A simple theory about writing


The first thing to say is this is not in any way supposed to be a complete or unified theory – I’m well aware that there are many other important strands to improving pupils’ writing and have written about many of them before. But I do think this theory (which has been bubbling away on my mental back burner for a while now) describes just one of the processes that can turn otherwise able pupils from poor writers into much more able ones. That said, I tend to get a bit over excited about these sorts of things and am often mistaken. Knowing how keen many readers are to point out flaws in my thinking I thought I’d test out its robustness by sharing it with you.

So, here goes:

1. We can only say what we can think.

2. We can only write what we can say.

3. But if we can say it, we can write it.

What does this tell us about the importance of speech?

Myhill and Fisher tell us that spoken language “forms a constraint, a ceiling not only on the ability to comprehend but also on the ability to write, beyond which literacy cannot progress.” This being the case, why isn’t talk taken more seriously in schools? Well, Robin Alexander suggests that “One of the reasons why talk is undervalued is that there is a tendency to see its function as primarily social… buttalk in classrooms is cognitive and cultural as well as social.”

Now that sounds excellent doesn’t it? But what does it actually mean? If my first statement, “we can only say what we think” is correct, and I’m pretty sure it is (try thinking about something you don’t know. Tricky, isn’t it?) then what we need to do is improve pupils’ thinking. But thoughts are hard to see. All we as teachers can do to evaluate our pupils’ thoughts is to examine what they say and do. Clearly if you can already write well then there’s no real need to make you speak – we can see your thoughts on paper. But often children struggle to turn their thoughts into words fit for the page.

Typically, we focus on the confidence and clarity with which pupils speak. We worry far less about the quality of what they say. The overwhelming majority of pupils (especially in secondary schools) can speak fluently, but many can only speak in ‘everyday’ language; they are unable to put their thoughts into a more academic register. This ability is something teachers usually take for granted: we mentally translate our ‘everyday’ utterances into academic language. Some of our pupils can do this too. But some can’t. Take for example the ‘verbally able’ pupil who has great ideas but seems incapable of writing any of them down. What’s with that? Contrary to what we may think, they’re not being awkward or lazy, they simply can’t translate their thoughts into the language of writing. They know what they’ve just said wouldn’t make sense if they wrote it down. So they don’t.

What we tend to do is give these pupils writing frames to enable them to structure their thoughts and get something written down. These kinds of constraints are stifling and result in stilted, plodded writing and dependent pupils. Instead, all we need to do is to use these scaffolds at the point of speech; simply ask pupils to rephrase what they’ve said so that they ‘speak like an essay’.

So, when conducting a classroom discussion, display prompts around the room and direct pupils to express their ideas using the thought stems. Here’s some history ones:

  • One effect of … was to …
  • This was because…
  • As a result …
  • The first reason for … was …
  • These factors led to …
  • The effect of this was…

Or how about these from English:

  • This might mean… because…
  • The writer used the phrase… to imply…
  • The writer’s intention is…
  • This could also suggest…
  • The word … is effective because…

Another advantage to using these prompts at the point of speech is that the scaffolding is a lot easier to remove as pupils are less likely to repeat cumbersome frames unnecessarily.

Using these prompts in combination with key words is even more effective. If we want pupils to ‘speak like scientists’ then we need them to not only use subject-specific keywords, but we also need them to use language in a way that a scientist would. So, in an experiment designed to demonstrate the principles of the conservation of mass, teachers would, as a matter of course, introduce the keywords ‘products’ and ‘reactants’. But what happens then?

Take this example:

Screen Shot 2014-02-26 at 13.24.53

Although the logs (reactants) seem bigger than their charred remains (products), the mass remains the same. When pupils have a go at this experiment, they naturally want to say things like: ‘The amount of reactants is the same as the amount of products.’ We need to scaffold their ability to talk about this chemical reaction using scientific language. We need to show that them that a scientist would be more likely to say: ‘The mass of the reactants equals the mass of the products.’ Pupils tend to say: ‘The mass has stayed the same.’ They need prompting to say: ‘The mass has been conserved.’

