Author: Truong Thi My Linh
Abstract
English two-consonant (CC) clusters play an important role in pronunciation, yet remain persistently arduous for EFL learners because of the structural mismatch between English and Vietnamese syllable structure. Under establishing models of Contrastive Analysis Hypothesis, the Markedness Differential Hypothesis and the Speech Learning Model, this explanatory sequential mixed-methods study investigated CC cluster errors and the perception-production relationship among first-year non-English majors at Nguyen Tat Thanh University. First, 110 participants administered a 24-item identification test covering 12 cluster contrasts distributed initially and finally. A volunteer sample of 11 freshmen, subsequently had controlled-reading tasks in isolated words and in sentences, totalling 528 tokens. These recordings were then evaluated for intelligibility on a four-rating scale by two native English-speaking raters and classified into error categories. Praat was also used to verify 60 of the 120 disagreement tokens, while the remaining 60 tokens were excluded, yet it did not change the error distribution by sensitivity analysis. 117 of the 468 classified tokens had errors, mostly concentrating in coda clusters, with deletion being the dominant type (67.5%), followed by substitution (22.2%) and epenthesis (5.1%). The results also show the positive correlation between perception and intelligibility at the learner level (r=.81), although this relationship was sensitive to one learner (ρ=.55). However, there was no significant contrast-level link, and several contrasts exhibited perception-production mismatches. The findings are theoretically predominant to support the predictions of the CAH and MDH and the co-evolution view of SLM-r. Overall, the findings reinforce the significance of cluster-focused instruction in pronunciation teaching, prioritizing coda clusters and combining perception training with explicit production practice in L2 speech learning models.
KEYWORDS: Consonant clusters, Intelligibility, Perception–production relationship, Phonological errors, Syllable position, Vietnamese EFL learners