Having considered how retrieval of knowledge that has been encoded to memory plays contributes to the teaching of reading, it seems logical to explore the role of knowledge and memory in the teaching of writing. In this post, I refer to stages of the Talk for Writing process, which my school has developed extensively over the last few years. I have not included the whole teaching sequence which makes up the Talk for Writing process, but you can find more information here. To focus my thinking, I also at times had to refer to specific examples. For this I have selected a unit of work on narrative, in particular work on warning tales.
Assessing Prior Knowledge / Skills
Before the unit of work, children would be set a short writing task so that the teacher could find out what it is that the children know and don’t yet know, and what they can do and can’t yet do. There’s a balance to be struck here: don’t give the children enough of a prompt and what they write could well be of a lower standard than they’re capable of. Similarly, give them too much of a prompt and what they write may not be a true reflection of their current stage of development. Any information garnered might well be pretty inaccurate, providing a false starting point. So, in an attempt to get that balance right, I’d suggest clarifying the intended effect of the writing. For example, the whole point of the writing is to make the reader think that something terrible will happen. Then, using image and discussion around what children already know, give them enough to write about. Doing this well before lesson one of the unit gives the teacher time to analyse the writing and plan accordingly, and it provides a bit of spacing which is known to be a desirable condition for learning.
Providing a Model Text
Once the teacher knows the needs of the children, a main text for the unit of work is selected. Most likely, this will need to be edited by the teacher to ensure children are shown excellence in the areas of writing that they need to develop. For example, the teacher may have noticed that children were not using complex sentences in their writing, or that they used them grammatically inaccurately. The teacher would edit in these features.
Children will memorise this text, so that they can retell it by heart. One reason for this is that effective writers have a wealth of language patterns to daw upon when choosing how to encode their thoughts on paper. Children who are read to regularly, read lots themselves and have quality conversations with adults will develop these language patterns. Unfortunately, there are a lot of children who don’t experience these things. As well as providing these experiences at school, memorising a model text can help children to encode the language patterns into memory. This is why it is important to edit in language patterns that you know children haven’t fully grasped. They immediately begin work on correcting misconceptions as they repeatedly talk an accurate text. Another reason for committing a text to memory is to aid the writing process (described in more detail later). One way of helping children to memorise a text is through repeated, structured retelling. To start with, the teacher talks the text. At each retelling, the children join in more and more and the teacher gradually says less and less until children can retell the story on their own. This is supported with a text map like the one below.
This is one paragraph from the warning tale text. Each sentence is in a box to help children who still haven’t mastered sentence demarcation, and each paragraph is on a separate sheet of flip chart paper for a similar reason. The text map plays the role of the retrieval cue – it’s used to prompt the retelling of the text. Its role as a retrieval cue also becomes important when children come to write their own versions of the text type (more on that later).
It goes without saying that some children will memorise texts quicker and more effectively than others. With this in mind, something I’m trialling from September is to make short videos which children can use to practise retelling, either at home or at school, like this:
So over the course of a week or two, children will have practised retelling the text many times, using the text map as a retrieval cue, to the point where they can retell it by heart. A quick point here about accuracy – I don’t think that they need to retell it word for word. There are many ways to put a message across and as long as it makes sense and is consistent with the original intended effect of the writing, I think that’s fine.
Reading as a reader
Children will have managed to commit the text to memory alongside various other activities and alongside reading more examples of the focus text type. I have already written about the teaching of reading here. The brevity of this paragraph does not reflect the importance of the content – the link between good reading and effective writing is obvious. Retelling from memory and reading comprehension, in my experience, are complementary.
Reading as a writer
This is essentially analysing writing to figure out what it is that the writer did in order to achieve the intended effect on the reader. These two posts go into more detail. Firstly, the teacher would develop a writer’s toolkit with the children. In the past, I have shown children how to analyse a text, collate the ideas and display the toolkit on the working wall. I’d refer to it when the children write and encourage them to do so to help them know what they could do to achieve the intended effect.
But I think I missed a trick here. The ideas gathered in a writers’ toolkit are what effective writers have stored away in memory. If children could encode these ideas into their long term memory, then when a suitable retrieval cue is presented they would practise retrieving this essential writers’ knowledge. This would be great if they could learn to recall it at the point of writing, so the retrieval cue that I use with children needs to be something that is natural in the writing process. For example, the teacher could model the writer’s thought process: “How could I make the reader think that something terrible will happen?” If at this point, children have practised recalling aspects of the writers’ toolkit, as well as some language patterns for specific ways of doing it, they would have taken a huge leap towards producing effective writing.
