The internal structure of chaos: Letter category determines visual word perceptual units

Abstract

a b s t r a c tThe processes and the cues determining the orthographic structure of polysyllabic words remain far from clear. In the present study, we investigated the role of letter category (consonant vs. vowels) in the perceptual organization of letter strings. In the syllabic counting task, participants were presented with written words matched for the number of spoken syllables and comprising either one vowel cluster less than the number of syllables (hiatus words, e.g., pharaon) or the same number of vowel clusters (e.g., parodie). Relative to control words, readers were slower and less accurate for hiatus words, for which they systematically underestimated the number of syllables (Experiment 1). The effect was stronger when the instructions emphasized response speed (Experiment 2) and when concurrent articulation was used (Experiment 3), and the effect did not stem from phonological structure (Experiment 4). Furthermore, hiatus words were more slowly and less accurately pronounced than control ones (Experiment 5). Finally, in lexical decision, opposite effects occurred as a function of word length, with shorter words producing a facilitatory effect and longer words showing interference (Experiment 6). Taken together, the results show that perceptual units extracted from visual letter strings are influenced by the orthographic status of letters. We discuss the implications of such findings in view of current theories of visual word recognition.Ó 2012 Elsevier Inc. All rights reserved. IntroductionWhat do we see when we read a word? What are the functional units of word perception? These questions opened a paper by Santa, Santa, and Smith in 1977, who stated that the issue 'has generated an impressive body of literature in the last 75 years, but there is little agreement on an answer' (p. 585). More than 30 years later, the situation has hardly changed, and the claim still holds true.The issue of orthographic coding in the perception of letter strings has been of interest since the earliest times of research on reading (Huey, 1908). The interest keeps on being high currently because understanding the basic processes of visual word recognition constitutes a keystone of any theory of reading. Visual word recognition is not performed letter by letter, but rather operates on larger letter chunks that are processed simultaneously. Hence, a recurrent question in the field is to determine what processing units are involved in the early steps of written word identification and how the perceptual processing system organizes letter strings into larger units. In the present paper, we report a set of studies aimed at exploring the role of letter category (consonant vs. vowel letters) in determining the internal structure of polysyllabic words.The issue of perceptual units has been approached from different angles according to periods and dominant trends in the field. Below, we present an overview of the major approaches and then discuss their relevance for the perceptual processing of polysyllabic words, which is of sp...

Orthographic and Morphemic Effects in the Written Syllable Counting Task

“…This replicates the findings reported by Chetail and Content (2012) according to which the CV pattern constrains the orthographic structure of letter strings. More importantly, CV pattern and morphemic structure 10 underestimation occurred both in prefixed (e.g., réunion) and non-prefixed (e.g., création)…”

“…Inhibitory effects of syllabic neighbourhood have also been confirmed in masked priming experiments (e.g., Carreiras & Perea, 2002;Dominguez, de Vega, & Cuetos, 1997; in Spanish; Mathey, Doignon-Camus, & Chetail, 2013, in French). For example, Mathey et al (2013) showed that a French word like rocher (/ʀɔ.ʃe/) which has a first syllable of low frequency was recognized more slowly when it was preceded by a pseudoword prime sharing the first syllable (e.g., robane, /ʀɔban/) rather than the first letters (e.g., roisie, /ʀwasi/).The proposition that polysyllabic words are parsed into letter clusters corresponding to phonological syllables contrasts with the recent work of Chetail and Content (2012, CV pattern and morphemic structure 4 2014 Chetail, Scaltritti, & Content, in press, in Italian) suggesting that the segmentation of words into small units is primarily driven by orthographic cues. According to this hypothesis, the organization of consonant and vowel letters within words (i.e., the CV pattern) constrains the perceptual structure of letter strings, with each vowel or vowel cluster underlying one orthographic unit.…”

“…This level of representations could be activated early during the time course of word recognition and may be engaged in tasks tapping both lexical and pre-lexical levels of processing (e.g., syllable counting task, lexical decision task, perceptual tasks, see Chetail & Content, 2012, 2014.…”

“…The mismatch between the two structures leads either to errors in favour of vowel-centred units, or to longer processing, necessary to solve the conflict. Accordingly, Chetail and Content (2012) showed that when the intentional resort to phonology was reduced (concurrent articulation), the influence of orthographic information increased, leading to a stronger hiatus effect. CV pattern and morphemic structure 5 Hiatus words are particularly appropriate to disentangle orthographic and phonological activation during visual word recognition, due to the mismatch between orthographic vowelcentred units and phonological syllables.…”

Electrophysiological markers of syllable frequency during written word recognition in French

“…Such representations may develop during learning to read via exposure to frequently cooccurring letter combinations, as frequent orthographic syllables naturally tend to correspond to frequent orthographic clusters (Adams, 1981;Seidenberg, 1987;see Conrad et al, 2010) and orthographic chunking of letter strings into orthographic syllablelike units may be driven by the distinction between consonant and vowel letters (Chetail & Content, 2012; see also Kandel, Hé rault, Grosjacques, Lambert, & Fayol, 2009). …”