Frontiers in Psychology (Apr 2014)
Can words be read without abstract letter identities?
Abstract
Most theories of reading and writing posit a level of representation that codes abstract letter identities (ALIs) independent of the font, case, and even modality in which the letter was presented or produced (e.g., Caramazza & Hillis, 1990). Most cognitive theories assume that ALIs are critical for mediating between peripheral (e.g., processing visual shapes, executing motor plans) and central (e.g., accessing semantic or lexical information) written language processes. We present a case study of an individual with a severe impairment in processing ALIs, in order to examine the role of ALIs in reading and writing. Case study CH was a 52 year-old male with 17 years of education who suffered a stroke five years prior to testing resulting in extensive damage to the left temporal, parietal and occipital lobes. Auditory word processing was intact, but he was impaired in spoken picture naming, producing many semantic or picture description errors. Single-letter processing tasks CH was administered tasks that probed his ability to process letter-shape and letter-identity information. His performance on these tasks is reported in Table 1. Tasks that tapped into letter-shape processing included direct copy, copying letters from a tactile presentation, within-case same-different matching, pseudoletter decision, and a mental rotation tasks that required deciding whether a letter at an odd orientation was facing the correct direction or was mirror reversed. CH was unimpaired on all of these tasks, indicating no impairment in processing the shapes of letters either in perception or production. Tasks that tap into abstract letter identity processing included naming letters from a visual or tactile presentation, letter writing-to-dictation, cross-case same-different matching, direct copy transcoding, and copy transcoding letters from a tactile presentation. CH was severely impaired on all letter-identity processing tasks. Given this pattern of performance, we conclude that CH has an impairment in processing ALIs but not letter-shape information. Whole word processing tasks CH was also given a series of whole word processing tasks. He was given a set of 80 words and 20 words to spell (both oral and written), read aloud and recognize via oral spelling. His impairment in many of these tasks was striking (Oral Spelling: 0/100, Written Spelling: 2/100; Recognizing Orally Spelled Words: 0/100). However, his single word reading was surprisingly intact; he read 47/80 words correctly (59%), but not a single nonword. Furthermore, many of his reading errors were semantic (e.g. LAST → “end”) suggesting that the approximate meaning could be extracted from the written word. Discussion CH’s acquired dyslexia and dysgraphia left him with a profound impairment in processing abstract letter identities. This impairment affected his ability to process strings of letters in a variety of tasks; for example nonword reading, spelling, recognizing orally spelled words. However, while impaired, his single word reading was surprisingly good given his single letter impairment, suggesting an additional route to word meaning from visually-presented familiar words that does not require abstract letter identities.
Keywords
