← All lessons

Writing Craft

Find something to say

  • Generate several candidate ideas quickly instead of waiting for one.
  • Choose the smallest, most specific version of a prompt.
  • Test an idea before committing a whole piece to it.
Free sample lesson — reading only

teach

Teach: find something to say

The blank page is not a failure of imagination. It is the absence of a method.

Every writer meets the empty page. The difference between one who stalls and one who starts is not talent — it is having a way in. Idea generation is a set of moves you run deliberately, in about ninety seconds, not a mood you wait to arrive.

The waiting trap costs more marks than any other habit. A student reads the prompt, stares at it, hopes something arrives, and burns the minutes that mattered most — then writes the first idea that turns up out of sheer panic, which is almost never the best one. Inspiration is not a plan. What you need is a repeatable way to produce three or four candidates fast, because the quality of a piece comes from having something to choose between, not from the first thing that surfaced.

Go small and specific rather than big and impressive. Asked to write about courage, most students reach for war, rescue or disaster — subjects they have no material for and cannot supply one honest detail about. The writer who chooses telling a friend an unwelcome truth has something real to work with. The useful question is never 'what is the most dramatic version of this prompt?' but 'what is the smallest version of this that I have actually seen?'

The three-angle move gives you candidates on demand. For any prompt, generate one personal angle — something you have lived; one observed angle — something you watched happen to someone else; and one invented angle — something plausible you can imagine completely. Three genuinely different starting points in under two minutes. Then pick the one where you can already see a specific detail, because that detail is your evidence that you have material rather than merely a topic.

Mine the prompt's own words before inventing anything. Prompts carry more than students notice. Take 'a door you were not supposed to open': every word is a lead. Which door, and where? Whose rule was it, and why did they make it? What happens when it is broken — and what does not opening it cost? Interrogating the prompt generates material without requiring a flash of inspiration, and it guarantees that what you produce is actually on task.

Test an idea before committing a whole piece to it. A workable idea passes two checks: can you name one concrete scene you can see, and can you say what it is about in a single sentence? If you cannot picture a scene, you will pad. If you cannot say what it is about, the piece will wander and stop rather than end. Twenty seconds spent on those two checks is the cheapest insurance available — far cheaper than discovering the idea was hollow in paragraph three with no time left to start again.

Waiting trap

Staring at a prompt hoping an idea arrives, instead of generating one.

Three-angle move

Producing a personal, an observed and an invented candidate.

Prompt mining

Taking leads from the prompt's own words before inventing.

Two-check test

Can you name one scene, and say what it is about in a sentence?
  • Four genuinely different candidate ideas.
  • Each contains at least one concrete, specific detail.
  • The final choice is justified by material, not by which sounds most impressive.

show

Show: the move done well

Study one clean execution before you try it yourself.

Read the walkthrough below. It shows the decisions behind the model, not just the model, so you can reuse the move rather than copy the words.

Ninety seconds before the first sentence

Exam prompt: 'The last time you changed your mind.' Task: You have ninety seconds before you start writing. Produce candidate ideas, test them, and choose one. Do not write the piece.

Nothing in this walkthrough is a paragraph. It is the ninety seconds before the first sentence, which is the part nobody watches anybody do. The near-miss is the idea most students would take into the exam, and it is not a weak idea — it is an idea that runs out in paragraph three.

  1. Step 1 — Mining the prompt before inventing anything

    Seven words, and the first three have already ruled out the idea most people are about to have.

    'The last time you changed your mind.' Before I generate anything, I read what those words allow.

    'The last time' is the restriction. Not the best time, not the biggest — the most recent. So the impressive answer, the one where I changed my mind about something enormous, is already off the table unless it happens to have been yesterday. The prompt has ruled out the dramatic version before I have thought of it.

    'Changed your mind' tells me where the interest is. It is not in the opinion I hold now. It is in the moving — which means whatever I choose has to have a before, an after, and something visible in between.

    And ninety seconds is not a hint that I should hurry. It is the claim that this is a procedure rather than a mood. Ninety seconds buys four candidates and two tests, and it is far less time than I would lose staring at the page hoping the good idea turns up by itself.

    • The last timeThree words doing the job of a rule. They ask for the most recent occasion, not the most impressive one, so the enormous answer is out before you have thought of it. Read a prompt for its restrictions first — that is the part a marker checks you noticed.
    • changed your mindThe interest is in the moving, not in the opinion. A piece about what you think now has no shape at all; a piece about the moment it shifted has a before, an after and something visible between them. Ask what a prompt's main verb requires you to show, then check your idea contains it.
    • ninety secondsNot an instruction to rush — an instruction to have a method. A minute and a half is enough for four candidates and two checks, and it is a fraction of what you lose sitting still waiting for the good idea. The students who stall and the students who start are not separated by talent.

