Friday, 19 April 2019

Whole Language Unicorns at home at the University of Melbourne

I was disappointed this week to see that the University of Melbourne, one of Australia’s (and indeed the world’s) most highly-esteemed universities, was promoting the ideas of David Hornsby on the important subject of how we teach children to read. If you haven’t already done so, you can listen to the interview I am referring to (it takes about 25 minutes) here

I have some familiarity with David’s work, having read publications such as this one and have no doubt about the depth of his commitment to supporting teachers and improving children’s progress in learning to read. It is unfortunate, however, that his key ideas about the role of phonics instruction remain uninformed by the last 20 years of cognitive science research on how best to get all children across the bridge to successful reading, writing, and spelling, in the first three years of school. Instead, David’s ideas are pretty Whole Language unicorns that persist in the imaginations of those who believe that “authentic children’s literature” is the key driver of successful reading. This does not align at all with the recommendations of three national inquiries into the teaching of reading, in the US, the UK, and Australia all of which recommended an explicit focus on systematic phonics instruction as the starting point for novice readers - alongside development of all of the other skills essential for successful reading. 

I did not find myself in complete disagreement with David, however, as I listened to this podcast. 

Here are the points on which David and I agree:

1.      Teachers are not nearly knowledgeable enough about the structure of language. In order to be able to teach the written code, teachers need to be able to understand that English is a morpho-phonemic language, and the full implications of what this means for how written and spoken language map to each other. I have blogged previously about concepts that teachers need to understand in order to be able to do their jobs properly, about the fact that they currently lack this knowledge, and the fact that responsibility for this sits with education academics who have shown a disgraceful disregard for the reading science, and the importance of ensuring this is transmitted to classroom teachers. 

I note that A/Prof Misty Adoniou has also written about low teacher knowledge stating that “All the participants in this study had patchy content knowledge of the English language, and very little knowledge of how to apply what content knowledge they had, although they were convinced this kind of knowledge was important”. So we seem to be in agreement, David, Misty, and me, that teachers are being sold short by their pre-service education. Where you choose to direct your rage about this is up to you, but I would start with the damage done to young children, particularly those who come from disadvantaged backgrounds, and need the turbo-charging impact of highly effective teaching by highly knowledgeable teachers in the first three years of school. You might also want to consider the indignation of teachers who are paying thousands of dollars for courses that do not deliver the goods they thought they were buying – competence to go into a classroom and teach children how to read. 

2.      Low teacher knowledge means that teachers jump around from one seemingly attractive quick-fix commercial product to another. There’s an awful lot that goes on in schools that either lacks an evidence base, or for which the published evidence indicates a lack of efficacy. My colleague Dr Caroline Bowen AO and I have written about this here. This damages the professionalism and accountability of teachers, and yes, sorry, teachers, do need to be accountable for their decisions and practices, though the extent to which this plays out in practice is questionable. When was the last time a teacher was held to account by a registration body for failing to apply evidence-based approaches in their practice? 

3.      Programs don’t teach children how to read, teachers do. This is a point that is often made by Dr Louisa Moats, and speaks to some of the points made above about the importance of strong teacher knowledge. I understand though, why many teachers seek the security of a program, because it instils some confidence, given the weak knowledge base conferred by their pre-service education. Can you see a vicious cycle here?

These are the points on which I disagree with David:

1.      He makes a blanket declaration that “Programs don’t work”. That’s about as silly as saying “antibiotics don’t work”. Sometimes they do, sometimes they don’t. It depends on a range of factors. My experience in working with teachers in schools is that programs often create the platform on which teachers go on to extend their knowledge and develop their practice. Programs definitely work better in the context of high teacher knowledge (in which case, perversely, there is less need for them) and when they are administered with fidelity.  When teachers deliver their own version of Program X and see only mediocre results in return, it is highly problematic that they then report “Program X does not work”. If we’re going to deliver programs, let’s first:
a.      Check their alignment with the current evidence on how best to teach reading
b.      Ensure that teachers receive appropriate training and ongoing coaching support so the program is delivered as intended, i.e. with fidelity. 

