The impact of dialect exposure on artificial literacy learning in transparent and opaque orthographies

In two miniature artificial literacy learning experiments we explored whether exposure to a dialect helps or hinders literacy acquisition. Participants trained to read and write 30 artificial words rendered by 13 invented graphemes. During testing, we assessed performance without feedback on the trained words and 12 novel words. Critically, in the dialect condition, half of the trained words were dialect variants while in the standard condition words remained unchanged. In Experiment 1 (transparent orthography), we found an overall benefit from dialect exposure during reading, presumably because a larger amount of lexical variants facilitates focus on grapheme-phoneme conversion. We also found impaired reading for words with a dialect variant, suggesting lexical competition effects. Overall, performance was worse in Experiment 2 (opaque orthography) where these trends did not reach significance due to task difficulty, suggesting that more training is required. Still, results suggest facilitation of literacy learning with dialect exposure.