How does dialect exposure affect learning to read and spell? An artificial orthography study.

Abstract

Willia m s, Gl e n n P., P a n a yo t ov, Nikol ay a n d Ke m p e , Ver a (2 0 2 0) H o w Do e s Dial e c t Ex p o s u r e Affec t Le a r ni n g t o R e a d a n d S p ell? An Artifici al O r t h o g r a p h y S t u dy. Jou r n al of Ex p e ri m e n t al P syc h olo gy: G e n e r al. IS S N 0 0 9 6-3 4 4 5 Do w nlo a d e d fro m: h t t

Exposure to dialect variation in an artificial language prior to literacy training impairs reading of words with competing variants but does not affect decoding skills

“…We pre-registered our hypotheses (available at https://osf.io/bxt87) with respect to the contrastive deficit and to the processing of untrained words. We hypothesised that, as in Williams et al (2020), a contrastive deficit would emerge in the Dialect condition such that reading is impaired for contrastive compared to non-contrastive words due to competition between the two variants during attempts at direct lexical retrieval. We expected no such competition in the No Dialect condition where no competing variants exist.…”

“…Subsequent studies using artificial literacy learning sought to clarify these issues. In Williams et al (2020), we taught adult participants a miniature artificial language consisting of 30 words, interleaved with literacy training using an invented script. In the dialect exposure condition, half of the words were learned as dialect variants instantiating similar phonological differences as the Brown et al (2015) simulation.…”

“…A dialect benefit in reading of untrained words is a counter-intuitive finding that may have arisen because the regimen of interleaved word exposure and literacy training in the artificial literacy learning experiment of Williams et al (2020) created a situation in which participants received literacy training before stable word form representations had been established in long-term memory. Such a training regimen is probably not typical for literacy learning in a first language where access to phonological forms of most words that children learn to read and spell is already well entrenched.…”

“…Such a training regimen is probably not typical for literacy learning in a first language where access to phonological forms of most words that children learn to read and spell is already well entrenched. Instead, the training regimen in Williams et al (2020) resembled concurrent word and literacy acquisition as is more common in adult second language learning where orthographic input is presented from the outset and reading is recruited for word learning. In this situation, the increased variation in word forms arising from interleaved presentation of alternative variants in the dialect exposure condition may have incentivised learners to employ phonological decoding because for contrastive words, there was no other information to determine which one of the two pronunciation variants to produce.…”

“…The first aim of the present study was therefore to determine whether the results of the artificial literacy learning study by Williams et al (2020) still hold under ecologically more valid conditions resembling first language literacy acquisition when learners are trained to read and spell words they already know. While keeping the overall amount of input identical to Experiment 3 from Williams et al (2020), we re-designed the training procedure so as to front-load the blocked exposure to artificial words prior to literacy training. We then compared a group that encountered the same variants during word learning and literacy training (No Dialect condition) with a group that learned different variants for half of the words prior to literacy training (Dialect condition).…”