British English, Canadian English, Jamaican English, New Zealand English, Scottish English and US English.
Key concept: form versus function
Common features of IDS include higher pitch, wider pitch range and slower articulation rate. These are all thought to support language learning for young infants.
But, these features, or special forms of speech, may not all serve the same function:
some may serve a socio-emotional role, by attracting the infant's attention and engaging them with the caregiver in social interactions;
others might be important in highlighting informative linguistic content, which is important for learning words and grammatical structures.
Early on, infants are attracted to speech that conveys positive emotion. As they get older and learn more about the surrounding language, they become more interested in speech that supports further language learning.
What forms of IDS occur across languages, and do these relate to changing functions of IDS as the infant develops?
Visualization of a meta-analysis approach, C. Cox (2022)
Methods: performing a meta-analysis
The study uses a meta-analysis to compare findings from 88 different studies that investigated 33 different languages between them. All the studies compared IDS to ADS in terms of at least one of the following five previously documented features of IDS:
Pitch: do caregiver produce IDS with a higher pitch than ADS?
Pitch variability: do caregivers produce IDS with more pitch variability than ADS?
Vowel space area: Are vowels in IDS more distinct than in ADS?
Articulation rate: Is speech produced more slowly in IDS than in ADS?
Vowel duration: Are vowels longer in IDS than in ADS?
Results from all 88 studies were pooled to see if they produced consistent findings, with a particular focus on how language and infant age affected the different features of IDS compared with ADS.
Try it yourself!
Do your own 'meta-analysis' (e.g. for a language investigation) by selecting a handful of studies that test the same question and comparing the findings.
Are the results consistent? Which results do you think are the most/least reliable, and why?
You could address commonly-tested research questions such as:
Do babies prefer IDS over ADS?
At what age do children successfully complete the Wug test?
Does a larger amount of IDS or CDS lead to larger vocabulary size?
Explainer: what is 'vowel space'?
In this explainer video we explain:
what vowel space is,
how vowel space can vary, and
how vowel space relates to the study of infant-directed speech.
If you want to learn about the technical details of how vowels are measured see this video by Richard Ogden on measuring the acoustic properties of vowels using the free software Praat. Learning to measure the acoustic properties of vowels is a skill you could develop in a Linguistics degree.
Explainer video (2 mins)
Across languages, differences were observed in some features, and universality in others:
Pitch, pitch variability and articulation rate showed similar differences across languages, with higher pitch, more pitch variability and slower articulation rate in IDS compared with ADS.
Vowel space area and vowel duration showed some cross-language differences.
This suggests that there are both universal tendencies and language-specificity in caregivers' use of IDS.
Across ages, change was observed in some features, and stability in others:
Pitch, articulation rate and vowel duration appeared to be adapted to infants' increasing language abilities, adjusting to a more ADS-like model over time.
Pitch variability and vowel space area were stable over time.
This suggests that caregivers' respond to infants' changing language abilities by adjusting some features of their speech to support further language development towards a more adult-like model. In other features, stability may reflect the fact that these features continue to support learning across the span of early development.
In more detail
Pre-workshop taster video (3 mins)
Workshop talk slides
Live Workshop talk video (20 mins)