Chapter 30 · Grammar Academy

Putting It All Together: Plan, Draft, Revise & Proofread a Whole Piece of Writing

A whole piece needs one governing idea that every paragraph serves, and four honest stages: plan, draft, revise and proofread. This guide covers planning for a reader and purpose, a beginning, a middle and an end, staying on one topic, brainstorming and outlines, introductions, body paragraphs and conclusions, the thesis statement, outlines for different genres, linking paragraphs, revising for voice, rhythm, punctuation and length, expository and persuasive essays, formal letters and reports, managing time and exam writing, peer review, proofreading a full draft, and judging when a piece is truly finished, with clear examples, for learners aged 5 to 19.

▶ Open the interactive putting it all together lesson

For beginners · ages 5–8

Think first, then write

Before the first sentence, a writer decides two things: who the piece is for, and what it is about. Those two answers are the plan.

Think first: For and About.

For: my grandmother.
About: the baby elephant at the zoo.

Now every sentence has one job: tell Grandma about the baby elephant.
The ideaDoes it fit "About: the baby elephant"?Keep it?
It was only a little taller than me.Yes - it tells how big the elephant wasKeep
It sprayed water over its back.Yes - it tells what the elephant didKeep
My new shoes are blue.No - it is about shoesSave it for another day
We ate rice and dal for lunch.No - it is about lunchSave it for another day
Its mother stood close beside it.Yes - it is still about the baby elephantKeep

Thinking first is the first stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. Every piece you make in this chapter, from a four-line story to a full essay, starts right here, with For and About.

The two answers do different jobs. The About line decides which ideas belong. The For line decides how you say them.

ForAboutA first sentence that fits
my grandmotherthe baby elephant at the zooGrandma, I saw a baby elephant today, and I thought of you.
my classthe baby elephant at the zooOur zoo has a new baby elephant, and we went to see it.
my little brotherthe baby elephant at the zooGuess what? A baby elephant can use its trunk like a hose!
my classhow our class grew beansOur class grew beans in paper cups by the window.

Notice three things. First, the same About line can open in three different ways. Grandma gets a warm sentence that talks to her. The class gets news. A little brother gets a surprise.

Second, when the About line changes, every idea changes with it. A piece about growing beans needs seeds, water and sunlight. It does not need an elephant, however much you liked the elephant.

Third, the About line should be small enough to finish. "Animals" is far too big for four sentences: you could write a whole book and still not finish. "The baby elephant at the zoo" is just right. A small topic is what keeps a short piece whole.

Picture packing a bag for a day out. You do not throw in everything in your room. First you ask: where am I going, and who am I going with? A day at the beach needs a towel and a hat, not a woolly scarf. Your About line is the place you are going. Each idea is something you might pack. Pack only what that day needs, and the bag closes easily.

This section gives you the first step of a whole piece: For and About. The next section gives the piece its shape: a beginning, a middle and an end. Section 3 checks that every part stays on the About line, and section 4 reads the whole piece aloud. By section 9 you will have a short checklist that tells you when a piece is finished. In Practice, the plan grows from two short lines into an outline with one line for every paragraph.

Here is a hard one, and it is fine to find it hard. Sometimes an idea seems to be off the topic but really belongs. Look at "Its mother stood close beside it." That sentence is about the mother, not the baby. Should it go? Ask one question: does it help the reader see the baby elephant? Yes. It shows how small the baby is and how it is cared for. So it stays. Here is a second hard one. The same idea can fit one plan and not another. "We ate rice and dal for lunch" does not fit About: the baby elephant. But it fits About: my day at the zoo very well, because lunch was part of the day. So an idea is never on or off the topic by itself. The plan decides. And a third. Sometimes you have two good topics and cannot choose. Ask which one you can say more about for this reader. If Grandma loves animals, the baby elephant wins. If Grandma has never seen a zoo at all, a piece about the whole day might suit her better. The plan is not a list of good ideas. It is a choice, made for one reader, about one thing.

Did you know that the word topic once meant a place? It comes from an old Greek word, topos, which means "a place". Long ago, teachers of speaking thought of good ideas as places in the mind where a speaker could go to find something to say. So your About line is the place your piece lives. Every sentence should stay in that place, the way a picnic stays in one garden. And here is a second surprise. Builders do the very same thing as writers. Before anyone lays a single brick, someone draws a plan of the whole house: where the doors go, where the kitchen goes, how big each room will be. Nobody starts a house by putting one brick down and seeing what happens. A writer's For and About lines are a tiny plan of the house. They take one minute, and they save you from building a kitchen in the middle of the garden.

A plan can change. If you find a better idea while you are writing, you may change your About line. But then check every sentence again against the new line. For and About do not need to be full sentences. A few words are enough: "For: my class. About: our bean plants." Not every piece has a reader you can name. The reader of a diary is you, and that is a real reader too. Thinking first takes about a minute for a short piece, not an hour. It is a small step that saves a lot of crossing out later.

  1. Stop for one minute before you write the first sentence.
  2. Write "For:" and name your reader in a few words.
  3. Write "About:" and name your one topic in a few words.
  4. Choose a topic small enough to finish in the space you have.
  5. Hold each idea next to your About line before you use it.
  6. Keep the ideas that fit and save the rest for another day.
  7. Think about what your reader already knows and what they need to be told.
  8. Choose a first sentence that suits your reader.
  9. If you change your About line, check every sentence again.
  10. Remember that true is not the same as on the topic.
Think first, then write. Before the first sentence, decide two things: 1. For: who is this piece for? 2. About: what is it about? Say each one in a few words. "For: my grandmother. About: the baby elephant at the zoo." The About line decides which ideas belong. "It sprayed water over its back" belongs. "My new shoes are blue" does not, even though it is true. The For line decides how you say it. Grandma, your class and your little brother each need a different first sentence. Choose a small topic. "Animals" is too big for a short piece. "Our cat's first bath" is just right. Thinking first does not slow you down. It tells every sentence where to go. It is the Plan stage of the Whole-Piece Check. Next: a beginning, a middle and an end.

A beginning, a middle and an end

Every whole piece has three parts: a beginning, a middle and an end. Each part has its own job.

Three parts, three jobs.

Beginning: start it. Who, where, and what it is about.
Middle: tell the most. What happens, in order.
End: finish it. How it turned out, or how it felt.
The partIts jobIn Aman's sack race piece
Beginningsays what the piece is aboutOn sports day I ran in the sack race.
Middletells what happened, in orderI hopped as fast as I could, and a boy fell in front of me. I jumped round him and kept going.
Endtells how it turned outI came second, and I got a silver ribbon.
No endleaves the reader waitingI hopped and hopped. A boy fell in front of me.

A beginning, a middle and an end is the shape you plan in the first stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. When you plan, you plan all three parts, not just the first sentence.

The three parts look a little different in a story and in a description, but every piece has them.

A story about losing a toothA description of my grandmother's kitchen
BeginningMy tooth had been wobbly for a whole week.My grandmother's kitchen is small and always warm.
MiddleAt dinner I bit into an apple, and the tooth came out in it. I held it up for everyone to see.Steel pots hang above the stove. Jars of pickle line the shelf, and the air smells of cumin.
EndThat night I put my tooth under my pillow and smiled at the gap.It is my favourite room in the whole house.

Notice three things. First, the beginning tells the reader what the piece is about straight away. My tooth had been wobbly says this is a tooth story. My grandmother's kitchen says this is about one room.

Second, the middle is the biggest part. It holds the most sentences, because it does the most telling. In a story, it tells what happened, in order. In a description, it shows the details, one after another.

Third, the end answers the beginning. The tooth story began with a wobbly tooth and ends with a gap. The kitchen piece began with a small, warm kitchen and ends with why it matters. A good ending feels as if it belongs to the same piece as the beginning.

Picture a sandwich. The top slice of bread is the beginning: it tells you a sandwich is coming. The filling is the middle: it is the biggest part, and it is the reason you are eating. The bottom slice is the end: it holds everything together so nothing falls out. Take away the bottom slice, and the filling drops into your lap. A piece with no ending is a sandwich with no bottom slice.

Section 1 gave you For and About. This section gives your piece its shape: a beginning, a middle and an end. Section 3 checks that all three parts stay on the same topic, and section 8 checks that the beginning and the end still match each other. In Practice, each part grows: the beginning becomes an introduction paragraph, the middle becomes several body paragraphs, and the end becomes a conclusion paragraph.

Here is a hard one, and it is fine to find it hard. How long should each part be? There is no fixed number of sentences. But the middle is usually the longest part, because it does the most work. A beginning is often one sentence. An end is often one sentence. The middle might be two, three or more. If your beginning is longer than your middle, the reader waits a long time for anything to happen. Here is a second hard one. Sometimes a story seems to end, but the problem is not solved. "The cat was stuck in the tree. We called and called. Then it got dark." Is that an ending? It sounds like one, because it is the last thing that happened. But the reader still wants to know: did the cat come down? A real ending answers the question the beginning asked. And a third. An ending does not have to be happy. "We never found the kite, but Papa says we will make a new one on Sunday" is a good ending. It tells the reader how things turned out. An ending only has to finish the piece, not fix everything in it.

Did you know that people have been telling stories with a beginning, a middle and an end for more than two thousand years? Long ago in ancient Greece, a teacher named Aristotle wrote about what makes a good play. He said that a whole story has a beginning, a middle and an end. The beginning does not need anything to come before it. The end does not need anything to come after it. The middle sits between them. That idea is still used today by people who write films, cartoons and picture books. And here is a second surprise. Many old folk tales open and close with the same kind of words, so listeners always know where they are. In English, many begin "Once upon a time" and many end "and they lived happily ever after". Those words are like the front door and the back door of a story. But remember: the magic words are not the ending. The ending is what happens to the characters.

The three parts do not need to be the same length. The middle is usually the biggest part. A beginning does not have to start with Once upon a time. It only has to tell the reader what the piece is about. An ending does not have to be happy. It has to tell how things turned out, or how the writer felt. The words The End can still go at the bottom of a story if you like them. They are just not the ending. The ending is the sentence before them.

  1. Draw three boxes before you write: Beginning, Middle and End.
  2. Put a few words in every box before you start.
  3. Make the beginning say what the piece is about.
  4. Make the middle the biggest part, because it does the most telling.
  5. Put the middle in order: first, next, then.
  6. Make the end tell how things turned out.
  7. Check that the end answers the question the beginning asked.
  8. Do not let The End stand in for an ending.
  9. Do not stop just because you reach the bottom of the page.
  10. Read all three parts together before you say the piece is done.
Every whole piece has three parts. 1. Beginning: it starts the piece and says what it is about. 2. Middle: it tells the most, in order. It is usually the biggest part. 3. End: it finishes the piece and says how things turned out. "I hopped and hopped. A boy fell in front of me." That piece stops in the middle. The reader is left waiting. "I came second, and I got a silver ribbon." That is an ending. It tells how the race turned out. The End is a label, not an ending. Reaching the bottom of the page is not an ending either. Plan all three parts before you write. Draw three boxes and fill each one with a few words. Next: staying on one topic the whole way through.

Staying on one topic the whole way through

A whole piece stays on one topic from the first sentence to the last. The beginning, the middle and the end are all about the same thing.

Check every sentence against the About line.

Not against the sentence just before it. Against the About line at the top.

About: my kite.
Still about the kite? Keep it.
Drifted to aeroplanes or Dubai? Cut back to the kite.
The sentenceDoes it follow the one before?Is it still about my kite?
I have a red kite with a long tail.It is the first sentenceYes
On Sunday I flew it high in the sky.YesYes
The sky was full of aeroplanes.Yes - the skyNo - it is about aeroplanes
My uncle flew in an aeroplane to Dubai.Yes - an aeroplaneNo - it is about the uncle
In Dubai there is a very tall building.Yes - DubaiNo - it is about a building

Staying on one topic is part of the Revise stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. When you read your piece again to make it better, the first thing to check is that every part is still about the one thing it set out to be about.

A piece can wander in three places. Each one needs a different look.

Where it wandersWhat it looks likeAbout: our class trip to the lake
In the middleone sentence slips in from another topicWe fed the ducks. My cousin has a new phone. We rowed a boat.
By drifteach sentence leads a little further awayWe rowed to the island. The island had a temple. My grandmother goes to a temple every week.
At the endthe last sentence is about something newWe fed the ducks and rowed a boat. Tomorrow is my birthday.
Nowhereevery part stays on the lake tripWe fed the ducks, rowed to the island and watched the sun set over the water.

Notice three things. First, a slip in the middle is the easiest to see. One sentence does not belong, and taking it out mends the piece.

Second, drift is the hardest to see, because every step looks sensible. The temple follows from the island, and the grandmother follows from the temple. You only catch drift by looking back at the About line after every sentence.

Third, an ending on a new topic is a special kind of wandering. Tomorrow is my birthday may be exciting news, but it opens a new piece instead of finishing this one. An ending must still be about the lake trip.

Picture walking a dog on a lead. The About line is the lead. The dog can sniff left and right, but the lead keeps it on the path. Drift is like letting the lead out a little more with every step: the dog follows one smell, then the next, then the next, and soon it is three gardens away. To stay on the path, keep the lead short and look back at the About line after every sentence.

Section 1 gave you an About line, and section 2 gave your piece a beginning, a middle and an end. This section checks that all three parts stay on that About line. Section 4 adds a second check: reading the whole piece aloud. In Practice, the same check runs across several paragraphs at once, and in Craft it runs against a thesis statement, the one sentence that a whole essay must serve.

Here is a hard one, and it is fine to find it hard. How far can a sentence go before it has drifted? Look at this piece. About: our trip to the lake. "We rowed to the island. On the island there was a small temple. We rang its bell and the sound rolled across the water." The temple is not the lake. Has it drifted? Ask the question: is it still part of the lake trip? Yes. The writer visited the temple on the trip, and the bell rang across the water. It belongs. Now look at the next sentence someone might add: "My grandmother goes to a temple every week." This one has drifted. It is not part of the trip at all. It only shares a word, temple, with the sentence before. So the test is not whether a sentence mentions the topic word. The test is whether it is part of the topic. Here is a second hard one. Sometimes a piece needs a short step outside the topic to explain something. "We rowed slowly, because Papa had hurt his arm the week before." The hurt arm happened another week. But it explains why they rowed slowly on the lake, so it serves the topic. A step outside is fine if it comes straight back.

Did you know that newspaper writers have a simple way to keep a story on one topic? Before a reporter writes, an editor often asks: what is this story about, in one sentence? If the reporter cannot say it, the story is not ready. Then, after it is written, the editor reads each paragraph and asks whether it still serves that one sentence. Anything that does not is cut, however interesting it is. That is the same check you are learning here, with a grown-up name: staying on the story. And here is a second surprise. The word digress means to wander away from the main topic. It comes from a Latin word meaning to step aside. People have been noticing drift for a very long time, long enough to give it its own word. Writers still use it today: "Let me not digress" means "let me stay on the topic".

A sentence that names something else can still belong, if it helps the reader see the topic. Its mother stood close beside it belongs in a piece about the baby elephant. A short step outside the topic is fine if it explains something about the topic and comes straight back. Do not fix drift by changing your About line to match it. Fix it by cutting back to the topic. A piece about a whole day can include many things, because the day is the topic. Check against the About line you actually wrote.

  1. Keep your About line where you can see it while you write.
  2. After each sentence, look back at the About line, not just at the sentence before.
  3. Ask whether each sentence is part of the topic, not whether it shares a word with it.
  4. Check the beginning, the middle and the end, all three.
  5. Mark the first sentence that drifts, because the ones after it may have drifted too.
  6. Cut back to the topic instead of changing the About line to fit the drift.
  7. Take out a stray sentence that slipped into the middle.
  8. Make sure the ending is still about the same topic as the beginning.
  9. Keep a sentence that names something else if it helps the reader see the topic.
  10. Save drifted sentences you like for another piece.
Stay on one topic the whole way through. Check every sentence against the About line, not just against the sentence before it. A piece can wander in three places: 1. In the middle: one stray sentence slips in. Take it out. 2. By drift: each sentence leads a little further away. Kite, sky, aeroplanes, Dubai. Cut back to the first sentence that drifted. 3. At the end: the last sentence starts a new topic. Write an ending about the same topic. The test is not whether a sentence shares a word with the topic. It is whether it is part of the topic. Do not change the About line to fit the drift. Bring the piece back to the About line. Next: reading it aloud at the end.

Reading it aloud at the end

At the end of writing, read the whole piece aloud, straight through. Your ears catch what your eyes skip.

Read it aloud, from the first word to the last.

Point to each word as you say it.
Tick the margin wherever you stumble or stop.
Fix each tick.
Read it aloud again.
What you hearWhere Sara heard itThe fix
a word said twiceI won a goldfish at the the ring game.take one the out
a missing wordThen we home.Then we went home.
a sentence that does not make senseMy goldfish is very loud.My goldfish is very orange.
a jump - a step is missingI won a goldfish. My goldfish is called Sunny.I won a goldfish. We carried it home in a bag of water. My goldfish is called Sunny.
no bumpOn Friday our school had a fair.nothing to fix

Reading aloud is part of the Proofread stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. It comes at the very end, when the whole piece is written, and it checks the whole piece, not just one sentence. Chapter 24, Sentence Rhythm, uses reading aloud to tune how a piece sounds; here it checks that the piece makes sense.

When you read aloud, notice the kind of stop you make. Each kind tells you something different.

The stopWhat it meansWhat to do
You trip over a worda word is doubled, or the wrong word is theretake it out, or change it
You pause, puzzled, and look backa word is missing, or the sentence does not make senseadd the word, or say the sentence a new way
You think, wait, how did we get here?a step is missing between two partsadd the missing step
You frown at a sentenceit does not belong to the topictake it out (section 3)
You reach the end and wait for morethe piece stopped instead of endingwrite the ending (section 2)

Notice three things. First, every stop is useful. A stumble is not a failure; it is the piece telling you where it needs help.

Second, not every stop is a small slip. Some stops are about the whole piece: a missing step, a stray sentence, a missing ending. Reading aloud from start to end is the only way to hear those, because you can only hear a jump when you read both sides of it.

Third, reading aloud only works if you read what is on the page. If you say what you meant to write, your voice will fix the slip without you noticing. Point to each word, and say exactly that word.

Picture walking along a path in the dark with a torch. Silent reading is like running along the path because you know it well: you never notice the stone in the way. Reading aloud is like walking slowly, shining the torch on every step. You find the stone, the hole and the place where the path is missing a plank. Then you can mend each one and walk the whole path again.

Sections 1 to 3 gave you a plan, a shape and one topic. This section gives you a check for the very end: read the whole piece aloud. Section 9 puts it on a short checklist for deciding whether a piece is finished. In Practice you will read aloud pieces of several paragraphs, and in Master it becomes part of proofreading a full draft with the Four-Pass Check from Chapter 23. Chapter 24 uses reading aloud for a different job: tuning rhythm.

Here is a hard one, and it is fine to find it hard. When you read your own writing aloud, your voice wants to help you. It knows what you meant, so it quietly adds missing words and skips doubled ones. That is why some writers ask someone else to read their piece aloud to them. A friend does not know what you meant, so they read exactly what is on the page, bumps and all. Here is a second hard one. Some stops are not mistakes. You might pause at a full stop, or at a long word, or because you are reading something exciting. Those pauses are fine. The stops that matter are the ones where you feel puzzled, trip, or have to read a sentence twice to understand it. And a third. A piece can sound smooth aloud and still have a problem. If every sentence is clear but the ending is missing, reading aloud will not trip you; it will just leave you waiting. So when you reach the last word, ask one more question: did that feel like the end?

Did you know that, for a very long time, people read aloud far more than they read silently? In many old libraries and schools, reading meant reading out loud. Silent reading was once so unusual that people noticed and wrote about it when they saw someone doing it. Today most of us read silently, but writers still go back to the old way when they want to check their work. And here is a second surprise about your own eyes. Read this line quickly: "Paris in the the spring." Many readers do not notice the second the the first time, because the eye jumps ahead to the words it expects. Your voice cannot jump ahead in the same way. It has to say each word in turn, so it bumps into the extra the. That is why reading aloud works. It makes you meet every single word.

Reading aloud catches a lot, but not everything. A word spelled wrong can still sound right, so spelling needs its own check later. Not every pause is a problem. Pauses at full stops and in exciting moments are fine. If you cannot read aloud, in a quiet library or an exam hall, read slowly in your head and move your lips silently, pointing to each word. It works almost as well. This section uses reading aloud to check that a piece makes sense. Chapter 24 uses it for rhythm, which is a different job.

  1. Wait until the whole piece is written before you read it aloud.
  2. Read in a clear voice, at the speed you would read to a friend.
  3. Point to each word with your finger as you say it.
  4. Say exactly what is on the page, not what you meant to write.
  5. Tick the margin wherever you stumble, stop or feel puzzled.
  6. Do not stop to fix things on the first reading; read to the end first.
  7. Go back to each tick and fix it.
  8. Listen for jumps where a step is missing between two parts.
  9. When you reach the last word, ask whether it felt like the end.
  10. Read the whole piece aloud once more after you fix it.
Read it aloud at the end. When the whole piece is written, read it aloud, straight through, from the first word to the last. 1. Point to each word as you say it. 2. Tick the margin wherever you stumble or stop. 3. Read to the end before you fix anything. 4. Fix each tick. 5. Read the whole piece aloud again. Your ears catch what your eyes skip: a doubled word, a missing word, a sentence that does not make sense, a jump where a step is missing, a sentence that does not belong, a piece that stops. Say what is on the page, not what you meant. Your voice wants to help you, so make it read exactly. When you reach the last word, ask one more question: did that feel like the end? Next: finishing a story.

Finishing a story

A complete story has someone it is about, a place, a problem or a wish, what happens, and an ending that answers the problem.

A story plan in three boxes.

Beginning: who, and where?
Middle: what is the problem or wish? What happens, in order?
End: how does it turn out? Does it answer the problem?
The boxWhat goes in itIn Meena's monkey story
Beginningwho and whereGrandfather feeds the birds on our roof every morning.
Middle - the problemwhat goes wrong, or what someone wantsA monkey grabbed the whole bag of grain.
Middle - what happensthe steps, in orderGrandfather put out peanuts on the other side of the roof. The monkey dropped the bag and ran to them.
Endhow it turns outGrandfather picked up the grain and smiled.

Planning a story in three boxes is the Plan stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. You then draft it box by box, check it stays on one topic, and read it aloud at the end, as sections 3 and 4 showed you.

Most short stories turn on a problem or a wish. The ending is the answer to it. When you know the problem, you know what the ending must answer.

The problem or wishA middle stepAn ending that answers it
Our kite got stuck in the neem tree.We tried to shake the branch, but it would not fall.Didi climbed the wall and freed it with a broom.
I wanted to win the sack race.I practised hopping every evening in the garden.On sports day I came first and won a gold ribbon.
My little brother was afraid of the dark.Grandma told him the stars are lamps for the night.Now he goes to sleep looking out of the window.
The power went out on Diwali night.We lit every diya we had.The street glowed more brightly than ever.

Notice three things. First, every ending answers its own problem. The kite story ends with the kite, not with lunch. The dark story ends with the brother asleep, not with a new toy.

Second, an ending does not have to be a win. I wanted to win the sack race could end with "I came last, but I hopped the whole way without falling." That still answers the wish: it tells how it turned out.

Third, the middle is where the story happens. One middle step is the least a story needs; two or three steps make it richer. But every step must lead towards the ending, not away from it.

Picture a ball thrown up into the air. The beginning is your hand holding the ball. The problem is the throw: the ball goes up, and everyone watches. The middle is the ball in the air. The ending is the catch. A story that stops at the exciting part is a ball thrown up and never caught. Readers keep looking up, waiting. Your ending is the catch that brings it safely down.

Sections 1 to 4 gave you For and About, a beginning, a middle and an end, one topic, and reading aloud. This section puts them together into one complete story. Section 6 does the same for a description, which has no problem to answer. Section 7 gives the finished story a title, and section 9 checks that it really is finished. In Practice, a complete narrative piece grows to several paragraphs.

Here is a hard one, and it is fine to find it hard. How big should the problem be? For a short story of five or six sentences, the problem must be small enough to answer quickly. A lost pencil, a stuck kite or a wish to win one race can be answered in a few sentences. A dragon that has burned down the whole city cannot. If your problem is too big, your story will stop before it is answered, just as Kabir's did. Here is a second hard one. Sometimes a story has two problems, and only one gets answered. "Our kite got stuck, and my brother was crying. Dad got the kite down." Is the story finished? The kite problem is answered, but what about the brother? You can answer both - "My brother stopped crying and flew it himself" - or leave out the second problem, so the story has only one thing to answer. And a third. An ending can be quiet. Meena's story ends with Grandfather smiling. Nothing explodes and nobody cheers. But the problem is answered, so the story is finished. A quiet ending that answers the problem is better than a loud one that does not.

Did you know that many old folk tales from around the world follow the same simple shape? Someone wants something, or something goes wrong. They try, and something happens. Then the problem is answered. That is the shape of the tale of the thirsty crow who drops pebbles into a pot until the water rises high enough to drink. It is also the shape of the tale of the hare and the tortoise, where the slow tortoise wins the race. These tales have been told and retold for hundreds of years, and they are still told to children today. The thirsty crow story is only a few sentences long. It has a crow, a problem, what the crow does, and an ending that answers the problem. It is a complete story in a nutshell. And here is a second surprise. Storytellers in many places learned their tales by heart, not from books. A clear shape - a problem and an answer - made a story easier to remember and easier to tell again.

A story does not always need a big problem. A small wish, such as wanting to see a peacock, is enough for a short story. The ending does not have to be happy. It has to answer the problem: how did it turn out? A surprise ending is fine, but it should still answer the problem the story set up. Some stories for very young readers are just a happy day with no problem at all. For this section, give your story a problem or a wish, so the ending has something to answer.

  1. Start with For and About, so you know who the story is for and what it is about.
  2. Plan three boxes before you write the first sentence.
  3. Put who and where in the Beginning box.
  4. Put the problem or wish at the start of the Middle box.
  5. Add one to three steps in the middle, in order.
  6. Write an End box that answers the problem.
  7. Choose a problem small enough to answer in a short story.
  8. Keep every sentence on the story's one topic.
  9. Do not stop at the most exciting moment; carry on to the answer.
  10. Read the whole story aloud and check the ending answers the problem.
A complete story has five things: 1. Who it is about. 2. Where it happens. 3. A problem or a wish. 4. What happens, in order. 5. An ending that answers the problem. Plan them in three boxes: Beginning (who and where), Middle (the problem, then what happens), End (how it turns out). A story is not finished just because something exciting happened. Kabir's monkey ran onto the roof, and the reader was left waiting. A story is finished when its problem is answered. Choose a problem small enough to answer in a few sentences. Endings can be happy, sad or quiet, as long as they answer the problem. Then keep it on one topic and read it aloud at the end. Next: finishing a description.

Finishing a description

A complete description names the one thing it is about, shows its details in a clear order using the senses, and ends with a closing line about the whole.

A description plan in three boxes.

Beginning: name the thing, person or place.
Middle: three or four details, in one order - top to bottom, outside to inside, or near to far.
End: a closing line about the whole, or how it makes you feel.
The boxWhat goes in itIn Isha's garden description
Beginningname the one thingBehind our school there is a small garden.
Middle - first detailthe nearest part, with a senseNear the gate, orange marigolds grow in a long row and smell sharp and spicy.
Middle - next detailsmoving on in the same orderIn the middle stands an old neem tree, with a bench in its shade. At the far end, a rusty tap drips into a bucket.
Enda line about the wholeIt is the quietest place in the whole school.

Planning a description in three boxes is the Plan stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. A story's plan has a problem and an answer; a description's plan has an order and a closing line. Both still need all four stages.

A description needs an order, so the reader can follow your eyes. Choose the order that fits the thing.

The orderGood forAn example
Top to bottoma person, a tree, a tall buildingMy grandfather wears a white cap. Below it, his eyes crinkle when he smiles. His long kurta reaches his knees, and his sandals are worn thin.
Outside to insidea house, a box, a fruitThe mango is green and speckled outside. Inside, the flesh is bright orange and drips with juice. In the very middle sits a big, hairy seed.
Near to fara view, a garden, a beachAt my feet the sand is warm. A little way off, children build a castle. Far out, a fishing boat bobs on the waves.
No ordernothing - the reader gets lostThere is a boat. My feet are warm. Some children. The sea is blue. Sand.

Notice three things. First, each order moves one way and keeps going that way. Top to bottom never jumps back to the top. Near to far never jumps back to your feet.

Second, good details use more than one sense. Sharp and spicy uses the nose. Drips with juice uses touch and taste. Bobs on the waves uses the eyes. A description that uses only the eyes is like a photo; one that uses the other senses too is like being there.

Third, the order tells you where the middle ends. When you reach the bottom, the inside or the far edge, it is time for the closing line. That is how a description knows it is finished.

Picture a camera moving slowly across a room in one direction, from the door to the window. It stops on each thing for a moment and shows it clearly. It never jumps back. When it reaches the window, it pulls back to show the whole room at once. That last wide picture is your closing line. A description is that slow camera: one direction, clear stops, and one last look at the whole.

Section 5 finished a story, with a problem and an answer. This section finishes a description, with an order and a closing line. Together they are the two simplest shapes of a whole piece. Section 7 gives each finished piece a title that fits. In Practice, a complete descriptive piece grows to several paragraphs, each one taking a part of the order, and in Craft you will choose an outline to fit any genre.

Here is a hard one, and it is fine to find it hard. Which order should you choose? There is no single right order. But one order will usually suit the thing best. A tall thing, like a person or a tree, suits top to bottom. A thing you can open, like a fruit or a box, suits outside to inside. A wide view, like a beach or a field, suits near to far. If you are not sure, ask: how would my eyes move if I were standing there? Here is a second hard one. Can a description ever mention something happening? Yes, if it shows what the place is like. "At the far end, a rusty tap drips into a bucket" is something happening, but it is part of the picture. It becomes a story only when one event leads to another and to an ending: the tap broke, we called the plumber, the garden flooded. And a third. How long should a description be? A short one needs only three or four details. More is not always better. Four well-chosen details in one order, each with a sense, will show a place more clearly than ten details in a jumble.

Did you know that people who make maps and plans of buildings describe places in a fixed order too? A builder's plan shows a house room by room, from the front door inwards. Guides who show visitors round an old fort or palace often walk the same path every time, from the gate to the courtyard to the towers, so that every visitor sees the place in the same order. The order makes the place easy to follow and easy to remember. A written description works the same way. The order is the path. Your reader walks it with you. And here is a second surprise about the senses. Smell is one of the strongest senses for memory. Many people find that a smell, such as rain on dry earth or a spice being fried, can bring back a place all at once. That is why adding a smell to a description can make a reader feel they are really there.

There is no single right order. Choose the one that suits the thing, and then keep to it. Not every detail needs a sense word, but most good descriptions use more than just the eyes. A description can show small things happening, like a tap dripping, without becoming a story. The closing line can sum up the whole thing or say how it makes you feel. Either is a real ending. Running out of details is not.

