RESUMO
Learning to spell is a challenging process, especially for young learners, in part because it relies on multiple aspects of linguistic knowledge, primarily phonological and morphological. However, alongside these universals, there are significant writing system specifics, namely, language-specific and script-specific factors that may also challenge young readers and writers (Daniels and Share, 2018). The current study focuses on the impact of four distinctive visual-orthographic features of the Arabic abjad on spelling, namely, (i) the similarity of many basic letter-forms, (ii) allography (the positional variants of the letter forms), (iii) ligaturing (the joining of letters), and (iv) non-linearity (extra-linear diacritic-like signs used to mark consonantal, short vowel and morpho-syntactic distinctions). We examined the distribution of visual-orthographic spelling errors across three grade levels as well as the developmental changes in these errors. We predicted that these errors would account for a significant proportion of children's spelling errors. Ninety-six Arabic-speaking pupils from three elementary grades (1st, 2nd, 4th grades) were presented with a sequence of six pictures and asked to write a story or several sentences about the events depicted. All spelling errors were analyzed and categorized according to two types of categories: six visual-orthographic categories and six additional categories that relate to the more traditional error types (e.g., phonological). The results showed that the visual-orthographic category was the second most common error category across the three grade levels, accounting for over one quarter of all spelling errors. Ligaturing and letter shape formation errors emerged as the two most prevalent types of errors in this category. These findings clearly demonstrate that visual-orthographic features of the Arabic abjad pose significant challenges in learning to spell.
RESUMO
The ability to overcome interference from the first-language (L1) is a source of variability in second language (L2) achievement, which has to date been explored mainly in same-script bilinguals. Such interference management, and bilingual language control more generally, have recently been linked to domain general executive functions (EF). In the current study, we examined L2 proficiency and executive functions as possible predictors of susceptibility to L1 interference during L2 processing, in bilinguals whose languages do not share an orthographic system. Seventy Arabic-Hebrew bilingual university students performed two tasks indexing cross-language interference (from L1 to L2). Lexical interference was assessed using a cross-modal semantic similarity judgment task in Hebrew, with false-cognates as critical items. Syntactic interference was assessed using a self-paced reading paradigm and grammaticality judgments on Hebrew sentences whose syntactic structures differed from those of Arabic. EFs were examined using spatial and numerical Stroop tasks, to index inhibitory control, and a task switching paradigm, to index shifting abilities. We found significant L1 interference across the lexical and syntactic domains, even in proficient different-script bilinguals. However, these interference effects were not correlated, and neither type of interference was related to domain general EF abilities. Finally, offline susceptibility to syntactic interference, but not lexical interference, was reduced with greater L2 proficiency. These results suggest at least partially independent mechanisms for managing interference in the two language domains, and raise questions regarding the degree to which domain general control abilities are recruited for managing L1 interference.
