Understanding English word accentuation

Abstract

We contrast two ways of representing English word stress or accent—that exemplified by Chomsky and Halle [Chomsky, N., Halle, M., 1968. The Sound Pattern of English. Harper & Row, New York] and by Garde [Garde, Paul, 1968. L’accent. Presses Universitaires de France, Paris], and that found in many English-language dictionaries. For Chomsky–Halle and for Garde, full vowels are always stressed; for the dictionary makers, full vowels can be either accented or unaccented. We argue for an analysis of English accentual patterns along the lines of the dictionary treatment of accents. The dictionary representations, with alternating accented and unaccented syllables, suggest a natural way of assigning foot structure to syllables. Feet can be ternary, binary, or unary, but the latter are permitted only word finally. All syllables are assigned to some foot, with the exception that there are some word-initial syllables that remain unfooted (and by implication, unaccented). For most English words neither foot structure nor the location of accents is part of the lexical entry. Hence, there must be a set of metrification rules for creating feet. The location of the accents then is easily derived from the generated foot structures. We show how the rules apply to various kinds of morphologically simple (underived) words. No additional mechanisms are required for morphologically complex forms. The metrification rules apply simultaneously to each morphological constituent. This manner of forming feet is not cyclic in accordance with the usual interpretation of that concept.