DUAL ROUTE
The hypothesis that the reader of an alphabetic language has two ways of matching the form of a word on the page to a mental representation. The first (the lexical route) seeks a match for the word as a whole; the second (the sub-lexical route) interprets the letters phonologically, by means of grapheme-phoneme correspondence (GPC) rules. In normal reading, the lexical route is faster and more efficient; but it is argued that readers need the sub-lexical route when encountering a word which they have not seen before in written form, an unusual proper noun, a neologism, a non-word etc.
The dual-route model faces problems in situations where a sequence of letters permits of more than one interpretation. The model assumes that the sub-lexical route offers access to standard GPC rules, while the lexical route handles ‘exception’ words. On this analysis, a reader would take the same time to identify GOPE as a non word as to identify HEAF. However, neighborhood effects are found to apply. Words like GOPE with only ‘friends’ (HOPE, ROPE) are identified faster than those like HEAF with ‘enemies’ (LEAF vs DEAF). One solution to this is found in analogy theory which suggests that words are interpreted phonologically by analogy with others, perhaps mainly on the basis of their rime. Another is to assume that the sub-lexical route contains information on all possible interpretations of a particular letter or digraph (-EA- being recognised as potentially both /i:/ and /e/); hence the slower reaction time.
The strongest evidence supporting a dual route model comes from studies of acquired dyslexia. One type, surface dyslexia, seems to involve impairment of the lexical route but permits use of the sub lexical. Subjects thus regularise irregular words. A second type, phonological dyslexia, seems to involve an intact lexical route but an impaired sub-lexical one. The subjects can pronounce familiar words, both regular and irregular, but are incapable of suggesting pronunciations for non-words.
The extent to which the sub-lexical route is employed may vary from language to language. A relatively opaque orthography like the English one may involve greater dependence upon the lexical route than the transparent orthography of Spanish. However, there is some evidence from Spanish dyslexics of semantic exchanges (when a word like APE is read as monkey); this suggests that both whole-word and phonological routes are employed.
See also: Grapheme-phoneme correspondence rules, Reading: decoding
Further reading: Balota (1994); Harris and Coltheart (1986)
