Stop-like articulation of voiced dental fricative \δ\ in 2-year-old children: An acoustic analysis

The voiced dental fricative \δ\ is frequently modified from its canonical form in American English, often becoming nasalized, lateralized, or stop-like in casual speech [Manuel (1995); Manuel and Wyrick (1999); Zhao (2010)]. Despite modifications in their production, \δ\ variants in adult speech were found to exhibit acoustic evidence suggesting a dental place of articulation, consistent with Stevens’ (2002) feature-cue-based model of speech processing. However, it is unclear whether children produce similar modifications for \δ\. This study compares the stop-like variant of \δ\ produced by 2 to 3-year-old children in the Imbrie Corpus [Imbrie (2005)] with adult productions. The data indicate that stop-like \δ\ frequently occurred in utterance-initial position, sometimes occurred when preceded by obstruent consonants, but rarely occurred when preceded by sonorant consonants or vowels in the children’s speech; these trends are similar to those found in adult data. Furthermore, stop-like \δ\ tokens had higher burst spectrum peak, lower normalized burst amplitude, and lower F2 at following vowel onset than \d\ tokens in the same vowel context; these acoustic differences are consistent with a dental place of articulation. The data suggest that children, like adults, may be preserving the dental articulation instead of substituting \d\ for \δ\, as is sometimes assumed.