Reading for pleasure: A reader replies to Michael Rosen Part 1

/, reading/Reading for pleasure: A reader replies to Michael Rosen Part 1

Back in July I wrote this post on how we might encourage children to read for pleasure to which children’s author Michael Rosen left a long & detailed comment critiquing my ideas. The comment included this statement:

When children are deemed to be ‘not reading’ i.e. being unable to pass the Phonics Screening Check, some teachers are being asked to do more of the same, rather than do anything different, nor to investigate whether there are multiple reasons for a) not passing the phonics screening check or b) finding out whether some children can read pretty well but fail the PSC anyway. Indeed, there is now some evidence to suggest (UKLA) that ironically some ‘middle class’ children how have learned to read at home with their parents using a variety of methods find the PSC confusing. Some have been observed ‘correcting’ the nonsense words, e.g. ‘strom’ to ‘storm’. So they do not pass.

A few weeks ago, another correspondent, Jacqui Moller-Butcher left what I thought was a very interesting and thoughtful response to the point above and has kindly agreed to my request to post her reply as a separate post. Here it is.

Dear Michael,

I work right now with lots of primary children aged 6-11 who’ve learnt to read using mixed methods at home and at school because mixed methods are still being taught alongside phonics, at least in some primaries. When I say mixed methods, the children read (by sight) words they recognise easily and guess those they don’t recognise based on letter clues.

You’re right. They do ‘correct’ strom to storm and other words like this, often. In exactly the way you describe, and for the same reasons, the same children ‘correct’ carry to cry, journey to joy and would to wild. I have a bank of hundreds of very interesting and informative examples. But, of course, for these words, this is not ‘correcting’ at all. The guessed word in each case is a real English word – but it is wrongly read. This is ‘mis-reading’ – just as much a ‘mis-reading’ as storm is of strom.

The children do this ‘lookaliking’ reading (for that’s how it seems to me) when reading words in context and when reading them in lists. It’s fascinating and frustrating in equal measure.

These children are practised guessers. They guess instinctively from shape and, usually, according to my bank of examples so far, from first, last and a random middle letter. This is exactly what they do in the case of strom. In context, the sentence itself rarely informs the guess, I’ve found. The children often guess a lookalike that makes no sense at all but carry on regardless. Many children I work with do this for a word every sentence or so, sometimes more often, sometimes much more.

Parents and teachers can (and do, I’m sure) easily investigate this for themselves, making a simple record of mis-read words and the ‘lookalikes’ from just 5-10 mins of reading aloud. Patterns emerge very quickly which are useful for informing teaching. As many teachers and parents realise, this ‘lookalike’ reading strategy shoots holes in a child’s comprehension.

However, the child can often get the overall gist to do well enough to at least talk about main threads – and answer some questions if a test. They can pull this off so well that they can even appear to be a ‘good’ or above average reader, sometimes, therefore, causing no concern at school. It’s the parents in these cases who’ve noticed the problem and who bring their child to me. I think these are the very children you are talking about in relation to the Y1 test.

But, the words they mis-read, with a quick guess and a lookalike, are often words they know very well and fully comprehend. That’s not right no matter how well they manage, whichever way you look at it. This is not effortless, accurate reading. We can’t let them move on to secondary reading like that, but we do. For these children, our teaching of reading has not equipped them adequately for secondary, where texts will be ever more difficult to read, with fewer pictures on which to rely.

Do we want any of our children to read like this, making incorrect guesses for any words they would otherwise know, after seven years at school? I don’t. This should be uncontroversial.

The students might get by at secondary, understanding enough to cope, and they might eventually self-fix their lookalike reading habit. I think many must, somehow. Perhaps if you do it for 5-6000 hours or so it starts to correct itself in time for GCSEs… or not. But whether they get by or they don’t, whether they are picked up and helped by brilliant teachers, parents or intervention strategies on the way or not (many are, of course), it still can’t be right to teach them to rely on luck and a bit of guesswork to read in the first place. No matter how lovingly and enthusiastically we do it. It’s a very odd strategy to deliberately encourage in our youngest, most impresionable pupils, given the flawed outcomes.

