On the role of statistical learning at the very beginning of learning to read

Overview & Synopsis

Based on a cognitive approach to the inner structure of writing systems and the existing literature on explicit instruction, statistical learning, and the syllabic bridge theory, we proposed a theoretical framework in which a level of associative units mediates between orthographic and phonological representations. The research presented in this thesis had two overarching aims: first, to examine whether statistical learning, at this level, allows prereaders to extract grapheme-phoneme correspondence (GPC) regularities in letters-to-syllable associative learning, and second, to determine whether phonological representation is influenced by statistical orthographic knowledge generated by associative units. To address the first aim, the research in Experiments 4.1 and 4.2 demonstrates that prereaders’ exposure to a greater variety of GPC regularities embedded in letters-to-syllable associations enhances their phonemic awareness and the generalisation of non-taught syllables. These findings suggest that the potential involvement of SL in learning letter-to-syllable associations might facilitate the acquisition of the alphabetic principle at the very beginning of learning to read. Addressing the second aim, the research in Experiments 5.1 and 5.2 showed that individuals’ responses to a phonological task are influenced by statistical orthographic knowledge, resulting from the interplay between orthography and phonology through associative units. These findings offer insights into early reading acquisition methods, underscoring the importance of tailored explicit instruction and SL in literacy development.