And because their ability to speak about the conservation of mass has shifted, so has their ability to think. If we explicitly teach pupils to talk in this way, then they will be able to think like scientists. But we need to give them further opportunities to speak like scientists and to explicitly use their language/knowledge to talk about experiments in terms of scientific equations.

And having expressed thoughts in academic language allows us to write in academic language. I’m not really sure why this is. It may have something to do with new synaptic connections being formed by having expressed thoughts in a particular structure, but then again, that might be gibberish. What I do know is that this works. Certainly there are a few caveats: this doesn’t take into account motor difficulties or even confidence, but providing you have learned the print code, can hold a pen and are willing to give it a go, it’ll work.

The downside is some pupils’ unwillingness to reframe their thoughts using your prompts. Why, they might truculently ask, should we change the way we speak? It’s important to explain that there is nothing wrong with ‘everyday’ language. They way they speak is fine for everything else but it’s not the language of the classroom and if they want to write like an essay, they need to be able to speak like an essay. Like anything else worthwhile, this will take time to embed; changing classroom cultures is simple but far from easy.

Here’s a presentation on this and other ideas for improving writing that I gave to Education Scotland’s National Literacy Network in March:

I go into more detail in my new book The Secret of Literacy: Making the implicit, explicit which should, barring further disasters, be available in June.

Related posts

How to get students to value writing
Developing oracy – it’s talkin’ time
Mind your language

29 Responses to A simple theory about writing

  1. […] Read more on The Learning Spy… […]

  2. Tim Taylor says:

    Excellent blog. I agree with every word.

  3. Very interesting David – and right at the heart of a major teacher-research tradition in history education. There has always been a great deal of interest in history about the relationship between historical knowledge/understanding and the language of history, but this increasingly dominated history teacher published discourse in the early 2000s following the work of Christine Counsell on history and literacy. She was very critical of writing frames etc. and emphasised that getting better at writing was actually about getting better at thinking about history – if we can organise our historical knowledge then we can write about it far better.

    There then followed a large number of articles published by history teachers in the journal Teaching History that looked at this very issue. James Woodcock’s 2005 piece ‘Does the linguistic release the conceptual’ is the most cited piece of history teacher research published in the last ten years, and his explicit consideration of the relationship between historical knowledge and historical language encouraged many other history teachers to write about the use of language. At the same time, history teachers (arguing against short-answer exam-style writing) were pushing hard for essays and arguing *against* writing frames, as these prioritised structure over knowledge, resulting in poor writing.

    Perhaps the best recent work on this (I would say this…) has involved teachers looking at the relationship between pupils *reading* works of academic scholarship and writing their own. I wrote a piece in 2007 on how pupils could not argue unless they had something to argue *against* (pupils are always told to be ‘argumentative’ but no one really told them what this meant) while other history teachers (say Bellinger in 2008, Foster in 2011, Laffin in 2013) got pupils (from KS3 to A-Level) looking in great detail at how historians structure knowledge in writing. In short, the thing that concerns history teachers (at least those doing research and publishing it) is the relationship between knowledge and literacy.

    In all we have a very rich research tradition on this in history education – if anyone’s interested in following this up with me I’d be happy to continue a conversation!

    • David Didau says:

      Thanks Michael – I’m a fan of Counsell and also of Caroline Coffin who’s done good work on academic language and genre pedagogy in history. I think the problem with writing frames is that they stifle, but some constraints are excellent for scaffolding thought. By teaching sentence structure we can not only make children but writers but also better subject specialists.

      • Absolutely agree, Michael. Take Edexcel’s History A. There are three two-mark questions that require one sentence. That one sentence needs the words in the question and a fact. The questioning forces students to write, and by extension me to teach, in this stilted, perfunctory manner. Whilst there are longer questions students are often not afforded the time to develop their explanations.

        The recent changes for current Y10s sitting this exam seem to be designed to increase the challenge of learning how to structure, rather the developing argument.