The only teaching solution for those already addicted to lookalike reading is to unpick the guessing habit and to teach children to read seemingly unfamiliar words by their letters, from left to right, accurately. (That’s what I spend my time doing now, teaching children to unlearn the habits that hinder their reading.) This may slow them down at first but speeds them up soon after if taught and practised thoroughly. But it’s not easy!

It would be much easier and more positive for all if, for those still early in the system and those just starting out, we unite, professionals and parents alike, in teaching children to read, with precision, the words they actually see on the page, not the ones they think they see.

Then they’ll be better equipped to read strom as the word it really is – alien, muggle, hobbit, wookiee, lilliputian, oompa-loompa or not.

With all best wishes

Jacqui is a secondary English teacher who worked as a literacy consultant and an Assistant Head in charge of literacy across the curriculum. She is also the mother to four young children all of whom she has encouraged to love reading and taught to read using phonics instruction.

2015-12-29T09:02:55+00:00December 28th, 2015|Featured, reading|


  1. Tom Burkard December 28, 2015 at 5:08 pm - Reply

    It’s really pathetic that we’re still fighting the same old battle. In 1995, Linnea Ehri wrote:

    “Studies of predictability of words in text indicate that on average 25 to 30 per cent of words can be guessed correctly. However the most important content words are the least predictable; only 10 per cent can be guessed…to guess words effectively, most of the surrounding words must be known. To read these accurately, a reader must use processes other than contextual guessing.”

    Jacqui’s post should be tattooed inside Michael Rosen’s eyelids.

  2. The Quirky Teacher December 28, 2015 at 8:27 pm - Reply

    I encounter this phenomenon all the time in upper primary and it does my head in. I think it is a massive, national issue. Unfortunately, those that ‘guess’ based on shape or first and last letter familiarity don’t seem to make any sense of the resulting ‘sentences’ whatsoever, yet when I mention that phonics knowledge gaps are the issues that need a few hours of intensive intervention, I just get told to ‘differentiate more’. What can I do with 15 minutes during assembly? What can I do during English input? The majority of my class have moved on. These children are at risk. I do what I can to teach the phonics during the times I hear these struggling children read, but it’s too little, too late.

    • Ljcowgill December 28, 2015 at 10:50 pm - Reply

      I’ve only been teaching since September but I’m sick of being told that I should differentiate more for pupils who can’t read!

  3. John Hodgson December 28, 2015 at 8:50 pm - Reply

    No-one knowledgable in teaching the reading of English would deny the value of a grasp of characteristic letter-sound correspondences. This is not the same as arguing that ‘phonics’ (a term that denotes a more or less intense focus on such correspondences) is the only important thing, and that children are being denied the gift of reading by those who advocate also using contextual and other cues. Let us take two words that come easily to hand: ‘Reading’ and ‘Didau’. Without context, I cannot tell whether the first word refers to a lexical activity or to a town in Berkshire. A correct reading of the grapheme depends on such knowledge. To read David’s name, which I have never heard spoken (perhaps I should get out more), I need to harness a range of resources. First of all I need to know where the morphemes join. Learning to recognise regular morphemes such as ‘ly’ is a very helpful aspect of reading, and my difficulty here is that I don’t know whether the morphemes are ‘Di’ and ‘dau’ or ‘Did’ and ‘au’. Reading phonetically, I’m tempted to say ‘Did-oh’ – but then I realise that I’m sounding as in French! My experience of foreign names makes me suspect that the is long and that the name is pronounced “Died-ow’. But until I hear the word spoken in the context of education blogs, I shall never know its morphemic and phonemic structure and will be unable to read it with confidence.

