For some time, it has been assumed that writing requires phonological mediation (orthographic word forms cannot be activated directly from semantics, but only indirectly from phonological word forms). More recently, phonological mediation has been questioned, due to the observation that on tasks that require the production of spoken and written responses on the same naming attempt, some aphasic subjects produce different words (e.g. a correct written response and a spoken semantic error, or vice versa; or, two distinct semantic errors). This pattern of performance cannot be accounted for by phonological mediation, but is easily accommodated by the orthographic autonomy hypothesis, which allows the direct activation of orthographic word forms from semantics. However, the observation that damage to word meaning or to word form usually results in comparable difficulties in spoken and written output, suggests that phonological and orthographic word forms, albeit autonomous, can interact. The available data are compatible both with a "weak" version of orthographic autonomy (orthographic word forms are activated both directly from semantics and indirectly from phonological word forms), and with a "strong" version of the same hypothesis (orthographic forms are directly activated from semantics; they do not receive direct activation from phonological forms; and they are allowed to interact with phonological forms via sublexical conversion mechanisms).
|Number of pages||32|
|Journal||Language and Cognitive Processes|
|Publication status||Published - Oct 1997|