  1. Start with For and About: one thing, person or place.
  2. Name the thing in your first line, so the reader knows what is described.
  3. Choose one order: top to bottom, outside to inside, or near to far.
  4. Keep to that order without jumping back.
  5. Pick three or four details, not every single thing you can see.
  6. Give most details a sense: sight, sound, smell, touch or taste.
  7. Stay on the one thing; do not slip into another room or another place.
  8. Do not let a description turn into a story halfway.
  9. End with a closing line about the whole, or how it makes you feel.
  10. Read it aloud and check that the reader could follow your eyes.
A complete description has four things: 1. The one thing, person or place, named in the first line. 2. An order: top to bottom, outside to inside, or near to far. 3. Three or four details that follow the order, using the senses. 4. A closing line about the whole, or how it makes you feel. Plan them in three boxes: Beginning (name it), Middle (details in order), End (the closing line). A description is not finished when you run out of things to say. Rohan's garden ended on a broom. It is finished when its closing line sums up the whole. A description has no problem to answer. If one event leads to another and to an ending, it has become a story. Next: a title that fits.

A title that fits

A title names what the whole piece is about. Choose it, or check it, after the piece is written.

A title is a promise.

It tells the reader what the whole piece will be about.

The Monkey Thief promises a monkey story.
Grandmother's Roof Garden promises what Neel's piece really gives.
The titleDoes it fit the whole of Neel's piece?Why
The Monkey ThiefNothe monkey is in only one sentence
PlantsNofar too big - it could be about any plants anywhere
My Grandmother Keeps a Small GardenNot quiteit is the first line, not a title - it is too long
Grandmother's Roof GardenYesevery sentence fits under it
Our RoofNot quiteclose, but the piece is about the garden, not the whole roof

Choosing a title is part of the Revise stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. Your About line from the Plan stage is often the best place to find it. If no title fits the whole piece, check whether the piece has drifted (section 3).

Titles go wrong in a few common ways. Each one promises the wrong thing.

What goes wrongAn exampleThe piece is really aboutA title that fits
Only one detailBlue Shoesmy first day at a new schoolMy First Day at Sunrise School
Only the exciting momentThe Big Splasha family picnic by the river, where someone fell inOur Picnic by the River
Far too bigAnimalsfeeding the ducks at the lakeFeeding the Ducks
About something elseMy Best Friendmaking a paper boat in the rainA Paper Boat in the Rain
The first line, not a titleLast Sunday We Went to the Fair and Rode the Big Wheelour day at the fairA Day at the Fair

Notice three things. First, a good title is short. A few words are enough. It does not need to be a sentence.

Second, a good title is the size of the piece. Animals is a whole library. Feeding the Ducks is one afternoon. Match the title to what is actually there.

Third, a good title often comes straight from the About line. If your plan said About: making a paper boat in the rain, then A Paper Boat in the Rain is already waiting for you. The title and the About line are two names for the same one thing.

Picture a label on a jar. A jar of mango pickle labelled Mango Pickle is helpful. A jar of mango pickle labelled Chilli, because there is one chilli in it, is confusing. A jar labelled Food is true but useless. A title is the label on your piece. Read what is inside the jar first, then write the label that tells people what they will find.

Sections 5 and 6 finished a story and a description. This section gives each finished piece a title that fits. Section 8 checks that the beginning and the end still match each other, and section 9 adds the title to a short checklist for deciding whether a piece is finished. In Craft, the same idea grows into the thesis statement: one sentence that names what a whole essay will show.

Here is a hard one, and it is fine to find it hard. Sometimes a title can be fun and still fit. The Monkey Thief does not fit Neel's garden piece. But if Neel's whole piece were about the monkey - how it came, what it took, how Grandmother chased it away - then The Monkey Thief would be a perfect title. So a title is never good or bad on its own. It is good or bad for one piece. Here is a second hard one. What if two titles both fit? Grandmother's Roof Garden and Grandmother's Tins of Tomatoes might both fit. Choose the one that covers more of the piece. The tomatoes are part of the garden, but the garden also has chillies and tulsi. So Grandmother's Roof Garden covers more. And a third. Should a title give away the ending? It is your choice. Our Lost Puppy Comes Home tells the reader the ending before they start. Our Lost Puppy keeps the surprise. Both fit the piece. Neither is wrong; they just do different jobs.

Did you know that the word title comes from an old Latin word, titulus, which meant a label or an inscription? Long ago, a titulus could be a label written on a jar, a sign on a building, or words cut into stone to say what something was. A title is still a label today: it tells the reader what they are about to find. And here is a second surprise. Many writers say they choose the title last. They write the whole piece first, then read it through and ask what it turned out to be about. Some newspapers have a special editor whose job is to write headlines - the titles of news stories - after the reporter has written the story. That way the headline matches what the story actually says.

A title is good or bad only for one piece. The Monkey Thief is wrong for a garden piece and right for a monkey piece. A title can be fun or plain, as long as it fits the whole piece. Giving away the ending in a title is a choice, not a mistake. This section checks whether a title fits. Which words in a title get capital letters belongs to Chapter 18, so it is not checked here.

  1. Choose the title after the piece is written, or check it then.
  2. Look at your About line first; the title often comes from it.
  3. Write two or three possible titles before you choose.
  4. Hold each title against the whole piece, not just one sentence.
  5. Do not name only the most exciting moment.
  6. Do not name only one small detail.
  7. Make the title the same size as the piece, not bigger.
  8. Keep it short: a few words, not a whole sentence.
  9. If two titles fit, choose the one that covers more of the piece.
  10. If no title fits the whole piece, check whether the piece has drifted.
A title is a promise. It names what the whole piece is about. Choose it, or check it, after the piece is written: 1. Look at your About line. 2. Write two or three possible titles. 3. Hold each one against the whole piece. 4. Keep the one that every part fits under. Titles go wrong when they name only one detail (Blue Shoes), only the exciting moment (The Monkey Thief), something far too big (Animals), something else (My Best Friend for a paper boat piece), or the whole first line. A good title is short and the same size as the piece. Grandmother's Roof Garden fits every sentence of Neel's piece. If no title fits, the piece may have drifted. Next: checking the beginning and the end.

Checking the beginning and the end

A beginning and an ending must belong to the same piece. Cover the middle, and read the first and last sentences side by side.

The ten-second check.

Cover the middle.
Read the first sentence, then the last.
Ask: do they belong to the same piece?
First sentenceLast sentenceDo they match?
Our class planted a neem sapling by the school gate.And that is why I love my new bicycle.No - two different pieces
Our class planted a neem sapling by the school gate.Now the sapling is taller than our teacher.Yes - the ending answers the beginning
My grandmother was afraid of the new phone.Now she sends me a photo of her garden every morning.Yes - the fear at the start is answered
Our street is noisy in the morning.The bus was late, so I missed maths.No - the ending has wandered to school

Checking the beginning and the end together is part of the Revise stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. It takes ten seconds, and it catches problems that a sentence-by-sentence check can miss.

An ending can belong to its beginning in three ways. Any one of them is enough.

The way it belongsFirst sentenceLast sentence
It answers the beginningOur kite got stuck in the neem tree.Didi climbed the wall and freed it with a broom.
It returns to the beginningEvery morning the old fisherman pushes his boat into the sea.And tomorrow morning, he will push it out again.
It closes what the beginning openedMy grandmother's kitchen is small and always warm.It is the warmest room in the whole house.
It does not belongOur kite got stuck in the neem tree.We had puris for dinner.

Notice three things. First, a story's ending usually answers its beginning: the problem at the start is solved, or at least settled, at the end.

Second, a description's ending usually closes what the beginning opened. The kitchen piece begins with a small warm kitchen and ends on the whole kitchen again.

Third, some endings return to the beginning, using the same picture or words. The fisherman pushes his boat out at the start and will push it out again at the end. That return makes the piece feel like a circle, complete and whole.

Picture the two slices of bread in a sandwich. The top slice is the beginning, the bottom slice is the end. If the top slice is brown bread and the bottom slice is a biscuit, something has gone wrong, however good each one is. The two slices must belong to the same sandwich. Lift the top and look at the bottom: do they match?

Sections 2 and 3 gave your piece a beginning, a middle and an end, all on one topic. Section 7 gave it a title. This section checks the two ends against each other in ten seconds. Section 9 puts it on a short checklist for deciding whether a piece is finished. In Practice, the same check runs between an introduction paragraph and a conclusion paragraph, and in Craft between a thesis statement and the essay's last lines.

Here is a hard one, and it is fine to find it hard. Sometimes the ending seems to be about something new, but it really does belong. Read this. First: "I was afraid of the deep end of the pool." Last: "Next summer, I want to learn to dive." Is diving a new topic? No. It shows how far the fear has gone: from afraid of the deep end to wanting to dive. The ending answers the beginning by showing the change. So the test is not whether the ending uses the same words as the beginning. The test is whether it answers, returns to or closes what the beginning opened. Here is a second hard one. When the beginning and the end do not match, which one should you change? Ask: which one is the piece really about? Look at the middle. If most of the middle is about the neem sapling, keep the sapling beginning and rewrite the ending. If most of the middle is about the bicycle, you may have found the piece you really wanted to write, and the beginning should change instead. Either way, all three parts must end up about one thing.

Did you know that many stories end in the same place they began? Storytellers sometimes call this a circle story. The story starts at home, goes out on an adventure, and comes back home again at the end, a little changed. Think of a child who leaves for school afraid in the morning and comes home happy in the evening. The ending lands where the beginning started, and that makes the piece feel complete. And here is a second surprise. Many songs do the same thing. A song often ends by repeating the line it began with, or its chorus. When the first line comes back, listeners know the song is ending. A piece of writing can do the same: when the ending echoes the beginning, the reader feels the circle close.

The ending does not need the same words as the beginning. It needs to answer, return to or close what the beginning opened. An ending that only repeats the beginning word for word adds nothing. It should show what changed, or how it turned out. The ten-second check does not replace reading the whole piece. After fixing the ends, read the middle again to check it joins them. When the two ends do not match, either one may need to change. Decide what the piece is really about first.

  1. Cover the middle when the piece is finished.
  2. Read the first sentence and the last sentence one straight after the other.
  3. Ask whether they belong to the same piece.
  4. Check that the ending answers, returns to or closes what the beginning opened.
  5. If the beginning set a problem, make sure the ending answers it.
  6. If the ending belongs to another piece, find where the middle changed direction.
  7. Decide what the piece is really about before you change either end.
  8. Rewrite the ending, or the beginning, so both serve the one topic.
  9. Read the middle again after fixing, to check it joins the two ends.
  10. Let the ending add something, not just repeat the first sentence.
Check the beginning and the end together. Cover the middle. Read the first sentence, then the last sentence. Do they belong to the same piece? An ending belongs to its beginning when it: 1. answers it (the kite is freed), 2. returns to it (the fisherman will push his boat out again), or 3. closes what it opened (the warmest room in the house). A good beginning and a good ending can still belong to two different pieces: a neem sapling and a new bicycle. To fix a mismatch, decide what the piece is really about, then rewrite the ending or the beginning. Read the middle again to check it joins them. The ending should add something, not just repeat the beginning. Next: is it finished?

Is it finished?

A piece is finished when every box on the checklist is ticked, not when you stop writing.

The Is-It-Finished checklist.

1. Beginning: does it say what the piece is about?
2. Middle: does it tell the most, on one topic?
3. End: does it answer or close the beginning?
4. Title: does it fit the whole piece?
5. Read aloud: have I read it once and fixed every bump?
The boxAditi's pieceKabir's piece
BeginningYes - the science fairYes - the robot at the science fair
MiddleToo short - one sentenceYes - the pencil, the drawing
EndNo - cut off: Then weYes - I want to build robots one day
TitleNo titleYes - The Robot That Could Draw
Read aloudNoYes
Finished?No - stoppedYes - finished

This checklist is the last question of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread. After all four, ask: is it finished? In Discover, the five ticks are how you answer. In Master you will ask the same question about a whole essay.

A piece can look finished for the wrong reasons. Each of these is a reason to stop, but not a reason to say finished.

The reason to stopWhy it does not mean finishedWhat to check instead
The bell rang.The piece does not know about the bell.Is there an ending? If not, write one quickly.
The page is full.A full page can still stop in the middle.Turn over and finish the ending.
I am tired.Tired is about the writer, not the piece.Run the checklist; fix one box at a time.
I wrote The End.The End is a label, not an ending.Does the last sentence answer the beginning?
It feels long enough.Length does not decide.Are all five boxes ticked?

Notice three things. First, none of these reasons is about the piece. Bells, pages and tiredness are about the writer's day. Only the checklist asks about the piece itself.

Second, when time is short, the checklist tells you what matters most. If you have one minute left, write the ending first. An ending matters more than a title, and a title matters more than a perfect word.

Third, finished is a real answer. When all five boxes are ticked, say finished and mean it. Adding more after that can make a good piece worse, not better.

Picture a pilot before a plane takes off. The pilot does not just say, I think it is ready. The pilot reads a checklist, one item at a time: fuel, doors, lights, wings. Only when every item is ticked does the plane take off. Your five boxes are a writer's checklist. Tick each one honestly, and when all five are ticked, your piece is ready to go.

Sections 1 to 8 gave you every part of a whole piece: For and About, a beginning, a middle and an end, one topic, reading aloud, finishing a story and a description, a title, and matching the two ends. This section ties them into one checklist. Section 10 practises everything in Discover together. In Practice the checklist grows to cover several paragraphs, and in Master you will judge for yourself when a demanding piece is finished, or merely stopped.

Here is a hard one, and it is fine to find it hard. What if one box cannot be ticked because you have run out of time? Then fix the most important box first. An ending is the most important, because without it the piece has stopped. Then the read-aloud, because it catches slips. Then the title. A piece with an ending and no title is closer to finished than a piece with a title and no ending. Here is a second hard one. How do you know a box is really true? Ask the question exactly. Not, is there a last sentence? but, does the last sentence answer or close the beginning? Not, is there a title? but, does the title fit the whole piece? A box with something in it is not the same as a box that is ticked. And a third. Sometimes a piece is finished, but you still want to change it. Maybe you think of a better word. That is fine, if it makes the piece better and keeps every box ticked. But if the change adds a new topic or a second ending, it unticks a box. Finished means the piece works, not that it can never be touched.

Did you know that pilots, surgeons and builders all use checklists, even after many years of doing their jobs? It is not because they forget how to do their work. It is because, when people are busy or in a hurry, small steps are easy to skip. A short list of questions, read one at a time, catches the step that was skipped. Many people who study safety say that a simple checklist is one of the most powerful tools there is. A writer's checklist does the same job. You know how to write a beginning and an ending. The checklist makes sure you did not forget one when the bell was about to ring. And here is a second surprise. The word finish comes from an old French word, finir, which means to end, and that comes from the Latin finis, meaning a boundary or limit. To finish something is to bring it right up to its edge, not to leave it in the middle.

A ticked box must be true, not just filled. A title that does not fit is not a ticked Title box. When time is short, the ending comes first, then the read-aloud, then the title. Finished does not mean the piece can never change. A better word is fine if every box stays ticked. Finished is a real answer. When all five boxes are ticked, saying so is right, not lazy.

  1. Do not say finished until you have run the checklist.
  2. Draw five boxes in the margin: Beginning, Middle, End, Title, Read aloud.
  3. Check one box at a time, in order.
  4. Tick a box only when the question is really true, not just when something is written.
  5. Make sure the ending answers or closes the beginning.
  6. Make sure the title fits the whole piece, not one part of it.
  7. Read the whole piece aloud and fix every bump before ticking the last box.
  8. If time is short, write the ending first, then read it aloud, then add the title.
  9. Do not let the bell, a full page or tiredness decide for you.
  10. When all five boxes are ticked, say finished and stop adding.
Is it finished? Run the five-tick checklist. 1. Beginning: does it say what the piece is about? 2. Middle: does it tell the most, on one topic? 3. End: does it answer or close the beginning? 4. Title: does it fit the whole piece? 5. Read aloud: have I read it once and fixed every bump? A piece is finished when all five are ticked. It is not finished just because the bell rang, the page is full, you are tired, or you wrote The End. A tick means the box is true, not just that something is there. If time is short, the ending comes first, then the read-aloud, then the title. When all five are ticked, say finished, and stop adding. Next: the first mixed review.

First mixed review

Discover gave you eight steps. Run them in order, on the whole piece, and the piece will tell you when it is finished.

The Discover route, in order.

1. For and About.
2. Beginning, middle, end.
3. One topic all the way through.
4. Read it aloud.
5. A story ends by answering its problem; a description ends by closing on the whole.
6. A title that fits.
7. First and last sentences match.
8. Five ticks: Beginning, Middle, End, Title, Read aloud.
The problem in Ishita's pieceThe step that finds itThe fix
My cousin has a new phone.Step 3: one topictake it out
Then weStep 5: a story ends by answering its problemWe found her eating cane and carried her home.
Title: The GoatStep 6: a title that fitsThe Evening Chameli Ran Away
First: would not come home. Last: (cut off)Step 7: first and last matchthe new ending answers the beginning
All fixed?Steps 4 and 8: read aloud, five ticksFinished

These eight steps are the Discover form of the Whole-Piece Check, this chapter's four stages: Plan (step 1), Draft (steps 2 and 5), Revise (steps 3, 6 and 7) and Proofread (steps 4 and 8). In Practice, each step grows to cover pieces of several paragraphs.

Every step finds a different kind of problem. If you know which step finds which problem, you can fix a piece quickly.

The problemThe step that finds itThe section
Ideas from everywhere, no topicFor and About1
A piece that stops in the middleBeginning, middle and end2
A sentence that drifts or slips inOne topic all the way through3
A doubled or missing word, a jumpReading aloud4
A story that stops at the exciting partFinishing a story5
A description that ends on a broomFinishing a description6
A title that names one detailA title that fits7
An ending from a different pieceChecking the beginning and the end8
Saying finished when the bell ringsThe five-tick checklist9

Notice three things. First, the steps go from big to small. For and About decides the whole piece; reading aloud catches a single word. Fixing the big things first saves time, because a small fix in a sentence you later cut is wasted.

Second, some problems can be found by two steps. A stray sentence about a phone is caught by step 3, but it can also make you frown when you read aloud. That is good: the steps back each other up.

Third, the five-tick checklist comes last because it checks that every other step was done. It is the final yes or no.

Picture getting ready for school in the morning. You do the big things first: get up, wash, dress. Then the smaller ones: pack your bag, tie your laces. Last of all, you look in the mirror and check everything once. Writing a piece is the same. Plan first, then shape it, then check the topic, the ending and the title, and last of all read it aloud and look in the mirror of the five boxes.

This section brings all of Discover together: For and About, a beginning, a middle and an end, one topic, reading aloud, finishing a story and a description, a title, matching ends and the five-tick checklist. Practice begins with brainstorming: choosing what to write from many ideas, before the plan. Then the plan grows into an outline, and each part of a piece grows into a whole paragraph.

Here is a hard one, and it is fine to find it hard. Sometimes fixing one step breaks another. Ishita took out the cousin's phone, so step 3 is fixed. But now her middle is shorter: does it still tell the most? Then she wrote a new ending, so step 5 is fixed. But does the new ending still match the title? Each fix can change something else. That is why the steps end with reading aloud and the five-tick checklist. They check the whole piece again, after all the fixes. If a fix broke something, the last check will find it. Here is a second hard one. What if you do not have time for every step? Then use the order. The plan and the ending matter most, because without them the piece wanders or stops. Then the topic check. Then reading aloud. The title can be added in the last few seconds. A piece with a plan, one topic and an ending is nearly finished. A piece with a perfect title and no ending is not. And a third. Different pieces need different checks at step 5. A story needs its problem answered. A description needs a closing line about the whole. Know which kind of piece you wrote before you check its ending.

Did you know that the word review comes from an old French word meaning to see again? Re means again, and view means to see. When you review a piece, you look at it again, with fresh eyes, one step at a time. That is exactly what this section asks you to do. And here is a second surprise. Many writers say they write a piece in one go, then put it away for a little while before they review it. Coming back later, even after a short break, lets them see it more like a new reader would. Their eyes no longer fill in what they meant to say. You may not always have time to put a piece away for a day. But even standing up, stretching, and coming back to it after one minute can help you see it again.

The steps go in an order, but you can always go back. If reading aloud shows a stray sentence, return to step 3. Not every piece has a problem at every step. A step that finds nothing has still done its job. When time is short, keep the plan, one topic and the ending above everything else. A story and a description end differently. Check which kind of piece it is before step 5.

  1. Run every step, not only the one you remember best.
  2. Start with the plan: For and About.
  3. Mark the beginning, the middle and the end before checking anything else.
  4. Check every sentence against the About line.
  5. Check the ending: a story answers its problem, a description closes on the whole.
  6. Fix big problems before small ones.
  7. Check that the title fits the whole piece.
  8. Cover the middle and read the first and last sentences together.
  9. Read the whole piece aloud after all the fixes.
  10. Finish with the five boxes, and when they are all ticked, say finished.
Discover gave you eight steps. Run them in order on the whole piece: 1. For and About. 2. Beginning, middle, end. 3. One topic all the way through. 4. Read it aloud. 5. The ending: a story answers its problem; a description closes on the whole. 6. A title that fits. 7. First and last sentences match. 8. Five ticks. Each step finds a different problem. Name the step when you find one. Fix big things before small things, because a small fix in a sentence you later cut is wasted. A piece is not finished because each part looks good. It is finished when every step has been run on the whole piece, and all five boxes are ticked. Next: Practice, brainstorming ideas.

Building accuracy · ages 8–12

Brainstorming ideas

A brainstorm has two halves, and they must happen in order. First you collect: write down every idea you can, fast, without judging any of them. Then you choose: look at everything on the page and pick the one topic that suits the task.

Collect first, choose second.

Collect: Eid morning, the school Christmas fair, kites at Makar Sankranti, the river mela, Diwali at Nani's house, my cousin's lantern.

Choose: Which fits the task and the reader? Which do I have the most to say about? Which is small enough to finish in three paragraphs?

Chosen: About: Eid morning at our house.
Brainstormed topicFits the task and the reader?Enough to say?Small enough?Choose it?
Festivals of IndiaYesFar too muchNo - it needs a whole bookNo
My cousin's lanternYesOnly one paragraph's worthYesNo - too thin for three paragraphs
My new cricket batNo - the task asks for a festivalYesYesNo - save it for another piece
Kites at Makar SankrantiYesYesYesA strong second choice
Eid morning at our houseYesYes - getting ready, prayers, foodYes - one morningYes

Brainstorming is the first move of the Plan stage of the Whole-Piece Check, this chapter's four stages: Plan, Draft, Revise and Proofread. Before you can decide For and About, you need something to choose from, and a brainstorm gives you that.

There is more than one way to collect ideas. Each one suits a different kind of thinker, and you can use them together.

Way to brainstormHow it worksWhat it looks like for "Eid morning at our house"
A listWrite ideas one under another, as fast as they comenew clothes, attar, the walk to the park, prayers, hugging everyone, sheer khurma, Eidi from my uncles
An idea webPut the topic in a circle in the middle; draw a line out for each new idea; ideas that belong together sit on the same branchMiddle: Eid morning. Branch 1: getting ready. Branch 2: the prayers. Branch 3: the food and the visitors.
The six question wordsAsk who, what, where, when, why and how about the topic, and answer each oneWho? my whole family. Where? the park near our flat. When? just after sunrise. Why? to celebrate the end of Ramzan.

Notice three things. First, the same brainstorm can happen twice. The first time you brainstorm topics and choose one. The second time you brainstorm details about the topic you chose. Farhan's first list was festivals; his second list was everything about Eid morning.

Second, the idea web does something a list cannot. It puts ideas that belong together on the same branch, and each branch can become one body paragraph. Getting ready, the prayers and the food are already three paragraphs waiting to be written. That is why the next section, the simple outline, grows straight out of a good brainstorm.

Third, the six question words work best when you are stuck. If your list stops at three ideas, ask "Why?" and "How?" and new ideas usually appear. A question is easier to answer than a blank page.

Picture a fruit seller at the market. Before he fills his basket, he looks along every crate: mangoes, bananas, guavas, oranges, a few bruised apples. He does not grab the first fruit he touches. He looks at everything, then chooses the ripest ones for the customer in front of him. Brainstorming is walking past every crate. Choosing is filling the basket. The fruit left in the crates is not wasted; it waits for another customer, another day.

In Discover you wrote For and About before your first sentence. Brainstorming now comes before those two lines: you collect many possible topics, choose one, and then write For and About for it. The second brainstorm gives you the details, grouped into branches. The next section turns those branches into a simple outline, one line for every paragraph, in order. Later in Practice, each line becomes an introduction, a body paragraph or a conclusion. In Craft, the chosen topic grows into a thesis statement.

This part is harder, so take it slowly. Sometimes the idea you love most is not the idea you have the most to say about. Nisha loves the Durga Puja pandal near her school, but when she tries the six question words, she can only answer two of them. She knows much more about the day her family made modaks for Ganesh Chaturthi: who helped, how the dough was shaped, why her grandmother pressed each fold with her thumb, what went wrong with the first batch. The pandal is the topic she likes; the modaks are the topic she can write. For a piece of three paragraphs, the topic with more to say usually wins. The pandal can wait for a short poem. Here is a second hard case. Sometimes the task already gives you the topic: "Write about your school library." Is there anything left to brainstorm? Yes. You still brainstorm, but now you brainstorm details and angles. The library on a rainy lunchtime, the librarian's rule about bookmarks, the shelf of comics everyone races to, how the room smells of old paper. A given topic only takes away the first brainstorm, never the second. And a third. Two ideas on your list may both fit, both be the right size, and both have enough to say. Then let the reader decide. A class magazine read by children your age might enjoy the kites at Makar Sankranti, with string races and rooftop shouting. A letter to a grandparent who lives far away might be better about Eid morning at home, because it shares the family with someone who missed it. Choosing is not about finding the one perfect idea. It is about finding the best idea for this task, this reader and this space.

Did you know that the word brainstorm did not always mean a good thing? When the word first appeared, more than a hundred years ago, a brainstorm meant a sudden fit of confusion, as if a storm had broken out inside someone's head. Later, people began to use it for the opposite: a sudden, bright idea. Later still, in the middle of the twentieth century, people working in advertising turned brainstorming into a way for a group to collect ideas together. They had one main rule, and it is the same rule this section gives you: while ideas are being collected, nobody is allowed to judge them. The judging comes afterwards. Here is a second surprise. The word idea comes from an old Greek word linked to seeing. An idea is something you can see in your mind. That is why an idea web, where you can see all your ideas on one page at once, works so well. And many writers keep a notebook for ideas they are not using yet. When an idea does not fit today's piece, it goes in the notebook instead of the bin. The ideas you save from a brainstorm can start a piece you have not thought of yet.

A brainstorm does not need neat handwriting, full sentences or correct spelling. It is a rough page only you will read, so a word or two for each idea is enough. You do not always need three ways to brainstorm. One quick list is often enough for a short piece. Use the idea web or the question words when the list runs dry or the piece is longer. In a timed test, the brainstorm must be short: a minute or two, not ten. A small brainstorm still beats none, but it must not eat the time you need to write. Sometimes the first idea really is the best one. Brainstorming does not mean you must reject it. It means you choose it after comparing, so you know why it won.

  1. Read the task carefully before you start: what kind of piece, for whom, and how long.
  2. Set a short time limit for collecting, such as two or three minutes.
  3. Write every idea down, fast, in a word or two.
  4. Do not cross anything out while you are still collecting.
  5. Aim for at least six ideas before you stop.
  6. If you get stuck, ask who, what, where, when, why and how.
  7. Choose by asking: does it fit the task and the reader, do I have enough to say, and is it small enough to finish?
  8. Brainstorm a second time on the topic you chose, to collect its details.
  9. Group the details that belong together; each group can become one paragraph.
  10. Save the ideas you did not use in a notebook for another piece.
Brainstorming ideas. A brainstorm has two halves, always in this order: 1. Collect: write down every idea you can, fast, without judging any of them. 2. Choose: read the task again and pick one topic. To choose, ask three questions. Does it fit the task and the reader? Do I have enough to say about it? Is it small enough to finish? "Festivals of India" is too big. "My cousin's lantern" is too thin for three paragraphs. "Eid morning at our house" fits all three questions. Three ways to collect: a list, an idea web, and the six question words. The idea web puts related ideas on one branch, and each branch can become a paragraph. Brainstorm twice: once for the topic, once for its details. Keep only the details that serve the chosen topic and save the rest. The first idea is only the first. You know it is the best only after you have compared it with others. Brainstorming is the first move of the Plan stage of the Whole-Piece Check. Next: A simple outline.

A simple outline

A simple outline lists what each paragraph will cover, in order, before you draft. One line per paragraph.

An outline, one line per paragraph.

Para 1 - open: what the piece is about.
Para 2, 3, 4 - one main idea each, in a sensible order.
Last para - close: sum up, or a final thought.

Then check: every line on the topic? No line said twice? A sensible order?
The outline lineThe question it answersIn Simran's annual day outline
Para 1What is the piece about?what annual day is, and when it was
Para 2What is the first main part?the Class 5 dance
Para 3What comes next?our play, The Clever Crow
Para 4What comes after that?the prize-giving
Para 5How does it close?why it was the best day of the year

An outline is the Plan stage of the Whole-Piece Check, this chapter's four steps: Plan, Draft, Revise and Proofread, grown to fit a bigger piece. Each line becomes one paragraph when you draft. Building that one paragraph well - its topic sentence, its details and its order - belongs to Chapter 27, Paragraph Architecture.

An outline can go wrong in three ways. Each one is much easier to fix in the outline than in the finished piece.

What goes wrongAn outline that shows itThe fix
Two lines cover the same ideaPara 2 - the dance. Para 3 - more about the dance.join them into one paragraph, or split the dance into two different parts
A line is off the topicPara 4 - my cousin's new phonetake it out; save it for another piece
The order jumps aboutPara 2 - the prize-giving. Para 3 - arriving at school. Para 4 - the dance.put the lines in the order they happened: arriving, dance, prizes
The best idea is missingno line for the play I was inadd a line for it in the right place
Nothing wrongPara 1 - what annual day is. Para 2 - the dance. Para 3 - our play. Para 4 - prizes. Para 5 - why it was the best day.no change needed - start writing

Notice three things. First, every outline line is short. It names the paragraph's one main idea in a few words. You will write the sentences later.

Second, the order depends on the piece. A day out goes in the order things happened. A description of a house might go room by room. An explanation of how to plant a seed goes step by step. Choose the order before you write the lines.

Third, an outline can change while you write. If you think of a better paragraph, add a line for it. But then read the outline again from top to bottom, and check that the new line still serves the main topic.

Picture the list of stops on a bus route, painted on a board at the bus stop. Before the bus leaves, every passenger can see where it will go, in what order, and where it will end. The driver does not decide each stop as the bus goes. An outline is that board for your piece. Each line is a stop, one paragraph long, and the reader travels them in order to the last stop.