    • David Didau December 28, 2015 at 8:56 pm - Reply

      Has anyone actually argued that phonics is “is the only important thing”? Once the basics of word recognition are mastered then we can crack on with he vastly more important business of reading comprehension. To this end, contextual general knowledge is crucial. Once you know Reading (with a capital r) is also a town in Berkshire then you’re ushered in to a world of hilarious punning.

      Trouble is, a significant majority (roughly 20%) really are being “denied the gift of reading by those who advocate also using contextual and other cues”.

      By the way, you guessed dead right: it is “died-ow” 🙂

    • Maggie Downie December 29, 2015 at 9:14 am - Reply

      With respect I would suggest that John is getting his morphemes and his syllables mixed up.Surely morphemes are word chunks which convey meaning and are mostly affixes. The process he describes is syllabification, which isn’t particularly necessary. With regard to ‘Didau’ a phonics taught child would recognize that with only one consonant following the ‘i’ it is most likely to be pronounced as /ie/. No need at all to worry about where the word ‘splits’. 🙂
      I do agree that the ‘au’ pronunciation would be guesswork.

      • John Hodgson December 29, 2015 at 9:25 am - Reply

        I meant morphemes as I find that helping children to recognise affixes is useful. Understanding how words are constructed is one method in the mix. I think.

        • Maggie Downie December 29, 2015 at 6:03 pm - Reply

          But you’re puzzling me, John, as there aren’t any morphemes in Didau. (Well, there may be in the original language the word came from, but none that an English speaker could discern.)
          I agree that learning about affixes/morphemes is very important but it’s not an ‘alternative’ to phonics instruction…

          • John Hodgson December 29, 2015 at 6:12 pm

            Yes, the foreign origin of many words complicates matters. I was just pointing out that many young children can learn about affixes like ‘ly’ in early stages of reading. As I understand it, the more purist line of SSP would forbid pointing out such things to children as this involves ‘mixed methods’. I’m happy to be proved wrong.

  4. John Hodgson December 28, 2015 at 8:55 pm - Reply

    In my last comment, I attempted to write phonemes in the correct manner (enclosed in forward slashes) but the system appears to have taken these as formatting instructions and the result is a bit garbled.

  5. Debbie Hepplewhite (@debbiehepp) December 28, 2015 at 9:01 pm - Reply

    Hi John,

    Part of teaching is that learners will always need to be told the ‘exact’ pronunciation of some words – particularly personal names and the names of places and words which are not in the spoken vocabulary of the learner. This is merely part of phonics teaching and should take place within the wider curriculum also. Further, of course it is part of reading instruction and reading experience that context supports or informs what the word means and how it is pronounced such as the ‘read/read’, ‘wind/wind’ type of words.

    Kind regards,


  6. Debbie Hepplewhite (@debbiehepp) December 28, 2015 at 9:02 pm - Reply

    Hi David, I would need to be told ‘ow’ as in ‘owl’ or ‘ow’ as in ‘bowl’!

    Best wishes,

    Debbie (and thank you for this posting – much needed as you know)

  7. Leigh December 28, 2015 at 10:44 pm - Reply

    Thank you Jacqui. I have worked in primary schools for over 20 years and have seen too many children leave year 6 getting by using this guessing method.

    Children are still being seriously let down by the mixed methods of ‘teaching’, despite Jim Roses’ research and the DfE’s Letters and Sounds documents that clearly demonstrate the need for systematic synthetic phonics. The phonics screening check was one of the best things that ever happened to the teaching of the English alphabet system, but there is still gross ignorance, by many Junior age teachers, about the need for learning the complex code which goes beyond phase 6. In my experience the 2014 National Curriculum with its Y3/4 and Y5/6 word lists doesn’t encourage or support a thoroughly phonics approach, alongside the morphology/ etymology understanding. It was a missed opportunity that could have really made a difference to ‘ rectifying’ literacy learning difficulties before secondary school.

    I’m looking forward to a generation that have had a thorough phonics grounding who, hopefully, teach the disbelieving and misguided teachers!

