Our language environment is rich in statistical structure. It is therefore natural to hypothesize that statistical learning (SL) —the ability to detect regularities in the environmental input— contributes to the efficient processing of written text and supports literacy skills. Some prior studies reported correlations between SL as a general ability and reading, while others found null effects. We propose that this link may be better revealed by measuring individuals’ learning of orthographic statistical patterns, a type of regularity more directly relevant to literacy. This study examined how sensitivity to bigram frequency in participants’ first language (L1, French) and the ability to learn novel bigrams through exposure relate to literacy outcomes. The measures showed moderate to good reliability (L1 sensitivity: r = .60; novel orthographic learning: r = .84) and were significantly associated with spelling, but not reading. Notably, sensitivity to bigram frequency was not significantly associated with orthographic learning performance and did not mediate the link between orthographic learning and spelling. These findings suggest that different aspects of sensitivity to existing and newly acquired orthographic patterns independently support spelling abilities in French, without extending to reading performance.

Regular print exposure is thought to benefit reading and language processes: those who read more have a larger vocabulary and better spelling and comprehension skills. Yet, there is little or no direct evidence that exposure to print facilitates reading. Here, we used an ecologically valid design to test the impact of print exposure on the early stages of reading in skilled adult readers. Participants read a novel at their own pace. Reading was followed by a lexical decision task, in which the positive trials were words that were exposed in the novel, and matched controls not exposed in the novel. If exposure during reading had a positive impact on subsequent word recognition, exposed words would be processed more efficiently than not-exposed words (exposure effect). This effect was obtained in three experiments. In addition, the effect was not modulated by the amount of exposure (1 vs. 3 occurrences in the text; Experiment 1), or the timing between reading and the exposure test (immediately after reading vs. on the following day; Experiment 3). However, the effect was present only in low-frequency words (Experiment 3). Interpretations of the exposure effect in terms of activation threshold and lexical quality are discussed.

PURPOSE:Children are assumed to acquire orthographic representations during autonomous reading by decoding new written words. The present study investigates how deaf and hard of hearing (DHH) children build new orthographic representations compared to typically hearing (TH) children.METHOD:Twenty-nine DHH children, from 7.8 to 13.5 years old, with moderate-to-profound hearing loss, matched for reading level and chronological age to TH controls, were exposed to 10 pseudowords (novel words) in written stories. Then, they performed a spelling task and an orthographic recognition task on these new words.RESULTS:In the spelling task, we found no difference in accuracy, but a difference in errors emerged between the two groups: Phonologically plausible errors were less common in DHH children than in TH children. In the recognition task, DHH children were better than TH children at recognizing target pseudowords. Phonological strategies seemed to be used less by DHH than by TH children who very often chose phonological distractors.CONCLUSIONS:Both groups created sufficiently detailed orthographic representations to complete the tasks, which support the self-teaching hypothesis. DHH children used phonological information in both tasks but could use more orthographic cues than TH children to build up orthographic representations. Using the combination of a spelling task and a recognition task, as well as analyzing the nature of errors, in this study, provides a methodological implication for further understanding of underlying cognitive processes.

This article presents the ENglish Reading Online (ENRO) project that offers data on English reading and listening comprehension from 7,338 university-level advanced learners and native speakers of English representing 19 countries. The database also includes estimates of reading rate and seven component skills of English, including vocabulary, spelling, and grammar, as well as rich demographic and language background data. We first demonstrate high reliability for ENRO tests and their convergent validity with existing meta-analyses. We then provide a bird's-eye view of first (L1) and second (L2) language comparisons and examine the relative role of various predictors of reading and listening comprehension and reading speed. Across analyses, we found substantially more overlap than differences between L1 and L2 speakers, suggesting that English reading proficiency is best considered across a continuum of skill, ability, and experiences spanning L1 and L2 speakers alike. We end by providing pointers for how researchers can mine ENRO data for future studies.

Lexical competition between newly acquired and already established representations of written words is considered a marker of word integration into the mental lexicon. To date, studies about the emergence of lexical competition involved mostly artificial training procedures based on overexposure and explicit instructions for memorization. Yet, in real life, novel word encounters occur mostly without explicit learning intent, through reading texts with words appearing rarely. This study examined the lexical integration of words learned through text reading. In Experiment 1, two groups of participants read a short book with embedded novel words. Only one group was asked to memorize the unfamiliar words. In the semantic categorization task, we found evidence for lexical competition with slower responses to existing orthographic neighbors (e.g., hublot) of the newly learned words (e.g., hubbot) than to a set of matched items. This effect was found independently of the group 24 h after initial exposure. In addition, a facilitation pattern was observed immediately after the reading session. However, post hoc analyses suggested that the competition effect was mainly driven by the data from the group receiving explicit learning instructions. Experiment 2 aimed to replicate the findings obtained in the group without explicit learning instructions. The results revealed the same pattern, characterized by a facilitatory effect immediately after the reading session and an inhibitory effect 24 h after the exposure. Overall, these results showed that lexical competition emerged from a naturalistic reading after a delay, regardless of whether participants were asked to learn novel words or not.

