The forest for the trees – or, what is the point of phonemes?

I have recently started re-reading Alan Watts’ The Way of Zen, and (as ever) a conversation on Twitter has prompted this little ramble. I’m going to quote a lot of Watts, I’m afraid. Early in Chapter 1, he says this:

The task of education is to make children fit to live in a society by persuading them to learn and accept its codes – the rules and conventions of communication whereby the society holds itself together. There is first the spoken language. The child is taught to accept ‘tree’ and not ‘boojum’ as the agreed sign for that (pointing to the object). We have no difficulty in understanding that the word ‘tree’ is a matter of convention. What is less obvious is that convention also governs the delineation of the thing to which the word is assigned. For the child has to be taught not only what words are to stand for what things, but also the way in which his culture has tacitly agreed to divide things from each other, to mark out the boundaries within our daily experience.

(NB as a Zen Buddhist, he doesn’t really think this is what education should be; he is commenting on how it appears to him to work in practice)

Rather than seeing ‘tree’ as your unit of definition, it’s possible (actually, maybe, more accurate) to imagine conceptualizing a whole forest as the complete organism, which then consists of parts: soil, roots, mycelium, understorey, canopy, fauna. In fact the view that a tree is a separate ‘thing’ is, ecologically speaking, rather problematic.

As with trees, so with words. This lovely study, by Amanda Seidl at Purdue University, looked at how babies learn to separate the sounds of words as discrete units of meaning:

“Parents may pause before saying an infant’s name, but they almost never do so for other words. This research explored whether touches could help infants to find where words begin and end in the continuous stream of speech. They need to find words before they can attach real meaning to their words,” Seidl said. “Because names of body parts are often the first words that babies learn and touching is often involved when caregivers talk about body parts, we speculated that touch could act as a cue to word edges.”

That is, babies may learn to distinguish words from the actual flow of sounds via the tickling of their toes.

Watts’ paragraph above continues:

Scientific convention decides whether an eel shall be a fish or a snake,and grammatical convention determines what experiences shall be called objects and what shall be called events or actions. How arbitrary such conventions may be can be seen from the questions, ‘What happens to my fist [noun-object] when I open my hand?’ The object miraculously vanishes because an action was disguised by a part of speech usually assigned to a thing! In English the differences between things and actions are clearly, if not always logically, distinguished, but a great number of Chinese words do duty for both nouns and verbs – so that one who thinks in Chinese has little difficulty in seeing that objects are also events, that our world is a collection of processes rather than entities.

So… my point is that yes, as many have said, the concept of the phoneme is an abstract one, an imposition upon the flow of sound which really makes up human speech. But then so too, in its way, is the concept of the word.

The separation of words in text is something that we take for granted – but many cultures in the past did not do this, and a few still do not do so today. To what extent are the unseparated words less conceptually separate, I wonder? (Or possibly more so, if no visual divider is needed?)

Just as babies need help separating words from the flow of speech, so new readers need help separating the chunks of sound that make up words. Current phonics practice is only one way of doing this, but in order to understand the mechanics of our language, it’s a pretty good starting point (as long as the words worked on are real ones, of course). The phoneme is an artificial distinction – but then, according to Alan Watts, so is everything else we think we can see.

ITT and synthetic phonics – collecting experiences

Questions relating to the efficacy of SSP in schools often return to the issue of how teachers were trained, and how they are trained now – as well as to the quality of CPD.

Many people who are in favour of the current approach to using SSP in schools seem to feel that it is not taught properly/enough/at all in ITT – or at least, their experiences suggest to them that this has been the case so far.

Yet apparently ITT programmes will fail Ofsted if they do not teach SSP, so presumably all are now doing so.

Is this discrepancy the result of a slight lag in the effect of all ITT programmes teaching how to use SSP? Or is it more complex (for instance, there could be different approaches to teaching the subject within different ITT courses, some of which prove more useful in the classroom than others). There are so many potential ways to teach SSP to ITT students, with potentially hugely varying results in how they might engage with it once they are teaching.

