Kathleen Currie Hall

Notes on: Scobbie, James M., Fiona Gibbon, William J. Hardcastle, and Paul Fletcher. 2000. Covert contrast as a stage in the acquisition of phonetics and phonology. In Michael B. Broe and Janet B. Pierrehumbert (eds.), Papers in Laboratory Phonology V: Acquisition and the lexicon (pp. 194-207). Cambridge: Cambridge University Press.

Summary:
In this paper, the authors describe the case of a particular 4-year-old who was classified as having phonological disorder (at least partially) because of a lack of voicing contrast in initial stops.  They show that in fact, he had a voicing contrast that was signalled by a difference in spectral tilt, and that he used VOT to consistently signal vowel quality.  Hence his phonological system was not so much disordered as incompatible with an adult phonetic output system.

Details:
1. Covert contrast -- very much key especially in clinical studies; often the case that adult listeners/transcribers may not be able to detect a contrast made by a child until significantly after the child begins to make it, leading to misdiagnosis or mistreatment.
2. In adult speech, spectral tilt (measured here as: dB at the first harmonic minus dB at second harmonic) tends to be greater after an aspirated stop than after an unaspirated one.
3. The child they measured had no audible difference between /t/, /d/, and /st/.  There were no significant differences in any of the duration measurements the authors took (including duration of VOT), when measured across following vowel context.
4. The child did produce significantly longer VOT before [i] than before [o], and significantly longer VOT before [o] than before [ae]; this mirrors adult speech where VOT before [i, u] tends to be longer than VOT before [E, a].
5. The child did produce a significant difference among the spectral tilts of the three onset consonants: /t/ had a much greater spectral tilt than /d/, and /st/ was somewhere in the middle.

Questions:
1. It's a bit misleading for the authors to list as many questions as they do in the last paragraph of 13.1 -- clearly, they will not be able to answer all of these questions (especially ones about acquisition timeline and broadly answering questions about the role of VOT in English acquisition) given that they have data from 28 words of one child from one session!
2. I'm surprised they don't make more of the fact that this child (even though "largely unintelligible") did in fact seem to have a good grasp of the adult phonological system in that voicing was cued by spectral tilt and vowel quality was cued by VOT.  Given that there _are_ in fact these differences in the adult system, it also seems surprising that more children don't do similar mis-analyses (why are there so many normally developing children?).  Are these potential sources of sound change?  What prompts the child to figure out the right pattern?
3. Given the discussion on p. 204 about breathy phonation not being a "necessary partner" to aspiration (is this part of the answer to the last question above?), I wonder about the input to this particular child.  Presumably he picked up on these cues from somewhere for them to be so regular.  What dialect of English do his parents speak (Scottish?)?  Is the English in the studies cited British, American, Scottis, Irish?  Would these differences make a difference in the child's acquisition?