Do orthographic and phonological peculiarities of alphabetically written languages influence the course of literacy acquisition?

Abstract

The most important cognitive models of alphabetic literacy acquisition, like those proposed by Uta Frith and Philip Seymour (cf. Frith 1985, and Seymour & Elder 1986), are presented as models of the way alphabetic orthographies in general are learned even though they are inspired by the study of the acquisition of a single orthography, needless to specify it, the English one. Indeed, it is not the fault of scholars working in English-speaking countries if the cognitive psycholinguistics of the written language has been much less pushed on in the other countries. However, everybody admits that it is important to check how and to what extent the orthographic peculiarities of alphabetic languages influence the course of literacy acquisition. Why could or should orthography-related differences in literacy acquisition arise? English presents a relatively opaque system: there are many alternatives to spell phonemes and many ways of sounding out graphemes, and, in addition, many of the correspondences cannot be predicted from contextdependent graphophonological rules. Other languages present a much higher degree of orthographic transparency, either in one conversion direction (from spelling to sound or from sound to spelling) or in both directions. Literacy onset in languages having a highly transparent orthography might exhibit a much more precocious and systematic use of phonological decoding than is the case in English. Besides, some characteristics of the spoken language, like complexity of the syllabic structure (for example, clusters of two and three consonants at the onset and offset of the syllable) and number of vowels, might also influence the acquisition of the corresponding written language. English presents both a high number of vowels and of complex syllabic structures. These two characteristics of the language might not encourage or facilitate the resort to phonological transcoding in reading and writing. Compared to English, access to phonological transcoding might thus be easier for languages having either a small number of vowels or relatively simple phonological structures or both. Recently, Wimmer & Goswami (1994) compared reading acquisition in English and German. In German, the mapping between graphemes and phonemes is largely consistent. The results of this study suggest a much more