Spelling-to-Sound Translation for English Disyllables: Use of Long and Short Vowels Before Single Medial Consonants

The English writing system is often seen as having rules that govern the choice between alternative pronunciations of letters but as having many exceptions to the rules. One postulated rule, the V¯ |CV rule, is that a vowel is pronounced as long rather than short when it is followed by a single consonant letter plus a vowel letter. We find, in an analysis of English disyllabic words, that exceptions to the V¯ |CV rule are not randomly scattered throughout the vocabulary. The use of long versus short vowels is systematically influenced by the identity of the vowel letter itself, the characteristics of the following consonant, and the identity of the word ending. In four studies with a total of 283 participants, we asked university students to pronounce disyllabic nonwords with medial vowel-consonant-vowel sequences and choose the better pronunciation of the first vowel, long or short. Participants were generally sensitive to the effects found in the vocabulary. However, participants used fewer long vowels than expected on the basis of the vocabulary statistics calculated for disyllabic words. We consider possible reasons for this difference and implications for statistical-learning views of reading. Given that the V¯ |CV rule is sometimes taught as a part of phonics programs, we also consider implications for reading instruction.