Vowel Reduction in Educated Isoko English

ABSTRACT

Vowel reduction, the weakening of strong vowels to the reduced /ə/ sound, is an essential phenomenon in the rhythm of Standard British English (SBE) and it is claimed to be absent in Nigerian English (NE). Some studies on the phonological features of Yoruba and Hausa Englishes, which are major Nigerian languages, have claimed that vowel reduction does not exist in NE. This is with little or no reference to the small group languages including Educated Isoko English (EIE). Therefore, this study investigated the existence or otherwise of vowel reduction in the rhythm of EIE, a minority sub-variety of Nigerian English, with consideration for age and sex variables. Prince and Liberman‟s Metrical Theory, and Grabe and Low‟s Pairwise Variability Index (PVI) served as the theoretical framework for the study. The subjects, who were EIE speakers accidentally sampled, comprised 50 male and 50 female native Isoko speakers, between ages 20-40 and 41-60 andwere residents in Isokoland. Five SBE speakers served as control. Praat speech analyser version 5.1.11 was used to record 10 controlled English sentences divided into sets A and B as uttered by the subjects. Set A consists of words with only full vowels while Set B has words with both full and reduced vowels. Spontaneous speech data of EIE and SBE were recorded to validate findings from the controlled uttered data. The perceptual analysis of grammatical and content words was done using percentages complemented with metrical, statistical and PVI measures. Instances of vowel duration were also subjected to a one-way ANOVA and Fisher‟s LSD Post-hoc test. Grammatical words showed 4.1% cases of appropriate reduction and 95.9% cases of inappropriate reduction. There were 98.8% instances of inappropriate reduction and 1.2% cases of appropriate reduction in content words. The vowels /a:/, /e/, /ɔ/, /a/, /u/, /o/, /i/, /ɔ:/ were substituted for /ə/ and /i:/, /i/, /e/ for /ɪ/.  /a:/ 40.3% and /i/ 71.6% were the highest occurring variants of /ə/ and /ɪ/, respectively. Planned comparisons showed a significant difference between both sets in SBE and between EIE and SBE in Sets B (F[3,16]=11.2, p