Section 11 gave you many ideas to choose from. This section turns the chosen ideas into an outline, one line per paragraph. The next three sections write the parts it lists: section 13 opens a bigger piece, section 14 builds the body paragraphs in order, and section 15 closes the whole piece. Section 16 checks every paragraph against the main topic. In Craft, outlines change shape for different genres.

Here is a hard one, and it is fine to find it hard. How many paragraphs should an outline have? There is no fixed number, but most school pieces of several paragraphs have between three and six: an opening, two to four body paragraphs, and a closing. The body is usually where the number changes. Ask: how many main parts does my topic really have? A day out might have three main parts. A description of a house might have one paragraph for each room you describe. Here is a second hard one. Sometimes two lines seem different but are really the same idea. "Para 2 - the food at the fair" and "Para 3 - the sweets we ate" are both about eating at the fair. Join them, or make them clearly different: one about food stalls, one about the games. And a third. Sometimes a line seems off the topic but belongs. In a piece about annual day, "Para 1 - how nervous I was that morning" is not about the show itself. But it opens the day and makes the reader care about what comes next. Ask the same question as always: does it serve the main topic?

Did you know that many writers plan long pieces with a list very like an outline before they begin? Writers of long books often make a list of chapters before they write, one line for each chapter, so they can see the whole shape of the book at once. Film makers do something similar with a storyboard: a row of small pictures, one for each scene, in order. If a scene does not work, they can move or cut it on the board before anyone films it. And here is a second surprise. The word outline first meant the line drawn round the edge of a shape, like the outline of a leaf on paper. An outline of a piece of writing does the same job: it draws the edge of the whole piece before you fill it in with sentences.

An outline is a plan, not a promise. It can change while you write, as long as you check the new outline from top to bottom. Outline lines are short notes. Writing full sentences in an outline uses up the time it was meant to save. There is no fixed number of paragraphs. Three to six is common for school pieces; the topic decides. This section plans which paragraphs a piece needs. How to build each paragraph well belongs to Chapter 27.

  1. Write For and About first, then the outline beneath it.
  2. Write one short line for each paragraph, down the side of a rough page.
  3. Give each line one main idea in a few words, not a full sentence.
  4. Make the first line open the piece and the last line close it.
  5. Choose an order that fits the piece: time, place, steps or importance.
  6. Check that every line serves the main topic.
  7. Check that no two lines cover the same idea.
  8. Check that the most important idea has a line.
  9. Fix the outline before you start drafting.
  10. Check each paragraph against its outline line as you write.
Before you write a piece of several paragraphs, make a simple outline. 1. Write For and About at the top. 2. Write one short line for each paragraph, in order. 3. The first line opens the piece; the last line closes it. 4. Each body line names one main idea. Then check the outline: - Does every line serve the main topic? - Do any two lines cover the same idea? - Does the order make sense? - Is the most important idea there? Fixing an outline takes seconds. Fixing a finished piece takes much longer. Aryan rewrote two paragraphs; Simran spent two minutes on her outline. An outline can change, but then check it again from top to bottom. Next: opening a bigger piece.

Opening a bigger piece

An introduction is the first paragraph of a piece that has more than one. It does not tell the story or explain the topic yet. It opens the door and shows the reader the rooms ahead.

The Three-Job Opening.

1. Name the main topic: what is the whole piece about?
2. Hook the reader: one small question, scene or surprising detail that makes them want to read on.
3. Signal the plan: tell the reader what the body paragraphs will cover, in the same order as the outline.
The sentenceThe job it doesWhere it comes from
On the night of our annual day, the hall was so full that parents stood along the back wall.Names the main topic: our annual dayThe About line of the plan
None of them knew that the twelve dancers on the stage had nearly given up a month before.Hooks the reader: why did they nearly give up?A small surprise the body will explain
This is the story of how we practised the dance, made our own costumes and got through the big night.Signals the plan: three parts, in orderThe body lines of the outline
Rohit's opening: The steps were very hard, and Vikram kept turning the wrong way.None of the three - it is a detail of paragraph 2Keep it for the body

The introduction is written at the Draft stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread: it is the outline's first line turned into a paragraph, and before you call the piece finished you check that it still matches the body. Opening one paragraph with a topic sentence is taught in Chapter 27, Paragraph Architecture; an introduction does the same job one size up, for a whole piece.

The three jobs stay the same in every bigger piece, but they look different in different kinds of writing.

Kind of pieceNames the topicHooks the readerSignals the plan
A story: the day the river floodedThe morning the river came into our laneThe water was already at the second step when Nani woke us.We did not know yet that by evening we would be sleeping in the school hall.
An explanation: how a beehive worksA hive is a busy town with no mayor.So who decides what thousands of bees will do each day?This piece looks at the queen, the workers and the drones, one at a time.
A description: my grandfather's cloth shopDadaji's cloth shop in the old marketYou smell it before you see it: starch, dust and sandalwood.Come in through the door, past the counter, and up to the shelves at the back.
A how-to piece: a paper lantern for DiwaliMaking a paper lantern for DiwaliAll you need is paper, glue and one hour before the lamps are lit.First the frame, then the paper, and last the light inside.

Notice three things. First, a story signals its plan more quietly than an explanation. The flood story does not list its parts. It hints at where the piece is going, so the reader knows there is a journey ahead without being told every turn, and without being told the ending.

Second, the signal follows the outline's order. The beehive piece promises the queen, the workers and the drones, so the body must meet them in that order. If the body begins with the drones, the reader feels a promise has been broken.

Third, the hook is small. It is one question, one smell or one surprising picture, never the best part of the whole piece. The best part belongs to the body, where there is room to tell it properly. A hook is a taste, not the meal.

Picture the front of a house at evening. The name on the gate tells you whose house it is: that is the main topic. A warm light in the window, or a smell of cooking, makes you want to step inside: that is the hook. And through the open door you can see a hallway with three doors off it, the rooms you are about to visit, in order: that is the signal. A good introduction is a name, a lit window and a hallway. It is not the living room.

Section 12 gave you an outline: one line for every paragraph, in order. This section turns the first line of that outline into a real paragraph that opens the whole piece and promises what the rest will cover. Section 14 builds the body paragraphs the introduction has promised, in the order it promised them, and section 15 closes the piece with a conclusion that looks back at the introduction. In Craft, the sentence that names the main topic grows into a thesis statement, the one sentence that says what a whole piece will show.

Here is a hard one. Take your time with it. Should the introduction be written first or last? Many writers write it first, straight from the outline, because it reminds them what each paragraph must do. Others write it last, once they know exactly what the body says. Both are fine. What is not fine is writing it first and never checking it again. If the body changed while you drafted, the introduction's promise must change too. Here is a second hard one. How loudly should an introduction signal? In an explanation, naming the parts is a help: "This piece looks at the queen, the workers and the drones." In a story, the same move can spoil it: "This story tells how I lost my dog at the mela, searched every stall and found him asleep under the balloon seller's cart." The reader now knows the ending. A story's introduction hints instead: "At the mela, I let go of Bholu's lead for one second. That was all it took." The kind of piece decides how much the signal says. And a third. Sometimes a hook seems to be off the topic but really belongs. Meena's piece about her grandmother's pickles began: "Ammamma's hands are always yellow in April." Hands are not pickles. But the next sentence explains that April is mango season and the yellow is turmeric, and the reader is pulled straight into the main topic. A hook may start a little to one side, as long as the same paragraph walks it to the topic. If it never arrives, it has wandered.

The word introduction comes from Latin. Intro means "inside" and ducere means "to lead", so to introduce something is to lead someone inside. That is exactly what your first paragraph does: it takes the reader by the hand at the door and leads them in. The same Latin verb, ducere, is inside the word conductor, a person who leads a band or an orchestra. News writers have their own name for the opening of a news story: they call it the lead. Journalists are trained to put the most important facts in the lead, so that a reader who stops after the first few lines still knows what happened, and a reader who goes on knows what the rest of the story will explain. Books do it too. A long non-fiction book usually has a contents page and often a short opening chapter that tells you what each later chapter will cover. Your introduction is that contents page, written as a paragraph, with a hook to make the reader turn the page.

Not every bigger piece needs a signal that lists its parts. A story usually hints at where it is going rather than naming every stop, so it does not give away the ending. A plain signal, such as "In this piece I will tell you about the soil, the seeds and the watering", is not wrong at this stage. It does the job. Smoother signals come with practice. An introduction can be written last, once the body is drafted. If you write it first, check it again at the end, because the body may have changed. A piece of only one paragraph does not need a separate introduction. Its topic sentence, from Chapter 27, is enough.

  1. Write the introduction from your outline, not from memory.
  2. Name the main topic in your first sentence or two.
  3. Choose one small hook: a question, a scene or a surprising detail.
  4. Make sure the hook belongs to the main topic, or leads to it at once.
  5. Signal what the body paragraphs will cover.
  6. List the parts in the same order as the outline.
  7. In a story, hint at the road ahead instead of giving away the ending.
  8. Keep every detail for the paragraph that owns it.
  9. Fit the opening words to your reader, as the For line asks.
  10. When the piece is drafted, check that the body keeps every promise the introduction made.
Opening a bigger piece. An introduction opens a piece that has more than one paragraph. It does three jobs: 1. Name the main topic: what is the whole piece about? 2. Hook the reader: one small question, scene or surprising detail on that topic. 3. Signal the plan: tell the reader what the body paragraphs will cover, in the outline's order. The key contrasts: A door, not a room. Starting straight with the first point skips the introduction. "The steps were very hard" is paragraph 2, not an opening. A taste, not the meal. An introduction that tells every detail leaves the body nothing to say. A promise the body keeps. If the introduction names the queen, the workers and the drones, the body meets them in that order. A story hints; an explanation may name its parts. The introduction is written at the Draft stage of the Whole-Piece Check, straight from the outline, and checked against the body before the piece is finished. Next: body paragraphs, in order.

Body paragraphs, in order

A body paragraph is one of the middle paragraphs of a bigger piece. Each one has a single job: to develop one part of the plan. Together, in a sensible order, they deliver everything the introduction promised.

One part each, in order.

1. One part: each body paragraph develops one line of the outline, and only that line.
2. Named at the start: its first sentence tells the reader which part this is, and points back to the part before.
3. A sensible order: the paragraphs come in an order you can give a reason for, and the same order the introduction promised.
Nisha's outline lineHer body paragraph beginsWhy it sits here
Growing the seedlingsIn March we sowed tomato and chilli seeds in yoghurt pots on the windowsill.It comes first in time: nothing can be sold before it grows.
Making the postersWhen the first leaves opened, we started on the posters.It comes second: posters only make sense once there is something to sell.
Selling dayOn the last Friday of April, sixty pots stood ready on two tables.It comes last: the sale is where the whole story was heading.
Farhan's stray sentence: Some seedlings died on the windowsill too.It sat in his posters paragraph.It belongs in the seedlings paragraph, with the other windowsill details.

The body paragraphs are the heart of the Draft stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread: each one is written to its own outline line, in order, and linked to the one before it. How to build one strong paragraph, with its topic, development, order and boundary, is taught in Chapter 27, Paragraph Architecture; this section puts several of them in a row and makes the row make sense.

What counts as a sensible order depends on the kind of piece. The outline chooses it at the Plan stage; the body paragraphs keep to it.

Kind of pieceA sensible orderExample body order
A story or a recountTime order: what happened first, next and lastLeaving home - the long bus ride - arriving in the hills
A how-to pieceStep order: each step needs the one before itGathering the paper - folding the frame - gluing and hanging the lantern
A description of a placePlace order: near to far, outside to inside, or one room after anotherThe gate - the courtyard - the kitchen at the back
An explanation of whyOne reason at a time, each in its own paragraph, often saving the strongest for lastTrees give shade - trees clean the air - trees hold the soil when the rain comes

Notice three things. First, every order has a reason the writer can say aloud. Time order follows the clock, step order follows the work, place order follows the reader's eyes, and reason order follows how strong each reason is. If you cannot say why a paragraph comes where it does, the order may be an accident.

Second, each paragraph keeps its details at home. A detail about the windowsill belongs with the seedlings, however late you remembered it. When a detail turns up in the wrong paragraph, the reader has to carry it backwards, and the paragraph it landed in stops being about one part.

Third, the first sentence of each body paragraph tells the reader where they are now. Simple time and place words, such as "When the first leaves opened" or "At the back of the house", also join the paragraph to the one before it. Choosing the linking word that names the right relation is the Bridge Test from Chapter 21; here it is enough that each paragraph points back to its neighbour.

Picture a goods train: an engine at the front, three wagons, and a guard's van at the back. The engine is the introduction and the van is the conclusion. Each wagon in between carries one kind of load: one holds rice, one holds cloth, one holds mangoes. You would not tip mangoes into the rice wagon, and you would not couple the wagons in a muddle if the first station is waiting for the rice. Each body paragraph is one wagon, loaded with one part of the plan and coupled to the next in the order the stations come.

Section 13 opened the piece and promised what the body would cover. This section keeps that promise: one body paragraph for each part of the outline, in a sensible order, each one pointing back to the one before. Section 15 closes the whole piece with a conclusion that looks back over every body paragraph, and section 16 runs the unity check across all the paragraphs at once, making sure none has wandered off the main topic. In Craft, the order of the body is chosen to fit the genre, and the links between paragraphs are strengthened across the whole piece.

Here is a hard one. Take your time with it. Sometimes more than one order is sensible. Rhea's piece on why her school should plant more trees has three reasons: shade, clean air and soil that does not wash away in the rain. Shade first? Soil first? For a story, the clock decides. For a set of reasons, there is no clock. What matters is that Rhea can say why she chose her order (perhaps she saved the reason her readers will care about most for last) and that her introduction promises the reasons in that same order. Sensible does not always mean the only right order. It means an order with a reason behind it. Here is a second hard one. Sometimes a detail seems to fit two paragraphs. "We watered the seedlings one last time" sounds like the growing paragraph. But if it happened on the morning of the sale, just before the pots were carried out, it belongs to selling day. Ask: which part of the plan does this moment belong to? The plan, and the moment, decide where a detail lives. And a third. A story sometimes starts in the middle on purpose, then goes back. A writer might open the body with the noisy sale, then write: "Six weeks earlier, none of this existed. There was only a packet of seeds." That is not a muddle, because the jump is signalled: the reader is told the clock has turned back. Farhan's jump was not signalled, so his reader was lost. At this stage, keep to time order unless you tell the reader clearly that you are jumping.

The word paragraph comes from the Greek paragraphos, from para, "beside", and graphein, "to write". In old Greek manuscripts, a paragraphos was a short line written in the margin, beside the text, to show where a new part began. Readers could run a finger down the margin and find each new part without reading every word. Later, scribes used other marks for the same job, among them a sign now called the pilcrow, which looks like a backwards P with a double stem. Much later still, printers stopped using a mark at all and simply began each new part on a new line, often set in from the edge. The empty space at the start of a paragraph is a descendant of that old mark in the margin. And here is a second surprise. Recipes number their steps for the same reason your body paragraphs come in order: in a recipe, the order is part of the meaning. Whisking the eggs after the cake has gone into the oven is not a different recipe; it is a mistake. A how-to piece is a recipe written in paragraphs.

Not every body paragraph needs to be the same length. A part that needs more room can have more, as long as it stays on its one part. A very big part may take two paragraphs, as long as both develop the same part and sit next to each other. A story may jump in time on purpose, if the writer tells the reader clearly that the clock has turned back or forward. For a set of reasons, more than one order can be sensible. What matters is that you can say why you chose yours, and that the introduction promises the same order.

  1. Number the body lines of your outline before you draft.
  2. Write one body paragraph for each numbered line.
  3. Choose an order you can explain: time, steps, place, or one reason at a time.
  4. Begin each body paragraph with a sentence that names its part.
  5. Let that first sentence point back to the paragraph before.
  6. Keep each detail in the paragraph whose part it belongs to.
  7. Jot a late-remembered detail beside its own part, not where you happen to be.
  8. Check that no part is told twice and no promised part is missing.
  9. If a story jumps in time, tell the reader clearly that it is jumping.
  10. Read only the first sentence of each body paragraph, in order, and check it walks the road the introduction promised.
Body paragraphs, in order. The body is the middle of a bigger piece. Build it in three steps: 1. One part each: every body paragraph develops one line of the outline, and only that line. 2. Named at the start: its first sentence says which part this is and points back to the one before. 3. A sensible order: an order you can give a reason for, and the same order the introduction promised. The key contrasts: Good paragraphs are not enough. Farhan's paragraphs were each good, but the sale came before the seeds, and the reader was lost. The order fits the kind of piece: time for a story, steps for a how-to, place for a description, one reason at a time for an explanation. Each detail lives at home. "Some seedlings died on the windowsill" belongs with the seedlings, not the posters. No part twice, no promise missing. The body is written at the Draft stage of the Whole-Piece Check, one outline line at a time. Next: closing a bigger piece.

Closing a bigger piece

A conclusion is the last paragraph of a bigger piece, and its work is to close all of it, not just the paragraph before it. It has three jobs, and one thing it must never do.

Three jobs of a conclusion.

1. Return to the main topic, the one the introduction opened.
2. Sum up what the body paragraphs showed, in a few fresh words.
3. Leave one final thought.

And never open a brand-new idea. A new idea needs its own paragraph, and the piece has no room left for one.
The jobWhat it doesFrom Aditya's garden piece
Return to the main topictells the reader the piece is coming home to where it beganThat bare corner is green now.
Sum up the bodygathers each body paragraph into a few words, without retelling itWe dug it, planted it and kept it alive through the rains.
Give a final thoughtleaves the reader with what it all meansA garden, we found, needs patience as much as it needs seeds.
Never: a brand-new ideaopens a topic the piece has not talked aboutNext term our class wants to start a school band.
Never: the introduction copiedhands back the first sentence as if nothing happenedIn conclusion, this year our class turned a bare corner of the playground into a kitchen garden.

Chapter 27 showed how to build one paragraph well; this section asks what the last paragraph must do for all the others. The conclusion is the final paragraph you write in the Draft stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread, and it answers the Check's last question out loud: does the piece still serve its one main topic?

Here is one piece and five ways to end it. Meenakshi's introduction says that getting ready for Onam takes her family three busy days. Her body paragraphs tell about the shopping, cooking the sadya, and making the pookalam, the flower carpet, at the front door.

The endingWhat it doesDoes it close the whole piece?
The end.stops after the last body paragraphNo - it has stopped, not finished
In conclusion, getting ready for Onam takes my family three busy days.copies the introductionNo - the body has vanished from it
And that is how we make our pookalam.closes only the pookalam paragraphNo - the shopping and the sadya are forgotten
My cousin's wedding is next month.opens a new topicNo - it starts a different piece
The shopping, the sadya and the pookalam took us three full days, but on Onam morning the whole house was ready. Getting ready, I think, is half of the festival.returns, sums up and gives a final thoughtYes

Notice three things. First, the ending about the pookalam sounds like a conclusion. It even begins with "And that is how". But it closes only the last body paragraph. A reader who began with the shopping is left wondering what happened to it. A conclusion names every part of the body, even if it gives each part only one word.

Second, the good conclusion uses words from the introduction, "three" and "days", but it does not copy the introduction's sentence. Coming back to the topic is not the same as saying the first sentence again. It is like coming home by a different road, carrying what you picked up on the way.

Third, the final thought, "Getting ready, I think, is half of the festival", is not a new idea. It grows out of everything the piece has shown. The wedding is a new idea: nothing in the piece has led to it, and the reader would need a whole new paragraph to understand it. That is the test. A final thought comes out of the piece; a new idea comes in from outside it.

Picture a guide leading a group round a museum. At the door she tells you what you are going to see. In each room you look at one part. At the end she brings you back to the door, and in a sentence or two she tells you what all the rooms added up to. She does not stop in the last room and walk away. She does not repeat her first words as if you had seen nothing. And she does not open a new room on the way out. Your introduction is the door, your body paragraphs are the rooms, and your conclusion is the walk back to the door.

Section 13 opened a bigger piece with an introduction, and Section 14 built body paragraphs in order. This section closes the piece, so you now have every part of the shape: introduction, body and conclusion. Section 16 checks that every one of those paragraphs stays on the main topic, and Sections 17 to 19 put the whole shape to work in a story, a description and an explanation. In Craft, the main topic becomes a thesis statement, and the conclusion returns to that one sentence.

Here is a hard one, and it is worth slowing down for. When is a final thought allowed, and when is it a new idea? Aditya might end his garden piece with "Next year we want to try pumpkins as well." Is that new? It talks about next year, which the piece never mentioned. But it is still about the garden, and it grows out of what the piece showed: the class learned to grow things there. A reader needs nothing more explained. So it is a fair final thought. Now compare "Next year we want to start a school band." The band has nothing to do with the garden, and the reader would need a whole new paragraph to understand it. That is a new idea. Ask two questions: does it grow out of the piece, and could the reader understand it without another paragraph? Two yeses mean it is a final thought. Here is a second hard one. Sometimes the last body paragraph already ends with a strong sentence, and it is tempting to stop there. Kunal's puppet show paragraph ended well. But a strong last sentence closes one paragraph. Only a conclusion closes the piece, because only a conclusion looks back over every paragraph, not just the one it follows. And a third. In a short piece of three paragraphs, the jobs can share a sentence. "The shopping, the sadya and the pookalam took us three full days" returns to the topic and sums up the body at once. That is fine. What matters is that all three jobs are done, not that each one has a sentence of its own.

The word conclude comes from a Latin word, concludere, which means "to shut up" or "to close in", the way you might shut animals into a pen. Inside it is another Latin word, claudere, "to shut". The English word close comes from the same family, and so, surprisingly, does a word you met in the grammar chapters of this Academy: clause. A conclusion, then, is the paragraph that closes the gate on the whole piece, so that everything the piece gathered stays inside. Here is a second surprise. Not every kind of writing has a conclusion. Newspaper reports are often written with the most important facts first and the least important last. That way, if an editor runs short of space, the report can be cut from the bottom and still make sense. A news report can simply stop. A school composition or an essay is built the other way. Its last paragraph is one of the most important, not the least, because it is where the reader learns what all the paragraphs meant together. Cut it off, and the piece is left with its gate open.

A conclusion does not have to begin with "In conclusion". If the relation is clear, you may leave the linker out, as the Bridge Test in Chapter 21 allows. Words like "So" or "In the end" can work too. A conclusion can be short. In a three-paragraph piece, two sentences that do all three jobs are enough. A story usually ends differently: it closes its events rather than summing up. Section 17 shows how a narrative piece ends. A final thought does not have to be a lesson or a moral. A feeling, a hope, or a look ahead that stays on the topic is just as good.

  1. Reread your introduction before you write the conclusion.
  2. Come back to the main topic in fresh words, not the introduction's exact sentence.
  3. Name every body paragraph in the summing-up, even in one word each.
  4. Keep the summing-up shorter than any body paragraph.
  5. End with one final thought that grows out of the piece.
  6. Ask of any last idea: could the reader follow it without another paragraph?
  7. Cut or save a brand-new idea instead of squeezing it into the ending.
  8. Never let "The end" stand in for a conclusion.
  9. Use a linker such as "So" or "In the end" only if it names the closing relation.
  10. Read the introduction and the conclusion together: one piece, two different sentences.
Closing a bigger piece. A conclusion closes the whole piece, not just the paragraph before it. It has three jobs: 1. Return to the main topic, the one the introduction opened. 2. Sum up what the body paragraphs showed, in a few fresh words. 3. Leave one final thought that grows out of the piece. And it never opens a brand-new idea. The key contrasts: - Returning to the topic, not copying the introduction. "That bare corner is green now", not the first sentence again. - Closing the whole piece, not only the last body paragraph. The shopping, the sadya and the pookalam, not just the pookalam. - A final thought, not a new idea. Pumpkins next year grow out of the garden; a school band does not. - Summing up, not retelling. A word or two for each paragraph is enough. - Finished, not stopped. "The end" is a label, not a conclusion. The conclusion is the last paragraph of the Draft stage of the Whole-Piece Check, and it answers the Check's final question aloud. Next: keeping every paragraph on topic.

Keeping every paragraph on topic

In a bigger piece, unity means that every paragraph serves the one main topic set out in the introduction. Each paragraph may be about a different part of that topic, but none may be about a different topic.

The unity check, run across the whole piece.

1. Find the main topic in the introduction.
2. Read the first sentence, the point, of each paragraph after it.
3. Ask of each one: does this paragraph serve the main topic?
4. Fix any that do not: cut it, re-aim it, or save it for another piece.

Never move the main topic to fit a paragraph that wandered.
The paragraphIts pointDoes it serve the main topic: keep the library open at lunch?
Paragraph 2Many students have no quiet place to read at home.Yes - it is a reason to stay open
Paragraph 3 (Anika)The library was built sixty years ago.No - it is history, not a reason
Paragraph 3 (Dhruv)Lunch is the only free hour for students on the early bus.Yes - it is a reason to stay open
Paragraph 4The lunchtime reading club would have to close.Yes - it is a reason to stay open

Chapter 27 showed how to keep the sentences of one paragraph on that paragraph's topic; this section checks that the paragraphs themselves stay on the piece's main topic. The unity check is part of the Revise stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread, and it is cheapest of all at the Plan stage, when each paragraph is still one line of an outline.

A paragraph can wander off a piece's main topic in four ways. Here they are in one piece. Rehan's introduction says: "The monsoon changes everything in our village."

How it wandersWhat it looks likeRehan's paragraph point
A stray paragrapha whole paragraph from another piece slips inOur cricket team reached the district final last year.
Drifteach paragraph follows from the one before, but the piece slides awayParagraph 3 is about the river rising; Paragraph 4 is about the bridge over the river; Paragraph 5 is about how the bridge was built.
Sharing a wordthe paragraph uses the topic's word but is about something elseMonsoon is also the name of my uncle's tea shop.
Too bigthe paragraph leaves the village for the whole worldRain falls on every continent on Earth.
On topicevery paragraph serves the village in the monsoonThe fields turn green, the lanes flood, and the school timetable changes.

Notice three things. First, a stray paragraph is the easiest to see, because it has nothing at all to do with the main topic. The fix is usually to cut it and save it.

Second, drift between paragraphs is the hardest to see, because each paragraph follows sensibly from the one before it. The river leads to the bridge, and the bridge leads to how it was built. You catch drift only by holding every paragraph against the introduction, never just against its neighbour.

Third, a paragraph that shares a word, or one that grows too big, can look on topic at a glance. Monsoon is right there in the uncle's tea shop paragraph, and rain is right there in the paragraph about every continent. Ask what the paragraph is about, not which words it uses. The main topic is not the monsoon or rain everywhere; it is what the monsoon does to this village.

Picture a train on one track. The introduction lays the track and says where it goes. Each paragraph is a carriage. Carriages can carry different things, passengers in one, letters in another, luggage in a third, but they must all be coupled to the same train on the same track. A stray paragraph is a carriage from another train that has been coupled on by mistake. Drift is a set of points that slowly switches the train onto a branch line. Before the train leaves, walk its whole length and check every carriage is on the same track.

In Discover, Section 3 checked that every sentence of a short piece stayed on one topic. Chapter 27 checked one paragraph's own sentences. This section lifts the same check up a level, to whole paragraphs inside one bigger piece. Sections 13 to 15 built the introduction, the body and the conclusion; this section makes sure they all serve the same main topic. In Craft, the main topic becomes a thesis statement, and Section 23 links the paragraphs that serve it so the reader feels them as one road.

Here is a hard one, and it is worth slowing down for. Sometimes a paragraph seems to be off topic but really serves it. Imagine Anika rewrote her history paragraph so that it began: "For sixty years, students have read here at lunch; closing the doors now would end something the whole school has shared." The history is still there, but now it does a job for the main topic: it gives a reason to stay open. That is re-aiming. A paragraph belongs not because of what it mentions, but because of what it does for the main topic. Here is a second hard one. What if more than one paragraph wanders, in the same direction? Suppose three of Varun's four paragraphs turn out to be about his hobbies, and only one is about the park. Then the honest question is not "which paragraph is stray?" but "which piece did I mean to write?" That is a Plan decision. Choose the main topic on purpose, write it into the introduction, and then check every paragraph against it again. What you must not do is quietly widen the main topic until everything fits. "My colony and my hobbies" is not a main topic; it is two. And a third. A paragraph can serve the main topic in its first sentence and still wander inside, halfway through. That is a problem for one paragraph, and Chapter 27 owns it: run the Paragraph Check on that paragraph. This section's question is the bigger one: is each paragraph, as a whole, serving the whole piece?

The word unity comes from the Latin unus, which means "one". The same word sits inside unit, unite, union and universe, which once meant everything turned into one. A unified piece is, quite literally, a piece that has been made into one. Here is a second surprise. Newspaper and magazine editors often have a short test for a story, especially a long one. They ask the writer to say, in a sentence, what the story is about. Then they read the first sentence of each paragraph and ask whether it serves that one sentence. A paragraph that does not is usually cut, however hard the writer worked on it, or saved for a different story. That is the very check this section teaches: one line for the whole piece, then one look at each paragraph's point. A paragraph cut this way can often become the start of a writer's next piece. Nothing good is wasted; it is only moved to the piece where it belongs.

A paragraph does not need to repeat the main topic's words to serve it. A paragraph about the early bus serves "keep the library open at lunch" without using the word library at all. The first sentence is a quick guide to a paragraph's point, but not a perfect one. If a paragraph's first sentence is vague, read the whole paragraph before you judge it. Cutting is not the only fix. Many wandering paragraphs can be re-aimed so that they serve the main topic. A story's paragraphs serve the main topic differently: each one moves the same story forward. Section 17 shows how unity works in a narrative piece.

  1. Underline the main topic in your introduction before you check anything else.
  2. Read only the first sentence of each paragraph and say what that paragraph is for.
  3. Hold every paragraph against the main topic, not against the paragraph before it.
  4. Ask what a paragraph does for the main topic, not which words it shares with it.
  5. Watch for a paragraph that grows from your topic to the whole world.
  6. Question your favourite paragraph as hard as the others.
  7. Re-aim a wandering paragraph if it can serve the main topic; cut it if it cannot.
  8. Save cut paragraphs for another piece, so nothing good is wasted.
  9. Never widen the main topic just to fit a stray paragraph.
  10. Run the check on your outline first, when a stray paragraph is only one line to cross out.
Keeping every paragraph on topic. In a bigger piece, every paragraph serves the one main topic in the introduction. The unity check, across the whole piece: 1. Find the main topic in the introduction. 2. Read the first sentence, the point, of each paragraph. 3. Ask of each one: does it serve the main topic? 4. Fix any that do not: cut, re-aim or save. The key contrasts: - A good paragraph, not always an on-topic one. Anika's history paragraph was well built and still served no part of the main topic. - Checked against the introduction, not against the paragraph before. Drift hides when you only compare neighbours. - What it is about, not which words it uses. A tea shop called Monsoon is not the monsoon in the village. - Fix the paragraph, not the main topic. "My colony and my hobbies" is two topics, not one. The unity check is part of the Revise stage of the Whole-Piece Check, and cheapest at the Plan stage. Next: a complete narrative piece.