A deeply important question is – what about the TAs? Are they confident and knowledgeable in their support of children’s literacy?

NB I have no idea at all of the truth (if there is one – there could be many); this post is basically a call for information, from teachers, trainers, ITT students, and anyone else with an interest in/knowledge of past and current situations.

Please comment – this seems to me to be an area in which greater clarity might prove to be extremely helpful in removing some urban myths.

Edited to add: Andrew Davies sent me this, which does seem to allow for great variability:

ofsted SSPAnd to add: there ‘s a thread in the RRF forum on the issue here.

And: I spoke to Keith Turvey (of the University of Brighton) on Twitter, who pointed out that ‘most ITT time is in school now’ and also said: ‘Most [ITT courses] set out expectations for SSP like we do but the execution on the job is down to schools and trainee, and the quality of trainee experience and opportunity varies significantly IMO. Thus partnership key.’

These seem some of the most important practical points, to me.

Look, Cover, Write, Check

Imagine if maths were taught the same way that the Look, Cover, Write, Check method ‘teaches’ spelling.

Each week, children would be given a list of 5-10 calculations, perhaps 10×45=450, 5×60=300, 4×39=156, 76×89=6764, etc, depending on how far up the school they are.

They have not been taught times tables, nor have they been taught how to do those calculations. They simply learn them by rote, so that in the test at the end of the week they can write out the whole of each calculation from memory.

How many children would be able to deduce, over time, the underlying patterns and rules and simple maths facts which are common to these calculations?

A few would, of course, and would be able to apply this knowledge to other similar calculations. Some would work out a bit of it, and so might make a decent stab at calculations which are close enough to the learned ones for some inferences to be made.

Many would have no clue.

If the specific calculations did not happen to turn up in other class work, I think most children would forget the list within weeks.

Yet this is what we expect from children in the Look, Cover, Write, Check method.

Without work on phonics and morphemes, rote-learning the spelling of ‘whether’, or ‘stupidly’, or ‘procrastination’ is no more useful than rote-learning 35×16=560 but never looking at 3×1, 30×10, 5×6, etc..

Memorization is crucial for some things, but pretty useless for others. When ‘rote’ learning is discussed, whether positively or negatively, it’s absolutely essential for this aspect to be clear.

See Improving Literacy by Teaching Morphemes for more on better methods.

How is reading being taught in the wild?

Reblogging for interest…

Filling the pail

Around the turn of the century, a U.S. panel reported on the evidence about how best to teach children to read. They were crystal clear; a systematic phonics programme was best. This was seen by many as the definitive end of the ‘reading wars’ that pitted whole-language advocates against promoters of a phonics-based approach. Whole-language was a theory of learning to read that emphasised whole words, ‘real’ books and students ‘constructing’ their own meaning. As such, it aligned with ‘constructivist’ views of teaching that remain fashionable in schools of education.

However, like Fukuyama’s declaration of the ‘end of history,’ hopes for an end to the debate on reading represented a false dawn. Whole-language advocates rebadged their approach as ‘balanced literacy,’ implying that phonics was now a part of it, but only one component part. Many people have come to accept their rhetoric that spending 5 hours per day doing nothing…

View original post 1,440 more words

Phonics and literacy – separate strands, or warp and weft?

One passage stood out for me in the recent report on the usefulness of the Phonics Screening Check:

The most frequently reported change by survey respondents in 2014 was increasing the pace of phonics teaching, and this finding was supported by data drawn from the case studies. As in 2013, an increased focus on pseudo-words was also reported by participants in the survey and case studies, as well as increased assessment of phonics. Analysis did not determine what form this increased focus took.

‘Analysis did not determine what form this increased focus took.’


Let’s just do a thought-experiment. Imagine that what’s going on in [some] schools in the name of phonics is actually not very helpful to wider literacy. Perhaps there are lots of nonsense-word games and very little 1-2-1 reading support. Then imagine that teachers have increased this extra phonics activity as a result of the PSC. Perhaps they are playing an extra hour of nonsense-word games per week. What else might they not be doing as a result? If this were the case, it would be unsurprising if a reported increase in ‘phonics’ would not result in an increase in literacy levels.