  2. Step 2 — Four candidates, then two questions that kill two of them

    Personal, observed, invented, mined. Then: can I see one scene, and can I say what it is about?

    Personal: the Tuesday I told Mrs Doyle I was leaving choir, and changed my mind in the ten seconds between saying it and her writing it down.

    Observed: my dad changed his mind about the neighbour's dog, over about four months, without ever saying so.

    Invented: a boy standing outside a birthday party he had already decided not to go to.

    Mined from the prompt's own words: the actual last time, which was yesterday, in the canteen queue, about a sandwich.

    Now the two checks. Can I name one scene I can see? The dad idea fails immediately — four months is not a scene, it is a period, and I would end up describing a change instead of showing one. Can I say what it is about in a single sentence? The birthday party fails that one; I can picture the boy perfectly and I have no idea what the piece would be for.

    The sandwich survives both checks and is thin. The choir idea survives both and has a person holding a pen in it, so that is the one.

    • over about four months, without ever saying soA true observation that fails the first check. Four months is a period, not a scene, so the writing would have to summarise the change rather than show it — and a summary of a change is exactly what makes a piece read as a report. If you cannot see one afternoon of it, put it down.
    • a boy standing outside a birthday partyPasses the scene check and fails the other one: the picture is sharp and I cannot say what the piece would be about. An idea with an image and no point looks strong for two paragraphs and then wanders, because nothing is telling you which detail to choose next.
    • the ten seconds between saying it and her writing it downChosen because the entire change fits inside one visible moment. Going small is not settling for less here — it is the requirement, since the prompt asked for a time. A change of mind you can watch beats a change of mind you would have to explain.

  3. Step 3 — The near-miss: a real idea that cannot be written in forty minutes

    Personal, honest and genuinely moving. It fails both checks, and it fails the prompt's first three words.

    Here is the idea most students would take in with them:

    'I used to think my grandad was boring, and then he died, and I realised he had a whole life I never asked about.'

    Nobody should sneer at this. It is personal rather than borrowed, it is true for a great many people, and there is a real change of mind inside it. Pieces that start here have been written well.

    It fails the scene check. The change happened across months or years, so there is no single moment to stand in — and the death is a second, much larger event that the piece would then have to carry as well. Forty minutes will not hold both.

    It fails the prompt too. 'The last time' asked for an occasion, and a slow realisation is not a time. The piece would be genuinely on task in its final two sentences and approximately on task before that.

    And there is a third problem, the quietest one. The idea arrives with its conclusion already attached: he had a whole life I never asked about. If the moral is available before the writing starts, the writing has nowhere to travel.

    • I used to think my grandad was boring, and then he diedA real idea, not a straw one, which is why it is worth taking apart. The change took years, so there is no moment to stand in — and the death is a second event far bigger than the one the prompt asked about. Two large events and forty minutes do not fit together.
    • he had a whole life I never asked aboutThe conclusion is already written, before a word of the piece exists. When you can state the point in advance, the writing becomes a walk towards a sentence the reader could have supplied themselves. Choose ideas whose meaning you expect to find, not ideas whose meaning you have already found.
    • a slow realisation is not a timeThe prompt asked for the last time, and this is a period of gradual change with no date on it. A piece built here is on task in its closing lines and approximately on task before them. Check your idea against the prompt's restricting words before you check it against your feelings about it.

  4. Step 4 — The chosen idea, tested (Selective standard)

    Not a paragraph — four lines of plan, each of which the two checks have already approved.

    Ninety seconds produces this, and it is enough to start writing:

    • Angle: personal.
    • Occasion: the Tuesday I told Mrs Doyle I was leaving choir.
    • Scene I can see: standing next to the piano while she found my name on the list and drew a line through it.
    • What it is about: finding out you wanted something at the exact moment you gave it away.

    The scene line is specific in the way the check demands. Not 'the choir room' — a position, an action, and one object. A place is not a scene; a place with somebody doing something in it is.

    The last line is not the topic. The topic is choir, and the topic would have produced a piece about singing. What it is about is a different question, and the answer is what will tell me which details to keep when I am three paragraphs in and running out of time.

    • standing next to the piano while she found my name on the listThis is the scene check passed properly. 'The choir room' would have been a place; this is a position, an action and an object, which is a moment. If your answer to 'what scene can you see' is a location, you have not passed the check — you have named the set.
    • finding out you wanted something at the exact moment you gave it awayThe second check, answered in one sentence, and notice it is not the subject. The subject is choir. This is the point — and it is the thing that decides, later, which details earn a place and which ones you cut when the clock runs down.