A complete narrative piece

A complete narrative piece is a story told in several paragraphs, where every paragraph does one job for the same problem. Its plan is an outline with one line for each paragraph.

A story outline, paragraph by paragraph.

Paragraph 1 - Opening: who, where, when, and what is normal or what someone wants.
Paragraph 2 - The problem: the one thing that goes wrong.
Paragraph 3 - What happens: what the characters do about it, in order.
Paragraph 4 - How it turns out: the answer to the problem, and how someone feels or what has changed.
The paragraphIts jobIn Nisha's sports-day story
Paragraph 1 - Openingsets up who, where and what someone wantsNisha runs the last leg of the relay, and her house has never won it.
Paragraph 2 - The problemone thing goes wrongRitu twists her ankle at the warm-up, and the team is one runner short.
Paragraph 3 - What happensthe steps, in order, all aimed at the problemNisha asks Tenzin to run, and they practise the baton pass behind the stands.
Paragraph 4 - How it turns outanswers the problem and closes the storyTenzin's pass is clean, they come second, and Ritu cheers loudest of all.
Not on the outlineserves no part of the problemwhat everyone ate at lunch, and who won the tug of war

This outline is the Plan stage of the Whole-Piece Check, this chapter's four stages: Plan, Draft, Revise and Proofread. You draft each paragraph to its line, revise and proofread the whole story, and then ask the last question: does it still serve its one problem?

The outline tells each paragraph what to do. The draft must also show the reader how each paragraph follows from the one before. In a story, most links are time links or content links.

The paragraphA weak startA start that links back
Paragraph 2 - The problemThen something happened.But at the warm-up, Ritu landed badly and grabbed her ankle.
Paragraph 3 - What happensThen we found someone.With ten minutes to go, Nisha ran to find Tenzin.
Paragraph 4 - How it turns outThen it was the race.When the whistle blew, Tenzin was ready with the baton.
A paragraph not on the outlineThen we had lunch.(cut it - lunch does not belong to the missing runner)

Notice three things. First, every strong start picks something up from the paragraph before it: the warm-up, the ten minutes, the baton. Choosing a link that names the real relation, here time, is the Bridge Test from Chapter 21. A paragraph that begins with "Then" every time names no relation at all, and the story starts to sound like a list.

Second, the problem is set up in one paragraph and answered in the last. If the closing paragraph is about something else - the prizes, the drive home, a new puppy - the story has answered a question nobody asked, and the real question is left hanging.

Third, the middle can grow. A story of five paragraphs may need two "What happens" paragraphs: a first try that fails and a second try that works. That is still one problem. The story has grown a longer middle, not a second problem, and the outline simply gains one more line.

Picture a train with four carriages. The engine is the opening: it pulls out of a station the reader can see. Each carriage behind it carries one part of the same journey - the problem, what happens, how it turns out - and each one is coupled to the carriage in front. The link at the start of a paragraph is the coupling. A paragraph about lunch is a carriage from a different train. It may be a perfectly good carriage, but hitch it on and the whole train heads the wrong way.

Sections 11 to 16 gave you each part of this story on its own: brainstorming, the simple outline, the opening paragraph, body paragraphs in order, the closing paragraph and the unity check across several paragraphs. This section joins them in one narrative, from plan to finished draft. Section 18 does the same for a description, which has no problem to answer, and section 19 for an explanation. Section 20 asks you to name the stage of the Whole-Piece Check that caught each slip. In Craft, a narrative becomes one of several genres whose outline you shape on purpose.

Here is a hard one, and it is worth slowing down for. How much should the opening paragraph hold? Sana's story is about leaving her violin on the bus on the morning of the school annual day. Her first draft spent two paragraphs on waking up, brushing her teeth and eating poha. Ask of each sentence: does the reader need this to understand the problem? The violin case, the bus and the annual day all pass the test. Breakfast does not. The opening shows only what the problem needs. Here is a second hard one. A new trouble appears in the middle. In Kunal's story, the kite is stuck on the temple wall, and in Paragraph 3 his little brother slips while climbing up after it. Is that a second problem, and so off the plan? Ask: does it make the first problem harder to solve? Here it does - now nobody can climb, so the boys must find another way to reach the kite. A trouble that makes the same problem harder belongs in the middle. A trouble that has nothing to do with the kite, such as a lost library book, is a second story. And a third. Is a sad or open ending finished? In Ines's story the lost cat is never found. Her last paragraph shows the family leaving a bowl of milk by the door every night, still hoping. The problem is answered: the reader knows how it turned out and how the family lives with it. A story has only stopped when the reader does not know how it turned out at all.

Did you know that the word narrative comes from the Latin narrare, which means "to tell"? Narrare is related to an older Latin word, gnarus, meaning "knowing". So a narrator is, in a sense, someone who makes things known: they know what happened, in what order, and they pass it on. And here is a second surprise. Before an animated film is drawn in full, its makers usually draw a storyboard: a row of small boxes, one rough sketch for each scene, set out in order like a comic strip. The team looks along the row and asks of every box whether it moves the story on. A scene that does not is cut at that stage, when it is only a quick sketch, rather than after days of careful drawing. A paragraph outline is a storyboard in words. Opening, problem, what happens, how it turns out: four boxes in a row. Cutting a paragraph about lunch costs one line of the outline. Cutting it after it is written costs a whole paragraph of work.

Four paragraphs is a starting shape, not a law. A very short story may fit in three, and a longer one may need two or three "What happens" paragraphs. A story need not open at the start of the day. It may open right at the problem, as long as the reader soon learns who and where. A link does not always need a linking word. Picking up an idea from the paragraph before works too, and Chapter 21's Bridge Test says to leave a linker out when the reader can already follow. An outline can change while you draft. If a better event turns up, use it, then check that the last line still answers the problem and update the outline to match.

  1. Write For and a main topic that names the story's one problem.
  2. Make an outline with one line for each paragraph before you draft.
  3. Give the opening only what the reader needs to understand the problem.
  4. Bring in the problem by the second paragraph.
  5. Put what happens in order, and let a failed try come before the one that works.
  6. Check that the last line of the outline answers the problem line.
  7. Start each paragraph with a link that picks up the one before it.
  8. Grow the middle with more steps, not with a second problem.
  9. Revise the whole story: read it aloud and cut any paragraph that does not serve the problem.
  10. Proofread every paragraph with the Four-Pass Check, then ask whether the problem is answered.
A complete narrative piece is several paragraphs about one problem. 1. Plan: For, and a main topic that names the one problem. Then an outline, one line per paragraph: Opening, Problem, What happens, How it turns out. 2. Check the plan: the last line must answer the problem line. 3. Draft: one paragraph per line, each starting with a link to the one before. 4. Revise: read it aloud and cut what does not serve the problem. 5. Proofread: the Four-Pass Check across every paragraph. 6. Ask: is the problem answered? If yes, it is finished. The key contrasts: more events is not more story - Farhan's six events made a list, Nisha's one problem made a story. A trouble that makes the same problem harder belongs; a different problem is a different story. "It was all a dream" and "The end" stop a story; an answer finishes it. Next: a complete descriptive piece.

A complete descriptive piece

A complete descriptive piece shows one subject in several paragraphs. Every paragraph shows one part of it, in an order the reader can follow, and every detail adds to one main impression.

A description outline, paragraph by paragraph.

Paragraph 1 - Introduction: name the subject, where and when, and its main impression.
Body paragraphs - one part each, in a clear order (entrance to far end, outside to inside, top to bottom), or one sense each.
Last paragraph - Closing: step back and sum up the whole, returning to the main impression.
The paragraphIts jobIn Dhruv's market description
Paragraph 1 - Introductionnames the subject and its main impressionEvery Sunday, the lane by the river becomes the busiest street in town.
Paragraph 2 - First partthe entranceFlower sellers thread marigold garlands, and the air smells sweet and sharp.
Paragraph 3 - Next partthe middle lanesOnions are piled into hills, and sellers shout their prices over each other.
Paragraph 4 - Last partthe far end, by the riverFish lie in rows on ice, and crows wait on the wall for scraps.
Paragraph 5 - Closingsums up the wholeBy noon the stalls fold away, and the busiest street in town is quiet again.
Not on the outlineserves no part of the marketthe car park next door, and which friend happened to be there

This outline is the Plan stage of the Whole-Piece Check, this chapter's four stages: Plan, Draft, Revise and Proofread. You draft each paragraph to its line, revise and proofread the whole description, and then ask the last question: does it still serve its one subject and its main impression?

A description can be planned in two main ways: part by part, or sense by sense. Both work. What matters is choosing one and keeping to it, so the reader always knows where they are.

The planParagraph 2Paragraph 3Paragraph 4
Part by part: the market, entrance to riverthe entrance and the flower sellersthe middle lanes and the vegetablesthe far end, the fish and the crows
Sense by sense: the marketwhat you see: garlands, hills of onions, silver fishwhat you hear: sellers shouting, scooters, a radiowhat you smell: marigolds, coriander, the river
A muddled planthe entrancethe sounds of the whole marketthe middle lanes

Notice three things. First, in the part-by-part plan each paragraph begins with a link that moves the reader on: "Past the flower sellers", "Further in", "At the far end". These place links work like the time links in a story. Choosing a link that names the real step is the Bridge Test from Chapter 21.

Second, the senses still belong in a part-by-part plan. Dhruv smells the marigolds at the entrance and hears the shouting in the middle lanes. The senses live inside each part. They are not a reason to jump.

Third, the muddled plan mixes the two ways. It walks to the entrance, then suddenly hears the whole market, then walks back into the middle. The reader loses their place. Pick one way of moving, and keep every paragraph to it.

Picture giving a friend a tour of a house with their eyes closed. You hold their arm and walk: the front door, then the hall, then the kitchen, then the garden at the back. At each step you tell them what is there - the creaky floorboard, the smell of toast. If you suddenly drag them from the kitchen to the roof and back to the door, they are lost. A description is that tour. The order is the route, each paragraph is a room, and the closing paragraph is standing in the garden, looking back at the whole house.

Section 17 planned and finished a story, which is held together by one problem. This section plans and finishes a description, which has no problem and is held together by one subject, one main impression and one order. Sections 13 to 16 gave you the introduction, body paragraphs in order, the closing paragraph and the unity check; here they work together in a new shape. Section 19 does the same for an explanation, and section 20 asks you to name the stage that caught each slip. In Craft, description becomes one of the genres whose outline you shape on purpose.

Here is a hard one, and it is worth slowing down for. Sometimes a detail seems to work against the main impression but really belongs. Mansi is describing the school library, and how quiet it is. She wants to include the ceiling fan, which clicks every time it turns. Is a noise off the plan in a piece about quiet? Ask: does this detail help the reader feel the main impression? Here it does. If the loudest sound in the room is a clicking fan, the room must be very quiet. A detail that sharpens the impression stays, even if it seems to point the other way. A paragraph about the library's late fees does not help the reader feel the quiet at all, so it goes. Here is a second hard one. When does a description turn into a story? In Sahil's description of the railway station, Paragraph 3 begins, "Then a porter dropped my suitcase, and it burst open." Now something has gone wrong, and the reader wants to know what happens next. The piece has started to become a story. A description can have people moving in it - a porter wheeling a trolley, a tea seller pouring chai - as long as they show the place rather than start a problem. And a third. What makes a closing paragraph more than a repeat? Chiara's introduction says the harbour is busy. If her closing paragraph only says "The harbour is busy" again, it adds nothing. A strong closing steps back and shows the whole in a new light: "By nightfall, the boats that left at dawn are tied up again, nose to nose, and the busiest place in town is asleep." It returns to the main impression without copying the first line.

Did you know that the word describe comes from the Latin describere? It is made of de, meaning "down", and scribere, meaning "to write". So to describe something was, first of all, to write it down. In Latin the word could also mean to copy, to sketch, or to mark out the shape of something, as you might mark out the outline of a field. That last meaning fits this section well. Before you write a single detail, you mark out the shape: an introduction, a paragraph for each part, a closing paragraph. And here is a second surprise. Many painters begin a landscape by roughing in the biggest shapes first - the line of the hills, the sky, the bend of a road - and only later add a single leaf or window. If the big shapes are in the wrong place, no amount of fine detail can save the picture. A description outline does the same job: it fixes the big shapes and the order of the parts, so that every small detail you add later has a place to sit.

Part by part and sense by sense are not the only orders. A description can also move through time, such as a street at dawn, at noon and at night, as long as it keeps to that order. Not every description needs five paragraphs. A small object, such as a seashell, may need only three: introduction, one body paragraph and a closing paragraph. A description may have people moving in it. It turns into a story only when something goes wrong and the reader starts waiting to see how it turns out. An outline can change while you draft. If a better part turns up, add it in its place in the order, then check that every paragraph still serves the main impression.

  1. Write For and a main topic that names one subject and its main impression.
  2. Choose one order before you draft: entrance to far end, outside to inside, top to bottom, or sense by sense.
  3. Make an outline with one line for each paragraph.
  4. Name the subject and its impression in the introduction.
  5. Give each body paragraph one part, and let the senses live inside it.
  6. Start each body paragraph with a link that moves the reader to the next part.
  7. Choose details that show the main impression, and leave out the ones that do not.
  8. Close by stepping back to the whole, not by adding one more part.
  9. Revise the whole piece: read it aloud and cut what does not serve the impression.
  10. Proofread every paragraph with the Four-Pass Check, then ask whether it still shows one subject and one impression.
A complete descriptive piece shows one subject in several paragraphs. 1. Plan: For, and a main topic that names one subject and its main impression. Choose an order. Then an outline, one line per paragraph: Introduction, one line per part, Closing. 2. Draft: one paragraph per line. Start each body paragraph with a link that moves the reader on. 3. Revise: read it aloud and cut what does not serve the main impression. 4. Proofread: the Four-Pass Check across every paragraph. 5. Ask: does it still show one subject and one impression, from introduction to close? If yes, it is finished. The key contrasts: more details is not more description - Ruhi's heap lost the reader, Dhruv's order walked them through. Part by part or sense by sense, but not both at once. A detail that sharpens the impression stays; one that wanders goes. One more part is not an ending; a closing paragraph steps back to the whole. Next: a complete explanatory piece.

A complete explanatory piece

A complete explanation tells a new reader how to do something, or how something works, in an order they can follow from start to end. Its plan has three parts, and the middle part does most of the work.

An explanation outline.

Introduction: what this piece explains, and why the reader might want to know.
Body paragraphs: one step or one part each, in a logical order - time order for how to do something; part by part, or cause to effect, for how something works.
Conclusion: the finished result, a final tip, or why it matters - never a new step.
The outline lineWhat goes in itIn Joel's nimbu pani piece
Introductionname the topic and why it is usefulOn a hot afternoon, nothing cools you down like a glass of nimbu pani. This piece explains how to make four glasses.
Body 1 - what you needeverything the reader must gather firstYou need two lemons, two spoons of sugar, a pinch of salt, a jug of cold water, ice and a knife.
Body 2 - the stepseach step in time order, linked with First, Next, After thatFirst, roll each lemon and cut it in half. Next, squeeze the juice into the jug and pick out the seeds.
Body 3 - finishingthe last steps, still in orderFinally, pour in the water, add the ice and stir.
Conclusionthe result and a tipTaste it before you serve it: if it is too sour, add a little more sugar.

Planning an explanation this way is the Plan stage of the Whole-Piece Check, this chapter's four stages: Plan, Draft, Revise and Proofread. The outline decides the order; drafting fills each outline line with one paragraph; revising and proofreading make sure a new reader can follow every step.

An explanation comes in two main kinds, and a third that is close to the second. All of them need an introduction, body paragraphs in order and a conclusion, but the order inside the body is different.

The kindThe order of the body paragraphsA main topicLinkers that fit
How to do somethingtime order: what you do first, next and lasthow to cover a school book with brown paperFirst, Next, After that, Finally
How something workspart by part, or cause to effect: this happens, so that happenshow a bicycle pump pushes air into a tyreWhen, So, This means, As a result
How something happens in naturestage by stage, in the order it happenshow rain forms from a puddleFirst, Then, Once, In the end
A story dressed as an explanationno order a reader can use - events, not stepsthe day I made nimbu pani and spilled itnone - this is a narrative, not an explanation

Notice three things. First, the kind decides the order. A how-to piece follows the clock: a reader doing the steps needs them in the order they will do them. A how-it-works piece follows the machine or the process: each part, or each cause, leads on to the next.

Second, the linkers come from the order, not the other way round. First, Next and Finally name a time relation, so they fit steps. So and As a result name a result, so they fit causes. The Bridge Test from Chapter 21 checks that each linker names the relation that is really there, and leaves it out when the order already shows it.

Third, an explanation is not a story. The day you spilled the nimbu pani may be a good story, but it does not tell the reader how to make the drink. If a memory helps the explanation - a warning or a tip - keep it to one sentence and make it serve a step. If it does not, save it for a narrative piece like the one in Section 17.

Picture giving a friend directions to your house over the phone. You cannot point. You cannot say "you know, the usual way". You have to start where your friend is standing and give each turn in the order they will meet it: out of the school gate, left at the temple, past the bakery, third house on the right. Miss one turn and your friend is lost in a lane you never mentioned. An explanation is directions on paper. The introduction says where you are going, each body paragraph is one stretch of the road, and the conclusion is the front door.

Section 17 planned and wrote a complete story, and Section 18 a complete description. This section completes the set with an explanation, whose body follows the order of steps or parts rather than the events of a story or the path of the eyes. All three use the same outline frame: introduction, body paragraphs in order, conclusion. Section 20 then reviews all of Practice and asks you to name the stage of the Whole-Piece Check behind every fix. In Craft, the explanation grows into a full expository essay with a thesis statement, in Section 26.

Here is a hard one, and some explanations really are harder to plan than others. The first hard case is a step the writer does without noticing. Joel nearly left out "roll each lemon on the table first". He always does it, so it did not feel like a step. The only way to catch an invisible step is to act the explanation out, or read it to someone who has never done the job, and do exactly what the page says - nothing more. If the reader gets stuck, a step is missing. The second hard case is choosing the order for a how-it-works piece. Mei is explaining how a bicycle pump works. Should she go part by part - the handle, the tube, the valve - or follow the air as it moves? Both are orders. But her reader wants to understand what happens when you push, so following the air is clearer: you push the handle down, the air in the tube is squeezed, it is forced out through the valve, and the tyre fills. Ask: which order lets the reader picture the thing working? The third hard case is a step that comes with a reason. "Pick out the seeds" is a step. "Pick out the seeds, because they make the drink bitter" is a step and a reason. Is the reason off the plan? No: it helps the reader see why the step matters, so they will not skip it. A reason that serves a step belongs. A memory that serves nothing does not. An explanation is finished when a new reader could follow it without you there. That is a harder test than "I wrote down what I do", and it is the one that counts.

Did you know that the word explain once meant to make something flat? It comes from the Latin word explanare, built from planus, which means flat or level. The same planus gives us the word plain. To explain something was to smooth it out and spread it open, so that nothing was folded up or hidden. That is exactly what a good explanation does. Every step is laid out flat, in order, where the reader can see it. And here is a second surprise. People who write instructions for a living - for machines, games or flat-pack furniture - often test them in a simple way. They give the instructions to someone who has never seen the product and watch, without helping, to see where that person gets stuck. Every place where the tester pauses or frowns is a step the writer forgot, or explained in words the tester did not know. You can do the same test at home. Give your explanation to a younger cousin or a friend, stay quiet, and watch where they stop.

Not every explanation needs a separate paragraph for what you need. In a short piece, the introduction or the first step can name the things to gather. Some steps can happen in either order: sugar first or salt first makes no difference. Then any order is fine, but choose one and keep to it. A how-it-works piece can often be ordered in more than one sensible way. Choose the order that lets the reader picture the thing working, and do not jump between orders. Linkers such as First, Next and Finally help a reader follow steps, but not every sentence needs one. When the order is already clear, a linker on every line only slows the reader down.

  1. Choose one process you really know or really understand, small enough to explain in a few paragraphs.
  2. Write For and Main topic at the top, and picture a reader who has never done this.
  3. Decide the kind: how to do something, or how something works.
  4. Write one outline line per paragraph: introduction, each step or part, conclusion.
  5. Put the body lines in a logical order: time order for steps, part by part or cause to effect for how it works.
  6. Act the outline out, with real objects or in your head, to check the order.
  7. Draft one paragraph per outline line, each giving one step or one part, and link each to the one before.
  8. Keep a memory or a reason only if it serves a step, and only in one sentence.
  9. End with the result, a tip or why it matters, never with a new step or The end.
  10. Read the whole piece as a new reader, then revise and proofread it before calling it finished.
A complete explanatory piece. 1. Plan: write For and Main topic, and picture a reader who has never done this. 2. Decide the kind: how to do something (time order) or how something works (part by part, or cause to effect). 3. Outline: introduction, one line for each step or part, conclusion. 4. Check the order by acting it out. 5. Draft: one paragraph per outline line, each linked to the one before. 6. Close with the result, a tip or why it matters. 7. Read it as a new reader, then revise and proofread. The key contrasts: knowing how is not the same as explaining how. An explanation follows steps or parts, not the events of a story. A reason that serves a step belongs; a memory that serves nothing does not. A conclusion gives the result; it never adds a step the reader needed earlier. The test: could someone who has never done it follow the whole piece without asking you a question? That is the last question of the Whole-Piece Check, asked of an explanation. Next: Explain your evidence.

Explain your evidence

Explaining your evidence means saying three things about every fix: what the problem is, the stage of the Whole-Piece Check that catches it, and the section or tool behind the fix.

Problem, stage, tool.

Plan - the outline is wrong: brainstorming (S11), the main topic and the outline (S12).
Draft - the piece does not match its outline: the introduction (S13), body paragraphs in order (S14), the conclusion (S15), links between paragraphs (the Bridge Test, Chapter 21), and the shape of a story, a description or an explanation (S17, S18, S19).
Revise - something is extra or reads badly: the unity check across every paragraph (S16), the Cut Test (Chapter 20), the Read-Aloud Test (Chapter 24).
Proofread - a slip: the Four-Pass Check (Chapter 23).
The problemThe stageThe section or toolThe fix
Main topic: everything about IndiaPlanBrainstorming ideas (S11), A simple outline (S12)choose one small topic: how our school celebrates Onam
Outline line: my cousin's wedding, in a piece about our school's OnamPlanA simple outline (S12)cut the line before drafting
Paragraph 1 starts in the middle of the action and never says what the piece is aboutDraftOpening a bigger piece (S13)an introduction that names the topic
Paragraph 3 comes before Paragraph 2, though the outline had them the right way roundDraftBody paragraphs, in order (S14)follow the outline's order
Last paragraph: The end.DraftClosing a bigger piece (S15)a conclusion that closes the whole piece
A sentence about the new library in a story about a lost dogReviseKeeping every paragraph on topic (S16)cut it
The garden is very, very, really quiet and peaceful and calm.Revisethe Cut Test (Chapter 20)The garden is quiet.
Strings of lights hangs above every stall.Proofreadthe Four-Pass Check (Chapter 23)hang
Every stage comes back cleanall fourthe Whole-Piece CheckFinished - no change needed

Naming the stage is the Whole-Piece Check run backwards. Instead of going Plan, Draft, Revise and Proofread in order, you start from a problem and ask which of the four stages owns it - and that answer tells you where to look first in your next piece.

Some problems look as if they could belong to two stages. They do not. The evidence - usually the outline - decides which stage the problem started in, and so which stage should have caught it.

The problemThe evidenceThe stage
Two paragraphs in the wrong orderthe outline itself lists them in that orderPlan (S12): fix the outline, then the draft
Two paragraphs in the wrong orderthe outline was right; the draft did not follow itDraft (S14): follow the outline
A whole paragraph off the main topicthe outline has a line for itPlan (S12): the outline line was off the main topic
A sentence off the main topicthe outline has no line for it; it crept in while draftingRevise (S16): the unity check cuts it
A skipped step in an explanationthe outline skipped it tooPlan (S12, S19): add the outline line
A skipped step in an explanationthe outline had it; the draft left it outDraft (S14, S19): write the paragraph the outline asked for

Notice three things. First, the stage is decided by where the problem started, and the outline is the evidence. Missing or misplaced against the outline means Draft. Wrong in the outline itself means Plan. Extra, with no outline line behind it, means Revise.

Second, earlier is cheaper. A problem caught at the Plan stage costs one outline line. The same problem caught at the Revise stage costs a whole paragraph, and sometimes the link before and after it as well. That is why naming the stage matters: it tells you to check earlier next time.

Third, finished needs evidence too. "It looks fine" is a feeling. "The main topic is single and small, every outline line has its paragraph in order, the introduction opens and the conclusion closes, nothing wanders off, and the Four-Pass Check found no slips" is evidence. When every stage comes back clean, say so, and say Finished - no change needed.

Picture a doctor who finds a child with a sore foot. A careless doctor puts on a plaster and sends the child home. A careful doctor puts on the plaster too, and then asks: where did this come from? A shoe that is too tight? A stone on the path? A twisted ankle at football? The plaster fixes today; the answer fixes next week. Explaining your evidence is the careful doctor's question. Plan, Draft, Revise and Proofread are the four places a writing problem can come from, and naming the right one stops it coming back.

This section closes Practice. Sections 11 and 12 built the plan: brainstorming and the outline. Sections 13 to 15 built the draft: the introduction, body paragraphs in order and the conclusion. Section 16 ran the unity check across a whole piece, and Sections 17 to 19 put everything together in complete stories, descriptions and explanations. This review joins them by naming the stage behind every fix. Craft begins with the thesis statement, which sharpens the Plan stage: one sentence that says what the whole piece will show.

Here is a hard one, and it is worth taking slowly: two problems that look identical but belong to different stages. Take two explanations of how to make a paper boat, each with the corner fold drafted before the half fold. In Kiran's, the outline says: Paragraph 2: fold the corners. Paragraph 3: fold in half. In Leela's, the outline says: Paragraph 2: fold in half. Paragraph 3: fold the corners. The drafts are the same. But Kiran's problem started in the Plan stage, and Leela's in the Draft stage. Kiran must fix his outline; Leela must follow hers. You cannot tell which stage owns the problem from the draft alone. You need the evidence: the outline. Here is a second hard case. Arjun's story about a lost goat has a paragraph about the village mela. Is that off the main topic? It depends. If the goat was lost at the mela, the paragraph sets the scene, and it is on the plan. If the goat was lost from the field behind the house, the mela serves nothing, and the Revise stage should cut it. The same paragraph is right in one piece and wrong in another. The main topic and the outline decide. And a third. Sometimes the honest answer is that nothing is wrong. That is the hardest answer to give, because it feels as if a reviewer should always find something. But a reviewer who invents a problem is as wrong as one who misses a real one. When every stage comes back clean, the evidence for Finished - no change needed is as strong as the evidence for any fix.

Did you know that the word evidence comes from the Latin evidens, which means clear or plain to see? It is built from videre, to see - the same root that gives us video and vision. So evidence is, at heart, what lets someone else see what you saw. "It sounds wrong" shows nothing to anyone else. "The outline has the corners before the fold" lets anyone look and see it for themselves. And here is a second surprise. Doctors, mechanics and pilots all use the same habit when something goes wrong: they do not only fix the fault, they record where it came from. A mechanic who replaces a worn tyre also checks why it wore unevenly. A pilot's checklist is used before every flight because a problem found on the ground, early, costs far less than one found in the air. Writers can do the same. A margin note such as "Plan - S12" is a tiny record of where a problem came from, and a row of them, across several pieces, shows you your own habits.

Not every piece comes with an outline. When there is none, judge from what the piece is trying to do, and say so: "no outline, so I judged against the main topic in the introduction". Some problems belong to more than one stage at once. A conclusion that both says The end and has a spelling slip needs two notes: Draft for the conclusion, Proofread for the slip. Section numbers are a help, not a test. If you can name the tool - the outline, the unity check, the Four-Pass Check - that is the evidence that matters. A clean result is real. If every stage comes back clean, Finished - no change needed is the right answer, not a lazy one.

  1. Read the outline first, if there is one: it is the evidence for most decisions.
  2. Check the Plan stage: one small main topic, and outline lines in a sensible order.
  3. Check the Draft stage: every outline line has its paragraph, in order, with an introduction, links and a conclusion.
  4. Check the Revise stage: anything extra, anything that drags, anything that does not earn its place.
  5. Check the Proofread stage: the Four-Pass Check on every paragraph, the last one included.
  6. For every problem, write the problem, the stage and the section or tool.
  7. Name the stage where the problem started, not where you happened to notice it.
  8. Use the outline to decide between Plan and Draft when a paragraph is in the wrong place.
  9. Give evidence for finished: what each stage found, not how the piece feels.
  10. When every stage is clean, write Finished - no change needed, and stop.
Explain your evidence. For every problem, say three things: what it is, the stage that catches it, and the section or tool behind the fix. 1. Plan - the outline is wrong: brainstorming (S11), the main topic and the outline (S12). 2. Draft - the piece does not match its outline: introduction (S13), body paragraphs in order (S14), conclusion (S15), links (Chapter 21), the shape of a story, description or explanation (S17 to S19). 3. Revise - something extra or something that reads badly: the unity check (S16), the Cut Test (Chapter 20), the Read-Aloud Test (Chapter 24). 4. Proofread - a slip: the Four-Pass Check (Chapter 23). The key contrasts: a fix is not complete without its reason. Name the stage where the problem started, not where you saw it. Wrong in the outline means Plan; missing or misplaced against the outline means Draft; extra means Revise. Finished needs evidence too, and when every stage comes back clean, Finished - no change needed is the honest answer. Next: Craft, the thesis statement.

Writing with intent · ages 12–16

The thesis statement

A thesis statement is the one sentence that states what a whole piece will show. A topic sentence states what one paragraph will show (Chapter 27, Paragraph Architecture). The thesis sits above them all, and each body paragraph's topic sentence serves one part of it.

Three questions for a thesis.

1. Whole: does it cover the whole piece, not just one paragraph?
2. Specific: could a reader roughly predict the body paragraphs from it?
3. Something to show: does it state an argument to defend or an explanation to give, rather than a bare fact or an announcement?

Fail one, and rebuild the sentence before you draft.
The sentenceWhat it isWhyA thesis?
The monsoon reaches our town in June.a bare facta calendar settles it; there is nothing to showNo
In this essay I will write about the monsoon.an announcementit names the topic but says nothing about itNo
The monsoon has many effects on people.a vague gestureit could promise anything, so it predicts no paragraphsNo
The underpass on Station Road floods every July.a topic sentenceone paragraph can finish itNo - it serves one paragraph
The monsoon is our town's hardest season and its most welcome one: it floods the roads, but it fills the wells and wakes the fields.a thesisit covers the whole piece, predicts its parts and needs showingYes

Every body paragraph then has one job: show one part of the thesis. The thesis is written at the Plan stage of the Whole-Piece Check (Plan, Draft, Revise, Proofread), and it is the governing idea that the check's last question returns to: does the finished piece still serve it?

A thesis does not have to be an argument. It has to say something the body will show. What the body does with it depends on the kind of piece.

