Do reading processes differ in transparent vs. opaque orthographies? A study of acquired dyslexia in Welsh/English bilinguals

Abstract

In English, the relationship between the written and spoken forms of words can be quite opaque so that it has been argued that skilled reading requires the use of two distinct procedures: (1) a sublexical grapheme/phoneme mapping process allowing the correct reading of regular words (CAT) and new words or pseudo-words (ZAT). (2) A lexical process that allows the correct reading of all words, and particularly of irregularly spelled words such as TWO. A common assumption is that English readers develop a sight vocabulary for all familiar words so that reading can take place either lexically or via sublexical grapheme/phoneme conversion. It is well established that each of these processes can be selectively impaired following brain damage. Selective damage to sublexical processes leads to specific difficulties in reading pseudo-words, while lexical damage affects real word reading (particularly of irregular words) and may also affect written word comprehension. The universality of this view of the reading process has been questioned.
Ardila (1991)
argued that reading in transparent orthographies such as Spanish could occur via sublexical grapheme–phoneme conversion alone. However, recent data contradict this view and rather suggests that a dual process may be needed even in transparent languages (
Basso and Paulin, 2003
,
Cuetos et al., 1996
,
Ferreres and Miravalles, 1995
,
Iribarren et al., 1999
).
In this study, we investigate this question in Welsh–English bilingual stroke patients. Welsh has a very transparent orthography. In theory, all words could be read (and spelled) successfully via a sublexical process and there is no obvious requirement to develop a sight vocabulary. This raises the possibility that patients with sublexical damage may have deficits restricted to pseudo-words in English, but extending to both words and pseudo-words in Welsh. If Welsh is read exclusively via a sublexical GP conversion process then one would expect that (1) patients with impaired Welsh pseudo-word reading should show comparable difficulties in reading Welsh words, (2) patients with impaired English pseudo-word reading may show relatively preserved English word reading, (3) patients with sublexical deficits should also show impaired lexical decision in Welsh, because correct LD would have to occur via (impaired) phonological mediation, (4) patients with sublexical deficits may show preserved lexical decision in English because correct LD could occur via direct access to the orthographic lexicon.

More information

Identifiers

ORCID for Elwyn Leek: orcid.org/0000-0001-9258-7504

Catalogue record

Export record

Altmetrics