  5. Step 5 — Aim higher: the same idea, put through one more question

    No new idea, no new scene — the 'about' line is rewritten and one question is added, and both cost about ten seconds.

    The same ninety seconds, run once more with one extra question:

    • Angle: personal.
    • Occasion: the Tuesday I told Mrs Doyle I was leaving choir.
    • Scene I can see: standing next to the piano while she found my name on the list and drew a line through it.
    • What it is about: the difference between wanting something and being able to say so in front of somebody who is already holding a pen.
    • What is still there at the end: the list, going back into the folder with the line already on it.

    The idea has not changed. Same angle, same Tuesday, same piano, same pen. Two lines are different.

    The first change is the 'about' line. Step 4's version — finding out you wanted something at the moment you gave it away — is a single realisation, and a single realisation gives a piece one beat and then nothing to do. This version has two sides to it: wanting the thing, and not being able to say so out loud. A piece can argue with itself for four paragraphs on that, and arguing with itself is what keeps a reader.

    The second change is an extra question, asked once the two checks are passed: what is in this scene that can still be there at the end? The answer was sitting in the scene already — a list with a line through a name — and it was never invented, only noticed. So the ending exists before the opening does.

    Selective standard produces an idea that will survive the whole piece. Scholarship standard produces one that already knows how it finishes.

    • the difference between wanting something and being able to say soStep 4's line named one realisation, which gives a piece a single beat and then leaves it standing there. This names a tension with two sides, so the writing has something to keep turning over. When you write your 'about' line, check whether it contains a conflict or only a discovery.
    • somebody who is already holding a penThe pen was in the scene from the start and had not been made to mean anything. Here it becomes the pressure: the decision is being recorded while it is still being felt. Look through the objects your scene came with before you invent anything — one of them is usually the reason the moment is difficult.
    • the list, going back into the folder with the line already on itThe ending, found before the beginning is written, and found by noticing rather than inventing. The list was already in the scene. Ask what object in your chosen moment will still exist when the piece stops — ten seconds of thinking, and you never have to hunt for a last line under time pressure.

compare

Compare: weak vs strong

Same material, same task. Only the move changes.

Both attempts respond to the same material. Read them side by side and name exactly what lifts the stronger one.

Find something to say

Weaker

For the prompt 'A risk worth taking', I will write about a soldier who saves his whole platoon during a battle. It will be very dramatic and exciting and it will show real bravery.

Stronger

For 'A risk worth taking': the week I told my swimming coach I wanted to quit — knowing she would argue, knowing I might not have the words. Scene: standing by the pool with my bag already packed. About: saying the true thing when silence is easier.

The weak version reaches for the biggest possible event and then describes it in adjectives — dramatic, exciting, brave — because there is no material underneath it; the writer has never seen a battle and cannot supply one honest detail. The strong version takes something small and lived, and already contains both a scene and a point, which is exactly why it will still be working in paragraph four.

Biggest event vs smallest true one

Weaker

For 'A time you were brave' I will write about rescuing someone from a burning building and being called a hero by everybody who was standing outside watching.

Stronger

For 'A time you were brave': the parents' evening where I had to tell my mother what my maths teacher was about to say, before she heard it from him.

The weak idea reaches for the largest event the writer can imagine and the one they have never been anywhere near, so the piece will have to run on borrowed drama rather than on detail. The strong idea is small, entirely plausible, and already contains what bravery actually requires — doing the harder thing when an easier one was available — in a scene the writer can picture completely.

Topic vs scene

Weaker

I am going to write about friendship and what it really means to be a good friend to somebody when things are difficult.

Stronger

The afternoon I waited forty minutes outside the sports hall for someone who had already gone home, and what I said the next day when she asked why I looked annoyed.

A topic names the subject and gives you nowhere to stand: 'friendship' offers no first sentence, which is why pieces that start there drift into generalities. A scene gives you a place, a duration and a conversation you now have to write, and the meaning tends to emerge from the writing itself. If your idea contains nothing you can see, it is still a topic.

guide

Guide: try the move

A quick warm-up before the writing task.

This is a short warm-up. Don't worry about getting it perfect—you'll use the same technique in the practice below.

For the prompt 'A promise that was hard to keep', generate THREE angles — one personal, one observed, one invented — in a single line each. Do not write the story.

Want feedback on your own answer? Get started to practise with instant marking.

  • Three candidates that are genuinely different, not rephrased.
  • Each names something concrete rather than a topic.