2.      David appeals to the romantic notion that all that is needed to get children interested in reading, is exposure to “authentic texts” (this is, in itself a problematic term as it implies a non-existent binary distinction, but I won’t get distracted about that here). Again, I have blogged previously about the “authentic illusion” in teaching children how to read. The logical fallacy that is occurring here is the conflation of books we read to children, with books we use to teach children to read. We want to expose children to a wide range of rich texts, to
a.      Help them understand the power and “magic” of books
b.      Extend their receptive and expressive vocabularies
c.      Expose them to increasingly complex sentences (and the increasingly complex ideas they represent)
d.      Help them understand narrative structure (and later, the structure of other discourse genres).


My three and a half year old grandson can be seen here "reading" to his two-month old sister. Yes, it warms my heart, at so many levels to look at images such as this. But no-one in my family is under any illusion that any actual reading instruction is taking place here. My grandchildren are engaged in precursors to becoming successful readers, but my daughter and son-in-law should be able to rely on their children's teachers to do the actual business of teaching them how to read. Schools should not rely on parents to be the critical link here. Some are able in this respect and some are not. Let's meet the needs of all children who enter foundation-year classrooms, not just the more advantaged ones. 

Immersing children in text does not teach them how to read, any more than playing hours and hours of Mozart concertos turns them into pianists or violinists who can play Mozart. Instead, the hours and hours need to be invested into learning and practising sub-skills.  Listening to music may be motivating and enjoyable, but it won't in and of itself teach the complex skill of playing music. Why do we "get" this in relation to music, but not in relation to reading?

Further, the incidental embedding of phonics instruction into the process of reading beautiful literature to children is inefficient (see Stephen Parker's free books on phonics instruction for further expansion on this) and fails to systematically address the knowledge of phoneme-grapheme correspondences that children need to learn in an efficient manner in the early stages of reading, in order to use their knowledge of the code to access more and more complex texts and their meanings. Sure, some children will be fine with the incidental approach, but we should not be focussing our teaching methods on those more advantaged children who will get across the bridge almost irrespective of the reading instruction they experience. 

A strong Tier 1 is the best insurance against the creation of instructional casualties, but is often not in place.

Ironically, in order to teach in accordance with the recommendations of the three international inquires referenced above, teachers need to be more knowledgeable about language, in exactly the way that David Hornsby and Misty Adoniou are advocating. 

3.      David relies on the (almost comically) flawed definition of decoding in the Australian Curriculum*:

A process of working out a meaning of words in a text. 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.

This may as well just be a definition of reading, as it fails to focus on the specific and essential sub-skill of mapping phonemes onto graphemes in order to first make sense of the black lines that we call letters on the page.

So – just because a definition has found its way into the Australian Curriculum, does not mean that it aligns with the science of reading. Anyone who is familiar with the Simple View of Reading, or with the Hollis Scarborough Reading Rope, will know how far short this definition falls of the skills beginning readers need to be taught in order to achieve decoding mastery. For my money, a child who can read aloud an isolated written word has decoded that word. Meaning comes next – the word is either in the child’s lexicon or needs to be added as a new word as a consequence of the oral language value-add of being able to read.

4.      David dials up the hoary old wind/wind chestnut to argue that “phonics alone isn’t enough”. Well, first of all, I am yet to meet anyone who claims that “phonics alone is enough”. 

       The debate as I see it, as around the role and type of initial instruction to get novices off the blocks as successful readers. Anyone familiar with the terrific Five from Five Project will know that no-one on the systematic phonics instruction side of the debate overlooks the importance of vocabulary, fluency, and comprehension – as well as inferencing and using contextual cues. The latter, however, should augment word identification rather than being the primary strategy relied upon by a novice who encounters an unfamiliar word.

My question for David, then is how does a child even get to the dilemma of wind Vs wind without first using phonics-based decoding skills?

5.      Finally, David makes the bizarre claim that systematic phonics instruction involves the teaching of nonsense words. Sigh. We don’t seem to be able to put this silly idea to bed. Firstly, the distinction between real words and non-words is dimensional rather than categorical. Just have a look at J.R. Tolkein, Lewis Carroll, J. K. Rowling, Dr Seuss, to name a few, if you don’t believe me. I am not aware of any evidence that says that novice readers approach unfamiliar words with a real-word – non-word filter.
Consider too, the ever-evolving nature of spoken and written language and the constant addition of what were once non-words, to the real-words lexicon. You can find examples of recent additions to English here. How can anyone (child or adult) read these new words if they are unable to decode in the truly get-the-word-off-the-page sense of decoding?

