Sunday 29 July 2018

An academic union: Perpetuating myths about phonics instruction




Poor old phonics. It really is the Cinderella of reading instruction. No-one gets up in the morning to have a debate about the importance of vocabulary, or comprehension, or fluency, or narrative language ability.  Even morphology, a long-overlooked, but vitally important part of learning to read a morpho-phonemic language such as English, is beginning to come in from the cold, and not before time. But phonics, it seems can only manage a conditional presence in some early years classrooms and initial teacher education programs; and even then, it is often relegated to metaphorical floor-scrubbing and never quite makes it to the ball.
This has been reinforced this week by the publication of a report authored by Professor Robyn Ewing of the University of Sydney, commissioned by the New South Wales Branch of the Teachers Federation[1]. That unions are investing skin in the pedagogical game is in itself notable, and an unabashed further policisation of the reading debate. You can download a copy of the Ewing-Union report from the NSW Teachers Federation website.

Let’s have a look at some of the arguments proposed in the Ewing-Union Report (direct quotes from the report are in red text like this).

Socio-economic status and early language exposure have long been known to be predictors of early reading success. This makes perfect sense, when we consider the fact that reading is a fundamentally linguistic task, and children’s pre-school language experiences sit on a social gradient. This does not mean, however, that reading is “speech written down”. Written text is typically more complex in terms of both vocabulary and syntax and has the pauses and hesitations that characterise spoken communication edited out. Reading has been described as a "biologically unnatural act", such that most children require years of specific instruction in its intricacies, in order to become proficient. This sets it apart from oral language, which is biologically natural, but is no set-and-forget skill, given its sensitivity to environmental exposure.

Reading requires a sophisticated grasp of spoken language, to resolve possible ambiguity and fully discern the writer’s intended (but sometimes obscured) meaning. This takes many years to develop, and like all complex skills that take learners from novice to expert, requires teachers to isolate small components of the task, and provide repeated practice for mastery, while building sequentially on emerging skills. Why learning theory applies to learning a musical instrument, or learning how to drive, but apparently not to reading is a mystery to me.

As has been noted by UK researchers, Roy and Chiat (2013), the discrepancy in early language exposure associated with SES means that schools actually need to accelerate the progress of children who start from behind with respect to early oral language skills. However, Professor Ewing offers no advice or strategies on how this acceleration might be achieved.  I agree with Professor Ewing’s contention that socio-economic status is an important variable that influences early reading instruction, via its influence on early oral language and text exposure. What I find disturbing, however, is the fact that this evidence is presented as some kind of thinly veiled apology for the sustained achievement gap that exists between children from less and more advantaged backgrounds. It's parents, you see, who are really responsible for teaching their children to read.

To improve reading for all Australian children, it is not constructive to assert undue pressure on educators (and teacher educators) to adopt only one way of teaching reading, as if it must and will answer all the difficulties that some students face.

Can Professor Ewing please point to even one reference in which anyone in the reading debate advocates the adoption of “only one way of teaching reading”? This is a straw man.

Teachers must be trusted with the responsibility of reflecting on and adjusting professional practice in the light of research evidence and theirs and other practitioners’ knowledge and experiences.

This translates as “It’s OK for teachers to choose their own adventure in early classroom instruction. Whatever they prefer is a fair thing. There’s no real science here”. If doctors, engineers, architects and pilots were encouraged to adopt professional practices in the same way, the Age of Enlightenment may as well not have happened. Professional trust has to be earned. It is not something that can be demanded just because an academic and a union say so.

Ironically though, while on the one hand advocating for this classroom free-for-all libertarianism, Professor Ewing levels a subtle criticism against her teaching colleagues who do opt for a systematic synthetic phonics approach, by playing the tired and flimsy “commercial programs card” (“Over the years many commercial programmes have argued that intensive phonics taught first is the key to helping all children learn to read”). Never mind that Whole-Language based levelled readers are commonly found in early years classrooms in Australia (and elsewhere) and are not exactly donated by the international publishing houses which promote and supply them.

“…human variation” must always be considered and that learning to read defied a prescriptive recipe for all children”.

Of course human variation must be considered in all learning contexts, but let’s not over-egg the human variation pudding. As Dr Caroline Bowen, AM  and I have noted previously, there are more similarities between children and their brains, than there are differences. If that were not so, teachers would not observe and benefit from patterns in children’s learning and use these patterns by adjusting their teaching, and forming and testing hypotheses about which approaches yield maximal return. This is another straw man distraction that is really just a fig-leaf for “whatever you think is a fair thing”. I was reminded of this way of thinking in education faculties a couple of years ago, when an education academic responded to my request for evidence supporting a pseudo-scientific approach that she was advocating, with “But you can find evidence to support just about anything in the peer reviewed literature’. Well, indeed you can. And that’s why we want graduates to be able to critique research and discern substance from fluff.

In decoding, readers draw on contextual, vocabulary, grammatical and phonic knowledge. Readers who decode effectively combine these forms of knowledge fluently and automatically, and self-correct using meaning to recognise when they make an error (The Australian Curriculum: English).

