Premium
The Modification of Onsets in a Markedness Relationship: Testing the Interlanguage Structural Conformity Hypothesis
Author(s) -
Carlisle Robert S.
Publication year - 1999
Publication title -
language learning
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.882
H-Index - 103
eISSN - 1467-9922
pISSN - 0023-8333
DOI - 10.1046/j.0266-8254.2003.03701.x-i1
Subject(s) - sonority hierarchy , linguistics , psychology , phonology , syllable , phonetics , vowel , syllabification , consonant , markedness , philosophy
This chapter reports the findings of a study investigating the possible influence of typological universals on the structuring of a second language phonology. Specifically, the study examines the frequency of modification of 2‐member and 3‐member onsets, the latter being more marked than the former (Greenberg, 1978). Though some evidence exists that longer onsets are modified more frequently than are shorter onsets (Anderson, 1987), the current study controls for a possible confounding factor—whether or not the onsets abide by the Universal Canonical Syllable Structure (UCSS). The UCSS states in part that the nucleus is the most sonorant constituent in a syllable and that segments comprising an onset will rise continuously in sonority from the most peripheral member through the nucleus (Cairns & Feinstein, 1982; Clements, 1990; Hooper, 1976; Kiparsky, 1979; Selkirk, 1984; Vennemann, 1988). All 1‐member onsets, by definition, abide by the UCSS, as any consonant will be less sonorant than the vowel constituting the nucleus. However, the UCSS is not an absolute universal, as some languages, including English, have complex onsets that violate the UCSS. In fact, all 3‐member onsets in English violate it. Previous research has demonstrated that onsets that violate the UCSS are more frequently modified than are those that do not, even when length is held constant (Carlisle, 1991b; Tropf, 1987). Consequently, any study that attempts to validly compare the frequency of modification of 2‐member and 3‐member onsets must examine only 2‐member onsets that violate the UCSS, given that all 3‐member onsets do. Such a control over the sonority relationships among the segments comprising the onsets guarantees that length is the actual variable being examined. The major contribution of this study is that it examines the production of onsets of different lengths while controlling for the sonority relationships among the segments in the onsets.