Poster de conférence
Résumé : Many studies of immediate serial recall (ISR) of verbal items have shown that the short-term storage of item information is supported by the activation of language knowledge stored in long-term memory, such as the lexical and semantic representations (for a review, see Majerus, 2009, in A. Thorn (Ed.), New York: Psychology Press). Similarly, as phonological and orthographic memory representations are highly connected in literate people, orthographic knowledge may influence memory performance in purely auditory ISR tasks. In adult participants, Pattamadilok, Lafontaine, Morais, and Kolinsky (2010, Language and Speech) showed that inter-item orthographic dissimilarity helps to reduce the deleterious effect of phonological similarity. They observed that, in a seven-word auditory ISR task, compared to words that shared neither the phonological nor the orthographic rhyme, performance was less affected when words rhymed but had different spellings than when they both rhymed and had the same spelling. However, the recall benefit due to orthographic dissimilarity was only observed at positions four to six of the word list. The present study provides converging children data. Using an ISR task of auditory five-word lists in 3rd grade children (mean age: 8.4 years old), we observed that inter-item orthographic dissimilarity (e.g., the list sport, bord, corps, nord, store, all ending with the same rhyme vs. the list passe, chasse, tasse, brasse, classe) assists the recall of item information (i.e., the proportion of items recalled independently of serial position) but not the recall of order information (i.e., absolute accuracy divided by item accuracy). In addition, only performance in the condition of phonological similarity and orthographic dissimilarity was significantly correlated with performance in the spelling task on the items used in the ISR, suggesting that the higher item accuracy obtained in this condition was partly due to the use of spelling knowledge. Consistently, irregular word reading (which is generally viewed as reflecting orthographic processing abilities) was highly correlated with the size of the orthographic effect. It worth noting that children did not notice that the spellings of the rhyming words could be dissimilar or not. These findings thus suggest that, in case of phonological similarity, spelling knowledge is automatically used to support the maintain of auditory items in memory, an effect that depends on the level of reading and spelling abilities.