According to the instance-based approach, each novel word encounter is encoded as an episodic trace, including different aspects of word knowledge (orthography, semantics, phonology) and context. Experiencing the novel word again leads to reactivating the previous instances to support word identification. Accordingly, once a link between orthography and meaning is established through several instances of co-occurrence, presenting the novel word form enhances semantic learning even if the contexts are uninformative about the meaning (Eskenazi et al., 2018). Here, we investigated whether informative contexts enhance orthographic learning in the absence of the novel word form. Participants read pseudowords in three definition-like sentences, followed by three unrelated filler sentences (baseline condition), three uninformative sentences (orthographic condition), or three informative sentences with synonyms replacing the pseudoword (semantic condition). After reading, participants were better at spelling pseudowords exposed in the semantic than in the baseline condition and recalled more definitions of the pseudowords exposed in the orthographic than in the baseline condition. Such results indicate that both semantic and orthographic learning benefit from the contexts where the target information is absent. Overall, this supports the instance-based approach and contributes to the understanding of the interplay between orthography and semantics in contextual word learning.

Encountering new words during reading is a common experience of any adult reader and it constitutes an important source of word knowledge. Despite the ubiquity of this phenomenon, studies remain scarce in adults. Here, we addressed new word acquisition in adults using a natural contextual exposure, and we focused on the influence of context informativeness on orthographic learning. Indeed, previous studies suggest that the availability of semantic information plays an important role in orthographic learning in adults, but no such advantage was found in children. We hypothesised that this discrepancy comes from the fact that new word learning was examined almost only through artificial settings in adults. On the contrary, in the present study, adult participants were exposed to new words by reading a book. Half of the new words were embedded in informative contexts (easy to infer new word meaning) and the other half in less informative contexts. Both recall and recognition tasks were used to assess orthographic and semantic learning. The results showed efficient learning of the orthographic form and no reliable effect of the context informativeness. Regarding semantic learning, we found that if a word was correctly spelled, its definition was more likely to be retrieved. This shows that the orthographic and semantic dimensions of a word co-occurring in context are likely to influence each other during lexical acquisition. More generally, the present experiment showed that using an ecological learning design is essential to understand the mechanism of new word learning.

Selon les résultats de l’enquête du Programme International pour le Suivi des Acquis des élèves (PISA) de 2018, un élève de 15 ans sur quatre est considéré comme un lecteur peu performant. La génération d’inférences joue un rôle majeur parmi les processus de compréhension en lecture et représente une source potentielle de difficultés. Il n’est donc pas étonnant d’observer un développement important d’interventions ciblées sur cette compétence. Leurs effets pourraient néanmoins être partiellement médiés par la motivation à lire. En conséquence, quelle que soit la nature des interventions, si les adolescents ne sont pas motivés à s’y engager, celles-ci n’auront probablement que peu d’effet durable. Cependant, la motivation à lire est rarement considérée dans l’élaboration de ces interventions. Nous discutons donc de la pertinence d’un outil prometteur, qui pourrait à la fois soutenir l’entraînement des compétences de compréhension en lecture et favoriser la motivation à lire des adolescents : le roman graphique.

According to the results of the 2018 Programme for International Student Assessment (PISA) survey, one in four 15-year-old students is considered a low-performing reader. Inference generation plays a major role in reading comprehension and represents a potential source of difficulties. It is therefore not surprising to observe a significant development of interventions targeted at this skill. However, their effects could be partly mediated by reading motivation. Therefore, whatever the nature of the interventions, if adolescents are not moti-vated to engage in them, they are likely to have little lasting effect. Nevertheless, reading motivation is rarely considered in the development of these interventions. Thus, we discuss the relevance of a promising tool, which could both support the training of reading compre-hension skills and foster reading motivation in adolescents: the graphic novel.

Chinese character recognition is based on a limited set of recurrent stroke patterns. Most Chinese characters are a combination of two or more of these components. To test whether readers are sensitive to combinations of components (or multi-component units [MCUs]) within a character, we conducted two probe detection tasks where participants had to detect the presence of a component in a target character. Critically, some targets contained an MCU that can stand as a character on its own, with its own meaning and sound, while other targets contained an MCU that only exists embedded within other characters (no associated meaning and sound). Participants had more difficulty detecting component probes that were a part of an existing MCU, compared to component probes that belonged to a non-existing MCU. These findings suggest that existing MCUs are a perceptual unit in Chinese character recognition.

