Exploration of the Phonetic Difference in Stop between Hakka Infant-Directed Speech and Adult-Directed Speech

This study was concerned with phonetic realizations of six oral stops /p, t, k, p(h), t(h), k(h)/ in Hakka collected from infant-directed speech (IDS) and adult-directed speech (ADS), and explored whether phonetic differences between them advantaged infants' language acquisition. Five mother-infant dyads took part in the recordings of ADS and IDS over the first six months after infants' birth. Phonetic-acoustic correlates of oral stops were investigated, including voice onset time (VOT), spectral peak frequency of the release bursts, and obstruent/syllable (O/S) ratio. The results of this research were listed as follows. First, regarding VOT, IDS had longer VOT than ADS in aspirated stops, but not in nonaspirated stops, no matter which factor (i.e. place of articulation of stops or following vowel context [i, a, u]) was taken into consideration. The expanded VOT difference between aspirated and nonaspirated categories in IDS might benefit infants' phonetic categorization and facilitate language learnability. Next, compared with ADS, spectral peak frequencies of stops produced in different places of articulation (PoAs) were significantly heightened in IDS, but stops with different PoAs remained spectrally separable from one another. Third, the O/S ratios in IDS were not larger than those in ADS, illustrating not only the articulatory difference between vowel and consonant in lengthening but also the role of proper onset-rime percentages in syllables for language acquisition. Finally, this study supported the concept of perceptually-based "phonetic enhancement" in IDS consonantal studies. To be conclusive, mothers modified Hakka stops addressed to infants systematically and strategically, and provided well-specified VOT distinctions among different stops.