Kind of pieceA thesis that worksWhat the body must do
Persuasive essayOur school should replace paper notices with one online board, because it would save paper, reach parents faster and end lost slips.defend each reason with support and answer the other side (Chapter 29)
Expository essayA pressure cooker cooks dal faster because it traps steam, and the trapped steam raises the temperature at which the water boils.explain each stage in turn: the trapped steam, the higher boiling point, the faster cooking
Reflective essayLosing the inter-school final taught me more about our team than winning the semi-final did.show what the loss revealed, one discovery to a paragraph
ReportThe Class 8 survey points to two causes of the morning queue at the gate: one narrow entrance and a late bell.set out the findings neutrally, one cause to a section (Chapter 22 owns the tone)

Notice three things. First, an explanatory thesis need not be debatable. Nobody sensible argues with how a pressure cooker works. But it still states something the body has to show, step by step, which a bare fact such as "Our cooker is red" does not.

Second, a good thesis often carries the plan inside it. The because-clause or the list after a colon names the body paragraphs in order, so the outline almost writes itself. That is useful, not compulsory: what matters is that a reader could predict the body, not that every part is listed.

Third, a thesis must be small enough to finish. "Technology has changed education" would need a book. "Recorded lessons help students who miss school to catch up, but they cannot replace questions asked in class" fits four paragraphs. The word limit decides the size of the thesis, just as a short piece in Discover needed a small About line.

Picture the lid of a jigsaw box. The picture on the lid shows what the whole puzzle will be. Each body paragraph is one area of the puzzle - the sky, the boat, the water - and every piece fits because the lid shows where it goes. A bare fact on the lid is a picture of one single piece. An announcement is a lid that says only "puzzle". Neither tells you where anything goes. A thesis is the full picture, shown before the first piece is placed.

In Practice you made an outline with one line per paragraph and kept every paragraph on the main topic (sections 12 and 16). This section places one sentence above that outline: the thesis. The next section shows how the outline beneath it changes shape with the genre. Section 23 links the paragraphs the thesis governs, sections 24 and 25 revise the whole draft against it, and sections 26 and 27 build complete expository and persuasive essays around one. In Master, the Whole-Piece Check's last question - does it still serve the governing idea? - is asked of demanding pieces under time pressure.

Here are the cases that take real judgement. First: thesis or topic sentence? Take "Cycling to school keeps students fit." Nothing in its grammar says which it is. The difference lies in the job it does. Ask how many paragraphs it could govern. If you can only develop it for one paragraph, it is a topic sentence, and it belongs inside the body. "Cycling to school keeps students fit, saves families money and clears the traffic at the gate" can govern three paragraphs, so it can be a thesis. The same idea, given a wider reach, changes level. Second: a fact, or an explanatory thesis? "Volcanoes erupt when pressure builds up under the ground" is true and found in any textbook. Does that make it a bare fact? Not if the piece's job is to explain how: the molten rock, the trapped gas, the rising pressure, the eruption. Then the body has something to show, stage by stage. Compare "The library closes at 4 pm": there is no how or why in it to explain. The test is not whether a sentence is true, but whether the piece has work to do in showing it. Third: the thesis that changes while you draft. Halfway through, your evidence may point somewhere else. Changing the thesis is not failure; it is honest. The danger is the piece that keeps the old thesis at the top while the body argues the new one. If the thesis changes, check every paragraph against the new sentence before you call the draft done. The lesson: a thesis is defined by its job - governing the whole piece - not by its length, its position or how clever it sounds.

The word thesis comes from Greek, where it meant "a setting down" or "a placing": something a speaker puts down in front of the listeners and then stands by. It shares its root with hypothesis, which is literally "a placing under" - an idea put underneath an investigation to be tested. That old meaning still shows in universities. A long piece of research written for a higher degree is called a thesis, because the whole document exists to set down and defend one position. A student may spend years on it, and every chapter must serve that one position, exactly as every body paragraph serves yours. Journalists have their own version. Many newspaper reporters are taught to include, near the top of a longer feature, a short paragraph that tells the reader what the story is really about and why it matters now. Editors often call it the nut graf, because it holds the story in a nutshell. Without it, a reader of a long feature can get three paragraphs in and still not know why they are reading. Essay, thesis, nut graf: three names for one habit. Say early, in a sentence or two, what the whole piece will show.

Not every piece states its thesis outright. Stories and many personal essays hold a governing idea without saying it in one sentence. The writer still needs one at the Plan stage, even if the reader never sees it written down. The end of the introduction is a common and useful place for a thesis in school essays, but it is not a law. Some pieces open with it; some build towards it. What matters is that the reader meets it early enough to follow the piece. A thesis that lists its parts is helpful but not compulsory. In a short essay a list can make the writing feel mechanical; the test is whether the body can be predicted, not whether every part is named. This section asks only whether a sentence can govern a whole piece. Whether a persuasive thesis is clear, specific and debatable enough to argue is tested with the tools of Chapter 29, Argument and Persuasion.

  1. Write the thesis at the Plan stage, before any body paragraph.
  2. Keep it to one sentence that covers the whole piece.
  3. Make it specific enough that a reader could predict the body paragraphs.
  4. State something to show: an argument to defend or an explanation to give.
  5. Replace a bare fact with what the facts add up to.
  6. Delete announcements such as "In this essay I will" and state the point instead.
  7. Size the thesis to the word limit: narrow enough to finish, wide enough for every paragraph.
  8. Write each body paragraph's topic sentence underneath it and check that each serves one part.
  9. If the thesis changes during drafting, check every paragraph against the new one.
  10. At the end, ask the Whole-Piece Check's last question: does the piece still serve this sentence?
The thesis statement. A thesis is the one sentence that states what a whole piece will show. A topic sentence states what one paragraph will show, and serves one part of the thesis. Three questions for a thesis: 1. Whole: does it cover the whole piece? 2. Specific: could a reader predict the body paragraphs? 3. Something to show: is it an argument to defend or an explanation to give? The key contrasts: a fact ("The monsoon reaches our town in June") or an announcement ("In this essay I will write about the monsoon") is not a thesis. A vague gesture ("Sport is important in many ways") predicts nothing. A topic-sentence-sized claim runs out after one paragraph. An explanatory thesis need not be debatable; it must still give the body something to show. Write the thesis at the Plan stage, check every outline line against it, and if it changes, check every paragraph again. Next: Outlining for different genres.

Outlining for different genres

Every outline lists what each paragraph will do, in order. What changes from genre to genre is the order and the kind of line. The task's purpose word tells you which shape to use.

Three outline shapes.

Narrative (to tell): in time order, or as situation, problem, attempts, turning point and resolution.

Explanatory (to explain): by parts, when order does not matter; by steps, when it does.

Persuasive (to argue): claim, reasons each with support, the other side answered, and a close. Chapter 29 owns this outline; here you only choose it.
Task wordsGenreOutline shapeA skeleton
Write a story about; narrate an incidentNarrativeproblem and resolution, in time orderOpening - Problem - Attempts - Turning point - Resolution
Explain how to; describe the processExplanatory, by stepssteps in the order a reader would do themThesis naming the steps - Step 1 - Step 2 - Step 3 - Close
Explain the features of; describe the partsExplanatory, by partsone part to a paragraph, in a sensible orderThesis naming the parts - Part 1 - Part 2 - Part 3 - Close
Explain whyExplanatory, by causesone cause to a paragraph, often strongest lastThesis naming the causes - Cause 1 - Cause 2 - Cause 3 - Close
Argue; do you agree; shouldPersuasiveclaim, reasons, the other side, closeClaim - Reason 1 - Reason 2 - Other side answered - Close

Choosing the outline's shape belongs to the Plan stage of the Whole-Piece Check (Plan, Draft, Revise, Proofread): once the reader, the purpose and the governing idea are decided, the shape follows from the purpose, and every later stage builds on it.

The same topic can be outlined three ways. Here is one school sports day, planned for three different tasks.

TaskShapeThe outline
Write a story about sports day.Narrative: problem and resolutionOpening: Meena laces her shoes for the relay. Problem: our fastest runner twists an ankle. Attempts: we try three others in practice. Turning point: Meena offers to run two legs. Resolution: second place, and the loudest cheer of the day.
Explain how our school organises sports day.Explanatory: stepsThesis: sports day takes four weeks of steps. Step 1: houses choose teams. Step 2: heats in PE lessons. Step 3: the track is marked the day before. Step 4: teachers and prefects run the events. Close: four weeks for one day.
Should sports day move to December?Persuasive: claim, reasons, other side, closeClaim: yes, it should. Reason 1: the cooler weather. Reason 2: fewer exams clash. Other side: December is busy with the annual day - answered by holding the two a fortnight apart. Close: a cooler, calmer sports day.

Notice three things. First, the topic stayed the same and the outline changed completely. The purpose chose the shape: the story needs a problem, the explanation needs steps, and the argument needs a claim and the other side.

Second, parts and steps are both explanatory, but they are not interchangeable. Use steps when the reader must follow an order: making a diya, organising a day, how water reaches a tap. Use parts when the order is yours to choose: the features of a library, the parts of a bicycle, the causes of a queue.

Third, each shape has a line the others do not. A narrative outline that has no problem is only a timeline. An explanation that has no thesis naming its parts or steps leaves the reader lost. A persuasive outline with no other side is not yet the argument Chapter 29 teaches. When you check an outline, look first for the line its genre cannot do without.

Picture three journeys. A story is a road trip: you travel in order, and somewhere in the middle the car breaks down and has to be fixed. An explanation is a museum tour: either room by room, in whatever order makes sense (parts), or along a fixed path where each room leads to the next (steps). An argument is a court case: the claim, the witnesses for it, the other side's best question, the answer, and the closing speech. Before you plan, ask which journey this piece is taking.

In Practice you made a simple outline with one line per paragraph (section 12) and planned complete narrative, descriptive and explanatory pieces (sections 17 to 19). Section 21 placed a thesis above the outline. This section makes the outline beneath it the right shape for the genre. Section 23 links the paragraphs the outline produces, and sections 26 and 27 build complete expository and persuasive essays on these shapes. In Master, section 37 returns to genre as a final synthesis, choosing outline, voice and structure together.

Here are the cases that take real judgement. First: a task that mixes purposes. "Describe a time you helped someone, and say what you learned." The first half asks for a story; the second asks for reflection. The answer is not to choose one and drop the other. Outline it as a narrative - situation, problem, what you did, the resolution - and give the reflection its own closing paragraph. The main purpose sets the shape; the second purpose earns one slot inside it. Second: time order inside an explanation. "Explain how our town grew from a village." Time order sounds like a story, and the outline will run in sequence: the railway, the mill, the college. But there is no problem and turning point, and the governing idea is a cause - each stage led to the next. So this is an explanatory outline by stages that happen to be in time order. Time order is a tool both genres use; what makes a story a story is the problem and its resolution. Third: a story inside an argument. Kabir's persuasive essay on safer crossings opens with the morning his little brother nearly stepped in front of a scooter. Does the story change the outline's shape? No. The story is the hook in the introduction, or the support for one reason. The outline stays persuasive: claim, reasons, the other side, close. A short narrative can live inside an argument the way a photograph lives inside a report. The lesson: name the piece's main purpose first, choose its shape, and only then decide where any smaller purpose sits inside that shape.

The names of the genres carry their purposes inside them. Narrative comes from the Latin narrare, "to tell". Argument comes from the Latin arguere, which meant "to make clear" or "to prove". And essay comes from the French essayer, "to try": the sixteenth-century writer Michel de Montaigne called his short prose pieces Essais, attempts, because each one tried out an idea. Journalists show how firmly purpose decides shape. A news report is usually built as an inverted pyramid: the most important facts first - what happened, to whom, where and when - and the less important details lower down, so that a reader who stops after one paragraph still has the story's core. A short story could never be built that way. Telling the ending first would empty the turning point of all its tension. Same newspaper, two shapes. The feature writer on the next page may well start with a scene and save the key moment for later, because that piece's purpose is different. The purpose decides; the outline follows.

These shapes are defaults, not cages. Skilled writers break them on purpose: a story told backwards, an argument that opens with the other side. Break a shape only once you know what it gives you. Many real pieces mix purposes. Choose the shape of the main purpose, and give the smaller purpose a slot inside it, rather than switching shape halfway. Descriptive and reflective pieces have shapes of their own too, such as spatial order for a description, or the experience followed by what it meant for a reflection. They are planned with the same care; section 37 in Master returns to them. The persuasive outline belongs to Chapter 29, Argument and Persuasion, which teaches how to build and test its claim, reasons and counter-argument. This section only asks you to recognise when a task needs that shape.

  1. Underline the task's purpose word before you plan anything.
  2. Decide the genre from the purpose, not from the topic.
  3. For a story, find the problem and the turning point before anything else.
  4. Do not mistake a timeline for a story: something must go wrong and be put right.
  5. For an explanation, decide whether order matters: steps if it does, parts if it does not.
  6. Order steps as the reader will do them, not as you happened to learn them.
  7. Keep one shape per outline: do not mix steps into a parts outline.
  8. For an argument, use Chapter 29's shape and make sure the other side has a line.
  9. If a task mixes purposes, use the main purpose's shape and give the other a slot.
  10. Before drafting, look for the one line the chosen genre cannot do without.
Outlining for different genres. The purpose chooses the outline, not the topic. Read the task's purpose word first. 1. Narrative (to tell): time order, or situation, problem, attempts, turning point, resolution. 2. Explanatory (to explain): steps when order matters, parts when it does not, causes for "explain why". 3. Persuasive (to argue): claim, reasons with support, the other side answered, close (Chapter 29). The key contrasts: a timeline is not a story until something goes wrong and is put right. Steps and parts are both explanatory, but steps follow the reader's order. A list of points is not an argument without a claim and the other side. When a task mixes purposes, use the main purpose's shape and give the other a slot inside it. Check every outline for the line its genre cannot do without. Next: Linking paragraphs across a whole piece.

Linking paragraphs across a whole piece

Every paragraph after the first opens with a link that does two jobs at once: it picks up the paragraph before, and it keeps the reader pointed at the thesis.

At every paragraph gap, three questions.

1. What did the last paragraph say, in a few words?
2. What is the relation between that point and this one? Name it with the Bridge Test from Chapter 21: adding, contrast, cause, result, time, example or conclusion.
3. Does this opener still point at the thesis?

Thesis: Our weekly market does more than sell vegetables; it holds the neighbourhood together.
The gapThe relationAn opener that only counts or signalsAn opener that names the idea
Paragraph 1 to 2contrast: more than foodSecondly, people meet there.But people come for more than tomatoes; they come to meet.
Paragraph 2 to 3adding: who gains mostFurthermore, it is cheap.Those meetings matter most to the people who live alone.
Paragraph 3 to the conclusionconclusionIn conclusion, the market is good.So the market is not only a place to shop; it is where the neighbourhood sees itself.
A story, morning to eveningtimeAlso, in the evening we went home.By the time the street lamps came on, the stalls were empty.

Notice that the strong openers do not always carry a linking word. 'Those meetings' needs no Furthermore: the words themselves pick up the paragraph before. The Bridge Test's last step still applies across paragraphs: leave the linker out when the road already runs.

Linking paragraphs completes the Draft stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread: each paragraph is written to its plan and linked to the one before it, and that chain of links is what lets the finished piece still serve its one governing idea.

A link can face backwards, forwards, or both. Only the openers that face both ways hold a whole piece together. Here are four openers for the third paragraph of the market essay, after a paragraph about people meeting.

The opener for paragraph 3Picks up paragraph 2?Points at the thesis?What the reader feels
Another way the market holds the neighbourhood together is its low prices.NoYesthe essay restarts at every paragraph, like a list
Those meetings remind me of my grandfather's village, where everyone met at the well.YesNolinked, but drifting away from the thesis
Thirdly, the market.Only by countingNoa number, not a connection
Those meetings matter most to the people who live alone.YesYesone argument, still moving forward

Notice three things. First, a chain can drift. Each opener can link neatly to the one before it, and still, by the fourth paragraph, the essay is about a grandfather's village instead of the market. Every link was local; none of them looked back at the thesis. That is why the slip of first sentences starts with the thesis written at the top.

Second, pointing at the thesis without picking up the last paragraph is not enough either. 'Another way the market holds the neighbourhood together' repeats the thesis at every gap, so each paragraph starts from zero. The reader never feels one point grow out of the last.

Third, the relation must be the true one. If paragraph 3 adds a reason, However tells the reader to expect a turn that never comes. If paragraph 3 answers an objection, Also hides the turn. The Bridge Test from Chapter 21 is the check: name the relation first, then choose the words. Across a whole piece, the words are often a short phrase that names the last point rather than a single linker.

Picture a relay race. Each runner takes the baton from the runner before, and every runner is heading for the same finish line. A runner who grabs the baton and sets off towards a different finish has made the handover and lost the race. A runner who ignores the baton and starts fresh from the track has kept the direction and broken the chain. The baton is the content link; the finish line is the thesis. A whole piece needs both at every handover.

Section 21 gave a whole piece one thesis, and section 22 shaped its outline to fit the genre. This section joins the outline's paragraphs into one chain, each opener picking up the last point and pointing at the thesis. Section 24 revises the linked draft for voice and rhythm, and section 25 for punctuation and length. In the complete essays, letter and report of sections 26 to 29, linking is part of every draft. In Master, section 33 reorders whole paragraphs, and every gap a moved paragraph touches must then be linked again.

Here is a hard one, and it is worth slowing down for: the paragraph that turns against the thesis. Sahil's essay argues that his school should keep its uniform. Paragraph 3 gives the other side: 'Some students say a uniform hides who they really are.' That opener needs a contrast link, and Some students say supplies it. But paragraph 4 must turn back again. If it opens 'However, a uniform hides only clothes', the reader meets two contrasts in a row and may lose track of which side the writer is on. Name each relation separately. Paragraph 3 turns away from the thesis; paragraph 4 answers and returns to it: 'That worry is fair, but a uniform hides only clothes; the debate club, the art room and the cricket field still show who we are.' The link into paragraph 4 must name the objection it is answering, not just signal a turn. Here is a second hard one: the jump in time. In a narrative told in order, the clock carries the relation and most openers need no linker at all. Tara's story of her first Onam away from home moves from morning to the sadya to evening without a single then. But her third paragraph goes back three years, to her grandmother's kitchen. Without a time link - 'Three Onams earlier, I had stood in Ammamma's kitchen' - the reader assumes the scene comes next, and the story stops making sense. Order is enough only while the order holds. And a third: when the content link is already enough. 'Those meetings matter most to the people who live alone' picks up the last paragraph by itself. Adding Furthermore in front of it stacks a second link on the first, which Chapter 21 already warns against. One link per gap, chosen to name the true relation, is the aim.

The word paragraph began as a mark, not a block of text. In ancient Greek manuscripts, a paragraphos was a short stroke written in the margin beside the line where a new section, or a new speaker, began. The word is built from para, 'beside', and graphein, 'to write': something written beside. Readers met no indents and no blank lines; the little stroke in the margin was the whole signal that one part had ended and the next was starting. The word transition has a story too. It comes from the Latin transire, 'to go across'. A transition is literally a crossing, which is exactly what a reader does at every paragraph break. And editors use a check that this section turns into a habit. After a draft is written, they note in the margin the one point each paragraph makes, then read those notes down the page as a list. Writing teachers often call it a reverse outline. If the list does not lead from one point to the next, no linking word in the draft will make it lead; the order has to change first. If it does lead, the openers only have to say so.

Counting linkers are honest when the points really are a list of equals, such as the steps of a method or three separate reasons in a short answer. The trouble starts when the points build on one another and the numbers hide how. Not every gap needs a linking word. A story told in time order, or an opener whose words already name the last point, carries the relation without one. A content link is a few words, not a summary. Retelling the last paragraph slows the piece and makes the reader read it twice. Links cannot rescue a weak outline. If two paragraphs do not belong in the same piece, no linker will join them; go back to the outline from section 22 first.

  1. Write the thesis at the top of the page before you link anything.
  2. At each paragraph gap, say in a few words what the last paragraph said.
  3. Name the relation to the next paragraph with the Bridge Test from Chapter 21 before choosing any words.
  4. Open the new paragraph with a short phrase that picks up the last point.
  5. Check that every opener still points at the thesis, not only at the paragraph before.
  6. Use Firstly and Secondly only when the points really are a list of equals.
  7. Leave the linker out when time order or the opener's own words already carry the relation.
  8. Give a turn against the thesis a contrast link, and give the return a link that names the objection it answers.
  9. Let the conclusion gather the chain and restate the thesis in new words.
  10. Copy the thesis and every first sentence onto one slip, read it aloud, and revise any gap that does not lead.
Linking paragraphs across a whole piece. At every paragraph gap, ask three questions: 1. What did the last paragraph say, in a few words? 2. What is the relation to this one? Name it with the Bridge Test from Chapter 21. 3. Does this opener still point at the thesis? A linker names a relation; a content link names the idea. 'Those meetings matter most to the people who live alone' links without a single linking word. Three weak links to watch for. Counting: Firstly, Secondly, Thirdly, when the points build on one another. Drift: each opener links to the one before, but the piece walks away from the thesis. Restarting: each opener repeats the thesis and ignores the paragraph before. Leave the linker out when time order or the opener's own words already carry the relation, and add a time link when a story jumps out of order. Test the whole chain by reading the thesis and every first sentence on one slip. It should sound like a short version of the piece. Linking is the second half of the Draft stage of the Whole-Piece Check. Next: revising for voice and rhythm.

Revising for voice and rhythm

Revising for voice and rhythm is two passes over the whole draft, each with a tool you already own. Neither pass starts at the first sentence and polishes downwards.

Two passes, whole piece first.

1. The voice pass - the Voice Check from Chapter 28. Name the piece's voice in a few words: its diction, sentence pattern, imagery and persona. Then mark every line that does not sound like that voice.
2. The rhythm pass - the Read-Aloud Test from Chapter 24. Read each paragraph aloud. Mark any stretch that drags or rushes against what its moment needs, and any ending where the weight lands on the wrong word.

Then change only what you marked, in the piece's own voice.
In Asha's draftThe pass that finds itWhat it hearsThe revision
The stall constitutes an indispensable nexus of communal discourse.voice: dictiona formal, abstract line in a plain, warm pieceHalf the depot stops there to talk.
Three long, winding sentences about four buses arriving at oncerhythm: drag or rushthe paragraph crawls through a moment that racesshort sentences: Four buses. Forty people. One kettle.
My grandfather opens his stall at five.voiceplain, warm, first person: it already fitskeep it
Paragraph 4 ends: and that is basically what it is like most of the time.rhythm: where the weight landsthe paragraph fades on its weakest wordsend on the image: and the kettle is already on.

Notice that the third row is a line the pass keeps. A whole-piece revision is mostly reading, and only partly changing. Most lines in a good draft already sound like the writer.

This pass is the first half of the Revise stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread: the draft is read whole for voice and rhythm before it is read for punctuation and length, and every change is judged by whether the piece still serves its one governing idea.

Voice and rhythm are separate passes, but a change made in one is heard in the other. Here is how four tempting revisions of Asha's draft fare once both passes are run.

The tempting revisionWhat it fixesWhat it breaksThe whole-piece verdict
Join four short rush sentences with 'whereupon' and a semicolonnothing: the rush needed short sentencesthe plain voice, and the speedreject: it fixes neither pass
Break a long, plain sentence about the quiet afternoon into three short onesnothing: the afternoon is slowthe slow pace the moment needsreject: the length was right
Change 'Half the depot stops there to talk' to 'A large proportion of commuters converse there'nothingthe voice it had just recoveredreject
Move the kettle to the end of the last sentencethe weight at the end of the piecenothing: same words, same voicekeep

Notice three things. First, run the voice pass before the rhythm pass. The voice pass changes words; the rhythm pass changes lengths and order. If you tune the rhythm first, a voice revision can undo it, and you have to listen again.

Second, a rhythm fix must be made in the piece's own voice. Joining two short sentences with 'whereupon' fixes nothing in a plain essay: it raises the diction and slows the rush. Join with the words the writer already uses - and, so, but - or split with a full stop.

Third, a line can sound wrong on its own and be right in the piece. Four words in a row - 'Four buses. Forty people.' - look like a fault if you read them alone. In a paragraph about the six o'clock rush, they are the point. The test is always the voice of the whole piece and the pace of the moment, never the sentence by itself.

Picture tuning a band before a concert, not one instrument in an empty room. A violin tuned perfectly to itself can still clash with the rest of the band. The voice pass checks that every instrument is in the same key; the rhythm pass checks that the band speeds up and slows down together, where the music asks. You only retune the instruments that are off, and you only know which ones are off by listening to the whole band play.

Section 23 linked the paragraphs of a draft into one chain. This section is the first revision pass over that linked draft: voice and rhythm, heard across the whole piece. Section 25 runs the second pass, for punctuation and length. The complete essays, letter and report of sections 26 to 29 use both passes in turn, and section 30 fits them into a limited time. In Master, section 33 revises a full draft more boldly, cutting and reordering, and section 32 holds one voice across a whole reflective essay.

Here is a hard one, and it is worth slowing down for: the line that breaks the voice on purpose. Nandini's story is told in a calm, plain, close first person. In the middle, her grandmother speaks: 'Beta, the rice will not cook itself while you stand there dreaming.' That line has different diction, a different rhythm and a different persona. The voice pass will mark it. But it is speech, and it is meant to sound like the grandmother, not the narrator. A voice pass checks the narrating voice; a character's speech is judged by whether it sounds like that character. Keep it. Here is a second hard one: the deliberate shift at the turning point. Faisal's account of a road accident he saw begins in long, steady sentences on a slow Sunday morning. At the moment of the crash, the sentences become short and bare: 'A horn. Glass. Then nothing.' After that, they lengthen again. The rhythm pass hears a sudden change, and so does the voice pass. Both are deliberate: the shift is the moment. The test is whether the shift serves the piece's governing idea and returns afterwards. If it does, it is a choice, not a slip. And a third: when the whole voice is wrong. Harry's application letter is chatty from first line to last. The voice pass finds nothing to mark, because nothing breaks the voice; every line is equally chatty. A whole-piece pass for consistency cannot catch a voice that is consistently wrong for its reader. That is a Plan decision, made with the Reader Test from Chapter 22, and it comes before line revision, not during it.

The word persona, now used for the character a writer speaks as, began on the stage. In Latin, persona was the word for the mask an actor wore in a play, and later for the character the actor played. When Chapter 28 asks who is speaking in a piece, it is asking which mask the writer has put on. A voice pass is a check that the writer has kept the same mask on from the first line to the last, unless a change of mask is the point. The word rhythm has a long history too. It comes from the Greek rhythmos, a word for measured movement or flow. It was used for music and dance before it was used for writing, which is why rhythm in prose is so often heard rather than seen. And here is a working habit you can borrow. Many writers and editors read a draft aloud before they call it finished, or listen to someone else read it to them, because the ear catches a stumble that the eye slides straight over. Many also revise in separate read-throughs, each listening for one thing, rather than trying to catch everything at once. This section's two passes are that habit, given a name.

A piece can change voice on purpose: a character's speech, a quoted letter, or a deliberate shift at the turning point. The pass marks it; you then decide whether it serves the piece. Varied rhythm does not mean every paragraph must mix long and short sentences. A slow moment may want only long ones, and a race may want only short ones. The two passes judge consistency and pace, not whether the voice suits the reader. A voice that is wrong for the task is fixed at the Plan stage, with the Reader Test from Chapter 22. A whole-piece pass finds the few lines that need changing. If it marks almost every line, the draft may need to be rewritten from its plan rather than revised.

  1. Read the whole draft through once before you change a single word.
  2. Write the piece's voice at the top in a few words, using the marks of the Voice Check from Chapter 28.
  3. On the first pass, mark only the lines that do not sound like that voice.
  4. On the second pass, read each paragraph aloud with the Read-Aloud Test from Chapter 24.
  5. Mark stretches that drag or rush against what their moment needs.
  6. Check where the weight lands at the end of each paragraph and of the whole piece.
  7. Run the voice pass before the rhythm pass, so word changes do not undo rhythm changes.
  8. Make every fix in the piece's own words: join with and, so or but, or split with a full stop.
  9. Keep deliberate effects: a character's speech, a run of short sentences at a sudden moment, a slow stretch where the scene is slow.
  10. Read the whole piece once more after revising, and ask whether it still serves its one governing idea.
Revising for voice and rhythm. Two passes, over the whole draft, before changing a word: 1. The voice pass, with the Voice Check from Chapter 28. Name the voice in a few words, then mark every line that does not sound like it. 'The stall constitutes an indispensable nexus of communal discourse' breaks a plain, warm voice. 2. The rhythm pass, with the Read-Aloud Test from Chapter 24. Read each paragraph aloud and mark any stretch that drags or rushes against its moment, and any ending where the weight lands wrongly. Three long sentences about four buses arriving at once crawl through a moment that races. Then change only what you marked, in the piece's own voice. The key contrasts. Polishing each sentence alone is not revising the piece: Kunal's grand opening became a new voice break. A fix for rhythm must keep the voice: whereupon does not belong in a plain essay. A line that sounds wrong alone can be right in the piece: Four buses. Forty people. Keep deliberate effects, and send a voice that is wrong for its reader back to the Plan stage. This is the first half of the Revise stage of the Whole-Piece Check. Next: revising for punctuation and length.

Revising for punctuation and length

The punctuation-and-length pass is one revision, run across the whole draft, in a fixed order: cut first, then choose the marks. There is no point polishing the punctuation of a sentence you are about to cut, and a cut often changes which marks the sentences around it need.

One pass, two questions, asked of the whole piece.

Cut (the Cut Test, Chapter 20): cover it. Is any meaning, tone or emphasis lost? Ask it of phrases, of sentences, and of whole paragraphs, remembering what the rest of the piece already says.

Marks (the Effect Test, Chapter 25): name the effect, check the rule, try two, read aloud, keep one. Then ask the whole-piece question: has this effect already been spent somewhere else?
In Farhan's draftThe whole-piece problemThe revision
As I said at the start of this article, the banyan tree by our gate is going.Repeats paragraph 1 almost word for wordCut it: the reader already knows
In this paragraph I am going to explain why the tree has to go.Announces instead of sayingCut to: The tree has to go because its roots have cracked the wall.
The roots have cracked the wall - the council says it is unsafe!The second dash and the first exclamation mark in a quiet pieceTwo full stops: The roots have cracked the wall. The council says it is unsafe.
Not everyone on our road agrees.Short, and looks cuttable, but it turns the piece towards the neighbours who want the tree keptKeep it: covering it loses the turn
I will miss it - the shade, the parrots, the roots we sat on after school.The one moment the piece turnsKeep the dash: it is the only one left

This pass belongs to the Revise stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread: after it, the draft should need only the Four-Pass Check from Chapter 23, and it should still serve the one governing idea it was planned around.

Both tests change when you ask them of a whole piece instead of one sentence. Here are six decisions from three different drafts.