Professor Ewing can ride here on the coat-tails of the fact that this Orwellian definition of decoding made its way, courtesy of the influence of Whole Language thinking and pedagogy, into the Australian Curriculum. She knows, as well as the next person, that the meaning of “decoding” has been appropriated here to fit with the popular, yet not scientifically-supported “Three-Cueing Approach” (see more here on this, courtesy of Alison Clarke’s Spelfabet website. This is a definitional retro-fit that fools no-one who has ever worked with a struggling reader who cannot get over the first hurdle of decoding text in the manner described by Gough and Tunmer when they proposed the Simple View of Reading.

There’s a peculiar irony in Professor Ewing noting, in a non peer-reviewed, union-commissioned report, that the so-called Clackmannanshire Study was “never peer reviewed”, while neglecting to mention this peer-reviewed 2012 follow-up study in the journal Reading & Writing which found “Overall, the group taught by synthetic phonics had better word reading, spelling, and reading comprehension. There was no evidence that the synthetic phonics approach, which early on teaches children to blend letter sounds in order to read unfamiliar words, led to any impairment in the reading of irregular words”.

Thus many of the youngest children, particularly boys, are labeled reading failures early in their school career.

This is a reference to England’s Phonics Screening Check, which is just what it says on the packet – it is a screen and a check. Professor Ewing provides no evidence for her claim that some children are labelled as “reading failures” if they do not pass the check. If true, this would be an outrageous response to a Year 1 screen and one whose proponents should be held to the highest level of account. 

Let's remember too, that in education and in health, we only screen in situations where something timely can be done to change an outcome. No children are "failures" for not meeting a screening criterion; this is merely data that indicates a need for further assessment and possible intervention.

The synthetic phonics check in England has not delivered improvements in long-term reading capabilities. The report also noted concern about the amount of class time now devoted to learning to read nonsense words rather than real words.

I have responded to the first part of this claim in a previous blog post

Interestingly, in the same week that the PIRLS results were released, the debate about introducing a PSC in Australia has intensified, with detractors trying to deflect focus away from the fact that England's first cohort of students to have experienced the PSC are the very ones at the centre of the up-lift described above. The PSC was a large natural experiment, with no other known mass interventions occurring at the same time. If England's PIRLS results had deteriorated in 2016, PSC-detractors would have pounced on this as evidence that "it doesn't work". But the thing about the evidence game is you can't have your cake and eat it too

As for the second part of Professor Ewing’s statement above, no class time should be devoted to the teaching of nonsense words. Class time should be devoted to the teaching of decoding skills, promoting vocabulary development, fluency and comprehension, an appreciation of narrative structure, and mastery of increasing syntactic complexity. Nonsense words can then be handled (and enjoyed) by children where these occur in children’s books, as they do frequently (think: Dr Seuss, J.K. Rowling, Roald Dahl, Beatrix Potter, Spike Milligan, J.R. Tolkien…and many more). This is just another straw man distraction Professor Ewing, and also conveniently side-steps the fact that all novel words may as well be pseudo-words to a child who has not encountered them before. It is only by decoding and checking a word against their lexicon that children can make educated guesses as to whether a newly-met word is “real” or not. This, surely, is part of the discovery part of learning to read?

Becoming a fluent and accurate reader means learning to use all the cue systems: semantic, graphophonic and syntactic cues.

This argument is also referenced above. However I'd like to explore its utility with you, the readers of this blog post. For this little experiment, I am going to assume that most readers of my blog have well-developed reading skills, of both the decoding and language comprehension varieties. So let's see how well the Three Cueing system serves you, as competent readers, in discerning ("decoding") from contextual cues (semantic, syntactic, and grapho-phonic), the meanings of the three highlighted words below, with which (I am guessing) at least some of you will not be familiar:

  • I’ve spent nearly three years studying this theory and would like to share my inchoate assessment of it with you.
  • His ideas about economic theory were intriguing but protean. This made for some "interesting" class debates. 
  •  Let me adumbrate the controversial findings in this scientific paper for you.
If you didn’t know the words inchoate, protean and adumbrate before, do you know and understand them now, having encountered them in context? I doubt it. To acquaint yourself with these words and their pronunciation and meaning, you probably tried first (not last) to say them aloud, i.e., to decode them, in the traditional sense of the word. Perhaps you looked for familiar roots and affixes for a morphological and syllabification cue. No doubt you were able to ascertain the syntactic function, or part of speech each fulfilled (adjectives x 2 and verb), but did that help you to perform the necessary self-check on your discernment of meaning? Perhaps you thought you knew what they meant. But did you? Does it matter if you were wrong?

Well yes it does. If you can't decode and derive meaning, as per the SVR, then your work as a reader is not yet done. Thank goodness, in this case, for Google, which will provide not only the meanings, but assist on the pronunciation front as well. The point here is not the obscurity of the words (necessary to illustrate my point), but the inherent complexity, cognitively and linguistically, of dealing with unfamiliar written words, an experience which novice readers encounter many times a day. With working memory still under construction, beginning readers need developmentally appropriate skills that increase the automaticity of the reading process, rather than taking them into a conceptual cul-de-sac.  

And no, dear reader, they are not pseudo-words

Stand easy everyone. 
Nothing was “exploded” in Professor Ewing’s union-commissioned paper. Rather, a number of tired though conveniently protean myths have simply been perpetuated; unhelpfully and uncritically so. 

(C) Pamela Snow (2018). 


[1] Yes, I know there should be an apostrophe after the “s” in “Teachers” but ironically the Teachers(') Union itself apparently doesn’t think so.