Fabienne Chetail

Understanding the front end of visual word recognition requires us to identify the processes by which letters are identified. Since most of the work on letter recognition has been conducted in...

Written symbols such as letters have been used extensively in cognitive psychology, whether to understand their contributions to written word recognition o

Recent studies in alphabetic writing systems have investigated whether the status of letters as consonants or vowels influences the perception and processing of written words. Here, we examined to what extent the organisation of consonants and vowels within words affects performance in a syllable counting task in English. Participants were asked to judge the number of syllables in written words that were matched for the number of spoken syllables but comprised either 1 orthographic vowel cluster less than the number of syllables (hiatus words, e.g., triumph) or as many vowel clusters as syllables (e.g., pudding). In 3 experiments, we found that readers were slower and less accurate on hiatus than control words, even when phonological complexity (Experiment 1), number of reduced vowels (Experiment 2), and number of letters (Experiment 3) were taken into account. Interestingly, for words with or without the same number of vowel clusters and syllables, participants’ errors were more likely to underestimate the number of syllables than to overestimate it. Results are discussed in a cross-linguistic perspective.

Although the majority of research in visual word recognition has targeted single-syllable words, most words are polysyllabic. These words engender special challenges, one of which concerns their analysis into smaller units. According to a recent hypothesis, the organization of letters into groups of successive consonants (C) and vowels (V) constrains the orthographic structure of printed words. So far, evidence has been reported only in French with factorial studies of relatively small sets of items. In the present study, we performed regression analyses on corpora of megastudies (English and British Lexicon Project databases) to examine the influence of the CV pattern in English. We compared hiatus words, which present a mismatch between the number of syllables and the number of groups of adjacent vowel letters (e.g., client), to other words, controlling for standard lexical variables. In speeded pronunciation, hiatus words were processed more slowly than control words, and the effect was stronger in low-frequency words. In the lexical decision task, the interference effect of hiatus in low-frequency words was balanced by a facilitatory effect in high-frequency words. Taken together, the results support the hypothesis that the configuration of consonant and vowel letters influences the processing of polysyllabic words in English.

Humans are known to continuously extract regularities from the flow of stimulation. This occurs in many facets of behaviour, including reading. In spite of the ubiquitous evidence that readers become sensitive to orthographic regularities after very little exposure to print, the role of orthographic regularities receives at best a peripheral status in current theories of orthographic processing. In the present article, after the presentation of previous evidence on orthographic redundancy, the hypothesis that orthographic regularities may play a prominent role in word perception is developed.

A thorough knowledge regarding monosyllabic word reading has been accumulated over decades, which contrasts with our understanding of polysyllabic word processing. One reason why modelling of polysyllabic word reading is lagging behind might be related to the issue of orthographic segmentation, parsing requiring the integration of two types of information, the number of units to be extracted and boundaries between these units. In the present study, we focussed on the effect of number of syllables, and we compared lexical decision latencies in French words and pseudowords as a function of syllabic length (two vs. three graphosyllables). An effect was found in pseudowords, low-frequency words and high-frequency words, items with three syllables being processed more slowly than items with two syllables. We discuss what processes of current models of visual word recognition may underlie this effect, and based on previous studies and analyses on word mega corpus, we propose a new interpretation of the effect in terms of number of orthographic vowel-centred units.

According to a recent hypothesis, the organization of letters into groups of successive consonants and vowels (i.e., CV pattern) constrains the orthographic structure of words. Here, we examined to what extent the morphological structure of words modifies the influence of the CV pattern in a syllable counting task. Participants were presented with written words matched for the number of syllables and comprising either one vowel cluster less than the number of syllables (hiatus words, e.g., création) or the same number of vowel clusters (control words, e.g., crépiter). Participants were slower and less accurate for hiatus than control stimuli, be it words (Experiments 1, 3) or pseudowords (Experiment 2). More importantly, this hiatus effect was present even when the stimuli had a morphemic boundary falling within the hiatus (e.g., ré-agir). The results suggest that the CV pattern of items more strongly influences performance in the syllable counting task than the morphological structure.