ABSTRACT
English spelling is opaque because the correspondence between graphemes and phonemes is not one-to-one. Because of borrowings, historical changes in English, scribal preferences, and so on, the orthographic system is complex. The relationship between graphemes and phonemes is not one-to-one, but consonant grapheme-to-phoneme relationships are fairly consistent. One current theory, Statistical Learning, is that English readers implicitly learn the probabilities that a certain grapheme will correspond to a certain phoneme and use knowledge of those probabilities to read. This chapter contains tables of the raw grapheme–phoneme probabilities for consonants and vowels. However, raw probabilities are adjusted by knowledge of context. Implications for English L2 learners are presented.
