The hyperarticulation hypothesis of infant-directed speech

Abstract

Typically, the point vowels [i,ɑ,u] are acoustically more peripheral in infant-directed speech (IDS) compared to adult-directed speech (ADS). If caregivers seek to highlight lexically relevant contrasts in IDS, then two sounds that are contrastive should become more distinct, whereas two sounds that are surface realizations of the same underlying sound category should not. To test this prediction, vowels that are phonemically contrastive ([i-ɪ] and [eɪ-ε]), vowels that map onto the same underlying category ([æ- ] and [ε- ]), and the point vowels [i,ɑ,u] were elicited in IDS and ADS by American English mothers of two age groups of infants (four- and eleven-month-olds). As in other work, point vowels were produced in more peripheral positions in IDS compared to ADS. However, there was little evidence of hyperarticulation per se (e.g. [i-ɪ] was hypoarticulated). We suggest that across-the-board lexically based hyperarticulation is not a necessary feature of IDS

Dog-directed speech: why do we use it and do dogs pay attention to it?

“…Accordingly, infant-directed speech has been shown to increase cerebral activity more than adult-directed speech [5], meaning that infants are more engaged in what is being said to them when they listen to this special speech register. Infant-directed speech has also been hypothetized to facilitate language learning [6] by supporting the construction of phonetic and vowel categories [7,8], the clearer production of consonants [3] and the acquisition of new words [9]. This role in language learning is consistent with the decrease in the use and acoustic specificity of infant-directed speech that follows the development of language skills during the first year of the child [10][11][12].…”

Unnatural pedagogy : a computational analysis of children's learning to learn from other people.

“…The model predicts that the statistically optimal data do not simply parrot the target distribution but modify it in ways that match infant-directed speech. Specifically, consistent with previous research (Kuhl et al, 1997;Cristia & Seidl, 2013;Burnham et al, 2002) the corner vowels are hyperarticulated. Figure 2.2 shows the predicted change in distance between all pairs of vowels.…”

“…Figure 2.2 shows the predicted change in distance between all pairs of vowels. The model predicts hyperarticulation in most vowel pairs and consistent with IDS, but contra previous arguments (Cristia & Seidl, 2013), the statistically optimal input includes hypoarticulation of some vowel pairs. Kuhl, 2003;Cristia & Seidl, 2013), but contra previous arguments (McMurray et al, 2013), the statistically optimal input includes increases in within-category variability…”

“…Infant-directed speech (IDS, also known as motherese) is characterized by reduced speed, elevated pitch and affect, and unusual prosody. At a less perceptible layer, Corner vowels (/A/, as in pot; /i/, as in beet; /u/, as in boot) are hyperarticulated in IDS (Kuhl et al, 1997;Cristia & Seidl, 2013;Burnham, Kitamura, & Vollmer-Conna, 2002), resulting in an increased vowel space. Hyperarticulation should improve the learnability of vowel categories; example clusters that are more distant are easier to identify.…”

“…This sparked the idea that IDS is optimized to teach phonetic categories (Kuhl et al, 1997). Further research demonstrated that the trend of hyperarticulation in IDS is not constant across all phoneme pairs; some pairs are hypoarticulated (Cristia & Seidl, 2013). Additionally, the within-phoneme variability of some phoneme categories increases (de Boer & Kuhl, 2003;Cristia & Seidl, 2013;McMurray, Kovack-Lesh, Goodwin, & McEchron, 2013).…”