What the draft hasAsked of the whole pieceDecision
In conclusion, as this essay has clearly shown,The last paragraph's place already tells the reader it is the endCut the phrase
For example, the stall sold out by ten on Saturday.Covering it leaves paragraph 2's point with no evidenceKeep it
Then, after that, next,Three linkers naming one relation of timeKeep one: Then
Grandma's laugh, loud enough to stop the cooker whistle,The voice pass chose this image; covering it loses toneKeep it
Three dashes in a 200-word pieceThe pause is spent before the moment that needs itKeep the dash at the turn; full stops elsewhere
A colon before the answer in the final lineThe one reveal the whole piece builds towardsKeep it

Notice three things. First, the Cut Test reads the whole piece. 'The banyan is very old' may pass on its own. But if paragraph 1 has already said 'The banyan is over a hundred years old', covering the second sentence loses nothing at all. A sentence is wordy or not relative to the piece it sits in, which is why sentence-by-sentence trimming misses the biggest cuts.

Second, the Effect Test reads the whole piece too. A dash makes a pause the reader feels. The first time, they feel it; the fourth time, they stop noticing. So the whole-piece question is: where is the moment this piece turns, and does the strongest mark sit there? In a short piece it is usually one place, and often near the end.

Third, the two halves of the pass work together. Cutting clears clutter away from a mark: once 'and so, as you can see,' is gone, the full stop before the last short sentence has room to land. And sometimes a cut needs a new mark: two sentences that said one idea twice become one sentence, and a colon or a semicolon may now join what is left. That is why the marks come second.

Picture a film editor with a rough cut that runs twenty minutes too long. First the editor takes out the scenes that repeat what the audience has already seen. Only then does the editor decide where the music swells: once, at the moment the story turns. Music under every scene would mean no scene stands out. Your cuts are the editor's first job; your strongest marks are the music.

Section 23 linked the paragraphs, and section 24 revised the whole draft for voice and rhythm. This section is the second revision pass: punctuation and length, across the whole piece. Together the two passes make up the Revise stage of the Whole-Piece Check. Section 26 runs every stage end to end on a complete expository essay, and section 27 on a persuasive one. In Master, section 33 revises a full draft by reordering as well as cutting, and section 34 proofreads it with the Four-Pass Check.

Here are three cases where careful writers disagree at first, and the tests have to be asked properly. The first is repetition that is a choice. Anjali's magazine piece on saving water ends each of its three paragraphs with 'Every drop counts.' A quick Cut Test says the second and third are already said, so cut them. But covering them loses emphasis: the repeated line is the piece's pattern, chosen in the voice pass (Chapter 28). The Cut Test protects emphasis as well as meaning, so the line stays. If her last paragraph also restates the whole introduction, that restatement still goes. A chosen echo is not the same as saying everything twice. The second is the word limit against the link. Vikram is six words over. The easiest cut is 'In the same way,' at the start of paragraph 3. Covering it, each paragraph still makes sense, but the reader may lose the sign that paragraph 3 works like paragraph 2. Ask the question from Chapter 21: does the road already run? If paragraph 3 opens on the parallel itself ('The canteen has the same problem.'), the linker can go. If not, cut elsewhere: 'at this moment in time' to 'now' saves four words, and 'each and every' to 'every' saves two more. The third is the mark that is correct and still wrong. Meher's report on a flood relief camp reads: 'Volunteers arrived at six; they unloaded rice, blankets and medicine; by noon the camp was ready!' Every mark passes the rule. But a report wants a calm, neutral voice (Chapter 22). The exclamation mark adds a cheer the report does not want, and the semicolons press three steps into one breathless sentence. Try full stops, read both aloud, and keep the calmer version. The rule tells you which marks are allowed; the piece tells you which one to choose.

Did you know that the oldest names for punctuation marks were really names for lengths? The word comma comes from the Greek komma, meaning 'a piece cut off', from a verb meaning 'to cut'. Colon comes from the Greek kolon, 'a limb' or 'a member', a longer section of a sentence. And period, still the American name for a full stop, comes from the Greek periodos, 'a going round': a whole sentence that comes back to its close. Ancient teachers of speaking used these words for the pieces of a spoken sentence, measured by where a speaker paused. Only much later did the words move from the pieces to the marks that end them. So cutting and punctuating began as one art: measuring how long each piece of a sentence should be, and where the reader should pause. Here is a second, more recent example. In the days of the telegram, messages were often charged by the word, so senders learned to cut every word that did not earn its place. Punctuation marks were awkward to send, so many senders spelled out the word STOP where a full stop would go. A telegram was a whole piece revised for length under a strict limit, which is exactly the pass this section teaches, with a price on every word.

Shorter is not always better. A cut that removes a reason, an example or a link makes a piece shorter and worse. No mark is banned and none has a fixed quota. A piece may need two dashes or a run of short sentences; the test is whether each effect earns its place. This pass does not replace proofreading. Fixing a mark that breaks a rule is the job of the Four-Pass Check (Chapter 23), and that check still comes afterwards. A word limit is a ceiling, not a target. A piece that says everything it needs in 210 of 250 words is not unfinished; do not pad it back up.

  1. Run this pass after voice and rhythm and before proofreading.
  2. Read the whole draft once without a pen before you change anything.
  3. Cut first, then choose the marks.
  4. Ask of every paragraph, not just every sentence: does the piece already say this?
  5. Cut announcing openers such as 'In this paragraph I will explain'.
  6. Keep the links, the evidence and the voice: covering them loses meaning, relation or tone.
  7. Circle every dash, colon, semicolon and exclamation mark, and name the effect of each.
  8. Spend the strongest mark once, at the moment the piece turns.
  9. Match every mark to the tone the piece needs, and leave rule-breaking marks for the proofread.
  10. Count the words against any limit, and treat the limit as a ceiling, not a target.
Revising for punctuation and length is one pass across the whole draft. 1. Read the whole draft once without a pen. 2. Cut: use the Cut Test from Chapter 20 on paragraphs, sentences and phrases, asking what the piece already says. 3. Keep what fails to cut: links, evidence and the voice the last pass chose. 4. Mark: use the Effect Test from Chapter 25 on every dash, colon, semicolon and exclamation mark. 5. Spend the strongest mark once, where the piece turns, and match every mark to the piece's tone. 6. Count the words against any limit, then hand the draft to the proofread pass. The key contrasts. Wordy in a sentence is not the same as wordy in a piece: a sentence that repeats another paragraph fails the Cut Test even when it reads well alone. A chosen echo is emphasis, not repetition. A mark used once is an effect; a mark used everywhere is a habit. A choice between two correct marks is revision; a mark that breaks a rule is proofreading. And a word limit is a ceiling, not a target. Next: a complete expository essay.

A complete expository essay

An expository essay explains: how something works, why something happens, or what something is made of. Its whole shape follows from one sentence, the thesis, which states what the essay will explain and names the parts it will explain it in.

The expository essay, stage by stage.

Plan: Reader. Purpose: to explain. Thesis: what will be explained, in its parts. One line per body paragraph, in order.
Draft: an introduction that ends on the thesis; one body paragraph per part, each opening on its topic sentence and linked to the one before; a conclusion that pulls the parts together.
Revise: voice and rhythm (section 24), then punctuation and length (section 25).
Proofread: the Four-Pass Check (Chapter 23), across the whole essay.
PartIn Anaya's essayWhat it does
IntroductionEvery monsoon, the road by the station turns into a river. Our city floods for three connected reasons ...Opens on a scene the reader knows, then states the thesis and names its three parts
Body 1Rain falls faster than paved ground can take it.The first cause, the one the reader can see: why the water collects
Body 2Even water that reaches a drain may not get away.The second cause, linked: it names how the drains add to the first
Body 3Where the water once had somewhere to go, there are now buildings.The third cause, the deepest: the lakes that were filled in
ConclusionFast rain, choked drains and lost lakes are one problem: the water has nowhere to go.Pulls the parts into one picture; adds no new cause and no blame

Writing an expository essay is the Whole-Piece Check run end to end, this chapter's four stages of Plan, Draft, Revise and Proofread, followed by its last question: does every paragraph still explain the one thing the thesis promised?

Most weak expository essays fail at the thesis, before a single body paragraph is written. Here are five drafts of a thesis, and what each would do to the essay.

Thesis draftWhat goes wrongThe revision
Our city floods every monsoon.A fact: it says what happens, not what the essay will explainOur city floods each monsoon for three connected reasons: ...
In this essay I will write about floods.An announcement: nothing the reader can hold the paragraphs againstName what will be explained, and its parts
The council must stop the floods now.A claim to defend: it belongs to a persuasive essay (section 27)Explain the reasons; leave the must for an argument
Floods have many causes, effects, histories and solutions.Too big: four essays in one sentenceChoose one: the causes
Recycled paper goes through four stages: collecting, sorting, pulping and pressing.Nothing: a process thesis with its parts in time orderKeep it

Notice three things. First, the thesis decides the body. Three causes means three body paragraphs, in the order the thesis names them. If the thesis names drains first and the body begins with lakes, the reader is lost before paragraph 2 ends. If the body has a fourth paragraph the thesis never promised, either the thesis or the body must change.

Second, the order must have a reason the reader can feel. A process goes in time order: paper is collected, sorted, pulped and pressed, and no other order makes sense. Causes often go from the most visible to the deepest: rain on concrete, then drains, then the lost lakes. Kinds and features often go from simple to complex. Choose one principle, say it in the thesis, and keep it.

Third, an explanation is not an argument. Anaya's essay may say that building over the lakes made the floods worse; that is a cause, and explaining it is the essay's job. It does not say who should be punished or what the council must do; that is a claim, and it needs the Argument Check from Chapter 29 and a different outline. Keeping the purpose plain - to explain - keeps the voice neutral (Chapter 22) and keeps every paragraph on its part.

Picture a museum guide leading a group through one exhibition of three rooms. At the door, the guide says what the exhibition will show and names the three rooms. In each room, the guide explains one thing only. Walking along each corridor, the guide says how the next room connects to the last. At the exit, the guide does not open a fourth room; the guide says what the three rooms showed together. Thesis, body paragraphs, links and conclusion are the door, the rooms, the corridors and the exit.

Sections 21 to 25 gave you each tool on its own: the thesis, the outline that fits a genre, the links across a whole piece, and the two revision passes. This section runs them all, in order, on one complete expository essay. Section 27 does the same for a persuasive essay, using the outline from Chapter 29, and sections 28 and 29 do it for a formal letter and a report. Section 30 fits all of it into a fixed time. In Master, section 31 runs the Whole-Piece Check in full on a demanding piece.

Here are three decisions that separate a competent expository essay from a strong one. The first is where explanation shades into argument. In the flood essay, 'Building over the lakes left the rain with nowhere to go' explains a cause. 'The builders should never have been allowed' judges it, and that is argument. A useful test: could a reader who disagrees about what should be done still accept the sentence? If yes, it explains. An expository essay may even report a debate - 'Some engineers favour wider drains; others favour restoring the lakes' - without choosing a side. Explaining a disagreement is still explaining; joining it is not. The second is choosing between two honest orders. Faisal is explaining how a lunch box travels from a home to an office in his city. He could order the body by stage (collection, sorting, the train, delivery) or by person (the collector, the sorter, the carrier). Both can work. But the thesis must promise one, and the body must keep it: a paragraph on the morning collection followed by a paragraph on 'the sorters' sends the reader back and forth. Ask what this reader most needs. Someone curious how one box finds one desk needs the stages, in time order. The third is the difference between a conclusion that lists and one that connects. 'In conclusion, there are three causes: rain, drains and lakes.' repeats the thesis, and the Cut Test from section 25 would remove it. 'Fast rain, choked drains and lost lakes are one problem seen three ways: the water has nowhere to go.' adds no new part, yet it shows how the parts fit. That is what a conclusion is for. The list was the thesis; the connection is the reason the reader read the essay.

Did you know that the word essay began as a word for a try? It comes from the French essai, 'an attempt' or 'a trial', from the verb essayer, 'to try'. In 1580 the French writer Michel de Montaigne published a book of short prose pieces called Essais: attempts to think a question through on paper. Every school essay since carries that name, and every one is still an attempt to make a reader understand. Expository comes from the Latin exponere, 'to set out' or 'to put forth', from ex, 'out', and ponere, 'to place'. To expound something is to lay it out in front of the reader, part by part, the way a seller at a market sets out goods on a cloth so that each one can be seen. And a third. Paragraph comes from the Greek paragraphos, from para, 'beside', and graphein, 'to write'. At first it was not a block of text at all, but a mark written beside a line to show where a new part began. A body paragraph is still exactly that: the place where one part of the explanation begins.

Not every expository essay has three body paragraphs. The thesis decides: two causes need two paragraphs; five stages may need five, or may be grouped into three. Expository does not mean dull. A scene or a question can open the essay, and the voice can be lively, as long as the purpose stays explaining. An essay can report a debate without arguing it. Saying that people disagree is explaining; choosing a side turns it into a persuasive essay. The plan can change while you draft. If a better part appears, change the thesis first, then the body lines, and check them against each other again.

  1. Write Reader and Purpose (to explain) before anything else.
  2. Write a thesis that states what the essay will explain and names its parts.
  3. Choose a topic small enough for the paragraphs you have time to write.
  4. Give each part one body paragraph, and give each body paragraph one part.
  5. Choose an order with a reason: time, visible to deep, or simple to complex.
  6. Open each body paragraph on a topic sentence that names its part.
  7. Begin each body paragraph after the first with a link that names how it connects.
  8. Keep the purpose to explain: report a debate if you must, but do not join it.
  9. End with a conclusion that shows how the parts fit, adding no new part.
  10. Revise with both passes, proofread the whole essay, then check it against the thesis.
A complete expository essay is the Whole-Piece Check, run end to end. 1. Plan: Reader, Purpose (to explain), Thesis, and one line per body paragraph in order. 2. Thesis: what the essay will explain, in its parts. Not a fact, not an announcement, not a claim. 3. Draft the introduction: a scene or question the reader knows, ending on the thesis. 4. Draft the body: one part per paragraph, each opening on its topic sentence and linked to the one before. 5. Draft the conclusion: show how the parts make one picture; add no new part. 6. Revise for voice and rhythm, then for punctuation and length. 7. Proofread with the Four-Pass Check, then ask: does every paragraph still explain what the thesis promised? The key contrasts. Everything you know is not an explanation. A thesis that names parts governs; a fact does not. Body paragraphs follow parts, not sources. Explaining a cause is not arguing a claim. A conclusion that lists repeats; a conclusion that connects completes. Next: a complete persuasive essay.

A complete persuasive essay

A persuasive essay is the Whole-Piece Check with Chapter 29's argumentative outline as its plan. The outline says what goes in each paragraph; the four stages say when you do each job.

One persuasive essay, four stages.

Plan: the reader, the purpose and one debatable thesis; then Chapter 29's outline - introduction, reasons with evidence and warrant, counter and answer, conclusion.

Draft: one paragraph per outline line; each reason paragraph opens with its reason; each paragraph linked to the one before.

Revise: the Argument Check across the whole draft, then voice, rhythm and the Cut Test.

Proofread: the Four-Pass Check, first line to last.
StageWhat Pooja does in her canteen essayThe slip it prevents
PlanThesis: the canteen should switch to steel plates and cups from next term. One line per paragraph.Seven reasons and no direction
Draft: introductionOpens on the overflowing bins behind the canteen and ends on the thesis.An opening about "the world today"
Draft: reason paragraphs"The first reason is the waste we can see." "Cost, surprisingly, is the second reason."Reasons with no evidence under them
Draft: counter paragraph"Some will ask who will wash four hundred plates a day." Then the answer.Silence about the obvious objection
Draft: conclusionRestates the thesis and its two reasons; adds no new claim."So plastic is bad." or a bigger claim than the reasons reached
ReviseRuns the Argument Check on the whole draft; cuts a paragraph about the town's plastic-bag rule, which the thesis does not need.Polishing sentences inside a weak argument
ProofreadRuns the Four-Pass Check on every paragraph, the last as carefully as the first.Slips in the conclusion, the part a reader meets last

Each row is a stage of the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread; in a persuasive essay the plan is Chapter 29's outline and the revision includes Chapter 29's Argument Check, so the final question becomes: does every paragraph still argue for the one thesis?

The outline is only useful if each drafted paragraph actually does its line's job. The test is the paragraph's opening: does it announce the reason, the objection or the conclusion that the outline gave it?

Outline lineAn opening that serves itAn opening that drifts
Reason 1: the bins overflowThe first reason is the waste we can see: by the end of lunch, the bins behind the canteen are full.Plastic is used almost everywhere in modern life.
Reason 2: costCost, which sounds like an argument against steel, is in fact a second reason for it.Steel is a strong metal used to build bridges.
Counter and answerSome will ask who will wash four hundred plates a day. It is a fair question.Some people will never agree, and they are simply wrong.
ConclusionA switch that cuts our waste and pays for itself within a year is worth making next term.One day, perhaps, the whole world will stop using plastic.

Notice three things. First, each opening that serves its line is also a link. "The first reason", "a second reason" and "Some will ask" tell the reader where they stand in the argument: the Bridge Test from Chapter 21, naming the relation (adding, then contrast) between one paragraph and the next.

Second, the drifting openings are not false. Plastic is used almost everywhere, and steel does build bridges. They drift because they are true about the topic area, not about the thesis. A persuasive essay keeps only the sentences that argue for its one claim.

Third, the drifting conclusion makes a bigger claim than the essay earned. Two reasons about one canteen cannot show anything about the whole world. A persuasive conclusion widens only as far as its evidence reached, and the Argument Check from Chapter 29, run in the Revise stage, is what catches it when it does not.

Picture a bridge you are building across a river to a doubtful reader on the far bank. The thesis is the point where the bridge lands. Each reason paragraph is a pillar, its evidence is the concrete inside it and its warrant is the bolt that fixes pillar to deck. The counter paragraph is an inspector who tests the weakest pillar out loud and shows that it holds. Seven thin poles do not make a bridge; three solid pillars in a line do. You draw the bridge before you pour any concrete, and you inspect it before anyone walks across.

Section 21 gave you the thesis statement, section 22 the outline shapes for each genre, section 23 the links between paragraphs, and sections 24 and 25 the two revision passes. Section 26 put them together for an expository essay, which explains. This section does the same for an essay that argues, using the outline Chapter 29 built for it. The one addition is a stage of judgement: in Revise, the Argument Check runs across the whole draft. Section 28 turns to a formal letter, a piece whose format is fixed before you write a word, and section 30 fits all four stages into a clock.

Three cases here take real judgement. The first is a strong reason for a slightly different thesis. Pooja's draft has a paragraph on plastic in the sea harming fish and birds. It is persuasive, and it is about plastic. But her thesis is about one canteen. Run the warrant: can she write the step from "plastic harms sea life" to "this canteen should switch to steel"? Only by assuming that this canteen's plates reach the sea, which she cannot show. So the paragraph argues a bigger, general claim. The honest move is to cut it, or to fold one sentence of it into the introduction as background, not to keep it as a reason. The second is where to put the counter paragraph. One draft answers the objection straight after the introduction, clearing the ground before the reasons. Another places it just before the conclusion, so the reader meets the doubt only after the case has been made. Both are sound shapes. What matters is that the objection is the strongest one, that it is answered, and that it is linked to its neighbours. The outline decides the place before drafting; moving it later means relinking both paragraphs on either side. The third is when revision changes the thesis. While revising, Pooja finds that her cost reason only holds if steel is used for lunch, not for the breakfast snacks sold in paper bags. Her thesis covered every plate and cup the canteen uses. The honest move is to narrow the thesis to lunch, then check every paragraph against the new wording, the conclusion included. A change at the top is a change everywhere; that is exactly the question the Whole-Piece Check ends on.

The word essay comes from the French essayer, "to try". When the French writer Michel de Montaigne published his collection in 1580, he called the pieces Essais: trials, attempts. An essay was first of all an attempt to think something through on paper. Persuade comes from the Latin persuadere: per, thoroughly, and suadere, to advise or urge. The same root gives us dissuade, to urge someone away from something. To persuade, then, is not to overpower a reader. It is to advise them so thoroughly that they come round. And here is a practice from competitive debating. In many formal debating formats, speakers do not choose their side. The side is assigned, often shortly before the debate begins. So debaters learn to build the strongest case for a position they might never have picked, and to imagine the strongest attack on a position they hold. That is exactly the work a counter-argument paragraph asks of you. Before you answer the other side, you must first state it as well as its best supporter would. An essay that does this is both a fair attempt and a thorough piece of advice, which is what the two old words meant all along.

The outline is a plan, not a cage. Two reasons or three, the counter before the reasons or after them: the shape can bend, but each of its jobs must still be done. Some tasks ask you to discuss both sides rather than argue for one. Then the outline changes, and the conclusion weighs the two sides instead of arguing for one. Read the task before you plan. A short timed answer may fold the counter into the last reason paragraph instead of giving it its own. The objection must still be named and answered. The Argument Check in Revise tells you whether the argument holds; it cannot make a weak position strong. Sometimes an honest revision ends with a narrower thesis than the one you started with.

  1. Read the task twice and decide your side before you plan anything else.
  2. Write the thesis as one clear, specific, debatable sentence.
  3. Outline one line per paragraph in the order of Chapter 29's outline.
  4. Keep two or three reasons, each with evidence and a warrant, rather than a long list.
  5. Choose the strongest objection a doubtful reader would raise, not the weakest.
  6. Open each reason paragraph with its reason, worded as a link to the paragraph before.
  7. Hold every paragraph's opening next to the thesis: is it a reason for this exact claim?
  8. Write a conclusion that restates the thesis and its reasons and claims no further.
  9. Revise in order: the Argument Check first, then voice, rhythm and the Cut Test.
  10. Proofread the last paragraph as carefully as the first, then ask whether every paragraph still argues for the thesis.
A complete persuasive essay is the Whole-Piece Check with Chapter 29's outline as its plan. 1. Plan: choose one side and write one debatable thesis. Outline the introduction, two or three reasons each with evidence and a warrant, the counter-argument and its answer, and the conclusion. 2. Draft: one paragraph per outline line. Each reason paragraph opens with its reason, and each opening links to the paragraph before it. 3. Revise: run the Argument Check across the whole draft, then revise for voice, rhythm and length. 4. Proofread: the Four-Pass Check, first paragraph to last. Then ask whether every paragraph still argues for the thesis. The key contrasts. More reasons is not more persuasive: two developed reasons beat seven bare ones. A true paragraph about the topic area is not a reason for this thesis. "Some people disagree, but they are wrong" is not a counter-argument. A conclusion restates; it does not grow the claim or stop mid-reason. Next: A formal letter or application.

A formal letter or application

A formal letter is the Whole-Piece Check run inside a format that Chapter 22 already teaches. The format gives the letter its parts; the four stages make sure every part serves one purpose.

One formal letter, four stages.

Plan: the reader and their designation, the one purpose, the format the task or school states; three lines - subject, details, request.

Draft: the format's parts in order; the body in three jobs - purpose, details, request; a close that matches the salutation.

Revise: the Reader Test (Chapter 22) for register throughout; the Cut Test (Chapter 20) for anything the reader does not need.

Proofread: the Four-Pass Check on the words, then a format check: names, dates, subject, close, signature.
Part, in a common orderWhat it does in Siddharth's letterThe slip it prevents
Sender's address and dateTells the principal who is writing and when.A letter nobody can reply to
RecipientThe Principal, with the school's name and address.A letter with no clear reader
Subject lineRequest for permission to hold a book sale in the school hall on 14 December.A purpose the reader must hunt for
SalutationDear Sir/Madam, or the form the school's format states.A greeting that does not suit the reader
Purpose paragraphStates the request in the first two sentences.Three sentences of warm-up first
Details paragraphTime, stalls, supervising teacher, where the money goes.A request the reader cannot act on
Request paragraphAsks for permission and a reply by a stated date.An ending that only says "Thanking you"
Close and signatureYours faithfully, then name and class.A close that does not match the salutation

The parts come from the format, but the order of work comes from the Whole-Piece Check, this chapter's four steps of Plan, Draft, Revise and Proofread; in a letter, the one governing idea is its purpose, named in the subject line, so the final question becomes: does every part of the letter help this reader act on that one purpose?

Formal letters come in many kinds, but the body of almost every one does the same three jobs. What changes is the purpose, and so what the details must give.

Kind of letterThe one purposeWhat the details must giveThe request
Application for leaveto be excused from school for stated daysthe dates, the reason, how missed work will be made upgrant leave for those dates
Complaint to a local officeto get a fault repairedwhere, since when, earlier reports, the effect on residentsrepair it by a reasonable date
Application for a post or placeto be chosen for itwhat the writer has done that fits the post, and when they are freeconsider the application; invite to an interview or trial
Letter to the editorto raise a public issuethe facts of the issue and why it matters to readerspublish the letter; urge a stated action

Notice three things. First, the subject line is the governing idea in a few words. Like a thesis, it decides what belongs: every sentence in the body should help the reader act on it, and a sentence that does not, however true, is off the plan.

Second, the details are what the reader needs to act, not what the writer feels. Dates, places, names, what has already been tried and what will happen next all help. A grandfather's love of books, however sincere, does not help a principal decide whether the hall is free on 14 December.

Third, the request is concrete: what you want done, and by when. "Thanking you" is a courteous close in many Indian letters, but it is not a request. The reader should finish the letter knowing exactly what they are being asked to do.

Picture a parcel sent to a busy office. Before anyone opens it, the label tells the clerk who sent it, when, who it is for and what is inside: that is the sender, the date, the recipient and the subject line. When it is opened, the first thing on top is a note saying what the parcel is for. Underneath, the contents are packed in order. At the bottom is a slip that says what to do and by when. A clerk can act on that parcel in a minute.

Section 27 built a persuasive essay for a reader you may never meet. A formal letter often persuades too, but it writes to one named reader, for one purpose, inside a format that is fixed before you begin: Chapter 22 taught its parts, its openings and closings and its register. This section adds nothing to that format. It assembles a whole letter inside it with the Whole-Piece Check. Section 29 turns to a report, a neutral factual piece with its own shape, and section 30 fits all four stages into a clock.

Three cases here take judgement rather than rules. The first is the letter with two requests. Imran writes to the ward office about a broken streetlight on Lane 4. While drafting, he wants to add that garbage on the same lane has not been collected for a week. Is that two purposes? Ask two questions: will the same reader act on both, and can one honest subject line name both? Here the answer is yes - "Complaint about a broken streetlight and uncollected garbage on Lane 4" - so one letter can carry both, with a details paragraph for each. But an application to the principal for leave and a request to join the football team go to different decisions, so they need two letters. The second is the true detail that does not help the reader act. Anjali applies for the post of student editor of the school magazine and wants to describe her grandmother, who once worked on a newspaper. One sentence may serve, if it shows why she cares about the work. A whole paragraph of family history does not: the teacher choosing an editor needs to know what Anjali has written, edited and organised. The test is not whether a detail is true or moving, but whether it helps this reader decide. The third is the complaint that must be firm without being rude. Joseph's first draft is angry. If he revises the tone first, he may soften the facts and the deadline away along with the anger. So revise in order: first check that the facts and the request are complete, then revise the register with the Reader Test. A calm letter that names the date of every earlier complaint and asks for collection by Friday is far firmer than an angry one.

The short "Re:" that sometimes stands in front of a subject line is not short for "regarding". It comes from the Latin phrase in re, "in the matter of". Lawyers and clerks used it for centuries to say what a document was about, and it survives in letters and emails today. The "cc" in an email has an older history still. It stands for carbon copy. Before photocopiers, a typist slipped a sheet of carbon paper between two sheets of paper, so that every key struck on the top sheet printed a copy on the one below. The copy is gone; the two letters remain. And salutation, the formal name for the greeting, comes from the Latin salutatio, a greeting, from salus, health or safety. Roman letter writers so often began by wishing the reader well that a common opening, si vales, bene est, ego valeo - "if you are well, that is good; I am well" - was often shortened to its initials. The forms change; the idea does not. Every formal letter, from a Roman tablet to an email, has first had to tell its reader who is writing, what it is about and why it matters to them.

Formats differ between schools, exams and countries: where the date goes, whether lines are indented, whether the salutation is "Dear" or "Respected". The format your school or task states always wins over any general pattern. Emails share the letter's logic - one purpose, the purpose first, a clear request - but have their own shape, with a subject line and no postal addresses. Chapter 22 covers both. Some applications, such as a job or a course application, travel with a form or a CV. The letter is then a covering letter: shorter, but still one purpose, its details and a request. Formal does not mean cold. One courteous sentence of thanks or context is welcome, as long as it does not delay the purpose.

  1. Name the reader and their designation before you write anything.
  2. Decide the one purpose and write it as a subject line.
  3. Use the format your school or task states, part by part.
  4. State the purpose in the first one or two sentences of the body.
  5. Give every detail the reader needs to act: dates, places, names, what has been tried.
  6. Leave out details that are true but do not help the reader decide.
  7. End the body with a concrete request: what, and by when.
  8. Match the close to the salutation, as Chapter 22 teaches.
  9. Revise the whole letter for register with the Reader Test, and cut what the reader does not need.
  10. Proofread the words, then check the format part by part, signature included.
A formal letter or application is the Whole-Piece Check run inside a format Chapter 22 already teaches. 1. Plan: name the reader, decide the one purpose and write it as a subject line, and check the format the task states. Sketch three lines: purpose, details, request. 2. Draft: the format's parts in order - sender, date, recipient, subject, salutation. Then the body: the purpose first, the details the reader needs, a clear request. Then a close that matches the salutation, and your name. 3. Revise: read it as the reader would. Is the purpose clear at once? Could they act today? Does the register hold throughout? Cut what they do not need. 4. Proofread: the Four-Pass Check on the words, then the format, part by part. The key contrasts. Politeness is in the register and the format, not in a long warm-up. True is not the same as useful to the reader. "Thanking you" is a close, not a request. Firm is not the same as angry. Next: A report.

A report

A report carries information back to a reader who asked for it. Its plan is its shape: four parts, each with one job, in the same order every time.

The four parts of a report.

Title: names the subject, plainly.
Introduction: who asked, what the report examines, and how the information was gathered.
Findings: what was found, grouped under headings, each finding with its evidence.
Conclusion or recommendation: what the findings mean, and what should be done, drawn only from the findings.
PartIts one jobFrom Kabir's library report
Titlename the subjectWhy Students Skip the Library at Lunch
Introductionstate the purpose and the methodThe student council asked us to find out why few students use the library at lunch. We asked forty students in Classes 8 to 10 the same three questions.
Finding 1one finding, with its evidenceOpening time: 24 of the 40 said the library opens too late in the break.
Finding 2a second finding, with its evidenceGroup work: 11 of the 40 said they cannot discuss projects there, because talking is not allowed.
Recommendationwhat should be done, drawn from the findingsThe library should open at the start of the break, and one table could be set aside where groups may talk quietly.

In the Whole-Piece Check, a report's governing idea is its purpose: Plan the four parts against it, Draft each finding to its heading, Revise for the neutral tone Chapter 22 teaches and cut what serves no part of the purpose (Chapter 20), then Proofread the whole report, figures included, with the Four-Pass Check from Chapter 23.

