Abstract
This paper examines the developmental patterns of #sC clusters in 40 English speaking children, ages 2;5 to 4;2, (mean age 3;1). The central focus is the production of different groups of #sC clusters and how these productions relate to sonority sequencing of sounds. We evaluated the patterns found in correct realizations of target clusters, as well as the ones in cluster reductions. Despite considerable individual variations in subjects' productions, a few general trends were identifiable. Results lend limited support to sonority sequencing but only for /s/ + approximant clusters. This pair of clusters, /sl-, sw-/, also reveal some implicational relationships. Continuancy of C2 in the cluster was significant only between /sn-/ and /sw-/. Homorganicity of the cluster members was not influential in the production patterns.
Keywords:
Notes
1 In this paper, we adopt the following ten point scale of sonority from Hogg and McCully (1987).
2 Wyllie-Smith et al. also examined the productions of children with speech impairment and found even less support for sonority in their reduction patterns of #sC clusters. Among these, /s/ + stop targets were the only ones that supported the sonority sequencing, while /sl-/ and especially /sw-/ and /sN-/ reductions showed overwhelming violations (−90%, and −93%, respectively) of the sonority hierarchy. Chin (Citation1996), on the other hand, found support for sonority in the productions of his subjects with speech impairment.
3 If the cluster was reduced to a non-target sound (e.g., /sw-/→[f]) it was excluded.
4 The reason that only a subgroup of subjects (i.e., 19 children) were looked at was due to the fact that the data for the remaining 21 children were collected earlier without their non-/s/-clusters' productions. Thus, the results of non-/s/-clusters and their comparisons with #sC clusters in these children come from a database which is more limited.