Think about this for a moment – every word is potentially a non-word to a novice, because their orthographic lexicon is still under construction. I have seen teachers “play” with non-words in Phonological Awareness activities and am completely comfortable with that – the children were experiencing rhyme and enjoyed talking about the meaning of the words. Consider this example: the teacher is using the words “wig, pig, and big” in a rhyming exercise and then introduces “rig” asking the class if it is a real word or a non-word. Chances are a lot of five-year olds will say it’s a non-word. What a great opportunity to wrap some vocabulary work into the Phonological Awareness activity. 

So can we be very clear about this please – non-words are extremely useful as a light-touch screening tool to determine the extent to which children are mastering essential phoneme-grapheme correspondences, so that they can decode in line with the scientific definition of decoding, not the politically sanitised version that is pleasing to Whole Language unicorns. 

I hope going forward that our esteemed educational institutions will apply careful quality and evidence-alignment filters over ideas they promote to teachers, who, more than anyone else, must be thoroughly tired of the contentious nature of something that they need to be able to do everyday - and across all year levels. 

In the last fifty years, we have put men on the moon, eradicated a number of terrifying childhood illnesses, and made major advances in engineering, aviation, construction, and manufacturing.  

But thanks to a massive knowledge-translation failure,  we cannot be relied upon to teach children how to read.  

*Postscript: I am pleased to advise that Australian Curriculum website definition of decoding has been updated since this blogpost was published, and now reads as follows:

A process of efficient word recognition in which readers use knowledge of the relationship between letters and sounds to work out how to say and read written words.

(C) Pamela Snow (2019) 


  1. Dear Pam, phonics is a form of instruction not a skill. The relevant skill is learning grapheme-phoneme correspondences or GPCs (what is sometimes called “orthographic phonology”) rather than “phonic decoding skills”. It is important not to use the phrase “phonic decoding skills” because when I and others criticize phonics it does not mean we are rejecting the importance of GPCs. There is, however, a question about how best to teach GPCs: In combination with other letter-sound correspondences? In combination with morphology? In the context of reading meaningful texts? First, before considering meaning (as is the claim with systematic phonics), etc.

    For a short description of what it wrong with the evidence in support of systematic phonics, and an alternative approach (NOT whole language), you can find a recent short article I co-authored with my brother in the Washington Post. See:

    It would be great if you would critique my work criticizing phonics instructions, both from the point of view of theory (see Bowers & Bowers, 2018), and data (Bowers, 2019). Papers can be downloaded from here:

    Bowers, J.S., and Bowers, P.N. (2018). Progress in reading instruction requires a better understanding of the English spelling system. Current Directions in Psychological Science, 27, 407-412

    Bowers, J.S. (2018). Reconsidering the evidence that systematic phonics is more effective than alternative methods of reading instruction. PsyArXiv.

    1. Hi Jeff,
      I have read an article of yours describing SWI and the importance of morphology, and completely agree, so thank you for your work! In your comment you stated that the phonics debate is about teaching GPCs. What are your thoughts on blending and segmenting? I find my problems with current classroom practice is less to do with how GPCs are introduced (I'm perfectly happy for them to be introduced in meaningful contextualised words), but with the fact that little work is put into teaching kids to pull words apart and put them together (both into phonemes and morphemes). Blending is not taught until after all basic single letter correspondences are taught, and is then taught to be used as a last resort in identifying an unknown word. To me, strong PA skills are the critical skills needed to be able to unpack the makeup of a word. Unknown words should first be sketched (sounded out) using current print knowledge, then adjusted to find a word that fits context. What are your thoughts?

    2. Hey Kathy,
      I hope you don't mind if I respond to your question to Jeff. Since I'm in schools all the time, I may be in a better position to respond.

      I suspect some key parts of my practice will seem similar to the blending or segmenting practices you reference. However, the context of how that happens may be different very different. That context takes a bit of explaining.