I’m sure this sounds like blaming teachers: I’m not, I’m just asking questions. I’m a firm believer that teachers’ motives are good in the vast majority of cases, but an equally firm believer that the contents of a test will skew what is taught. You’d be crazy not to consider the test, really, wouldn’t you?

In the past, I’ve seen many claims of problems for good readers within phonics schemes, where very regimented work apparently prevents children from progressing, or treats children who can read as if they can’t, but until now nobody has been willing to answer my questions about what actually happened.

Around the time of the PSC check this year, though, one parent, on condition of anonymity, agreed to share her daughter’s story. It follows a number of patterns that have been claimed elsewhere as well, so I think it’s very much worth looking at to help unpick what’s actually going wrong in these kinds of case.

Since the report into the check didn’t look at what was actually going on in schools, anecdotes like this one represent the only comparison evidence available.

NB This post is not meant as a criticism of any particular scheme but a question about implementation and effects, and about how schools use and understand the synthetic phonics material that they are teaching.

Here’s the account of her experience:

Hi, my daughter who is now 7 and in y2 didn’t read before school. She was tested at the end of reception by an ed psych (at 5.5) and given an IQ on the 99.4th percentile and a reading age of just under 8. In reception she had a great teacher who skipped her through the [commercial phonics scheme stages] as her reading required. They had separate reading groups and N___ was always in the top group for that (the top group had 3 gifted children including N___ and were way above the rest).

In year one they stuck rigidly to the [commercial phonics scheme], the teacher was inflexible but in the reading group N___ remained in the top group. However in [commercial phonics scheme work] she was overtaken by others who she could read better than. I thought at first it was because of her writing – but no it was explained to me that in order to progress they had to know certain sounds.

They were given in isolated sounds for example what sounds does ou make, however give her an ou word such as group, soup, troupe, she had no difficulty reading it. She also had no difficulty sounding out new words – so wasn’t learning whole words rather than phonics.

She had to repeat [commercial phonics scheme stages], switched off and started to hate [commercial phonics scheme] lessons. In the end I got a copy of the sounds off the Internet and took her through them. We learned them by rote each half term to get to the next stage.

When she started y2 she had 3 stages left. She got her reception teacher again, she was still in the top reading group but one of the middle [commercial phonics scheme] groups, the top reading group was reading novels in class she was then going into [her commercial phonics scheme group] and doing easy books. Her teacher took her off the scheme within a few weeks.

Four things stand out for me here.

1. The mismatch between the child’s assessed reading age and progress, and her assessed progress with phonics sounds;

2. The fact that with a few videos from the internet (provided by the same scheme, I think) her mother was able to teach her daughter sounds which the school had apparently been unable to teach her for close to two years.

3. The different ways in which the various teachers approached the mismatch in progress.

4. The actual lived experience of this child. Imagine how baffling this must have been for her.

1. The mismatch with the reading age might be explained because a test of reading age based on common vocabulary would be testing something different from the ability to use phonics sounds to work out words, BUT I think it’s very interesting that the child could sound out and read words containing sounds which, in isolation, she did not know.

There might be various reasons for this: she may actually have known all the words by sight (though her mother notes that she could also sound out unfamiliar words, so this does not appear to be the reason); or (more likely) she might not have felt able to assign a specific sound to ‘ou’ if she knew that there was more than one possibility. Or there might be some other reason – unfortunately, we can’t know.

2. The mismatch between the child’s apparent inability to learn isolated sounds in class, and her learning of them at home with her mother, suggests that there was no deep learning issue attached to the child herself with regard to this. But it does suggest that something in the way these sounds were being taught at school did not click with her (and this is an intellectually very able child).

3.The problems for the child appear to have arisen as a result of inflexibility on a particular teacher’s part: a flexible approach to the system was much less problematic.

4. It seems to me pretty awful that this child was put into such mismatched groups for sounds and for reading actual books, when she was apparently able to learn the sounds perfectly easily and quickly.

