Contribution of orthography to vocabulary acquisition in a second language: Evidence of an early word-learning advantage in elementary-school children.

Recent studies have shown that children benefit from orthography when learning new words. This orthographic facilitation can be explained by the fact that written language acts as an anchor device due to the transient nature of spoken language. There is also a close and reciprocal relationship between spoken and written language. Second-language word learning poses specific challenges in terms of orthography-phonology mappings that do not fully overlap with first-language mappings. The current study aimed to investigate whether orthographic information facilitates second-language word learning in developing readers, namely third and fifth graders. In a first experiment French children learned 16 German words, and in a second experiment they learned 24 German words. Word learning was assessed by picture designation, spoken word recognition, and orthographic choice. In both experiments, orthographic facilitation was found in both less and more advanced readers. The theoretical and practical implications of these findings are discussed.

Investigating L2 reading aloud and silent reading in typically developing readers and dyslexic adolescents from grades 6 to 9.

The goal of this study was to examine the reading performance of French typically developing readers and dyslexic adolescents from grades 6 to 9 in English as a second language (L2) learned in a school context. Lexicality effects and the impact of two sub-lexical variables, that is cross-language orthographic markedness and congruency of grapheme-to-phoneme correspondences (GPCs), were investigated in three tasks: L2 reading aloud and lexical decision, and L2-to-L1 translation. English words and nonwords were divided into three conditions: (a) marked condition in which items have an L2-specific orthographic pattern (e.g., town), (b) unmarked congruent condition in which items have an L1/L2 shared orthography and similar GPCs across languages (e.g., fast) and (c) unmarked incongruent condition that contains incongruent GPCs across languages (e.g., dirt). The results yielded a significant deficit in dyslexic readers in all three tasks, suggesting poor decoding but also poor lexical orthographic representations in L2 and difficulties in connecting form to semantic representations. This deficit was mostly observed for the unmarked incongruent conditi-on, highlighting the need to carefully manipulate the sub-lexical features of items when examining L2 reading. The results are discussed in relation to the cross-language transfer hypothesis and to mono- and bilingual models of reading.

Investigating pseudohomophone interference effects in young second-language learners.

This study aimed to investigate phonological activation during silent word reading in French adolescents learning English as a second language (L2) at secondary school. Grade 6 and Grade 8 adolescents performed lexical decision tasks in English, where we compared processing of nonwords that were homophonic to real L2 words (i.e., pseudohomophones [PsHs]; e.g., grean) with that of orthographic control pseudowords (OCs; e.g., greun). In Experiment 1, PsHs were constructed so that they sounded like L2 words when using cross-language (L1) grapheme-to-phoneme correspondences (GPCs) only (e.g., grine), whereas PsHs were constructed with within-language (L2) GPCs (e.g., grean) in Experiment 2. Results showed a PsH interference effect as reflected by higher error rates and/or longer rejection times for PsHs compared with OCs whether using within-language or cross-language GPCs and at both grade levels. Evidence of this PsH interference effect was also observed in Experiment 3, which used PsHs that sounded like real L1 words when using L2 GPCs (e.g., droal for the French word drôle [funny in English]). We suggest that young L2 learners automatically activate both L1 and L2 GPCs during L2 silent reading in favor of strong cross-language interactions at the orthography-to-phonology interface. The results are discussed in relation to bilingual and developmental models on visual word recognition.

Grapheme coding during sublexical processing in French third and fifth graders.

This study aimed to investigate grapheme coding during silent word reading in French developing readers from Grades 3 and 5. Children performed a letter detection task in which three conditions were used; the letter to detect was (a) presented as a single-letter grapheme (simple condition; A in phare), (b) embedded within a multi-letter grapheme that is considered as a unit or not depending on context (weakly cohesive complex condition; A in chant where "an" is a unit but not in other words such as cane), or (c) embedded within a multi-letter grapheme that is systematically considered as a unit (highly cohesive complex condition; A in chaud). Results showed a grapheme condition effect in Grade 5 children only. In this group, both complex grapheme conditions were processed more slowly than the simple condition, but this complexity effect was much stronger for the highly cohesive condition. We suggest that graphemes can be coded as sublexical orthographic units from Grade 5 and that such orthographic fine-grained coding is affected by the degree of grapheme cohesion.

