Computational Models Of Visual Word Recognition Research Articles

Does bold emphasis facilitate the process of visual-word recognition?

The study of the effects of typographical factors on lexical access has been rather neglected in the literature on visual-word recognition. Indeed, current computational models of visual-word recognition employ an unrefined letter feature level in their coding schemes. In a letter recognition experiment, Pelli, Burns, Farell, and Moore-Page (2006), letters in Bookman boldface produced more efficiency (i.e., a higher ratio of thresholds of an ideal observer versus a human observer) than the letters in Bookman regular under visual noise. Here we examined whether the effect of bold emphasis can be generalized to a common visual-word recognition task (lexical decision: "is the item a word?") under standard viewing conditions. Each stimulus was presented either with or without bold emphasis (e.g., actor vs. actor). To help determine the locus of the effect of bold emphasis, word-frequency (low vs. high) was also manipulated. Results revealed that responses to words in boldface were faster than the responses to the words without emphasis -this advantage was restricted to low-frequency words. Thus, typographical features play a non-negligible role during visual-word recognition and, hence, the letter feature level of current models of visual-word recognition should be amended.

Does Viotin Activate Violin More Than Viocin?

The vast majority of neural and computational models of visual-word recognition assume that lexical access is achieved via the activation of abstract letter identities. Thus, a word's overall shape should play no role in this process. In the present lexical decision experiment, we compared word-like pseudowords like viotín (same shape as its base word: violín) vs. viocín (different shape) in mature (college-aged skilled readers), immature (normally reading children), and immature/impaired (young readers with developmental dyslexia) word-recognition systems. Results revealed similar response times (and error rates) to consistent-shape and inconsistent-shape pseudowords for both adult skilled readers and normally reading children - this is consistent with current models of visual-word recognition. In contrast, young readers with developmental dyslexia made significantly more errors to viotín-like pseudowords than to viocín-like pseudowords. Thus, unlike normally reading children, young readers with developmental dyslexia are sensitive to a word's visual cues, presumably because of poor letter representations.

On the role of the upper part of words in lexical access: Evidence with masked priming

More than 100 years ago, Huey (1908/1968) indicated that the upper part of words was more relevant for perception than the lower part. Here we examined whether mutilated words, in their upper/lower portions (e.g., ), can automatically access their word units in the mental lexicon. To that end, we conducted four masked repetition priming experiments with the lexical decision task. Results showed that mutilated primes produced a sizeable masked repetition priming effect. Furthermore, the magnitude of the masked repetition priming effect was greater when the upper part of the primes was preserved than when the lower portion was preserved-this was the case not only when the mutilated words were presented in lower case but also when the mutilated words were presented in upper case. Taken together, these findings suggest that the front-end of computational models of visual-word recognition should be modified to provide a more realistic account at the level of letter features.

The time course of orthography and phonology: ERP correlates of masked priming effects in Spanish

One key issue for computational models of visual-word recognition is the time course of orthographic and phonological information during reading. Previous research, using both behavioral and event related brain potential (ERP) measures, has shown that orthographic codes are activated very early but that phonological activation starts to occur immediately afterward. Here we report an ERP masked priming experiment in Spanish that investigates this issue further by using very strict control conditions. The critical phonological comparison was between two pairs of primes having the same orthographic similarity to the target words but differing in phonological similarity (e.g., conal-CANAL vs. cinal-CANAL vis à vis ponel-PANEL vs. pinel-PANEL), whereas the critical orthographic contrast was between pairs of primes that had the same phonological similarity to the target but differing in orthographic similarity (e.g., conal-CANAL vs. konal-CANAL). Orthographic priming was mainly observed in the 150-250-ms time window whereas phonological priming occurred in the 350-550-ms window.