This is not the only example I have heard of first-hand of unhelpful placing of children for phonics work. That’s not the fault of the alphabetic code.

If, as Debbie Hepplewhite has said (I think), there are potentially as many different ways of teaching phonics as there are schools, but many of them think they are doing the same thing, examples like this become essential, in order to allow comparison. In this instance, within the same school and one overarching system, there were clearly substantial differences in implementation (and the learning and well-being of the child).

The report into the PSC was necessarily inconclusive because the writers had no decent evidence – this is what they mean when they say they had no comparison (i.e. control) group, no ability to separate effects of PSC from effects of other phonics practices, and no knowledge of what actual methods ‘increased’ in schools. Absence of evidence is not evidence of absence, of course. Policy makers should be very careful about how they respond to this report.

What’s needed in order to resolve this debate, it seems to me (looking at it from a research perspective) is a much more finely-grained investigation, looking at correlations between different methods and outcomes. Do schools which are heavy on explicit and standardized phonics techniques do better (on a value-added score, not just the PSC) than schools which teach the sounds quite simply but then just do lots of their own kinds of literacy work to embed them? Which changes lead to improved PSC results? Which changes seem to lead to improved literacy results? Are these the same changes? Is it about methods in the abstract, or the quality of teaching regardless of method? Or (as seems most likely) something in between? This investigation would need to be at the level of classes within schools, since as the example shows, a school’s results may mask great variation of approach, even within an outwardly very standardized system.

If this report leads to greater division between phonics and ‘anti-‘phonics people that would be really unhelpful – most ‘anti-‘phonics people are worried more about methods and effects than about the content itself, when it comes down to it. And those methods and effects are not dealt with in this report.

The report does emphasize the clear and obvious benefits of teaching a SSP-style alphabetic code in some form. But what form? In the above example, the method seems actually to have undermined the child’s confidence when rigidly applied, and not even to have taught her the basics very effectively. Is this because of the method, the teacher, or a change in the child? The latter seems unlikely since she picked up actual reading so quickly and so early. So something wasn’t right in the classroom, for her at least.

I have expressed my concerns with the quality of the PSC previously here – though my understanding has moved on since then – and also here. But I find it worrying that a number of people seem to be cheering the inconclusiveness of the report – some even seem to have interpreted it as the PSC itself having no effect, rather than the authors of the report not having access to data which would have demonstrated what sort of effect it has had.

It has clearly had one effect, which the report itself acknowledges: schools are ‘doing more phonics’. But I don’t think anyone’s any closer to knowing what that means in practice.

One more thought experiment:

If your phonics is badly integrated with your broader literacy curriculum, isn’t it most likely that doing more (but dislocated) phonics work is going to improve your PSC results but not your overall literacy ones?

Since this is exactly what the report has found, shouldn’t we really be asking about the issue of how (or even if) phonics knowledge is embedded in good, complex, deep literacy work? And how the DfE can create drivers for this to happen more?

While the PSC has been shown to be a decent proxy for literacy despite its weaknesses, it’s not going to answer questions about how improvement happens. It’s not going to tell you why literacy has not improved in certain schools. It’s not even a perfect test of phonics: an improvement in PSC score could conceivably be accounted for by an increase in the real words score, which could be the result of other aspects of literacy work, not explicit phonics at all (an extreme and unlikely case, but a logical possibility).

Have any schools had their literacy results go down following the introduction of the PSC? Now those schools would be really worth looking at, if there are any.

Phonics knowledge seems to me to be the warp on which the whole of literacy can be woven. It’s integral to reading, writing, and understanding language, and like the warp in a piece of cloth it needs to be strong, simple, carefully laid but ultimately invisible. And if you don’t weave in the weft, or you expect it to just happen by itself, or you try and put the warp and weft onto separate looms, you’ve got nothing useful at all.

Weaving it in means never forgetting what phonics is for. It means never being so proud of your warp that you forget its real purpose is for weaving something far more complex and beautiful. Is this beyond the DfE to comprehend, I wonder?