    • Michael Rosen December 31, 2015 at 6:48 am - Reply

      I know people around here are sticklers for correctness so I thought I’d point out that Jim Rose’s name is Jim Rose not Jim Roses.

      • Jacqui MB January 1, 2016 at 12:31 am - Reply

        I’m no stickler! I’d have to self nit-pick. I make all manner of errors when I’m trying to type at top speed. I knew exactly which Jim we were talking about, and I was too interested to read the comment to notice. Jacqui

  8. Debbie Hepplewhite (@debbiehepp) December 28, 2015 at 11:45 pm - Reply

    David – your post is much appreciated – so I’ve flagged it up here:

    Warm regards,


  9. PStone December 29, 2015 at 8:57 am - Reply

    Hard to respond to this because we are not told how many children are reading inaccurately, how many errors they are making, nor what they are being asked to read, but I’ll try. Apologies that there will not be reams of academic papers to support this response – I just get on with it. If you want me to write about why I think some children who can read do not read fully accurately, I will, let me know?

    I too am often asked to work with children, usually in year 2, some older, whose teachers tell me they can ‘.. read well (as per year 2) , but the accuracy is not there and they can’t do comprehension.’ I’ll leave the ‘comprehension’ for now. And I do know that this inaccurate reading is very hard to fix when the child has got to year 5 or more and has been doing it for upwards of 3 years.

    The first thing I want to know is ‘Why are they given this level of book if someone thinks they can’t read it?’ It usually comes out that they can read their book, but are not bothering to be as accurate as they might. There is no great mystery to this. In brief, they have gone from FS and year 1 where adults at school and at home were giving almost daily focused attention to “Oh, you can read now – I’ll leave you to it.”

    What they need is to go back to having focused attention.

    And this is what I do:
    – we work in a group of no more than 6, with multiple copies of the same book
    – I tell them in advance what I am looking for – no skimming over words, read each word properly / listen to yourself, it has to make sense / read longer words by syllable
    – The 2 biggest things I tell them are:
    If you make a mistake, you have to notice, and you have to fix it.
    You are not allowed to jump on someone else if they make a mistake – we have to fix our own mistakes.

    (Is this still best practice, to tell children what we are looking for, before we start? I do hope so.)

    Once children know that somebody is closely interested in their reading, expects them to do it properly, knows they can do it properly, and that they are not going to be embarrassed by others hovering like vultures, waiting for the slightest mistake, it takes no more than 2 sessions – 2 sessions, tops – before they do it automatically and read accurately, independently.

    Then we can get on with the real work of enjoying our books, playing with them, talking about them and analysing their content.

    Let’s have a bit more “What’s going on for this child?” And a bit less “OMBGG! What the hell is going on in the previous key stage(s)?!”

  10. […] idea that phonics is not the only important ascpect of teaching children to read. Indeed not. Take this comment from John Hodgson for […]

  11. PStone December 29, 2015 at 10:05 am - Reply

    Re: phonics. All reception and year 1 and 2, and many nursery children, have been getting daily phonics lessons for at least the last 4 years.
    Each child entering year 3 will have had 550 phonics lessons, or 275 hours of phonics teaching.
    This is statutory.
    Why are the phonics fans still fighting about it – you are winning. You have won! You have phonics being taught at every child.
    RR has never been statutory and has never reached more than 12,000 children at any one time.
    Maybe it’s a blip or teething troubles that the DfE find this year that phonics teaching has slightly improved the Year 1 phonics screener pass rate but has had no effect at all on improving KS1 reading? Have the airy fairy ‘mixed methods’ (I have no idea what they are) really been able to counteract 275 hours directed phonics lessons? They must be pretty powerful.
    The real fruits of all this phonic labour will presumably come to pass with the 2017 KS2 SATs, when you will be vindicated for all to see.
    I can’t wait.

