English Stress Rules and Native Speakers

Abstract

In order to aid in the selection of rules to be included in an error analysis of Brazilian students of English, an experiment was undertaken comparing Chomsky and Halle's (1968) Main Stress Rule (divided into 16 smaller rules) with Guierre's (1970) stress rules, to discover which rules lead most frequently to the same stress assignment that native English speakers would give to unfamiliar words, nonsense words being used to guarantee lack of familiarity. Only five of Chomsky and Halle's rules (one of these coinciding with one of Guierre's) were as consistently followed as Guierre's suffix rules. The principle factors causing non-rule stress placement were a different interpretation of vowel quality (unknown underlying representation) and irrelevant consonant clusters and tense vowels. The high scores of three Alternating Stress Rules, compared to rules that are supposedly applied before them, leave doubts as to Chomsky and Halle's linear ordering.