        As such I have this year focused only on writing extended arguments. I’ll admit to being a little worried that shorter exam-style answers might then trip some up. However, the evidence so far suggests a growing ability to handle argument and rhetoric. On the other hand, this won’t necessarily be rewarded under either the current or future specifications.

  4. As a pedantic scientist: the logs are the reactants, the ash the products, and the mass of the logs is not the same as the mass of the ash (there has been liberation of carbon dioxide/monoxide that has left the system). There is an overall conservation of mass, if you take into account the mass of the gaseous products of combustion.

    That aside, I agree that thinking/speaking in the appropriate form is important when trying to write in the appropriate form. It’s a framework of understanding, thing, I think. Possibly the same effect as why skills are not particularly transferable (don’t exactly know why, but there you go). We make connections, and then re-think along the same lines. In the same way that it is really difficult to play Taboo…..

  5. Excellent post, very interesting and thought provoking.

    Just a partially-formed thought on Point 1 ‘We can only say what we can think’. I’d agree that this is the case most of the time – thought leads to speech. But I wonder whether there are also occasions where the sequencing of ‘internal thought’ and then ‘external speech’ are less clearly defined and less clearly linear. Certainly, when children (or adults) are engaged in collaborative activities, particularly where the talk is ‘exploratory’, then speech very much is thinking. We all must be familiar with the times when you’ve been discussing x or y and have said something but only in the process of speaking has the ‘thought’ become apparent (this, of course, can be revelatory or career-ending!).

    So, while I agree with Point 1, there’s something about this process that’s complicated and intriguing (and, as ever with language, we are only just beginning to unravel it’s mysteries).

    Anyway, great post, it’s got me thinking.

    • David Didau says:

      Aren’t all thoughts by their nature put into language?

      • Actually, not sure they are. Is possible to have an idea that can’t be adequately expressed, but the individual feels the meaning……Wittgensteinian….

        • David Didau says:

          I guess you can have a vague hunch or a sense of nameless dread, but are these really thoughts? I suspect that our memories only ever go back to the point at which we acquire language because until then we can only experience, rather than think.

          And didn’t Wittgenstein say, “the limits of my language are the limits of my world”?

          • Maybe he did (I have just started reading Ray Monk’s biography and haven’t reached that bit yet)…but he was wrong…I often ‘feel’ a complex truth, but struggle to find the language to express it….eventually arriving at a form of words that only captures it in a luke warm fashion; I know that the real thing is bigger…..

          • David Didau says:

            Wittgenstein wrong? Gasp! :)

            We can have a sense of something that we can’t explain, but can we be aware of that sense without thinking in words? I just don’t think we can. Any attempt to separate thought from language becomes meaningless; while they may be in same way distinct, they are inextricable.

          • Wittgenstein would be the first to encourage dissent and fresh thinking…….I think we can sometimes think in images (mental/visual) which can then be expressed in words, but sometimes not expressed….it is often these that move us emotionally….

            OK, this is getting a bit out of hand; although I think it is true, it can’t be very relevant to the everyday in school….or could it?

          • David Didau says:

            We may be able to ‘think’ in images, but the act of becoming aware of these ‘thoughts’ translates them into words, doesn’t it? Anyway, I think we agree in principle?

          • On the substantive point, that speaking in appropriate language form helps writing in that form, yes, we agree. On the tangent, I think we can think (very fast) visually, in a way that doesn’t use language…expressing it, though, is difficult…need to slow down the thought and capture it….especially when thinking about very complex situations…..? Maybe enough, for the time being.

        • I’m sure we’ve all experienced the feeling of not being able to put something into words. We know what we want to say, but don’t have the language to say it. Either because we’ve don’t know, or have forgotten, the words required to express the thought.

          I think that it’s possible to have a conceptual understanding of something, without being able to adequately describe it using either speech or prose. You might however be able to draw it.

      • No. I think most artists and musicians would fundamentally disagree with that.

        • Last comment was in reply to ‘Aren’t all thoughts put into language?’ I’d say language is the primary and most universal way of communicating thought, but there are other forms of expressing thought that don’t involve language – music, art, dance etc.

