I think I was wrong about phonemic awareness

Why thinking like a scout rather than a soldier when assessing new ideas on the teaching of phonemic awareness has made Mark Anderson change his mind.

By Mark Anderson

Download PDF

What we know from research is that phonological awareness in English mostly develops in a manner that moves from large grain size (syllable, onset-rime) to small grain size (phonemes). Furthermore, we also know that phonemes are at a greater level of abstraction – they are harder to hear and speak – than something at a larger grain size like an onset or syllable, which is relatively easy to hear.

So it certainly makes sense that instruction should follow the same trajectory in order to support that developmental progression towards greater abstraction. It’s a compelling idea that unfortunately does not appear to be backed up by anything other than anecdotal evidence. I know it’s compelling, because that’s what I believed.

There’s a lot of romanticism in our field, and we all have a tendency to fall for ideas that look, sound and feel right. One of the ideas I’ve fallen for is that learning must always progress from concrete to abstract, from easier to harder. Furthermore, like so many others, I am easily taken up by the idea that learning progresses in stages; each stage must be mastered in order to progress to the next.

These ideas may be accurate for learning in some domains, concepts or tasks, but are not universal. We can see this point more clearly when we consider phonemic awareness instruction in English.

Instead of teaching first the syllable level, then next the onset-rime level, and finally the phonemic awareness level, the reading instruction that appears to be most effective for accelerating phonological awareness starts with the smallest grain size – at the phoneme level.

Why would this be? It could be that our neat and tidy theories (learning moves from concrete to abstract, and progresses in stages) mislead us. Sometimes, it may be that aiming first for what is seemingly more difficult and complex can be what enables us to develop underlying skills. And as we will see in a moment, we may inadvertently be making phonological awareness tasks more difficult and complex than they need to be.

Because there’s yet another facet of phonological awareness instruction where I seem to have been mistaken: I believed that practising saying and manipulating the sounds without letters can be a valuable activity. I’ve argued in the past that a phonological awareness program without immediate application to graphemes, such as Heggerty, could be beneficial, and I argued this because I thought that 1) it certainly wouldn’t do any harm, and 2) it could be of great benefit to students who struggle to hear and speak the sounds, thus facilitating phonological sensitivity. So in a school with a large number of students struggling to learn to read, it seemed like a win-win: a short amount of instructional time (10–15 minutes daily), an easily deliverable set of routines and lessons that require little planning nor training, and a potentially large pay-off for students who need it the most.

But it seems my priors – or simply my own biases – again misled me. I assumed that phonology = important to reading and language, and extra practice = good, so therefore: additional phonological sensitivity practice is a net positive.

Why wouldn’t this be good? Because thanks to the tireless advocacy of others (Twitter isn’t all bad, I swear!), I’ve had my assumptions challenged, and have since been exposed to research that suggests, on the contrary, that our energy in the earliest grades should be laser focused on connecting sounds to spelling. And that in fact, written letters are a scaffold for hearing and speaking phonemes! In other words, we may be increasing the cognitive burden on students when we ask them to conduct phonological tasks without connecting them to letters.

It took me a while to truly hear this and turn the corner in my own thinking. I found myself digging in my heels and even arguing for the benefit of adding in this additional phonological sensitivity practice. At the same time, I was arguing out the other side of my mouth that schools needed to resist adding more and instead pare down and focus on what is most critical, as we recover from the pandemic!

At some point, we need to look at the evidence and acknowledge when it is substantive enough to challenge the neat theories we hold about learning. And here’s the thing about something as complex as reading: even the ‘experts’ have their own neat theories and biases and will cling to them even as disconfirmatory evidence begins to accumulate.

Here’s some of the evidence that tipped the scales for me:

• The National Reading Panel’s Report synthesis on phonemic awareness instruction (an oldie but a goodie, and not as well-known as it should be)

• Susan Brady’s ‘A 2020 Perspective on Research Findings on Alphabetics (Phoneme Awareness and Phonics): Implications for Instruction’ (Expanded Version)

• Articles by Ukraintez et al. (2011), Cary & Verhaeghe (1994) and Hohn & Ehri (1983) that support the idea that larger phonological units do not improve phoneme level skills

• Gersten et al.’s (2020) meta-analysis that found significantly smaller effect sizes if a reading intervention included phonological awareness, yet significantly larger effect sizes if they included encoding or writing

• Møller et al.’s (2021) RCT that found adding spelling instruction to reinforce phonics instruction for students at risk for reading difficulties improved phonological awareness, spelling and reading skills over and above teaching phonics and letter-sound correspondences, in the same amount of time

• Results from numerous studies that have compared instruction based on invented spelling against phonological awareness instruction without letters and found substantially greater effects for teaching sound-spelling connections to students based on their current levels of understanding (Pulido & Morin, 2017; Sénéchal et al., 2012; Ouellette et al., 2013).

Rationalist Julia Galef recently came out with a book in which she introduces the concept of a “scout mindset”, in contrast to a “soldier mindset”. I’ve found this distinction useful, because we have quite a number of soldier mindsets when it comes to theories of reading, and I find myself falling into that mindset when I am challenged in my own thinking. But by consciously adopting a scout mindset – an attitude of curiosity and an openness to revising my thinking based on the evidence – I can ward off my tendency to dig my heels in.

I realised recently as I defended some of my original positions on phonological awareness that I was taking on a soldier mindset.

The more I have learned, the more I have realised that almost every source of expertise on matters of literacy holds ideas that must be questioned in light of the evidence. That’s all part of the journey of knowledge, man. No one person holds all the pieces of the puzzle.

So here’s where I’m revising my thinking: phonological awareness practice without pairing sounds to spelling is inefficient and unsubstantiated by current research. Instead, the body of evidence points to the greater robustness of pairing sounds to print from the beginning of reading instruction. This, in turn, then leads to greater phonological awareness.

Phonology is important. It’s important to both language and to literacy. And it’s that reciprocal relationship between print and speech that develops skilled reading.

So let me state my revised thinking as clearly as I can: we should focus our classroom instruction in the earliest grades – and in spaces of intervention in later grades – on supporting students in connecting sounds to letters in print, and core instructional time should not be spent practising sounds without print.

Time and money will be best spent on enhancing a core school-wide explicit and systematic phonics program through training and ongoing coaching supports and peer feedback, oriented around ensuring that speech sounds are connected to spelling in every lesson, with sufficient opportunities to practise these connections in reading and writing.

I still think there is a place for phonological practice outside of letters, but only when wielded by a knowledgeable practitioner or interventionist, who uses it for specific students as a bridge back to application with letters. Otherwise, pending any research that shows it is effective as a core instructional move, it appears to be a waste of time.

I admit I was wrong – or at least, I seem to be as of now, pending any further studies.

In terms of the fundamental language connection of phonology before and beyond print – I still think it’s critically important. But what I realised is that the place to do that kind of work is in interactive read-alouds, rather than isolated phonological practice. In other words, as we read text aloud to students, we can pause and amplify the sounds of words and sentences, ask students to repeat them after us like an echo, choral read them together, and savour their sounds, prosody and meaning.

Embedding phonological sensitivity practice in the course of authentic reading experiences will be more powerful – and most importantly – will not take time away from core instruction.

This article appeared in the Dec 2021 edition of Nomanis.

A similar version of this article originally appeared on the author’s blog, Language & Literacy.

Originally from California, Mark Anderson [@mandercorn on Twitter] taught in the Bronx for seven years and served children receiving special education services in an elementary and middle school. He then supported other teachers and schools in the Bronx with English language arts instruction for four years. Mark Anderson currently supports schools across New York City with instruction for multilingual learners.