| Sumario: | The article focuses on the effect of orthographic transformation upon speed and accuracy of semantic categorizations. A study found support for each of the competing models of word recognition. Both serial and parallel processing seem to occur, but under different conditions. When fluent readers encounter familiar word presented in regular orthography, they seem able to chunk that information holistically, at least within the upper bounds of 3 to 6 letter words. On the other hand, when fluent readers encounter words which tend to pose a decoding problem, evidence for serial letter-by-letter processing is found. Fluent readers adopt different strategies of word recognition, depending upon factors having to do with the ease of decoding the visual input.
|
|---|