NB I’m sorry this is so opinionated this time! But really. How on earth do people maintain their patience for years when the real questions needing an answer are so obvious….? If phonics is so embedded in schools, if it is properly integrated into literacy work, why are all the parents I speak to (from various different schools, all at the just-missed-‘outstanding’ end of ‘good’) still in a fog about it?


Books for confident readers, 5 yrs and up

[Disclaimer: this post is not a value judgement on children from whatever circumstances who are not reading at this level at this age. It’s not that children should be doing this, it’s just that these ones are; this is just a response to my own experience of trying to find the right level of book for a number of children aged 5-6 over the past year, and, through that process, finding that modern books for children seem much easier than books aimed at the same ages in the past. I’m also trying to find my way to books they might read once they’re in Year 2 (and 3).]

Children who are confident readers at a young age have probably always tended to find that the books that are written for their age group are way too easy and therefore dull, while the books that actually get their minds working are too old for them from the point of view of emotional maturity.

However (take that, Mr. Gove) as a rule of thumb, from what I’ve seen so far, modern books aimed at, say, 8-9 year olds seem at least two or three reading years’ easier than older books (ie pre-1970s-ish) aimed at the same age. The characters in Beast Quest and Jack Stalwart are very similar in age to the Famous Five or the children in the Narnia books, yet the font size, length of text and complexity of plot are miles apart. I imagine there are all sorts of reasons for this, but it means that for children with a high relative reading age, the disparity is even greater than in the past it might have been, and modern age-appropriate books that actually give them a challenge seem harder to find.

That is, if you are a 6-year-old with a reading age of 9 or10 (for instance), the modern books of the right reading level will be aimed approximately at 12 year olds, and therefore be a bit bemusing. Since there’s a deluge of newly-published children’s books to wade through every year, it can be difficult to find a path through it all.

I’ve estimated that the more modern books listed below are readable in an hour or two by the children I’m talking about. The books are not challenging to them in terms of difficulty, but they do enjoy them (which is at least part of the point…). A major aspect of the speed with which may of these children are reading, though, is the cost, and the importance of libraries and of schools having a decent book budget – buying them all at £8.99 or so is not an option when they’ll go through four or five a week at least during term, given the chance, and more in the holidays.

A Blyton or longer Dahl, by contrast, can often work well spread over a couple of weeks, and have more in it to discuss.

A lot of the things in the list will be familiar, but what I’m hoping to do is collect other suggestions as well and include those in the future. Any ideas welcome.

NB I’ve had some people suggest Michael Morpurgo, but ours just aren’t keen yet. No monsters (all-important at the moment), and the books of his with the right level of vocabulary etc seem just too emotionally advanced for them. It’s a tricky balance. UPDATE: One has just expressed an interest in a MM book. We’ll see how it goes.

Our next books will be some E. Nesbit, so I’ll add some notes on those as we go.

Anyway. The list so far is pretty short, just a drop in the ocean, but here it is. I hope it will grow. I’ve divided it into three sections:

1. Right level, some old-fashioned content; 2. OK content,  more modern books; and 3. Other Solutions. (Our own ranking at the moment is based on the number and kind of monsters, but that’s just us.) I suppose I’m trying to do two things – find some good books, and start to understand why the current situation has arisen (and whether the situation is what I think it is).

1. Right level, some old-fashioned content

Anything by Roald Dahl

Dahl is an obvious choice – but some children find some of his writing just too spooky when they’re 6. Charlie and the Chocolate Factory goes down well though. Some aspects of the books are showing their age. But he’s brilliant, basically irreplaceable and a lot of it is timelessly surreal.

Anything by Enid Blyton

Great in terms of level in many ways – no complex character motivations – but there’s always the problem with old-fashioned attitudes to women and to class. I think most people remember how Anne is treated in the Famous Five – but rereading recently I’ve also noticed the assumed superiority of the children over the servants – as well as the fact that there are servants at all, and that their speech is represented as inferior via phonetic spellings rather than the normal ones. So, again, explanations often needed – a middle-class readership, and certain out-of-date attitudes on their part, are assumed.