  6. As I said in tweet, I like this and it is SEN friendly. We recommend prompts a lot: verbal and visual (doesn’t always mean images). Verbalising can work wonders – both teacher’s thoughts and student’s – it allows them to reinforce information received and think out loud. For those with poor working memories, gentle prompting and questioning can help them remember and understand what they have to do.

    I have worked and interviewed hundreds of students with specific learning difficulties and the one thing which is said over and over again is ‘I know what I want to say, I just can’t put it down on paper’.

    Another problem with writing inertia though can be sequencing difficulties – this might be the type of student who has loads of ideas but essays are either rambling, constantly going off on tangents or the student becomes so overwhelmed with the amount of information she is trying to get down she becomes frustrated and either does nothing or masks problems with behaviour.

    Your prompts could be used to help with sequence if they can have them to move around possibly?

    Concept mapping (mind maps or spider diagrams) can help with this but useful on post its to move round so they can eventually order ideas into a linear format. Computer concept mapping packages such as ‘Inspiration’ are great because with a press of a button the student can switch between the map and linear – this takes away the stresses of having to rewrite the concept map.

    Sorry, didn’t mean to write this much – like I said, a great post and inclusive :-)


  7. jonpatrick says:

    Great blog and, on the whole, I agreed throughout.

    As a primary teacher, I use “Think it. Say it. Write it. Read it.” to help emerging writers; this is a common strategy.

    The reason that this is helpful is because where children have made expected progress in speaking and listening, but not in writing, their speech ability can act as a hindrance to their writing. They are able to articulate complex sentences, but lack the written grammar to put these on to paper. So if I were to say, “Tell me a sentence about the Romans” they might reply, “The Romans lived a long time ago and they were really good fighters because they had sharp swords and they worked as a team and fought together.”

    This sentence makes sense when spoken out loud, and could be written down if the child was a great writer. But their spoken proficiency frustrates their writing, as they will forget their brilliant sentence by the time they have written the second word.

    What I really want them to do is think: “The Romans were good fighters.”, say that out loud and then write that down before moving on to the next thought. But encouraging talk encourages a stream of consciousness that the pupil is unable to put down on paper.

    I suppose that this is a roundabout way of refuting your third assertion, and suggesting that an emphasis on talk may (in some cases) make writing more difficult.

    • David Didau says:

      Thanks Jon

      I take your point – bad talk will not help pupils write. Of course not. That’s why I wrote about using academic language prompts to scaffold speech so that it is then possible to write in this way. Being asked to rephrase your thoughts is hard. It takes thought. So in doing this you will inevitably have to slow down and that may be one of the reasons why this seems to work so well.

      Cheers, David

      • Dan Ingman says:

        Interesting points, David.

        Well done to Robin Alexander for pointing out the primacy of confidence, clarity, interaction etc. in classroom oracy (in English at least) over the quality of talk. This often means that a meek student with the ability to say wise, intelligent, original things (and to let their words stand for themselves) is trumped by an overbearing presence, waving their hands about like Tony Blair without saying much. Speaking and Listening Criteria (or should I say the pressing need to have to teach to success criteria) hasn’t helped, specifying as a requirement a ‘range of non-verbal strategies’ or whatever it was (how quickly I forget!) Of course someone who is saying something intelligent may be expressing it in a verbally/expressively brilliant way (i.e. it’s easy to set up a false dichotomy here) but give me reasoned wisdom and complexity over hubris and hand gestures any day if it’s a choice between the two! Anyway, I’m straying from the point of your blog. What I mean to say is that the Robin Alexander study was excellent and long overdue.