      For example, when a student is reading with me and they substitute a word for another with the same first letter, or get stuck for any reason, my most common first provocation is to ask the student if they could "spell-out" the word they got stuck on. I have only my own anecdotal experience with this, but I find it remarkably effective. When a child is reading with you, asking them to spell-out a word is risk free. If they are reading to you, they most likely know the letter names automatically. If they don't it's important to know that too!

      When asked to "sound-out" a word, students know they have a good chance of getting it wrong. The stress of performing must take up so much working memory. Further, once a first phonological hypothesis has been announced, I imagine it can be hard to think of other options, especially in a stressful situation.

      It seems to me that after getting stuck on a word, it can help students to let their eyes take their mind through the orthographic structures of the word before committing to one pronunciation. In one of countless examples, I was once reading with a Gr. 2 student who misread the word "created" as "carried". I asked him to "spell-it-out" and he only got as far as "c-r" before he self-corrected. My guess is that once he spelled the "cr" combination, he was able to falsify the word he had guessed at first and then process the rest of the orthography (including phonological cues of graphemes) to get to the word.

      If spelling-out didn't get them to the word, I might simply tell them the word -- especially if I am uncertain that her or she even knows the word. From a known word, we can then analyze the grapheme-phoneme correspondences within their morphological constraints.

      Another practice that I use all the time is one of "spelling-out word structure" which often involves "tapping out" the orthographic structures of words also. We play games, trying to identify one tap for every grapheme or orthographic marker in a base. We announce any affixes as units without tapping. This process is used regularly in group investigations. So the spelling-out "t-ea-ch----ing" reflects understanding of the graphemes in the base. The pause between morphemes and the lack of tapping reflects awareness of the "-ing" suffix. By contrast spelling-out "t-e-a-c-h-i-n-g" shows lack of recognition of those structures. To understand the graphemes, we need to identify the phonemes. So this basic SWI practice involves rigorous phonological analysis.

      To see more detail on the conventions for "spelling-out-loud" and also "writing-out-loud" of orthographic structure I use that demands rigorous phonemic analysis you can go to this link:

      That includes videos to consider.

      In this video of me reading with my son when he was just starting to read, you can see ways I apply this practice during informal little reading sessions.

      My apologies that I'm not able to respond to your question more succinctly. I think that the core of what you are getting at with "blending and segmenting" is addressed in SWI, but in a context that may be quite different. I don't have many good videos of what this practice can look like. If you take a look at those links, I'd love to hear any concerns about what might be missing here for you compared to what you do in your practice of blending and segmenting.

      Thanks for the great question!

  2. Peter thanks for your comment, which I understand, and have made a small edit to the text accordingly. In everyday usage, we do tend to refer to "phonics skills", so I think this may be a regional usage issue as much as anything.

    Peter I have engaged at some length with you and have been positively influenced by your thinking regarding morphology. I'm not in a position to elaborate more fully right now (this blogpost took up a huge bite of time I didn't have this weekend), but will remain interested in your thinking and writings.

    Cheers, Pam

  3. Hello Pam,
    Just to clarify, it was my brother Jeff Bowers who responded to your post. He and I have written a number of times together, so it's easy to confuse us. Jeff just alerted me to the post, so I thought I'd share just a couple thoughts and illustrations of structured word inquiry (SWI) instruction to give context of the instruction Jeff is addressing.

    I hope you will read and critically respond to Jeff Bowers (2019) paper on the data about systematic vs. unsystematic phonics. Regardless, I'd like to emphasize that while SWI does not provide a sequence of grapheme-phoneme correspondences to teach, the instruction of orthographic phonology in SWI is not what people might think of when they think of "incidental phonics" instruction in a whole language context.

    In this account ( from Rebecca Loveless (see her great website, she shares a tale of working with a Grade 1 student who is seriously struggling with literacy. A key example in this story is when the student notices the link in spelling and meaning between the words “magic” and "magician" and the the different pronunciations of the "c" grapheme. Rebecca’s instruction that follows the “4 questions of SWI” (see here: however, allows her to capture this student’s observation to build understanding of WHY the “c” grapheme is needed in the spelling of “musician” instead of the “sh”. Only the “c” grapheme can represent the phonology of the morphologically related words that allows the morphemes to mark the connected meaning with the same spelling. A central fact of English orthography is that its grapheme-phoneme correspondences need to be able to represent the phonology of any pronunciations of the morphemes in related words. Teachers are guided to understand this basic feature of English spelling when they follow those same 4 questions to investigate and understand the spelling-meaning connections in words like “do/does”, “act/action” “sign/signal” and countless others.

