The Pronunciation of the English <NG>: A Case Study in Phoneme-Grapheme Relationships

Abstract

English spelling is frequently described as irregular, in the sense that several phonemes may sometimes be represented by one grapheme, or that the same phoneme may, on occasion, be represented by several different combinations of letters. In a recent study, however, it was demonstrated that the degree of phoneme-grapheme correspondence in English is actually in excess of 80%-more predictable than is ordinarily believed. The greatest inconsistencies occur between vowel graphemes and phonemes; among the consonants, the digraph is a problematic case, as it may stand for four different phonemes or phoneme combinations: /J, ug, ng, nj/. This article shows that the phonemes represented by the digraph can, with few exceptions, be predicted when the structure of the word in which it appears is taken into account. The phonemic value of the digraph in a word depends on certain features in the underlying morpheme structure of that word, as well as various subsidiary processes, e.g., assimilation. The example of the digraph suggests that English spelling, while not phonemic, should not be dismissed as hopelessly irregular without a careful investigation of all aspects of the complex relationship between orthography, phonology, and morpheme structure. Suggestions for the use of the findings in classroom situations conclude the article. It is a commonplace among teachers of language that speech is prior to writing, and that the written language is merely a visual surrogate for the spoken language. This view has deeply influenced the methodology of language teaching, especially in regard to the ordering of those activities through which the student comes into contact with the language being studied: first, he learns the sound or spoken word; only then is he introduced to the visual written represenation of that sound in the form of the letters and signs of the traditional orthography. Recently this view has been subjected to a re-examination, the result of which has been an adjustment, not so much of the principle that language is fundamentally speech, as in the inferences which are being drawn from it. As far as the written language is concerned, there has been a renewed appreciation of the role and value of literacy, and on the more pragmatic level an increased recognition of the validity of the fundamental principle that one first learns to read and then reads to learn (Saville-Troike 1973). This