I love words. Which is why I’d love to be on the side of the phonics-in-context debate that gets to define literacy as something like, ‘an ongoing interplay between a reader and text whereby the reader is empowered to construct knowledge through a dynamic exchange of meaning-making processes’. Brilliant.
By Nicola Bell
Instead, my repertoire must be jammed full of phrases like, ‘the correspondences between phonemes and graphemes’. Such verbiage is, it cannot be denied, absolute yawnsville.
Don’t get me wrong – the ‘interplay’ stuff really is great. I think that, as a skilled reader, I probably am empowered to construct knowledge and all that jazz. But we don’t necessarily work out how children become literate by reflecting on our own experiences of being skilled readers. For young children, the ability to interact with full passages of text relies fundamentally on skills in decoding individual words. As such, I am content to sacrifice the pretty wordplay, if it means sticking with the science.
According to a ‘contextualised phonics’ approach, written text offers readers three ‘cueing systems’: meaning, grammar and sound-letter relationships. The assumption is that good readers have all three systems working in sync. So it doesn’t matter whether the target word was pronounced correctly because the reader applied decoding strategies, or because they predicted it from the preceding sentence context. From this perspective, all’s well if the meaning is intact.
In actual fact though, early research by Keith Stanovich shows that such a belief does not hold true. When attempting to identify an unknown word, it is poorer readers who use sentence context more so than skilled readers. Use of context is therefore not indicative of successful text-level reading, but of compensating for unsuccessful word-level reading.
As such, it seems pretty counter-productive to base reading instruction on the end goal of syncing up readers’ ‘cueing systems’ to obtain text meaning. Instead, the goal should be to develop word reading skills to a point where recognition happens automatically, so that the reader does not need to devote effort to relying on context and is actually free to think about what’s going on in the story.
As stated by Stanovich (1984, p. 15):
[T]here is no substitute for automatic, efficient data-driven processing at the word level. Capacity must be freed for the all-important comprehension and text integration processes. The only contextual mechanisms that are ultimately advantageous are those that can facilitate word recognition automatically, without depleting the amount of cognitive resources that can be allocated to text-level processing.
More recently, an Australian study was conducted to look at the nonword reading strategies of Year 2 students. Specifically, the researchers sought to test whether good readers were more likely than poor readers to substitute real words (e.g., ‘rent’) for nonwords (e.g., ‘rint’), the reasoning being that those with strong literacy skills might then be unfairly disadvantaged on measures like the Phonics Screening Check. The results showed the opposite pattern: it was the weaker readers who made more frequent nonword-word substitution errors.
As with Stanovich’s research, this study indicates that poor readers tend to attempt word recognition by relying on context or guessing strategies. Good readers, in contrast, can decode well at a single-word level.
All this is not to say that language skills contribute nothing to word reading development. There is plenty of evidence to the contrary. Vocabulary is strongly – and reciprocally – related to word-level literacy skills, and at a broader level, spoken language skills certainly contribute to a reader’s comprehension of print.
There are also exceptions to letter-sound conventions, whereby it’s impossible to reliably identify a given word without sentence context. For example, the word ‘close’ in ‘close the door’ has a different meaning and pronunciation to ‘he stood close to the door’, which has a slightly different meaning again to ‘she was close with her mother’. Homographs (and homophones) are inescapable parts of the English language, and they will of course pop up in children’s books or writing activities.
However, it doesn’t follow that the best way to teach reading is to encourage kids to guess their way through written text. Using word-level knowledge, a reader can at least approximate a word like ‘close’, and – from there – settle on the right pronunciation and meaning.
Yes, it is important that children are exposed to texts that are meaningful, relevant and motivating, but that shouldn’t discount how critical the reading building blocks are. Try to imagine how unmotivating it must be for children who never got that basic training, and who are still trying to guess their way through One Flew Over the Cuckoo’s Nest. Words matter.
This article was published in the October 2019 edition of Nomanis.
Nicola Bell works in the MultiLit Research Unit as a postdoctoral research fellow. She has a PhD from the University of Queensland on the topic of literacy development in children with cochlear implants, and her research interests extend to language and literacy development in all school-age children.