Basic processes in reading : is the orthographic depth hypothesis sinking ?