    Here are a couple other illustrations of rigorous orthographic phonology instruction in SWI classrooms:

    A Grade 2 classroom teacher sharing their work:

    A video of a Gr. 1 teacher in his first year or two of introducing orthographic phonology through SWI to young grades

    Finally, you emphasize that nobody says that phonics on its own is enough with the statement "Anyone familiar with the terrific Five from Five Project will know that no-one on the systematic phonics instruction side of the debate overlooks the importance of vocabulary, fluency, and comprehension". This is a key point, but the 5 components of quality literacy instruction that you point to does not draw any explicit attention to instruction about the interrelationship of morphology and phonology, let alone the important role of etymology that is required to understand English orthography - including how grapheme-phoneme correspondences work.

    The closest this list of components gets to the workings of orthography is under the “vocabulary” component where it is stated: "Knowing the meaning of a wide variety of words and the structure of written language." But even here the term morphology and etymology are not referenced. So this list of components of a high quality literacy program fail to explicitly point teachers to the inherent interrelation of morphology and phonology (let alone the crucial role of etymology).

    My apologies for the perhaps overly long comment. Structured word inquiry is not phonics instruction nor is it whole language instruction. It is difficult to be brief when trying to convey a new way of thinking about literacy instruction that is significantly different from what people are familiar with. I hope that this response and the linked examples helps provoke a constructive discussion on these important issues.

    1. Hi Peter (and Jeff!) apologies about my error and thanks for the further thoughts. I have now watched the videos you shared and agree (as I have indicated previously)that more emphasis on morphology is long-overdue, for decoding, spelling and vocabulary development. I still don't agree that morphology is the starting point with novices (and have been over that ground several times in previous communications), but I am completely on-board with the idea that it needs more emphasis once some basic phoneme-grapheme principles have been mastered and a few key "sight" words (however one want to describe these) are bedded down to aid with fluency. I am spending more time talking about morphology when I deliver teacher PD, so you have definitely developed and influenced my thinking - thank you.
      Cheers, Pam

  4. Whatever the nomenclature, the science remains.

  5. Yes the science remains the same whatever terms we use. And as I detail in the links below, the science does not support the claim that systematic phonics is more effective than other common alternatives. I know that is what people keep saying, but check out my papers and then respond.

  6. This was a great article and the follow up conversation was equally enlightening. I've had the pleasure (and some anxiety) of helping my own two children learn to read. I also work with English Language Learners where gaps in first language literacy clearly show up as students try to learn English. My experience is that students must develop phonetic awareness and that this is an important part of learning to read. I've written about it here with specific reference to some of my Chinese students:
    Some students are too frightened of "big words", don't know how to break them down, might not even know that spelling is consistent, or have just memorized words. This is a stop-gap measure that works in proportion to how motivated they are, or how much pressure they have from home, or perhaps even their own cleverness.

    I've had bright, articulate students, who can read, but stumble through reading English. Others are academically under-achieving, but they can hammer through an unfamiliar text almost effortlessly, even if they don't know all the words they are reading. And I've had young students who are under-achieving, but are able to read through an unfamiliar English passage, and able to read it clearly, even if they don't understand all the content. This is where I agree with the bias towards an emphasis on phonetic awareness that comes out in Pam's post.

    But as Jeff and Peter point out, it's more than just the sounds. Strong readers can make all kinds of connections within the words they read. They break the words down, link them like the "magic" and "magician" comment above, and hypothesize new uses. I've seen this with my ELLs. An approach that I have developed, which I suspect is a variation of what Jeff and Peter have discussed as Structured Word Inquiry is described here, but which I call developing word consciousness

    Many students learn to read in spite of our efforts, but many more would learn to read so much better with more direct input and more opportunities to analyze words and their appearances, sounds, meanings, and even etymologies.

    Privileged kids are having this done for them at home (as several have mentioned here working with their own kids and grandkids), in all countries and in all languages. This privilege needs to extend to everyone. And it can.

    1. Well said and thanks for your contribution. You'll get no argument from me on any of the above.
      Cheers, Pam