  12. Dr. P December 29, 2015 at 10:12 am - Reply

    I have recently taught my young daughter (then two, now three) to read. We used phonics (including the alphablocks show) and it was very successful. However, as fluency of word recognition increased, and speed of reading increased on higher stage text, occasional look alike errors appeared; here, I remind her to use her sounds, she decodes the word and we move on. She has only been taught phonics, so the look alike error is showing something that has developed internally as reading fluency increased.

    I’m not entirely sure but what this appears to show is that she has used phonics to learn what words sound like and what they look like but when reading fluently she is using this experience of words as recognisable pictures rather than decoding each word. She has the tools to correct herself and read unfamiliar words when stuck through phonics though. The next time she sees that word, she reads it straight off having now acquired a better understanding of what the word looks like. I don’t think there is guessing involved in the original error; just a lack of mastery.

    I think sometimes when people refer to a mixed method they mean this. Word recognition when reading quickly and phonics to decode when necessary.

    • Tom Burkard December 29, 2015 at 10:30 am - Reply

      You are quite right–once children have decoded words or spelling patterns often enough (this varies enormously from child to child), recognition is automatic. However, skilled readers still process virtually every letter in a word–take the sentence “The athlete suffered a compound fraoture”–the deliberate error will almost invariably be noticed, even though the letter o is visually very similar the letter c, and the world is highly predictable from context.

      Of course, children will encounter new words all the time in their reading. After the age of about 9, the great majority of new vocabulary is learned this way. For this, the child needs good decoding skills and–perhaps even more importantly–a good understanding of morphology, especially morphemes with Greek or Latin origins. Alas, this is seldom understood by phonics advocates, who tend to favour a syllabic approach to longer words. This is severely limited by the lack of any regular pattern that tells us which syllables should be stressed.

      • Onlyamanatee December 30, 2015 at 5:39 pm - Reply

        The phonics people I’ve spoken to seem pretty keen on morphemes as well as syllables – and I agree that morphemes should be emphasized, especially in teaching spelling 🙂

    • David Didau December 29, 2015 at 10:47 am - Reply

      That sounds sensible.

  13. Mark Bennet December 29, 2015 at 11:02 am - Reply

    Isn’t guessing in the way that young readers are guessing – from first and last letter and shape and a bit of context – a major component of adult fluent reading? In other words, adults decode text for meaning, rather than for words. That is not to say that words are unimportant, but I think it requires a case to be made if learners are to be discouraged from using the reading strategies of fluent adults. The best way of attaining adult fluency may be attention to separate words – but again, there is a case to be made, rather than an assumption.

    Writers know this too – when text is translated into a different language with emphasis on words it often ends up longer, and longer again if it is translated back to the original language. So language does not work perfectly at word level.

    So is “expert performance” in reading being correctly identified? And having been identified, are teaching methods supporting the formation of expert performance over time?

    On the other hand it was the ability to guess simple words in this way which masked the need for glasses in one young child I know. And the strategy adopted at that stage was too crude to be effective and needed to be unlearned. I would suggest that the diagnostic focus of early reading interventions is also important.

    • PStone December 29, 2015 at 11:53 am - Reply

      What people call ‘guessing’ is actually deductive reasoning. Wouldn’t we all love children to have deductive reasoning under their belts at all times.
      Re: diagnostics, agreed again. Simplists (and reading is not simple) tend to notice problems and attribute the same cause and cure for each one. When they don’t work, they like to say the child has special needs. Their real special need is that nobody has been able to work out how to teach him.

      • Maggie Downie December 29, 2015 at 6:20 pm - Reply

        Interesting, Pat, your analysis of why children are labelled ‘special needs’.

        I worked for over a decade with KS3 children, many of whom came to me in Y7 with the ‘special needs’ label. The ‘special need’ most of them had was the ‘need’ to be taught that you work out what a word ‘says’ by decoding and blending, not by looking at it and guessing, or by saying that they didn’t know the word and expecting me to tell them what it ‘said’. This came as a complete surprise to most of them.

        It is to be hoped that more children are now reaching KS3 with better phonic knowledge but I suspect that there won’t be much improvement until the children who have taken the PSC start arriving.

