Print-sound regularities are more important than print-meaning regularities in the initial stages of learning to read: Response to Bowers & Bowers (2018).

We previously reported an artificial language learning study designed to compare methods of reading instruction that emphasise learning the relationship between spelling and sound versus learning the relationship between spelling and meaning. Behavioural and neural data supported emphasis on spelling-sound knowledge, and we therefore advocated use of phonics in the initial stages of learning to read. Bowers and Bowers argue that these conclusions are not justified because we (a) mischaracterised the English writing system and (b) mischaracterised the meaning-based instruction used in schools. In this article, we respond to the first point by showing that the novel words used previously were a good approximation to the types of written words that children are exposed to in the first year of reading instruction. We respond to the second point by showing that while enhancements to meaning-based instruction can assist pupils to infer the meanings of unfamiliar words, these methods actually disadvantage long-term learning of those words. We conclude by suggesting that reading instruction should be based on an understanding of the writing system, properly characterised across the trajectory of learning. This means emphasis on spelling-sound regularities in the initial stage of learning to read and increasing emphasis on spelling-meaning regularities as children gain greater experience with text.