New insights into bilingual visual word recognition: State of the art on the role of orthographic markedness, its theoretical implications, and future research directions.

In the past decade, research on bilingual visual word recognition has given rise to a new line of study focusing on a sublexical orthographic variable referred to as orthographic markedness, derived from the comparison of the two orthotactic distributions known by a bilingual reader. Orthographic markers have been shown to speed up language decisions but also, to some extent, to modulate language nonselectivity during lexical access (i.e., the degree of co-activation of lexical representations of the two languages). In this review, we (1) describe the results available in the literature about orthographic markedness on language membership detection and lexical access and discuss the locus of these effects, which leads us to (2) present theoretical extensions to the bilingual interactive activation models and discuss their respective adequacy to the data, finally leading us to (3) propose future research directions in the study of orthographic markedness, such as extension to different reading tasks and contexts as well as considering developmental and learning dynamics.

Do both WRAP and TRAP inhibit the recognition of the French word DRAP? Impact of orthographic markedness on cross-language orthographic priming.

We investigated lexical and sub-lexical orthographic coding in bilingual visual word recognition by examining interactions between orthographic neighbourhood and markedness. In three experiments, French/English bilinguals performed a masked lexical decision task in French (L1) in which orthographically related prime words could be either marked or unmarked English (L2) words, compared to unrelated primes (e.g., wrap, trap, gift-DRAP, meaning sheet). The results yielded an overall inhibition priming effect, which was unexpectedly more robust in the marked condition than in the unmarked one. This result highlights the need to integrate both lexical competition and orthographic markedness in bilingual models such as BIA/+ and determine how the latter may modulate lexical processing in bilinguals.

Disentangling cross-language orthographic neighborhood from markedness effects in L2 visual word recognition.

Previous research has reported that lexical access in bilinguals is language non-selective. In the present study, we explored the extent to which cross-language orthographic neighborhood size (N-size) effects, an index of language non-selectivity, should be dissociated from markedness effects, a sub-lexical orthographic variable referring to the degree of language- shared (unmarked) versus specific (marked) orthography. Two proficiency groups of French/English bilinguals performed an English (L2) lexical decision task with three word and non-word conditions: (1) English words with large French N-size/unmarked orthography (price), (2) small French N-size/unmarked orthography (drive), and (3) small French N-size/marked orthography (write). Evidence was found for orthographic markedness effects, albeit with a different pattern for word and non-word processing: while marked words were facilitated (responded to faster and more accurately) compared to unmarked words, the opposite pattern emerged for non-words. The pattern of results was comparable in both proficiency groups. No evidence emerged for the influence of first language (L1) neighborhood on L2 word or non-word processing. Thus, the results emphasize the need to integrate orthographic markedness as a relevant psycholinguistic variable in bilingual models of visual word recognition such as BIA/+ and to take it into account when investigating cross- language effects and the issue of language non-selectivity during visual word recognition.

The use and nature of grapheme coding during sub-lexical processing and lexical access.

This work aimed to investigate grapheme coding during sub-lexical processing and lexical access. Using the letter detection task in Experiment 1, we compared letter pairs that could be considered as a grapheme unit or not depending on context (referred to as weakly cohesive complex, e.g., an in chant vs cane) to real two-letter graphemes (highly cohesive complex, e.g., au in chaud) and single-letter graphemes (simple, e.g., a in place). Three experimental conditions were used, one of which was designed to prevent phonological influences. Data revealed that only highly cohesive complex graphemes were processed as units, not the weakly cohesive ones. The same pattern was found across experimental conditions in favor of an orthographic mechanism. In Experiments 2 and 3, a primed lexical decision task was used with two stimulus onset asynchronies (SOAs) and two different ranges of lexical frequency. We manipulated the number of graphemes removed from partial primes ( d**che vs do**he-DOUCHE) and relatedness. In contrast to Experiment 1, no evidence was provided in favor of a role of graphemes during lexical access. We suggest that graphemes can be conceived as sub-lexical orthographic units per se but can only be captured within a sub-lexical route to reading.