The current study investigated the extraction of orthographic and phonological structure of written words in adults with dyslexia. In adults without learning difficulties, Chetail and Content showed that orthographic structure, as determined by the number of vowel letter clusters, influences visual word length estimation. The authors also found a phonological effect determined by the number of syllables of words. In the present study, 22 French-speaking students diagnosed with dyslexia in childhood and 22 students without learning disabilities were compared. All participants performed the task of estimating word length. The pattern of results obtained by Chetail and Content was replicated: length estimates were biased by both the number of syllables and the number of vowel letter clusters. The study showed a significant interaction between phonological bias and group. The phonological effect was less important in students with dyslexia, suggesting reduced sensitivity to phonological parsing in reading. In contrast, the orthographic effect did not differ significantly between groups, suggesting that the sensitivity to the orthographic structure of written words is preserved in students with dyslexia despite their low-quality orthographic representations. We conclude that there is no systematic association between sensitivity to the structure of representations and quality of their content.

Readers capture statistics about letter co-occurrences very rapidly. This has been demonstrated with artificial lexicons and/or with restricted sets of orthographic regularities. The aim of the study was twofold: To examine the learning of new orthographic regularities in a more incidental exposure paradigm, and to investigate the impact of the diversity of letter contexts in which new orthographic regularities appear. For 2 months, participants played detection games for 20 min per day and were exposed to a large set of pseudowords, some of which included new bigrams (e.g., GK). Half of the new bigrams occurred in eight different items (high contextual diversity) and the other half were presented in only two items (low context diversity). At six time points, the participants performed a "wordlikeness" task in which they chose between two new pseudowords the one that was more similar to the items previously exposed (e.g., PUGKALE vs. PUGZALE). The results showed that the participants very rapidly developed a preference for items with a frequent new bigram and that this sensitivity increased steadily over the 2 months. Furthermore, the sensitivity to these new orthographic regularities was higher in cases of high letter contextual diversity. The latter result parallels what is observed at a lexical level with semantic contextual diversity.

In the last decades, repeated evidence for graphemic effects has been reported in skilled readers. For example, a letter is more easily detected in a word when it corresponds to a simple grapheme (e.g., A in PLACE) than when it is embedded in a complex one (e.g., A in BEACH). Such effects have been taken as a demonstration that graphemes are processed as perceptual units by the reading system. However, this conclusion has been recently challenged by studies using different experimental designs. In the present study, we used four experimental situations to get a clearer picture of the reliability of graphemic effects. We used four types of tasks: letter detection (Experiment 1), length estimation (Experiment 2), mixing case lexical decision (Experiment 3), and primed lexical decision (Experiments 4 and 5). In each task, the processing of words with complex graphemes (e.g., BEACH) was compared to the processing of words with simple graphemes (e.g., PLACE). Overall, we found no reliable grapheme effect, supporting the claim that graphemes are not perceptual units in skilled visual word recognition. An alternative interpretation of 'grapheme effects' previously reported with the letter detection task is discussed.

Resume Au cours de la derniere decennie, de plus en plus de travaux concernant les processus de lecture prennent en compte le niveau en orthographe des lecteurs adultes. Cet interet repose sur l’idee que la connaissance precise de l’orthographe des mots serait un indice de la qualite des representations lexicales, elle-meme essentielle pour assurer l’efficacite des processus de lecture. Cependant, il manque a l’heure actuelle un outil pour la langue francaise qui permettrait d’evaluer le niveau en orthographe lexicale specifiquement chez l’adulte normo-lecteur. L’article presente les etapes de construction, de validation et de creation de normes du test Brussels Orthographic Quality Scale (BOQS), qui evalue le niveau en orthographe lexicale dans la population adulte. Au final, ce test s’avere etre simple d’utilisation, rapide, fiable et apporte une plus-value par rapport aux tests existant dans les batteries d’evaluation des troubles du langage ecrit.

Recent findings demonstrated readers' sensitivity to the distinction between consonant and vowel letters. Especially, the way consonants and vowels are organised within written words determines their perceptual structure. The present work attempted to overcome two limitations of previous studies by examining the neurophysiological correlates of this perceptual structure through magnetoencephalography (MEG). One aim was to establish that the extraction of vowel-centred units takes place during early stages of processing. The second objective was to confirm that the vowel-centred structure pertains to the word recognition system and may constitute one level in a hierarchy of neural detectors coding orthographic strings. Participants performed a cross-case matching task in which they had to judge pairs of stimuli as identical or different. The critical manipulation concerned pairs obtained by transposing two letters, so that the vowel-centred structure was either preserved (FOUVERT-fovuert, two vowel letter clusters) or modified (BOUVRET-bovuret). Mismatches were detected faster when the structure was modified. This effect was associated with a significant difference in evoked neuromagnetic fields extending from 129 to 239 msec after the stimulation. Source localization indicated a significant effect in the visual word form area around 200 msec. The results confirm the hypothesis that the vowel-centred structure is extracted during the early phases of letter string processing and that it is encoded in left fusiform regions devoted to visual word recognition.