Three kinds of sentence appear in a report, and each belongs in one part only. Two other kinds do not belong at all. Mixing them up is the commonest reason a report is not trusted.

The sentenceWhat kind it isWhere it belongs
24 of the 40 students said the library opens too late.a finding, with its evidenceFindings
Most students we asked want the library to open earlier.a summary of the findingsConclusion
The library should open at the start of the break.a recommendationRecommendation
The librarian opens late on purpose because she does not like noise.an opinion, with no evidencenowhere in the report
On Tuesday it was so hot we nearly gave up.part of the story of the surveynowhere in the report

Notice three things. First, a finding always carries its source with it: 24 of the 40 students said. Remove the source, and "the library opens too late" becomes the writer's own opinion. The words that say who said it, or who counted it, are what make a sentence a finding.

Second, the conclusion may only say what the findings have already shown. If the findings are about opening time and group work, a conclusion that suddenly mentions the price of new books is a new idea, and nothing in the report stands behind it.

Third, the story of the survey is not lost; it is shrunk. Three days, two classes and a games lesson become one sentence of method in the introduction: We asked forty students in Classes 8 to 10 the same three questions. That is all the reader needs to know about how the information was gathered.

Picture a doctor's report after a check-up. It does not describe the drive to the clinic or the wait in the corridor. It names the patient, says what was tested and how, lists each result with its measurement, and ends with what to do next, and every piece of advice points back to a result above it. A report is that form: what was asked, how it was checked, what was found, and what follows. Anything else stays in the waiting room.

Section 28 assembled a formal letter: a fixed format written to one reader. This section assembles a report: a fixed format written to carry findings back to a reader who must act on them. The four parts sit on top of what you already have. The thesis statement from section 21 becomes the report's purpose. The genre outline from section 22 becomes its four parts. The paragraph links from section 23 become the logic that ties each recommendation to a finding. Section 30 then asks how to fit all four stages of the work into a limited time.

Here is a hard one, and most report writers find it hard at first: the line between a finding and a conclusion. "24 of the 40 students said the library opens too late" is a finding. "The library opens too late" is a conclusion. Both may be true, but they are different claims. The first reports what people said; the second says the writer agrees with them. A report may reach the second, but only in the conclusion, and only by pointing back to the first. If conclusions appear among the findings, the reader cannot tell where the evidence stops and the writer begins. A second hard case is the unwelcome finding. Nisha's survey for the sports committee found that most students liked the new sports timetable, but 9 of the 30 said the girls' football slot clashed with music lessons. Nisha liked the new timetable herself and was tempted to leave the clash out, because it spoiled the picture. That would not be neutral. A neutral report states every finding that bears on its purpose, the welcome and the unwelcome, in the same plain way. Neutral does not mean soft, and it does not mean selective. A third: should a report recommend at all? Only if the reader asked. If the council asked "Find out why students skip the library", the report may end with a conclusion that sums up the reasons. If the council asked "What should we do about the library?", it needs a recommendation. Reading the request carefully at the Plan stage decides which ending the report has.

The word report carries its own definition inside it. It comes, through Old French, from the Latin reportare, made of re-, meaning back, and portare, meaning to carry. A report is something carried back: the messenger goes out, finds out, and returns with what the person who sent them needs to know. That is why a report begins with who asked. The request is the reason the journey was made. Scientists use a fixed shape for this journey too. Many scientific papers follow a pattern often shortened to IMRaD: Introduction, Methods, Results and Discussion. The introduction says what question was asked; the methods say how it was tested; the results say what was found, kept apart from what the writers think it means; and the discussion says what it means. It is the school report's shape, grown up: the same strict separation of what was found from what the writer makes of it. So when you keep "24 of the 40 said" in your findings and "the library should open earlier" in your recommendation, you are doing exactly what a research team does when it keeps its Results apart from its Discussion.

Report formats vary. A science practical report, a school survey report and a newspaper report use different headings; if a task names a format, follow it, and use Chapter 22 for the format lines and the register. Not every report recommends. If the reader asked only for findings, end with a conclusion that sums them up. Neutral does not mean lifeless. Plain, exact words make a report clear, not dull; Chapter 22 owns how formal to be. A small survey supports only a small conclusion. Forty students in one school cannot tell you what all students everywhere think, and an honest report says so.

  1. Read the request first: who asked, and what do they want to know or decide?
  2. Write the purpose in one sentence before you gather or draft anything.
  3. Give the report a title that names its subject plainly.
  4. Put the method - who was asked or what was counted, and how - in the introduction, in a sentence or two.
  5. Sort the findings by subject, under headings, not by the order of the days.
  6. Attach the evidence to every finding: how many, out of how many, and who said or counted it.
  7. Keep opinion out of the findings; let the conclusion draw meaning from them.
  8. Make every recommendation point back to a finding, and add no new idea at the end.
  9. Revise for the neutral tone Chapter 22 teaches, and cut whatever serves no part of the purpose.
  10. Proofread the whole report, figures included, with the Four-Pass Check.
A report. A report carries information back to a reader who asked for it. It has four parts, in order: 1. Title: names the subject plainly. 2. Introduction: who asked, what the report examines, and how the information was gathered. 3. Findings: grouped under headings, each with its evidence. 4. Conclusion or recommendation: what the findings mean and what should be done, drawn only from them. The key contrasts. A report is sorted by finding, not by time: the story of the survey shrinks to one sentence of method. A finding carries its source ("24 of the 40 said"); an opinion does not. A conclusion says only what the findings have shown; a recommendation points back to a finding. Neutral means every relevant finding, welcome or not, stated plainly. In the Whole-Piece Check, the purpose is the governing idea: plan the four parts, draft each finding to its heading, revise for neutral tone and length, and proofread the whole report. Next: Managing your time while writing.

Managing your time while writing

A limited time does not remove any stage of the Whole-Piece Check. It shrinks each one. The skill is deciding, before the first sentence, how much of the clock each stage gets, and what each stage must still do in its smaller slice.

Divide the clock first.

Read the task and plan: a short, fixed slice at the start, roughly an eighth of the time.
Draft: most of the time, with one checkpoint at the halfway mark.
Revise: a protected slice near the end, for the biggest fixes only.
Proofread: the last few minutes, across the whole piece.

In forty minutes, that might be 5 to plan, 28 to draft, 5 to revise and 2 to proofread.
StageIn 40 minutesWhat it must still doWhat it can drop under pressure
Planabout 5 minutesread the task; state the thesis; list one line per paragraph, in orderfull sentences, a decorated mind map, a polished title
Draftabout 28 minuteswrite each paragraph to its line, linked to the one before; reach the conclusiona perfect first paragraph; a third example where two will do
Reviseabout 5 minutescut what wanders; add a missing link; check the thesis still matches the bodyrewriting whole paragraphs
Proofreadabout 2 minutesrun the Four-Pass Check quickly across every paragraphcopying the piece out again

This is the Whole-Piece Check under a clock: Plan, Draft, Revise and Proofread all still happen, each given the share of time it needs, and the final question - does it still serve the one governing idea? - is still asked, even if there is only half a minute to ask it.

Checkpoints tell you early that you are behind, while there is still time to adjust. The harder question is what to give up. The answer is the same every time: give up length before you give up a stage.

The situationThe tempting moveThe better move
Halfway through the time, still on the first body paragraphwrite faster and hopedrop one planned body paragraph and move on to the next line of the plan
Ten minutes left, two body paragraphs unwrittenwrite both, then a conclusion if there is timewrite one short body paragraph, then the conclusion
Three minutes left, no conclusion yetkeep adding to the last body paragraphwrite a two-sentence conclusion that returns to the thesis
Two minutes left after the conclusionstart an extra paragraphproofread the thesis, the conclusion and anything written in a hurry
Ahead of timehand in earlyuse the spare minutes to revise and proofread again

Notice three things. First, every better move protects the end of the piece. A shorter essay with a conclusion is a finished piece; a longer one that stops mid-sentence is stopped, not finished.

Second, the plan is what makes cutting possible. Meera could drop a paragraph in seconds, because each paragraph was one line in her plan and she could see which line mattered least. Rohan, with no plan, had nothing to choose between: all he could do was keep writing.

Third, time is usually lost early, not late. Twelve minutes polishing an introduction are twelve minutes taken from the conclusion and the proofread. Draft the introduction plainly, push on, and come back to sharpen it only if the revise slice allows.

Picture packing for a journey with a strict weight limit at the airport. You do not stuff the case until the zip bursts and then find that you have left out your passport. You put in what you cannot travel without first, the passport and the ticket, then the clothes you need, and then extras if there is room. In a timed piece, the plan, the conclusion and the proofread are the passport: they go in whatever happens. It is the extra example and the second anecdote that stay at home.

Sections 21 to 29 gave you every part of a complete piece: the thesis, the genre outline, the links, two revision passes, and the essay, letter and report shapes. This section adds the one thing they all assumed: enough time. It teaches no new stage. It fits the four you already have into a fixed window, and decides in advance what gives way. In Master, section 31 runs the whole Whole-Piece Check on a demanding piece, and section 36 runs the full process against a stated exam-style prompt and a strict clock.

Here is a hard one, and experienced writers still get it wrong: you are behind, and the plan has a paragraph you love. Siddharth has eight minutes left. His plan has two more body paragraphs: one on safety, his strongest reason for learning to swim, and one on swimming as exercise. He has no conclusion yet. He cannot write all three well. The right choice is to write the safety paragraph, briefly, then the conclusion, and to drop the exercise paragraph completely. The test is not which paragraph is easier to write; it is which one serves the thesis most. And the conclusion is not optional. A second hard case is the better idea that arrives halfway. Halfway through a timed essay, Pooja realises that her thesis would be sharper if it argued something slightly different. With a whole evening, she would replan. With fifteen minutes left, she should not start again. The honest move is to finish the piece her plan describes and, in the revise slice, adjust the thesis sentence so that it matches what the body actually shows. A matched, finished piece beats a better idea left half-built. A third: how much revising is too much? In the last five minutes, a change you cannot finish cleanly is worse than no change. Crossing out a whole paragraph you have no time to replace leaves a hole in the middle of the piece. Under a clock, revise only what you can complete: cut a sentence, add a linking phrase, correct the thesis, fix a slip. Larger changes belong to pieces with more time.

The words we use to measure time began as a way of slicing a whole into parts. The word minute comes from the medieval Latin pars minuta prima, the first small part, made by dividing an hour into sixty pieces. Dividing each of those again by sixty gave pars minuta secunda, the second small part, which is where our word second comes from. Both words began as a way of dividing a whole into slices, which is exactly what this section asks you to do with the time you are given. Chess players know the same lesson from the other side of the board. In timed chess, each player has a clock holding a fixed total of time for all their moves. A player whose time runs out usually loses the game, however strong their position on the board. So good players budget the clock from the first move: they think hard at the moments that matter and move quickly where the choice is simple, because a brilliant plan that is never finished counts for nothing. A timed essay works the same way. The best position in the world is worth little if the clock runs out before the conclusion.

The time split is a starting point, not a law. A long or tricky task may need more planning time; a short note may need almost none. Practice changes the split. The more often you plan quickly, the less time it takes, and the more is left for the draft. In an exam, the writing task usually shares the session with other questions. Decide its share of the total before you start, and keep to it. A rushed proofread is better than none, but it will not catch everything. If you know your own common slips, look for those first.

  1. Read the whole task first, and underline the format, the reader and any word limit.
  2. Note the finishing time and divide it between the four stages before you write.
  3. Write the checkpoint times in the margin.
  4. Plan in a few minutes: a thesis sentence and one line per paragraph is enough.
  5. Draft the introduction plainly; come back to polish it only if there is time.
  6. At the halfway checkpoint, compare where you are with where the plan said you would be.
  7. If you are behind, drop a body paragraph before you drop the conclusion.
  8. Always write a conclusion, even if it is two sentences that return to the thesis.
  9. Protect the last minutes for revising and a proofread across the whole piece.
  10. In the final minutes, make only the changes you can finish neatly.
Managing your time while writing. A limited time shrinks every stage of the Whole-Piece Check. It never removes one. 1. Read the task: format, reader, word limit. 2. Divide the clock before you write: a short slice to plan, most of it to draft, a protected stretch to revise and proofread. In forty minutes, perhaps 5, 28, 5 and 2. 3. Plan quickly: a thesis and one line per paragraph. 4. Draft to the plan, with a checkpoint at halfway. 5. When behind, give up length, not a stage: drop a body paragraph, keep the conclusion. 6. Revise only what you can finish; proofread across the whole piece. The key contrasts. A rushed plan beats no plan, but a fourteen-minute mind map costs the essay. A shorter piece with a conclusion is finished; a longer one that stops mid-sentence is stopped. In the last minutes, checking beats adding. Next: Master, the Whole-Piece Check, in full.

Thinking like a writer · ages 16–19

The Whole-Piece Check, in full

The Whole-Piece Check has four stages and a final question. Run in full, each stage asks its own question and calls on tools you already own; this section's job is to put them in order across one whole piece.

Ananya's plan, before a single sentence.

For: students and parents who read the school magazine.
Purpose: to persuade them the remaining lakes are worth defending.
Governing idea: A filled lake is not land gained; it is a flood defence lost.

1. The lake that was, and the claim.
2. What the lake did in every monsoon.
3. What happened in the streets after it was filled.
4. The fair counter: the city needs land for homes - and the answer.
5. What readers can do, returning to the claim.
StageThe question it asksTools, and their home chaptersWhat goes wrong if it is skipped
PlanWho is it for, what is it for, and what one idea governs it? Which paragraphs, in what order?the Reader Test (ch22); for an argument, the essay outline (ch29)good paragraphs about four different things
DraftDoes each paragraph do its one job from the plan, and follow from the one before?the Paragraph Check (ch27); the Bridge Test (ch21)paragraphs that sit side by side with no road between them
ReviseDoes the whole piece hold one voice, move at the right pace, earn every sentence, and, if it argues, hold up?the Voice Check (ch28), the Read-Aloud Test (ch24), the Effect Test (ch25), the Cut Test (ch20), the Argument Check (ch29)a correct piece that is uneven, padded or unconvincing
ProofreadIs every sentence, in every paragraph, free of slips?the Four-Pass Check (ch23): verb, match, mark, sensea strong piece that loses the reader's trust in one line
Then askDoes it still serve the governing idea from the plan?the plan itselfstopping mistaken for finishing

This is the Whole-Piece Check, run in full: Plan, Draft, Revise and Proofread, then the one question that decides whether the piece is finished. Every other section of Master is this check, applied to a harder piece or a harder condition.

Here is what each stage actually found in Ananya's draft. No stage found everything, and no stage found what another stage was built to find.

StageWhat the pass foundWhat Ananya did
PlanHer first governing idea was "Lakes are important." That is a topic, not a claim: nothing could disagree with it.She sharpened it into a claim a reader could test: a filled lake is a flood defence lost.
DraftParagraph 3 opened with "Also," but it reports what followed from the filling, not one more point.The Bridge Test (ch21) named the relation as result: "Within two monsoons of the filling, ..."
ReviseParagraph 4 slipped into an official voice: "It is humbly submitted that the authorities may kindly reconsider." Two sentences on herons served nothing. The counter was a weak version nobody holds.She restored the magazine voice (ch28), cut the herons (ch20), and stated the counter at its strongest before answering it (ch29).
ProofreadParagraph 5: "Each of the three lakes were" - a slip added during revision.The verb pass (ch23) fixed it to was. She ran all four passes on all five paragraphs, not just the opening.
Then askDoes every paragraph, including the new ending, still serve "a flood defence lost"?Yes. The piece is finished.

Notice three things.

First, the order runs from the largest decision to the smallest: the whole piece, then each paragraph, then each sentence, then each slip. That order is not a habit; it saves work. Every larger change can undo smaller work below it. Polishing the herons would have been wasted the moment they were cut, and the slip in paragraph 5 did not exist until revision created it. That is why the proofread comes last.

Second, each stage asks one question. A writer who checks commas while drafting slows the draft and still misses the drifting claim, because a comma check cannot see a claim. Keep the questions apart and each one gets a clear answer.

Third, the final question can send you back, but only to the stage that failed. If the conclusion now argues something new, the problem is a Plan problem: either the conclusion changes or the governing idea does. You do not redraft paragraphs that already serve the idea.

Picture building a house. First the architect draws the whole plan: how many rooms, where each one goes, why. Then the walls go up, room by room, each joined to the next by a doorway. Then the finishing: paint, light, the right handles, nothing left that does not belong. Last, the inspector walks every room with a checklist, not just the hall. Nobody paints a wall that the plan has not yet decided to keep. And when the inspector finds a crack, the builder fixes that wall; the house is not knocked down and started again.

Craft gave you each stage on its own: the thesis, the genre outline, linking, and the revision passes. This section runs all of them, in order, on one demanding piece, and adds the discipline of the final question. The sections that follow push the same check further. Section 32 holds a reflective essay in one voice; sections 33 and 34 take the revision and proofreading stages to full depth; section 35 brings in another reader; section 36 runs the whole check against an exam clock; section 38 makes the finished-or-stopped judgement on its own.

Here is the hard part, and it deserves care: three situations in which the four stages do not run in a straight line. The first is when drafting changes your mind. Farhan Qureshi planned an essay arguing that school should start an hour later. While drafting, he found that his strongest paragraph was really about phones in bedrooms at night, and that his real claim had become "Sleep is lost after lights-out, not before school." He now has two honest choices: rewrite the plan around the new claim and check every paragraph against it, or cut the new material and keep the old claim. The dishonest third choice, keeping the old thesis in the introduction while the body argues the new one, produces a piece that contradicts itself. A changed idea sends you back to Plan; it does not let you skip it. The second is a stage that finds nothing. Tom Hughes runs the Four-Pass Check on all five paragraphs and finds no slip. Was the pass wasted? No. "No change needed" is a result, and it is only worth anything because the pass was really run. The opposite trap is declaring a piece finished because you are tired of it; the answer may be the same, but only one of them is a judgement. The third is when two tools seem to disagree. In Ananya's paragraph 3, the sentence "The water had nowhere to go." repeats what the sentence before it implies, so the Cut Test from Chapter 20 seems to say cut it. But the Cut Test asks whether meaning, tone or emphasis would be lost, and here the emphasis would: read aloud (Chapter 24), the short sentence lands the paragraph's weight at its end. Covered, the paragraph goes soft. So the tools agree after all, once each is asked its full question. When they still seem to pull apart, the governing idea and the reader decide.

Two of the four stages carry their method inside their names. Revise comes from the Latin revisere, "to look at again" (re-, again, and visere, to look at). Revising is not fixing sentences as they come; it is coming back and seeing the whole piece again, as a reader would, once the draft is done. Proofread comes from the printing house. Before a book or newspaper was printed in full, the printer pulled a single trial copy from the type that had been set, called a proof, and someone read it slowly against the original and marked every slip. Only after the proof had been corrected did the full run begin. Proofreading was, from the start, the last check before the reader saw the page, never the first. Publishing still keeps the same order. Many publishers and newspapers separate the work: one editor looks at the structure and argument of a whole piece, a copy editor goes through it line by line, and a proofreader checks the final pages. Large decisions first, small ones last, so that no careful work is undone by a later change. The Whole-Piece Check gives one writer all of those jobs, in the same order.

In practice the stages overlap. Writers do fix an obvious typo while drafting, and that is fine, as long as it does not replace the full proofread at the end. Running the check does not guarantee a good piece. It guarantees that the piece has been checked for the problems that most often spoil good ideas; the ideas themselves still have to be good. The size of each stage scales with the piece. A short email needs a plan of one line; a long essay needs a plan of a page. Under an exam clock every stage shrinks, which section 36 treats on its own. Some pieces genuinely change shape in the writing. A plan is a working decision, not a contract; when it changes, the new plan must be checked against every paragraph, as Farhan's was.

  1. Write the plan before any sentence: reader, purpose, one governing idea, and one line for each paragraph in order.
  2. Test the governing idea: could a reasonable reader disagree with it or learn something from it? If not, it is a topic, not yet an idea.
  3. Draft fast and whole; do not polish a paragraph whose place in the piece is not yet certain.
  4. As each paragraph begins, name its relation to the one before (the Bridge Test, Chapter 21).
  5. Revise the whole piece in one sitting, read aloud, so that voice and pace are judged across it, not paragraph by paragraph.
  6. Cut before you polish: run the Cut Test (Chapter 20) on whole paragraphs before single words.
  7. If the piece claims anything a reader could dispute, run the Argument Check (Chapter 29) as part of revision.
  8. Proofread last, after the final change, with all four passes of Chapter 23 on every paragraph.
  9. Answer the final question in writing, paragraph by paragraph: does this still serve the governing idea?
  10. When the answer is no, return to the stage that failed, not to the beginning.
The Whole-Piece Check, in full, runs four stages and a final question. 1. Plan: reader, purpose, one governing idea, paragraphs sketched in order. Tools: the Reader Test (ch22), and the essay outline (ch29) if it argues. 2. Draft: each paragraph to its plan (ch27), linked to the one before (ch21). Do not polish yet. 3. Revise: the whole piece for voice (ch28), rhythm (ch24), punctuation and pace (ch25), and length (ch20); the Argument Check (ch29) if it argues. 4. Proofread: the Four-Pass Check (ch23) on every paragraph, last of all. 5. Then ask: does it still serve the governing idea? Yes means finished. No means back to the stage that failed. Key contrasts: good paragraphs are not yet a piece; a topic is not a governing idea; revising is seeing the whole again, not fixing sentences as they come; proofreading before revising wastes the proofread; finding nothing is a result only if the pass was run. Next: A complete reflective personal essay.

A complete reflective personal essay

A reflective personal essay is still a whole piece with one governing idea. The difference is what that idea is: not what happened, but what the writer now understands because it happened.

Meher's plan.

For: readers of the school's anthology of student essays.
Purpose: to reflect.
Governing idea (the insight): Teaching my grandmother to use a phone showed me that I had forgotten how to be a beginner.

1. The scene: a Sunday, the phone, my impatience.
2. What happened: Nani's eleven wrong buttons, and her laughter.
3. The turn: the moment I saw it was me who could not bear to be bad at something.
4. Reflection: what being a beginner asks of a person.
5. What changed: the tabla class, and the laugh I am learning.
PartIts jobIn Meher's essayThe common slip
ExperienceShow one moment concretely enough that the reader is therethe phone in Nani's hands, the wrong button, the laugha chronicle of the whole day, every detail given equal weight
ReflectionThink about the moment from where the writer stands now"I had been good at things for so long that I had stopped trying anything I might be bad at."a moral pasted on: "This taught me never to give up."
What changedShow the insight at work in the writer's life, or honestly in progressthe tabla class, where she is the worst in the room"I learned a lot from this experience."
VoiceOne voice, first line to last: the Voice Check (Chapter 28)first person, plain, a little wry, throughouta lurch into essay formula or into slang in one paragraph

In the Whole-Piece Check for a reflective essay, the Plan's governing idea is the insight, and the Revise stage leans hardest on the Voice Check. Draft and Proofread run as for any piece; then the final question asks whether every paragraph still serves what the moment showed.

The commonest planning mistake in this genre is to write an event where the governing idea should be. Here are four events, each turned into an insight that could govern an essay.

The event (not yet a governing idea)An insight the event could show
the day I lost the state chess finalLosing the final showed me I had been playing to win, not playing chess.
moving from Patna to Pune in Class 9Moving cities taught me that I had mistaken being known for having friends.
my first job, at my uncle's sweet shop during DiwaliPacking boxes for strangers showed me how much of a festival is other people's work.
the week our house had no electricityA week without power showed me that my family talks most when there is nothing else to do.

Notice three things.

First, an insight is a claim about the writer, and like any claim it could be wrong. That is what makes it worth an essay. "Moving cities was hard" cannot be wrong; "I had mistaken being known for having friends" can, and the scene must show it is true.

Second, a reflective essay lives in two times. The experience happened then and is told in the past tense: "Nani pressed the wrong button." The reflection happens now and is told in the present: "I think I had stopped trying anything new." Moving between them is a deliberate structural choice, not a tense slip, and the Four-Pass Check from Chapter 23 judges each sentence by its own time. What must not move is the voice: the person thinking now is the same person who was impatient then, and the essay sounds like one person throughout.

Third, a reflection is not a moral. "Hard work always pays off" could be pasted on to a thousand essays, and on to Rehaan's it is not even true. A reflection is specific to this moment and this writer. It may even be uncertain - "I am still not sure I have learned it" - and an honest uncertainty, precisely stated, is a stronger ending than a confident slogan.

Picture a photograph and the person holding it. The photograph is the experience: one moment, sharp and fixed, from then. The person holding it is the writer now, looking, noticing something they did not see when it was taken. A reflective essay shows the photograph clearly enough that the reader sees it too, and then lets the reader hear what the person holding it has come to understand. Too much photograph and it is an album; too much talk and there is nothing to look at.

Craft gave you genre outlines for narrative, explanatory and persuasive pieces, and S31 ran the Whole-Piece Check in full. This section adds the reflective outline, experience, reflection, what changed, and makes the Voice Check the centre of revision. Section 33 goes deeper into revising a full draft, cutting and reordering; section 37 returns to the reflective essay alongside the other genres, as a final synthesis of choosing an outline and a voice for a task.

Here is the hard part, and it deserves care: three questions this genre raises that have no mechanical answer. The first is the unresolved insight. Aditi Rao's essay is about the summer her parents separated. Her last paragraph says: "I used to think a family was the people under one roof. I do not know yet what I think it is, only that it is not that." Is that finished? Yes. The governing idea is the change in her understanding, and the essay shows it honestly. What would not be finished is the same essay ending on a confident moral it had not earned. An uncertainty can govern an essay if it is precise: she knows exactly what she no longer believes. The second is the two voices. Sam Lewis quotes what he thought at nine: "I was sure the dog would come back if I was good enough." That line is a nine-year-old's thought. The voice of the essay is his now, framing that thought, and it stays steady around it. Holding one voice does not mean flattening every quoted thought into the narrator's register; it means the narrator, the person reflecting, is recognisably one person from the first line to the last. The third is honesty. A reflective essay selects: it leaves out most of what happened. That is shaping, and it is the writer's right. It does not invent. If Meher made Nani press the wrong button eleven times when it was three, or added a tabla class she never joined, the essay would stop being a personal essay and become fiction presented as fact. A reader of a personal essay trusts that it is true; the shaping happens in what is chosen, not in what is made up. When the truth is too small to carry the insight, the insight is the thing to change.

The word essay began as a modest claim. It comes from the French essai, "an attempt" or "a trial", from the verb essayer, "to try". When Michel de Montaigne published the first books of his Essais in 1580, the title promised attempts, not verdicts: short pieces in which he tried out his thoughts on whatever caught his attention, very often on himself. In a brief note to the reader at the start, he says that he himself is the subject of his book. So a reflective personal essay is, in the oldest sense of the word, an attempt: a writer trying to understand an experience in public. That is why an honest "I am still not sure" can belong in one, and why a confident moral that settles everything often rings false. Reflect has an old meaning to match. It comes from the Latin reflectere, "to bend back": re-, back, and flectere, to bend. A reflection bends the writer's thought back on to the experience, the way a mirror bends light back to the eye. An essay that never bends back, that only reports the event, has not yet reflected.

Not every reflective essay needs a dramatic change at the end. Sometimes what changed is only how the writer sees something, and that is enough, if it is shown. The three-part shape is a strong default, not a law. Some fine reflective essays open with the reflection and then go back to the moment; the plan decides, but every part must still be there. Personal does not mean private. You choose what to share. An essay about a hard experience can stay with what you are willing for a stranger to read. Some school tasks call a piece reflective when they really want a narrative. Read the task: if it asks what you learned or how you changed, it wants reflection; if it asks only for what happened, it wants a story.

  1. Choose one moment small enough to show, not a year or a whole journey.
  2. Write the insight first, as one sentence beginning "It showed me that...", and make sure it could be wrong.
  3. Sketch the essay as experience, reflection and what changed, one line per paragraph.
  4. Keep only the details of the scene that show the insight; cut the rest, however true.
  5. Tell the experience in the past tense and the reflection in the present, and move between them on purpose.
  6. Write the reflection in words only you could write about this moment; if a sentence could end anyone's essay, rewrite it.
  7. End on what changed for you, or honestly on what is still changing, not on advice for the reader.
  8. Run the Voice Check from Chapter 28 across the whole essay, read aloud, and bring any stiff or slangy paragraph back into your voice.
  9. Select from the truth but never invent it; if the facts cannot carry the insight, change the insight.
  10. Proofread last, every paragraph, then ask whether every paragraph still serves what the moment showed.
A complete reflective personal essay runs the Whole-Piece Check with one difference: its governing idea is an insight, not an event. 1. Plan: one small moment; the insight as "It showed me that...", something that could be wrong; paragraphs sketched as experience, reflection, what changed. 2. Draft: the scene concretely, in the past; the reflection from now, in the present; each paragraph linked to the one before. 3. Revise: the Voice Check (ch28) across the whole essay, read aloud; cut every detail the insight does not need (ch20). 4. Proofread: the Four-Pass Check (ch23) on every paragraph, judging each sentence by its own time. 5. Then ask: does every paragraph still serve what the moment showed? Key contrasts: an event is evidence, an insight is the governing idea; a reflection is specific, a moral could end anyone's essay; then in the past, now in the present, one voice in both; selecting from the truth is shaping, inventing it is fiction; ending on what changed for you, not on advice for the reader. Next: Revising a full draft.

Revising a full draft

Revision works from the largest decision to the smallest, because a large change can delete a small one. There is no point perfecting the rhythm of a paragraph you are about to cut.

Revise top down: four levels, in order.

1. Thesis: does the draft argue what its introduction says it argues?
2. Structure: does every paragraph serve the thesis, once? Cut, merge or add.
3. Order: does each paragraph need the one before it? Move, then relink.
4. Paragraphs and sentences: one job per paragraph (ch27), a link at each boundary (ch21), then voice, rhythm, punctuation and length (ch28, ch24, ch25, ch20).
The levelThe question to askWhat the revision doesIn Rehan's canteen essay
ThesisIs the claim the body proves the claim the introduction makes?rewrites the thesis to fit what the draft showsmoves cheaper and healthier, starting this term into the first paragraph
StructureDoes each paragraph serve the thesis, and only once?cuts, merges or adds a paragraphcuts the canteen's history; folds the Thursday market into the cost paragraph
OrderDoes each paragraph need the one before it?moves a paragraph and relinks both its edgeskeeps cost after freshness, since the sharpened thesis names health then cost
ParagraphsDoes each do one job, developed, linked to the one before?sharpens a topic sentence, rewrites a linkthe cost paragraph now opens on its claim, not on the market
SentencesVoice, rhythm, punctuation, lengthvaries, sharpens and cutstwo flat openings rewritten; a padded sentence halved
Not yetSpelling, agreement, stray marksnothing: note it for proofreadingwere to was waits for the Four-Pass Check

Revision is the third stage of the Whole-Piece Check, after Plan and Draft and before Proofread, and it ends with the check's own last question: does the revised piece still serve one governing idea from first line to last?

