Do we need more than 50.000 words in the orthographic lexicon to be able to read?

Abstract

The goal of this paper is to explain the reading decoding process. I begin at the moments preceding it (choosing what to read, pre-reading, eye-fixation, up to the phase when stains are sprayed and then turned into invariant features). A convergence then occurs, and the invariant features of both left and right V4 areas are sent to the left ventral occipitotemporal cortex. Recycled neurons, then, recognize which, how many, and how the invariant letter features combine for differentiating one letter from another. I emphasize a rather neglected difference in reading processing models, between letter recognition and graphemes, demonstrating that letter recognition is language independent, while the grapheme value, or phoneme, is its raison d’être, for the grapheme represents the phoneme in the writing system. I present the orthographical context independent values many graphemes have, in the written Brazilian Portuguese (BP), as well as several cases of predictability of the values depending on the orthographical context or through metalanguage and/or the morphosyntactic and semantic textual context. Several arguments falsify the existence of an orthographic lexicon for the decoding of words, as an obligatory condition for their recognition.