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 English, letter perception modeling has been limited to the 26 letters of the Latin alphabet. However, many writing systems include letters with diacritic marks. In the present study, we examined whether diacritic letters are a mere variant of their base letter, and thus share the same abstract representation, or whether they function as separate elements from any other letters, and thus have separate representations. In Experiments 1A and 1B, participants performed an alphabetical decision task combined with masked priming. Target letters were preceded by the same letter (e.g., a–A), by a diacritic letter (e.g., â–A), or by an unrelated letter (e.g., z–A). The results showed that the primes sharing nominal identity (e.g., a) facilitated target processing as compared to unrelated primes (e.g., z), but that primes that included a diacritic mark (e.g., â) did not, with reaction times being similar to those in the unrelated priming condition. In Experiment 2 we replicated these results in a lexical decision task. Overall, this demonstrates that as long as diacritics are used in scripts to distinguish between lexical entries, the diacritic letters are not mere variants of their base letters but constitute unitary elements of the script in their own right, with diacritics contributing to the overall visual shape of a letter.

Individuals rapidly become sensitive to recurrent patterns present in the environment and this occurs in many situations. However, evidence of a role for statistical learning of orthographic regularities in reading is mixed, and its role has peripheral status in current theories of visual word recognition. Additionally, exactly which regularities readers learn to be sensitive to is still unclear. To address these two issues, three experiments were conducted with artificial scripts. In Experiments 1a and 1b, participants were exposed to a flow of artificial words (five characters) for a few minutes, with either two or four bigrams occurring very frequently. In Experiment 2, exposure took place over several days while participants had to learn the orthographic and phonological forms of new words entailing or not frequent bigrams. Sensitivity to these regularities was then tested in a wordlikeness task. Finally, participants performed a letter detection task, with letters being either of high frequency or not in the exposure phase. The results of the wordlikeness task showed that after only a few minutes, readers become sensitive to the positional frequency of letter clusters and to bigram frequency beyond single letter frequency. Moreover, this new knowledge influenced the performance in the letter detection task, with high-frequency letters being detected more rapidly than low-frequency ones. We discuss the implications of such results for models of orthographic encoding and reading.

In the domain of working memory, recent theories postulate that the maintenance of serial order is driven by position marking. According to this idea, serial order is maintained though associations of each item with an independent representation of the position that the item constitutes in the sequence. Recent studies suggest that those position markers are spatial in nature, with the beginning items associated with left side and the end elements with the right side of space (i.e., the ordinal position effect). So far however, it is unclear whether serial order is coded along the same principles in the verbal and the visuospatial domain. The aim of the current study was to investigate whether serial order is coded in a domain general fashion or not. To unravel this question, 6 experiments were conducted. The first 3 experiments revealed that the ordinal position effect is found with verbal but not with spatial information. In the subsequent experiments, the authors isolated the origin of this dissociation and conclude that to obtain spatial coding of serial order, it is not the nature of the encoded information (verbal, visual, or spatial) that is crucial, but whether the memoranda are semantically processed or not. This work supports the idea that serial order is coded in a domain general fashion, but suggests that position markers are only spatially coded when the to-be-remembered information is processed at the semantic level.

According to a widespread view on functional units in word reading, the perceptual structure of printed words is constrained by print-to-speech mappings. Here, we examined the hypothesis that the organization of consonant and vowel letters (the CV pattern) determines the perceived structure of letter strings. Skilled readers were presented with two kinds of bisyllabic French words. Half of the words included a silent E between two consonants (e.g., gobelet, /goble/) thus entailing three orthographic vowel groups, while the other half were control words with two vowel groups (e.g., cremeux, /kRemo/). Participants had to decide on the number of units or reproduce the physical length of the stimuli. Silent E words were consistently estimated to be longer (more units, longer lines) than control words, despite being matched in number of letters and phonemes. The effect was present both in the written modality (Experiments 1A-1B) and in the spoken modality (Experiments 2A-2B). When prereaders and beginning readers with limited knowledge of the orthographic form of words performed the tasks (Experiments 3A-3B), no bias was found, confirming its orthographic nature in skilled readers. The study provides a clear confirmation of the predictions based on the CV pattern hypothesis according to which the number of vowel letter clusters determines the perceived units of letter strings. (C) 2017 Elsevier Inc. All rights reserved.