        • PStone December 29, 2015 at 6:38 pm - Reply

          September 2017 – everyone awaits with baited breath.
          Thank you for supporting my point that children need someone to work out how to teach them.

          • Maggie Downie December 31, 2015 at 12:10 am

            It’s ‘bated’breath, Pat.As in ‘abate’,to diminish. We’re holding our breath, not putting something attractive in it in order to entice and entrap something.
            Just saying…

          • PStone December 31, 2015 at 8:15 am

            Maggie. Shall I go through your writings and find all your predicter text errors for you? There are three spaces missing in your last comment to me, for a start.
            Grammar police allus done it rong they selfs.

    • Tom Burkard December 29, 2015 at 11:55 am - Reply

      A lot of the confusion in this debate comes from the differing definitions of ‘reading’. Of course we read text for meaning–there would be no point to it otherwise. We also drive a car to get from one place to another, but this doesn’t obviate the fact that we aren’t going to get anywhere at all unless we know how to operate the controls efficiently. With skilled drivers, these mechanical functions operate without attention or effort.

      In the US, beginning drivers often begin driving in empty car parks, where there is no traffic to distract the learner from the essential task of operating the controls. This is far more effective than the standard procedure in the UK, where the learner has to cope with traffic well before any kind of automaticity in basic functions has been achieved. This is perhaps one reason why there are far fewer accidents per driver in the US than here.

      With reading, the desire to get ‘meaning’ is almost irrelevant when children first learn to decode. The desire to please parents and teachers is a far more salient factor than any meaning that may be contained in a beginning text. The simple fact remains: all good readers can decode print to sound effortlessly and automatically, and for all children, these skills are more efficiently learnt when they are explicitly taught.

      • PStone December 29, 2015 at 1:21 pm - Reply

        I like your driving in the car park analogy. But phonics is not about a clear road. It is about a lesson in switching on the ignition, then a lesson in releasing the handbrake, then a lesson in finding first gear. Lesson 4 would involve doing 1, 2 and 3 together and smoothly. No lesson in mirror checking or putting your foot on the accelerator, until you can do 1,2,3 automatically. And then it’s the weekend, so you probably need a refresher on 1,2,3 on Monday.

      • ian batten December 30, 2015 at 6:37 pm - Reply

        “This is perhaps one reason why there are far fewer accidents per driver in the US than here.”

        Are there? The USA has about four times as many road deaths per head of population, about twice as many road deaths per vehicle-mile. Given it has a lower general speed limit and (outside a small number of very dense conurbations) lower traffic density, that doesn’t seem consonant with a lower level of accidents (although there are some confounders such as the lower rate of seat-belt wearing). “Far fewer” seems a very strong claim: do you have a source?

  14. 1ofthe40 December 29, 2015 at 7:12 pm - Reply

    Whether or not you agree with it the PSC is here to stay but the problem is that because it is a mix of real and fake words, it is an unfair test and that’s one of the reasons it doesn’t fairly test reading. Either make it a decoding test or a vocabulary test. Fake words are fine to test decoding skill but real words should have a context (such as a simple sentence) to aid the reader’s understanding of what the word should say.
    One word from a previous year was ‘fuel’. Not many 6 year old use the word fuel and a lot of them read it as ‘fool’ (using the ‘ue’ from ‘blue’) which would been a pass if it was a fake word but not because it is a real word. Similarly a lot of kids read ‘crowds’ as ‘crow-ds’. Again, a pass if fake but a fail just because they weren’t familiar with that word and there was no context to make it make sense.
    A simple sentence along the lines of ‘Dad put fuel in the car.’ or ‘We stood in a crowd.’ would go some way to rectifying this.

    In cases like the above, knowing the word is more important than just decoding.
    So test the kids if you must but the test in its current form is flawed.

  15. […] Others have blogged about reading for pleasure here: Alex Quigley David Didau […]

Constructive feedback is always appreciated

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: