Reading aloud: Evidence for the use of a whole word nonsemantic pathway.

Abstract

Abstract It is widely assumed that the presence of an associative priming effect during the oral reading of orthographies with consistent spelling - sound correspondences signals the use of an orthographic code for lexical access (the addressed routine). Relatedly, the failure to observe such a priming effect has been taken to indicate the use of a routine that relies on subword spelling - sound correspondence knowledge (the assembled routine). This logic depends on the assumptions that (a) only the addressed routine (whole word orthographic knowledge) can produce priming, and (b) that it necessarily does so (i.e., is automatic). The present experiments show that, taken alone, neither the presence nor absence of priming effects in oral reading permit an inference as to whether the addressed or assembled routine is used. Converging operations which do permit such an inference are reported. The data support the view that (i) components of the word recognition system operate interactively such that use of the assembled routine yields priming under certain conditions, and (ii) normal readers of a shallow orthography use a nonsemantic, whole word pathway to name words.Resume Il est generalement admis que, dans le cas d'orthographes peu profondes (ou il y a correspondance systematique entre la graphie et la prononciation), la presence d'un effet d'amoreage associatif durant la lecture orale marque l'utilisation d'un code orthographique pour acceder au lexique (programme adresse). Par ailleurs, on considere que l'absence d'un effet d'amoreage indique l'utilisation d'un programme qui repose sur la connaissance des correspondances entre la graphie et la prononciation a un niveau inferieur a celui du mot (programme assemble). Cette logique depend des hypotheses suivantes: a) seul le programme adresse (connaissance orthographique des mots complets) peut produire l'amoreage et b) il le produit necessairement (c'est - a - dire automatiquement). Les experiences que nous avons menees montrent qu'on ne peut inferer, uniquement de la presence d'effets d'amoreage ou de leur absence durant la lecture orale, que le programme adresse ou bien le programme assemble est utilise. Les operations convergentes qui permettent une telle inference sont exposees dans le rapport. Les donnees recueillies montrent que i) les composantes du systeme de reconnaissance des mots fonctionnent en interaction de sorte que l'utilisation du programme assemble produit un effet d'amoreage dans certaines conditions et ii) les lecteurs normaux qui utilisent une orthographe peu profonde empruntent une voie d'acces non semantique aux mots complets pour nommer les mots.Until recently, research on visual word recognition has focussed on questions concerning how words printed in English are read. This analysis has yielded a remarkable consensus concerning some of the underlying processes (e.g., see reviews by Carr & Pollatsek, 1985; Patterson & Coltheart, 1987; see also Paap, Noel & Johansen, 1992) and has resulted in the dual route model(f.1). An extension of this work to cross - orthography investigations has resulted in a large body of research. One position was articulated in the strong version of the Orthographic Depth Hypothesis, which held that the consistency of spelling - sound correspondences within an orthography dictates which of the routines of the dual route model are used during reading. Shallow orthographies (i.e., those with consistent spelling - sound correspondences) were argued to be always read aloud prelexically (e.g., Allport, 1979; Bridgeman, 1987; Hung & Tzeng, 1981; Morton & Sasanuma, 1984; Turvey, Feldman, & Lukatela, 1984). This position has since been tempered by the linguistic observation that syllabic stress can not always be derived on the basis of prelexical phonology (e.g., Katz & Frost, 1992). Consequently, some researchers assume that prelexical phonology serves to activate a phonological lexicon which then serves to mediate naming or semantic access (e. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 37 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.