Abstract
Letter-by-letter (LBL) dyslexia is characterized by slow and laborious reading where reading latency increases markedly with the number of letters in a word. Interestingly, reading rate is also affected by high-level factors (i.e. imageability and lexical frequency), which suggests an implicit lexical/semantic access prior to conscious word identification. Fiset et al. (2006) recently proposed that the critical spatial frequencies for reading (between 2.5 to 3 cycles per letter) may be unavailable in LBL dyslexia and that implicit lexical/semantic access is mediated by lower spatial frequencies, which would fail however to offer information that is sufficiently accurate for explicit word recognition. To compensate, LBL readers would use high spatial frequencies for the sequential explicit identification of individual letters, which is the diagnostic feature of the disorder. The aim of the current study is to further investigate the special role of the medium frequencies for word recognition in normal readers. The critical medium spatial frequencies were filtered out from the words presented for overt reading, which therefore comprised only high and low ([[lt]]2 and [[gt]]6 cycles per letter) spatial frequencies. The results replicate the main features of LBL dyslexia. Reading latency increased linearly as a function of the number of letters in the word while being affected by imageability and lexical frequency. Error rates were relatively low, as in most LBL dyslexic cases. We thus caused letter-by-letter dyslexia in normal readers by depriving them of medium spatial frequency information. These findings are consistent with the crucial importance of this information and strongly suggest that dyslexics are truly unable to process these medium spatial frequencies for reading.
This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) to Martin Arguin and Frédéric Gosselin and by a graduate scholarship from the Fonds Québécois de Recherche en Nature et Technologies (FQRNT) to Karine Tadros