Cognitive processes underlying learning to read

As noted, reading is a complex learned skill. Children are not born with the facility to read, but rather must apply a range of basic perceptual, cognitive, and linguistic processes to the task of acquiring this skill. The Reading Program sought to uncover these basic underlying processes, and to understand how they are applied to the task of learning to read.

Of our many discoveries over the life of the CCD, we highlight here just two. First, we made major inroads in understanding the relationship between oral vocabulary and learning to read. It is well established that children with larger oral vocabularies, who know the pronunciations and meanings of many words, tend also to be better readers, but the cognitive mechanism supporting this link has not been well understood. We have developed and found evidence for a new theory of this mechanism, which proposes that oral vocabulary supports reading acquisition from the point in time, before written words are seen. In a training study, and using eye movements to record looking times during silent reading, we found that when a child knows the pronunciation and meaning of a word, and they have some knowledge of how sounds in spoken words map onto written letters, they form an expectation about the spelling of that word. This spelling expectation then supports children’s reading of the word the first time it is seen in print. Our paper reporting on these spelling expectations, which we refer to as ‘orthographic skeletons’, was published in a top journal, Developmental Science (Wegener, et al., 2018).

Second, we have explored in detail the nature of the association between paired-associate learning (PAL) and reading acquisition. PAL is a dynamic measure of the ability to learn new links between two items (e.g., learning that the symbol Y is called “vay”). In experimental learning studies, we have found that PAL ability is associated with success in learning new written words, and that it predicts this learning above and beyond nonlexical decoding ability, and existing written word knowledge (Wang, Wass & Castles, 2017). The results suggested that PAL may be one of the underlying mechanisms of written word learning, facilitating the connection between the phonology and the written representation of a word.

Another major legacy of the Reading Program was to bring the latest knowledge on learning to read together into a single paper that was accessible to educators and the broader public (Castles, Rastle, & Nation, 2018). The paper, “Ending the Reading Wars: Reading acquisition from novice to expert”, was published in the high impact journal Psychological Science in the Public Interest, and has had more than 73,000 downloads. This paper also generated significant media attention and has been the subject of a number of podcasts and a front-page feature in the UK Times Education Supplement.