The question is, then, how do we get children to encode the information in a writers’ toolkit so that it can be accurately recalled at the point of writing? What follows are some ideas that I haven’t tried with my class but will do so in September.
Taking for granted the modelling explained in this previous post, I’ll focus more on what I’ll get children to do. One option is to get children to reconstruct the toolkit information in a different way. For example, if I model the thinking about generating a toolkit by annotating a text, I could get them to reconstruct that information in a thinking map. I’d remove the annotated model that I showed them, give them the plain texts and get them to practise retrieving the information. That still won’t be enough for them to commit the toolkit to memory though. They’d need repeated, spaced practice for that. So I’d leave it a day or two, then drop in some multiple choice questions. I’d think carefully about the question because it has to be a good retrieval cue and one that I’ll use with children when they come to write. Something like:
To make the reader think that something terrible will happen, you could:
A. Make the setting dark
B. Make the main character do something relaxing
C. Reveal only part of the threat
D. Make the setting deserted
There is some playing around to do with the content of the options. It has been reported that ‘negative suggestion effect’ can be avoided through the use of explicit feedback. I think I’d also follow this retrieval practice with further study of the text type, perhaps by looking at a different text and finding either further ways to achieve the intended effect, or different language patterns for parts of the toolkit that children have already seen.
It may also be worth doing something similar with the language patterns for certain parts of the toolkit. For example:
Ways of making the setting sound deserted:
A. The place was empty.
B. Not a soul could be seen or heard.
C. Darkness hung in the corridor.
One way of providing feedback would be confirming options A and B as ways of making the setting sound deserted because it is clear that there is nobody else around, but that option C was more about making the setting sound dark.
Alternatively to the multiple choice model of retrieval practice, low stakes testing seems to be effective. A simple: “Write down as many ideas as you can… To make the reader think that something bad will happen, you could…” would work. Children could be encouraged to reconstruct the information in another different format, this time, say, a list.
Now, if children are going to encode a toolkit to memory and deliberately practise retrieving it, spaced over time, there are implications. Previously, units of work would follow this pattern over a few weeks:
Assess prior knowledge
Work on vocabulary and required knowledge
Learning the text to retell by heart
Reading other texts
Revision of previous text types using the content of the current text
Reading as a writer (developing toolkits, planning an adaptation of the main text to write independently)
Writing own version of the main text
Writing text type in a different context
Following this model, there is very little time between generating the toolkit and children writing. Certainly not enough time for encoding to memory and some spaced, deliberate practice of retrieving the toolkit. So, one solution is to develop the toolkit cumulatively over the course of the imitation stage rather than have the work massed at the end of the stage.
Innovation and Invention Stages
When the teacher is confident that children know enough to write well, we move into the stage where children will write an adapted version of the text that they have memorised. This is where the text map’s role as a retrieval cue becomes important. If the children can look at the text map and recall the model text, it can easily be tweaked to support them to write an adaptation. For example, if the children have learned the warning tale, which is set at a canal, the setting could be changed to, say, an abandoned warehouse. Part of the planning process at the end of the imitation stage would be to amend the text map to support this. Every time there is a reference to the canal, I’d cover it with a post it note and draw a quick symbol for its warehouse equivalent that children will need when they write. When children see the text map at the point of writing, they’ll be recalling the general pattern of the text, the key language patterns, and also changing it to a different context.
When I model the writing, I’ll also be referring to the writers’ toolkit that we developed and the accompanying possibilities for writing a particular idea. This bit is very important, because the next time the children write, I’d want to remove a layer of structure to teach independence – I’d remove the text map so that the children are left with the writers’ toolkit as well as any images or saved ideas that relate to the further change of context.
When the unit of work comes to a natural end, I’d plan in some revision sessions as part of the next units of work which give children more, yet further spaced practice of recalling both the text (Remember we learned this warning tale? Let’s see if we can still tell the story…) and the writers’ toolkit (Remember we learned how to make the reader think that something terrible will happen? Work with your partner to write down as many ways as you can that could have this effect…). But they would also be given chances to apply this. For example, at some point within the warning tale unit of work, we could revise some persuasive writing. After practising recalling the persuasion toolkit, children could then write in role – What would you say to the main character to persuade him to heed his mother’s warning?
Memorisation and retrieval practice could be effective in children’s development of writing, particularly when it comes to learning texts and internalising what it is that writers could do to achieve a certain effect. After all, knowing a range of texts and language patterns, plus having a secure idea of what could be written are two domains of knowledge that effective writers benefit from having.