        In terms of saying/writing: They’re obviously different things in themselves. David Crystal points out that speech is ‘time-bound, transient, dynamic’ (a phrase that’s always stayed with me from ‘Revisiting Grammar’ (?)…the book on every English undergraduate’s reading list) and in this sense, writing has its advantages. While writing we can ‘edit’ our thoughts more easily, rejecting words and sentence structures for better, clearer, more precise ones; experienced and skilled writers continually do this, often without thinking on too-conscious a level about it, almost as if a ghost has taken over their hand (brain?)My school has a whole-school literacy target at the moment of having students ‘speak in full sentences’ (often taken falsely as having students give answers that are more extensive than one word) in the hope that this will transmit to their writing, but my feeling is that this is a chicken-and-egg matter. Isn’t the reality more holistic – that writing and talking (I haven’t even mentioned reading!) are inter-dependent? We’ve all taught students who are reluctant to speak aloud, especially in the kind of wholly-realised sentences academic writing demands, for a number of reasons, one being the unwillingness to ‘play ball’ and use the kinds of words and speech patterns their teacher does, but who nevertheless write very competently.

        I think you’re right in saying that, lexically and semantically, pupils have to be able to use words in speech, in the correct context, to be really be able to think with/understand them, or at least to have had enough exposure to others doing so. This is why it’s worth breaking down the language and grammar of any subject being taught and, yes, insisting that students use the language of your subject in your lessons so it becomes the language of the classroom, helping them to internalise the kinds of talk and writing (and indeed thought) you want them to produce and, in terms of syntax, to gain the understanding of how particular words fit into sentences.

        I do find that, at another level of syntax, particularly the level of complex sentence structures for academic writing we (as English teachers anyway) want students to achieve, that this is better achieved through practise of writing, at looking at models of good and bad writing and so on. I liked the starters for analysis in English you cited that give examples of really useful ‘analytical’ and modal verbs, which are among those I’ve collated here:


        Take one of the examples for English you give:

        This could also suggest…

        We might have students consider what exactly the pronoun ‘this’ is referring to. It might be as complex a noun phrase/nominalisation as:

        His being so reluctant to talk about his childhood experiences could also suggest…

        Or, extending it even further:

        His being so reluctant to talk about his childhood experiences, a recurring feature of his interactions with others in the early part of the novel, could also suggest…

        In these examples I’ve given (invented solely to illustrate my point – I don’t have a text/character in mind!), I’m not sure that this kind of linguistic complexity occurs very often in classroom talk – or much talk at all for that matter – for reasons to do with time and economy, but nevertheless we may encourage it in writing. I’m reminded of your earlier posts around Hallidayan models of language/grammar you’ve written about elsewhere – of seeing language in terms of relatively simple building blocks, ‘blocks’ that we can extend into complex constructs to improve students’ writing.

        Anyway, thanks for the blogs, and keep up the good work – always food for thought and I agree with much of what you say. I have pre-ordered your book (the publication date has been put back I gather?) and eagerly await it!

      • nmurphy2013 says:

        Hi David,

        my Year 6 have found speaking frames extremely useful over the last few years and are later able to use the new structures in written work. Along side we do a lot of talk about register, formal, informal language structures etc. As a school, each year group, also learns a set of stories by heart, demonstrating how to adapt the memorised structures to create new stories etc. I’ve recently tried the same with short non-fiction text types and look forward to finding out whether this acts as ‘mental speaking frame’. With a high proportion of EAL learners we feel we have to ‘bathe them in language’ if not force feed them. ‘Talk for writing’ and talking about writing has been a huge focus for us – mainly because we sympathise with your ‘simple’ view of writing. Clearly it’s not the whole story but it contains essential truths.

  8. dbarlex says:

    Revisiting the points made by chemistrypoet – visual thinking is crucial in chemistry, that’s why chemists draw particle pictures and write equations. Talking about these and adding notes enables pupils to share their developing understanding and reveal conceptual problems and develop counterintuitive understanding – of course the ash weighs less that the original wood but if you take into account the oxygen that combined with the wood AND the gaseous products of combustion then overall the total mass is conserved and amazingly the total mass of the products of combustion is greater than the mass of the starting wood.

  9. […] what pupils think, but to change it. By asking them to express their ideas in academic language. I’ve blogged about this before, but by doing this we can have a surprising impact on pupils’ ability to write in academic […]

Constructive feedback is always appreciated

%d bloggers like this: