Vocal Pitch in Infants

Review of   ‘Vocal pitch in infant-directed speech is similar across Western and traditional cultures.’

The reported study investigates pitch, pitch variation and pitch range in North American, Fiji and Kenyan mothers. The results show that when these mothers speak to their 2-16 months infant, they elevate their pitch and extend their pitch variation and pitch range. Furthermore, differences in ID speech are accounted for by age and education level rather than culture. These findings are taken to support the idea that the vocal characteristics of ID speech are a human universal.

This study is a welcome addition to the relatively sparse cross-cultural data on child-directed speech, especially since it compares mothers from Western and traditional societies. However, in order to present the full picture to the reader I would like to see some revision of the analyses as well as a more thorough discussion of how these findings fit with the existing literature on CDS.

1. My main concern with the analyses has to do with the fact that it is not the relative change in vocal parameters from AD to ID speech that is being analysed with the regression models but only ID speech. This is because AD speech was not available for many of the participants. However, ID speech is, of course, also dependent on indexical features and can therefore only be analysed in comparison to AD speech from the same speaker. Thus, the regression analyses should be rerun using the relative difference between ID and AD speech in the dependent variables. This will inevitably lead to a loss of statistical power but is still a necessary step to derive general conclusions about how speakers change their speech when addressing infants.

2. The authors report a main effect of speech type and a marginal effect of culture on pitch but do not mention whether there was an interaction between culture and speech type. In table 1 and figure 1, it certainly looks like the marginal effect of culture is...