Anything by Arthur Ransome

One boy I know, now in Year 2, has been reading these avidly since the end of Year 1, and adores them; I loved them too, but I haven’t tried mine on them yet, the lack of monsters currently being a barrier. Again, though, there’s the problem that it’s not set in the same world that modern children inhabit, even middle-class rural ones like ours.

The Narnia series

Also brilliant, obviously, but also assuming post-war-era middle-class knowledge.

Any suggestions as to recently-published books of this sort of reading level, with this kind of depth, would be very gratefully received.

2. OK content, more modern books

The major problem is that they’re mostly so short, but I’m including them because somebody, somewhere, has loved them. I’m sure there are hundreds, possibly thousands of others in this category – Horrid Henry etc….

Greek Beasts and Heroes – Lucy Coats

A collection of very short books, each a very quick read for children reading at this level – but they’re getting some deep culture in a beautifully presented and appealing form – stories which will underpin all sorts of later reading (including J. K. Rowling, and, you know, Shakespeare…). Nice little books with lovely pictures – good for stretching the imagination. Lots of monsters.

Beast Quest, Sea Quest, etc.

Again, a quick read though longer than the above; lots of monsters, very pulp fiction really, but well done of its kind – lots of dialogue, lots of use of syntax, punctuation etc to create a sense of action in the descriptions. No depth to speak of, but lots of Adventures. Male central character, male viewpoint, but the girls aren’t silly and actually get to do stuff, which is something. One of my boys says there are some tricky words once in a while.

Dinosaur Cove

Lovely books, modelled on Beast Quest but with some paleontology for good measure and less fighting. Shorter & easier than BQ, and with entirely male protagonists, which is unnecessary. I wish OUP would produce some longer more challenging DCs with some girls in! They’re apparently aimed at 8-9 year olds, but they’re too easy for the confident Year 1 readers I know, who were whizzing through them early in the school year. The less confident but capable ones are motoring through them and similar at the moment (ie end of Year 1, mostly 6 year olds, some are still 5).

Rainbow Fairies

Argh. Some of the girls love these as initial chapter books, and whizz through them. Female point of view, obviously, but it’s all a bit Lego Friends for my liking. Luckily they mostly seem to get bored after 5 or 6 of them(!) and move on to Dahl and Blyton.

Jack Stalwart

Very similar level to Rainbow Fairies, very Boy Boy Boy; a smidgeon below Beast Quest in terms of difficulty. No depth, and, again, male protagonist. But fun if you like that sort of thing.

Anything by John Dougherty

Completely crazy – readable in an hour or two, but lots to spark the imagination after the book is finished. Good creative stuff, but so short!

The Worst Witch series

Similar to Beast Quest in level, but female protagonists. A wry sense of humour, but very boarding-school plotlines: the Harry Potter books owe a lot to these in terms of atmosphere and setting, though these are far less spooky, scary or complex. Very un-pink.

Dick King-Smith

His books aren’t massively long, but there’s a wryness, a craziness, a love of his characters, which creates depth. I like them. More to them than meets the eye.

3. Other Solutions

Comics, graphic novels, stories derived from these

Tintin a big favourite with some; DC Comics do a series of short ‘Super Heroes’ chapter books (again, an hour’s read at most, but great for kids who are into that stuff); the Phoenix Comic (a weekly comic with beautiful and very cool comic art, a range of levels of story, no advertisements, no weirdly oversexualized ladies…)

Horrible Histories

They like to dip in and out of these. Not for the squeamish child, though…

Usborne Beginners, Lift-the-flap reference books, etc

I’ve just recommended these for a confident reader in Foundation – the Beginners are book band 9+, beautifully illustrated, lots to think about. The Usborne non-fiction stuff is great, and I’ve found our two come back to them and get different things out of them at different stages.

Branded encyclopedias of Lego, etc.

Not a fan of tie-ins really, but our kids and their friends just love this stuff and will pore over them for hours together. They’re learning about reference books, too, and hierarchies of information, in spite of the made-up-ness of the ‘facts’. Great, actually, for feeding into complicated imaginative games in the playground.