Age-Specific Preferences for Infant-Directed Affective Intent

Mommy, speak clearly: induced hearing loss shapes vowel hyperarticulation

Talkers hyperarticulate vowels when communicating with listeners that require increased speech intelligibility. Vowel hyperarticulation is said to be motivated by knowledge of the listener's linguistic needs because it typically occurs in speech to infants, foreigners and hearing-impaired listeners, but not to non-verbal pets. However, the degree to which vowel hyperarticulation is determined by feedback from the listener is surprisingly less well understood. This study examines whether mothers' speech input is driven by knowledge of the infant's linguistic competence, or by the infant's feedback cues. Specifically, we manipulated (i) mothers' knowledge of whether they believed their infants could hear them or not, and (ii) the audibility of the speech signal available to the infant (full or partial audibility, or inaudible). Remarkably, vowel hyperarticulation was completely unaffected by mothers' knowledge; instead, there was a reduction in the degree of hyperarticulation such that vowels were hyperarticulated to the greatest extent in the full audibility condition, there was reduced hyperarticulation in the partially audible condition, and no hyperarticulation in the inaudible condition. Thus, while it might be considered adaptive to hyperarticulate speech to the hearing-impaired adult or infant, when these two factors (infant and hearing difficulty) are coupled, vowel hyperarticulation is sacrificed. Our results imply that infant feedback drives talker behavior and raise implications for intervention strategies used with carers of hearing-impaired infants.

Maternal interactions with a hearing and hearing-impaired twin: similarities and differences in speech input, interaction quality, and word production

PURPOSE

This study examined a mother's speech style and interactive behaviors with her twin sons: 1 with bilateral hearing impairment (HI) and the other with normal hearing (NH).

METHOD

The mother was video-recorded interacting with her twin sons when the boys were 12.5 and 22 months of age. Mean F0, F0 range, duration, and F1/F2 vowel space of the corner vowels /i/, /u/, and /a/ were compared in her infant-directed (ID) and adult-directed (AD) speech. The interactions were also coded for emotional availability, and vocabulary size was collected at 17 and 22 months.

RESULTS

Acoustic analyses revealed no difference among mean F0, F0 range, and duration between the twins. In contrast, when the corner vowels were plotted in F1/F2 vowel space, the results showed a diminished vowel space in speech to the HI twin compared to the NH twin. Ratings of emotional availability were lower for the HI than the NH twin, but the HI twin had a larger expressive vocabulary on both occasions, albeit in the lower percentile.

CONCLUSIONS

The mother appears more focused on maintaining the attention of the HI infant using the typical ID exaggerations to prosody and overlooking linguistic features such as the hyperarticulation of her vowels. The results have implications for early intervention strategies.