The reverse outline is the tool that makes top-down revision possible. You write it after the draft, not before, and it shows the skeleton the reader will actually meet.

ParagraphReverse-outline lineThe revision decisionWhy
P1 introductionwe should serve more local foodsharpen: our canteen should switch to local seasonal food this term, because it would make lunch healthier and cheaperthe conclusion had found the real thesis
P2local food is fresher, so students eat more of itkeepserves the healthier half of the thesis
P3the canteen was built in 1987 and had a fountaincut, and save in a separate filetrue, but serves no part of the thesis
P4vegetables in season cost lesskeep, and take in P5's evidenceserves the cheaper half
P5farmers sell at the Thursday marketmerge into P4not a new reason; it is P4's proof that the plan is practical
P6 conclusionso cheaper and healthier, starting this termrewrite its first lineit no longer needs to announce the thesis; it can close on what starts this term

Notice three things. First, every line in the outline is written in your own words, as a claim. If you cannot say what a paragraph claims in one line, that is itself a finding: the paragraph is doing two jobs, or none, and it is the first one to rebuild.

Second, the paragraph that was cut was a good paragraph. Revision does not ask whether a paragraph is true, interesting or hard-won; it asks whether it serves this thesis for this reader. Keep a separate file of cut paragraphs. Writers who know their good material is safe find it far easier to cut.

Third, the changes cascade. A sharper thesis made the conclusion's first line redundant; a merge changed what P4 opens with; a cut put P2 and P4 side by side, so the boundary between them needed a new link. Every structural change sends you back to the paragraphs it touched. That is why sentence work comes last: it is the only level that disturbs nothing above it.

Picture rearranging a room before you dust it. First you decide what the room is for. Then you take out the furniture that does not belong, however handsome it is, and move what stays so that the room works: the desk by the window, the lamp by the desk. Only when nothing is going to move again do you polish the surfaces. Dust a cupboard you are about to carry out of the room, and the work goes out with it.

Section 31 ran all four stages of the Whole-Piece Check end to end, and section 32 held a reflective essay in one voice. This section zooms into the third stage, Revise, and gives it an order: thesis, structure, order, then paragraphs and sentences. Section 34 zooms into the fourth stage, Proofread, and runs the Four-Pass Check across the revised piece. Section 35 adds a second reader, and sections 36 to 38 ask how much revision fits under a clock and when a revised piece is truly finished.

Here is a genuinely hard case. The best paragraph in the draft is off the thesis. Nandini Iyer's essay argues that her city should pedestrianise its old market street. Her fourth paragraph, about the street's history as a cloth market, is the most alive thing in the essay: specific, moving, beautifully paced. It does not serve the argument. Should she cut it, or change the thesis so that it fits? There is no single answer, but there is a single question: what is this piece for? If the task is set - a persuasive essay on pedestrianisation, for a council consultation - the task fixes the territory, and the paragraph goes to the cut file, or shrinks to one sentence that serves the argument (the street was built for people on foot). If the piece is her own and the history is what she actually cares about, the draft may have discovered its real subject, and the honest revision is a new thesis and a new plan. What she may not do is keep both: a persuasive thesis with a nostalgic paragraph in the middle is two pieces sharing a title. A second hard case: two orders that both work. Daniel Okafor's essay gives three reasons for a later school start. He could put sleep first and attendance second, or the reverse. Neither paragraph depends on the other, and the conclusion names them in neither order. Moving them would be change, not revision. Here the honest judgement at the order level is no change needed, and the time is better spent elsewhere. A third: the draft that found its thesis at the end. This is so common that many editors check the last paragraph first. Moving the conclusion's sharp claim into the introduction is right, but it leaves the conclusion with nothing to do. It must now close the piece another way - by drawing out a consequence, or by returning to the opening image - rather than repeating the sentence that has just moved.

The word revise comes from the Latin revisere, 'to look at again' or 'to visit again', built from re-, 'again', and visere, 'to look at', a form of videre, 'to see'. Revision is literally re-seeing. The word itself says that the job is to look at the whole piece with new eyes, not to correct it line by line. Writers have long known that this takes time. Blaise Pascal, the seventeenth-century French mathematician and writer, ended one of his published letters with an apology that it was so long, explaining that he had not had the time to make it shorter. Cutting is harder and slower than adding, and a long draft is often a draft that has not yet been revised. Film editors have a phrase for revision's hardest decisions. Scenes that are shot, paid for and sometimes very good, but which do not serve the finished film, are said to end up on the cutting-room floor, from the days when cut lengths of film were literally discarded there. Editors are judged as much by what they leave out as by what they keep. And editors of long pieces commonly use the reverse outline described in this section: after a draft is written, they list what each section or paragraph actually says, then read the list as if it were the plan. Gaps, repeats and paragraphs in the wrong place show up on that one sheet long before they show up in the prose.

Top-down is an order of priority, not a ban. If you notice a sentence slip while revising structure, mark it in the margin; just do not stop to fix it yet. Not every draft needs a structural change. If the reverse outline shows a clear thesis, every paragraph serving it once, in an order that works, the honest result at those levels is no change, and you move straight to sentences. Under an exam clock, revision shrinks but does not vanish: a two-minute reverse outline in the margin and one decision about the weakest paragraph is still revision (section 36). A set task limits revision. You may sharpen a thesis or change its direction within the prompt, but you may not revise your way into a different question from the one that was asked.

  1. Revise only a complete draft; finish the last paragraph first, even roughly.
  2. Rest the draft before you revise it, overnight if the deadline allows.
  3. Write the thesis as the draft actually argues it, then compare it with the introduction's.
  4. Write a reverse outline: one line per paragraph, each line a claim in your own words.
  5. Ask of every paragraph which part of the thesis it serves; cut or rebuild any that serve none.
  6. Merge two paragraphs that make one point; add a paragraph where the thesis promises a point no paragraph delivers.
  7. Order paragraphs so that each needs the one before it, or so that the order matches the thesis.
  8. After any move or cut, run the Bridge Test (ch21) on every boundary that changed.
  9. Only when the structure is settled, revise for voice (ch28), rhythm (ch24), punctuation (ch25) and length (ch20).
  10. Note surface slips for proofreading, and say no change needed honestly when a level is already sound.
Revising a full draft means re-seeing the whole piece, not correcting its mistakes. Revise from the largest decision to the smallest: 1. Thesis: does the draft argue what its introduction says? If the real claim is in the conclusion, move it up. 2. Structure: write a reverse outline, one line per paragraph. Cut what serves no part of the thesis, merge what repeats, add what is missing. 3. Order: each paragraph should need the one before it. After a move, relink every boundary it touched (ch21). 4. Paragraphs and sentences: one job per paragraph (ch27), then voice, rhythm, punctuation and length (ch28, ch24, ch25, ch20). Key contrasts. Revising is not proofreading: Tanvi fixed every mistake and still had a fountain in a canteen essay. Good is not the same as on the thesis: a fine paragraph that serves nothing goes to the cut file. Change is not the same as revision: two orders that both work need no move. Then ask the Whole-Piece Check's last question: does the revised piece serve one governing idea from first line to last? Next: Proofreading a full draft.

Proofreading a full draft

Proofreading a whole piece uses the same check you already know from Chapter 23. What changes is the direction: each pass runs across the whole piece before the next one starts.

The Four-Pass Check, across a whole piece.

1. Verb pass: every verb in every paragraph. Does it agree with its true subject? Does it hold the time the piece set in its first lines? (ch12, ch13)
2. Match pass: every pronoun back to its noun, even across a paragraph break; every article and preposition. (ch04, ch05, ch07)
3. Mark pass: every capital, end mark, apostrophe and quotation mark, including the title and headings. (ch11, ch17, ch18)
4. Sense pass: sound-alikes, missing and doubled words, words left behind by revision, names and numbers that disagree. Start from the last paragraph.
Then decide: fix the slip, keep the choice, or Finished - no change needed.
The slip in Ishita's reportThe pass that finds itWhy one careful read missed itThe fix
By midnight the water rises to our doorsteps. (paragraph 3)verb pass - ch13correct alone; paragraph 1 set the report last Julyrises - rose
It promised a new drain by May. (paragraph 3, after 'the ward councillors')match pass - ch04the noun it must match is in the paragraph beforeIt - They
Three meeting's were held in June. (paragraph 2)mark pass - ch17a plain plural; the eye slides over a familiar wordmeeting's - meetings
...and there anger grew. (paragraph 4)sense pass - ch17a spellchecker passes it; attention was lowest at the endthere - their
Rs 40,000 in paragraph 2, Rs 4,000 in paragraph 4sense pass - names and numbersthe two figures are two paragraphs apartcheck the source; make them agree
A second report, all four passes cleanall four passesnothing to missFinished - no change needed

Proofreading is the fourth stage of the Whole-Piece Check, the last before its closing question: once the passes come back clean, ask whether the piece still serves the one governing idea it began with.

The Four-Pass Check finds slips. It does not judge choices, and a whole piece is full of choices made on purpose in earlier stages. Chapter 23's rule holds at the scale of a whole piece: fix the slips, keep the choices.

In the pieceSlip or choice?Why, and whose chapter
And then the lights went out. (opening a sentence in a story)choice - keepstarting a sentence with And is standard (ch21)
Silence. (a one-word paragraph in a story)choice - keepa deliberate fragment, chosen for rhythm in revision (ch24)
Nani said, 'We was never hungry.'choice - keepa quotation is never corrected (ch23)
We can't wait to see you. (in an informal letter to a cousin)choice - keepa contraction suits the register (ch22)
Everyone brought their own torch.choice - keepsingular their is standard (ch04)
My brother and I was soaked. (the writer's own narration)slip - fixverb pass: a plural subject takes were (ch12)

Notice three things. First, every choice in the table was made in an earlier stage: the And and the fragment in revision, the register in the plan, the quotation by the speaker. Proofreading does not reopen those decisions. A proofreader who 'corrects' them makes the piece worse and teaches a false rule.

Second, proofreading sometimes finds a problem that is not a slip at all. A sentence that begins 'As the third paragraph showed' when the third paragraph was cut, or a conclusion that names four reasons when the body now gives three, is a revision problem that surfaced late. Go back a stage, revise it, and then proofread what you changed.

Third, a clean proof is a result, not a failure to look hard enough. When all four passes come back clean across the whole piece, Finished - no change needed is the honest answer. Inventing a correction to show effort turns a correct piece into an incorrect one.

Picture checking a house before guests arrive. You do not walk through once and glance at everything. You go round once for the lights, once for the taps, once for the doors and windows, and once for anything out of place, and each round covers every room, the hall and the stairs included. One kind of check per round is what lets you notice the one bulb that has gone in the back bedroom.

Section 33 revised a full draft from the top down: thesis, structure, order, then sentences. This section takes the revised piece through the fourth stage, Proofread, using the Four-Pass Check exactly as Chapter 23 taught it, now run across a whole piece. Section 35 adds a second reader, whose kind, specific feedback catches what your own passes miss. Section 36 fits all four stages under an exam clock, and section 38 asks whether a proofread piece is truly finished.

Here is a hard one: a tense shift that is right. Ayesha Siddiqui's reflective essay is told in the past: 'The summer I was fifteen, my grandfather gave me the key to his workshop.' Its last paragraph says: 'I still keep that key in my desk drawer.' The verb pass sees a past-tense piece and a present-tense verb. Is it a slip? No. The verb pass does not ask whether the tense changed; it asks whether the time changed. Keeping the key is true now, so the present is right. A proofreader who changes it to 'I still kept' has introduced a slip. A second hard one: a pronoun that is not wrong, only unclear. Ethan Cole's report says, at the end of paragraph 2, 'The council asked the contractor to inspect the drain.' Paragraph 3 opens: 'It was found to be blocked in three places.' It can grammatically be the council, the drain, or even the inspection. The match pass flags it because its noun is in another paragraph and has rivals. The fix is to name the noun - 'The drain was found...' - which is a proofreading fix that also protects the sense. A third: a slip that changes the governing idea. Anushka Rao's thesis reads: 'A later start time will effect attendance.' Effect as a verb means to bring about, so the sentence now claims that the later start will create attendance, which is not what she argues. Her claim is that the start time will affect, or change, attendance. The sense pass catches it, and it matters more than any other slip in the piece, because it sits in the sentence the whole essay serves. That is why the Whole-Piece Check ends by asking, after the proofread, whether the piece still serves its one governing idea.

The word proof comes, through Old French, from the Latin probare, 'to test'. A printer's proof was exactly that: a test copy, pulled from the type before the full run, so that mistakes could be found while they were still cheap to fix. Proofreading is reading the test. In the days of hand-set metal type, the lines of a book or newspaper were assembled in a long tray called a galley, and the first test prints taken from it were called galley proofs. Publishers still use the word galleys for the advance copies of a book that are checked before printing. Proofreaders also have a mark for undoing their own work. If a correction turns out to be wrong, they write stet in the margin, a Latin word meaning 'let it stand'. It is the professional's version of this section's rule: fix the slips, keep the choices, and be ready to take back a correction that was not one. And in many printing houses, important text was proofread by two people together: one read the original aloud, word by word and mark by mark, while the other followed the proof. The method worked for the same reason the four passes work: it stopped the eye from reading what it expected, and made it see what was actually there.

The four passes are a routine, not a guarantee. A careful writer still misses some of their own slips, which is why a second reader (section 35) is worth having for anything important. Not every difference is an error. House style - a serial comma, -ise or -ize, Rs or Rs. - is a matter of consistency within the piece, not of right and wrong. Make it consistent; do not call either form a slip. Under an exam clock the passes shrink, but they do not vanish: a fast verb pass and a sense pass on the last paragraph first still catch the slips most likely to be there (section 36). Proofreading cannot rescue a piece that has not been revised. An essay can be correct in every sentence and still argue the wrong thing.

  1. Proofread only after the last revision, never while a paragraph may still be cut.
  2. Before the first pass, note the time, the key names and numbers, and the register the piece sets.
  3. Run the verb pass through every paragraph: agreement with the true subject, and the time the piece set (ch12, ch13).
  4. Run the match pass across paragraph breaks: follow every pronoun back to its noun (ch04, ch05, ch07).
  5. Run the mark pass on everything, including the title, headings and captions (ch11, ch17, ch18).
  6. Run the sense pass from the last paragraph backwards, reading for meaning, not spelling.
  7. Check every name and number against its first appearance in the piece.
  8. Fix the slips and keep the choices: never correct a quotation, a deliberate fragment or a register choice.
  9. Send any revision-level problem back a stage, then proofread whatever you changed.
  10. When all four passes come back clean, say Finished - no change needed, and do not invent a correction.
Proofreading a full draft is the Four-Pass Check from Chapter 23, run across the whole piece, one pass at a time. 1. Verb pass: every verb agrees with its true subject and holds the time the piece set in its first lines. 2. Match pass: every pronoun points back to its noun, even when that noun is in the paragraph before. 3. Mark pass: every mark, including the title, headings and captions. 4. Sense pass: sound-alikes, missing and doubled words, words left by revision, names and numbers that disagree. Start from the end. Then decide: fix the slip, keep the choice, or Finished - no change needed. Key contrasts. One careful read is not a method: Ishita's caught three slips and missed four, all in the second half. A sentence that is correct alone can be wrong in its piece: 'the water rises' in a report set last July. A choice is not a slip: a quoted dialect, a fragment for effect and a sentence-opening And all stay. A revision problem found while proofreading goes back a stage. Proofread after the last revision, then ask the Whole-Piece Check's last question: does the piece still serve its one governing idea? Next: Peer review, at the whole-piece level.

Peer review, at the whole-piece level

Peer review at the whole-piece level starts with the whole piece. Comment on the governing idea and the structure before the sentences, and on the sentences before the commas.

Feedback from the top down.

1. The governing idea: is it clear, and does every part serve it?
2. Structure and order: does each paragraph do one job, in a sensible order?
3. Unity and links: does anything wander, or jump?
4. Voice, rhythm and length: only once the shape is sound.
5. Proofreading: last, and only if asked.

Each comment: where it is, what you experienced as a reader, and a question or a suggestion.
The commentLevelSpecific, kind and usable?
Confusing.none namedNo - vague: where, and what kind of confusion?
Paragraph 4 felt like a different essay: the thesis is about learning, this is about parking. Is it one of your three reasons?unityYes - names the place, the reader's experience and a question
This is boring.none namedNo - judges the writer and gives no direction
Your thesis is clear, and I agreed with it by the end of paragraph 2.governing ideaYes - specific praise the writer can keep
I would put a comma after However in line 12.proofreadingFine, but only after the bigger points

Peer review sits inside the Revise and Proofread stages of the Whole-Piece Check, this chapter's four stages: Plan, Draft, Revise and Proofread. A reader shows a writer where to revise. The way to phrase a line-level suggestion kindly belongs to Chapter 25, Punctuation and Editing for Effect, in its section on editing someone else's work; this section scales that kindness up to a whole piece.

Receiving feedback is half of peer review, and it is harder than giving it. Every comment can be sorted into one of three piles; a comment you cannot yet understand needs a question first.

The pileWhen a comment goes thereWhat the writer does
Act on itit points to a real problem with the governing idea, structure, unity or clarityrevise, starting with the largest point
Think about itit is a reasonable suggestion, but you are not yet sure it serves your purposetest it: try the change and reread, or ask a second reader
Set it aside, with a reasonit would change your purpose, your reader or your voice into the reviewer'skeep your version and know why
Ask firstyou cannot tell what the comment means, such as Confusingask: where exactly, and what did you expect to find?

Notice three things. First, act on it is not the same as agree with every word. A reader who says paragraph four lost them is always right about their own experience, even when their suggested fix is not the best one. Take the problem seriously; choose the fix yourself.

Second, set it aside needs a reason. If a reviewer wants your reflective essay to sound like a report, you can keep your voice, because your reader and purpose need it. Without a reason, setting a comment aside is just defending.

Third, the order matters for the writer too. Fix structure and unity first, because a sentence you polish today may be cut tomorrow. Proofreading notes wait until the shape is settled.

Picture a doctor examining a patient. A good doctor does not start by commenting on a small scratch while the patient's arm is broken. They look at the whole patient, find what matters most, treat that first, and explain it in words the patient understands. A good reviewer works the same way on a draft: the whole piece first, the most serious problem first, in words the writer can use. The scratches can wait until the arm is set.

Section 31 ran the Whole-Piece Check in full; sections 33 and 34 revised and proofread a full draft alone. This section adds a second reader. Section 36 asks how much of the process fits under exam time, when there is no reviewer. Section 38 asks the hardest judgement, whether a piece is finished, which is easier with a reader's honest response. Chapter 25 taught how to suggest a line-level change kindly; this section applies that kindness to a complete draft.

Here is a hard one, and it is worth taking slowly. What if the reviewer is wrong? A reader can misread, miss a point, or prefer a style that does not suit your purpose. But even a wrong suggestion usually points at something real. If a reader says your opening is boring, they may be wrong that it needs a joke, but right that it took too long to reach your thesis. Separate the reader's experience, which is always real information, from their proposed fix, which is only one option. Here is a second hard one. Two reviewers disagree. One says paragraph two is the strongest; the other says it should be cut. Neither comment cancels the other. Go back to your governing idea and your reader. Does paragraph two serve the thesis? If yes, keep it and ask what the second reviewer found weak about it. If not, the first reviewer's pleasure in it may not be enough to keep it. And a third. How honest should a reviewer be? Kind does not mean soft. If the thesis is missing, saying so clearly is the kind thing to do, because the writer can still fix it. What makes it kind is aiming the comment at the text, saying what works as clearly as what does not, and giving the writer something they can act on.

Did you know that publishing and science both rely on a version of peer review? Many journals that publish research send a paper to other experts in the same field before it is accepted. These reviewers read the whole paper and report on whether its main claim is supported, whether its method holds up and whether anything important is missing, as well as on smaller points. The author then revises in response, and often writes a reply explaining which comments were acted on and why others were not. That reply is the same sorting you are learning here: act on it, think about it, or set it aside with a reason. And here is a second surprise. Many writing teachers suggest a simple pattern for feedback that keeps it both kind and honest: say one specific thing that works, ask one real question, and make one specific suggestion. It is short, it starts from the text, and it gives the writer both something to keep and something to do.

A reviewer's experience is real information even when their suggested fix is not the best one. Take the problem seriously and choose the fix yourself. Kind feedback is not soft feedback. Saying clearly that a thesis is missing is kind, because the writer can still fix it. Line-level and proofreading notes are useful, but they belong after the larger points, and only if the writer asked for them. The writer owns the piece. A reviewer suggests; the writer decides, and should be able to say why.

  1. Read the whole draft once before writing any comment.
  2. Say in one sentence what you think the governing idea is, and compare it with the thesis.
  3. Comment from the top down: idea, structure, unity and links, then style, then proofreading.
  4. Choose the two or three points that matter most, not every point you could make.
  5. Make each comment specific: name the paragraph or line.
  6. Describe your experience as a reader, not the writer's failings.
  7. Name what works as precisely as what does not.
  8. Suggest or ask; do not rewrite the piece in your own voice.
  9. As the writer, read every comment before replying, and ask about anything unclear.
  10. Sort each comment: act on it, think about it, or set it aside with a reason.
Peer review at the whole-piece level starts with the whole piece. As a reviewer: 1. Read the whole draft first, without marking. 2. Comment from the top down: governing idea, structure, unity and links, then style, then proofreading. 3. Choose the two or three points that matter most. 4. Make each one specific, kind and usable: where it is, what you experienced as a reader, and a question or suggestion. 5. Name what works, precisely. 6. Suggest; do not take over the writer's voice. As a writer: listen, ask, and sort each comment - act on it, think about it, or set it aside with a reason. Do not defend every line; do not obey every note. Thirty-one comma marks helped Kavya less than three comments about her thesis and paragraph four. Next: exam writing, under time pressure.

Exam writing, under time pressure

Under exam time, every stage of the Whole-Piece Check still happens. Each one simply gets smaller.

A 45-minute piece, stage by stage (adjust for your own paper).

Read the prompt and plan: about 4-5 minutes - task, topic, form, reader, length; a thesis; one line per paragraph.
Draft: about 30 minutes - each paragraph to its line.
Revise: about 4-5 minutes - does every paragraph serve the thesis? Does the ending answer the opening? Does the form match the prompt?
Proofread: about 3-4 minutes - the Four-Pass Check (Chapter 23) across the whole piece.
The stageWhat it shrinks toWhat it must still do
Plana thesis and four or five paragraph lines in the marginmatch the task, the form and the reader the prompt names
Draftsteady writing to the lines, no restartskeep each paragraph on its line; link it to the last
Reviseone read for the whole, not each sentencecatch an off-thesis paragraph, a missing ending or the wrong form
Proofreadone pass with the Four-Pass Checkcatch the slips a fast writer makes: agreement, tense, a missing word

This is the Whole-Piece Check, this chapter's four stages - Plan, Draft, Revise and Proofread - run against a clock. The times above are a sensible starting point for a single 45-minute piece, not any paper's rule; read your own paper's instructions and divide its time the same way. Proofreading uses the Four-Pass Check from Chapter 23.

When time runs short, protect the parts that decide whether the piece works at all, in this order.

If you have this much time leftProtect this firstLeave this out
5 minutes, body unfinisheda short final body point and a real conclusion that answers the thesisa third example in the unfinished paragraph
2 minutes, no conclusiontwo sentences that return to the thesis and closeany polishing of earlier sentences
1 minute, piece completea quick proofread of the first and last paragraphs, where readers look hardestrewording sentences that are already clear
10 minutes, piece completea full revision read, then a full proofreadstarting a new draft

Notice three things. First, the conclusion is protected before almost anything else. A piece with no ending has stopped, however good its middle; examiners and readers notice an ending more than an extra example.

Second, the plan is what makes rescue possible. With paragraph lines in the margin, a writer who runs short knows exactly which point to compress and how the piece should close. Without them, running out of time usually means running out of piece.

Third, revising under time means reading for the whole, not for each word. The questions are large ones: is this the form and task the prompt asked for, does every paragraph serve the thesis, and does the ending answer the opening? Word-level polish belongs to the proofread, if there is time.

Picture a cook with one hour to make a full meal for guests arriving at seven. A good cook does not skip the shopping list or the final taste because time is short. They make a quick list, cook to it, taste before serving and wipe the plate's edge as it goes out. Each step is fast, but none is skipped, and if the clock runs tight they serve a simpler dessert rather than no dessert. An exam piece is that meal: plan, draft, check and proofread, each one fast, none left out.

Section 30 in Craft fitted the four stages into a limited time; this section applies that to exam-style prompts and to the larger, more demanding pieces of Master. Sections 33 and 34 revised and proofread a full draft with no clock; here each shrinks to a few minutes. Section 37 asks which outline and voice each genre needs, so that the planning minutes are spent well. Chapter 23's Four-Pass Check is the proofreading tool, and Chapter 29's outline is the plan when the task word is argue.

Here is a hard one, and it deserves care. How should time be divided when a paper has more than one piece? Divide by what each piece is worth, not by how much you have to say about it. If two tasks carry equal weight, give them roughly equal time, even if you could write far more about the first. A brilliant first piece and an unfinished second usually do worse than two complete, sound pieces. Write the finishing time for each task at the top of the paper before you begin. Here is a second hard one. What if the plan turns out to be wrong halfway through? Small problems - a weak example, a paragraph that needs trimming - can be fixed as you go. A large problem - you realise your thesis does not answer the task - is worth a quick repair, but rarely a restart: rewrite the thesis sentence to match what your paragraphs actually argue, if they still answer the task, and make the conclusion fit. Restarting at minute 20 almost always costs the ending. And a third. Is it ever right to skip proofreading? Almost never. Even one minute, spent on the first and last paragraphs, catches slips where readers look hardest. A piece with no proofread is a piece where the Four-Pass Check did not happen at all.

Did you know that the word deadline once had a far more literal meaning? In the nineteenth century the word was used for a line around a prison camp that prisoners were not allowed to cross. By the early twentieth century, newspaper offices were using it for the time by which a story had to be finished for the paper to be printed. Journalists still work this way: a reporter who must file a story in an hour does not skip the plan or the check, but shrinks each of them to fit. And here is a second surprise. Many people who mark timed writing say that the pieces which lose the most marks are not the ones with small slips, but the ones that answer a different question from the one set, or that stop without an ending. Both of those are prevented in the first and last few minutes: reading the prompt carefully, and keeping time to close.

The times in this section are a sensible starting point for one 45-minute piece, not a rule from any exam. Read your own paper's instructions and divide its time in the same proportions. A plan under time is notes in the margin, not a neat outline. A thesis and one line per paragraph is enough. If the plan goes wrong, repair it rather than restart; restarting usually costs the ending. The Four-Pass Check belongs to Chapter 23 and the argument outline to Chapter 29; this section only fits them into the clock.

  1. Underline the prompt's task word, topic, form, reader and length before planning.
  2. Write the finishing time for each stage in the margin before you start.
  3. Spend about a tenth of the time on a plan: a thesis and one line per paragraph.
  4. Draft to the plan, linking each paragraph to the last.
  5. Stop drafting at your planned time, even mid-paragraph, and close.
  6. Protect the conclusion before adding another example.
  7. Revise for the whole: task, form, thesis, unity and the ending.
  8. Proofread last with the Four-Pass Check, starting with the first and last paragraphs.
  9. Divide time between tasks by what they are worth.
  10. Repair a faulty plan rather than restart.
Under exam time, every stage still happens; each one gets smaller. 1. Read the prompt: task word, topic, form, reader, length. 2. Plan (about a tenth of the time): a thesis and one line per paragraph. 3. Draft to the plan. 4. Revise for the whole: task, form, thesis, unity, ending. 5. Proofread with the Four-Pass Check. Write each stage's finishing time in the margin. When time runs short, protect the ending and the thesis before polish. Aditya started at once and wrote the wrong form with no real ending. Nisha planned for four minutes, checked for eight, and finished. A rushed plan and a rushed proofread still beat none. Next: writing across genres, one more time.

Writing across genres, one more time

A genre is a kind of writing with a job its readers already recognise. The genre decides the outline's shape and the voice; it never decides whether the four stages happen.

Seven genres, seven shapes.

Narrative: an opening moment, a rising complication, a turn, an ending that shows a change.
Descriptive: one dominant impression, details in a spatial or sensory order, a closing impression.
Expository: a thesis that explains, one paragraph per part, step or cause, a conclusion that draws them together.
Persuasive: a claim, reasons with evidence, the strongest objection answered, a close that asks for something.
Reflective: the experience, told briefly, then what it meant then and what it means now.
Formal letter: the purpose first, the details, the request, a formal close.
Report: a title, a purpose, findings under headings, then conclusions or recommendations.
GenreOrdered byIts governing ideaIts voice
Narrativetimewhat changeda steady narrator
Descriptivespace or the sensesone dominant impressionprecise and concrete
Expositorylogic: parts, steps, causesa thesis that explainsclear and neutral
Persuasivethe strength of the reasonsa claimfirm, reasoned, fair to the other side
Reflectiveexperience, then meaningan insighthonest first person
Formal letterpurpose, detail, requestthe requestformal and courteous
Reportheadingsthe findingsneutral and impersonal

This is the Plan stage of the Whole-Piece Check, this chapter's four stages - Plan, Draft, Revise and Proofread - with one decision added at its start: name the genre, and let it choose the outline. The argument outline belongs to Chapter 29, letter and report formats and register to Chapter 22, and voice to Chapter 28; this section only chooses among them.

Real tasks are often hybrids. The skill is deciding which genre is the frame and which works inside it.

The taskThe frame (outside)The genre insideWhat the reader must find
A letter to the editor arguing for more evening busesformal letterpersuasivethe claim in the first paragraph, reasons, a closing request
A report on sports day that recommends changesreporta short persuasive closeneutral findings first; recommendations only at the end
A reflective essay about the night of the floodreflective essaynarrativethe story told briefly, then what it meant
A magazine article about a hill town for visitorsarticledescriptive, with some expositionone dominant impression, plus the facts a visitor needs

Notice three things. First, the frame decides the outside shape - the salutation and formal close of a letter, the headings of a report - and the genre inside decides how the body is ordered. A persuasive letter keeps the letter's format and carries the argument outline inside it.

Second, the inner genre must stay in proportion. A reflective essay that tells its story for four paragraphs and reflects for one sentence has turned into a narrative with a moral attached. Tell the event only as fully as the reflection needs.

Third, the voice follows the frame wherever a reader would notice. A report that recommends may grow firmer in its last section, but it does not switch to I feel very strongly. A story inside a reflective essay may be vivid, but it is told in the same first-person voice that will reflect on it.

Picture a builder given three jobs on one street: a house, a shop and a footbridge. The same tools come out for each - the drawing, the foundation, the walls, the final inspection - but nobody builds a footbridge with bedrooms. Before the first line is drawn, the builder asks what the building is for and who will use it, and that answer sets the shape. A genre is that answer for a piece of writing: the tools are the same in every genre, and the shape is not.

Craft built the genres one at a time: section 22 outlined narrative, explanatory and persuasive pieces, and sections 26 to 29 assembled an expository essay, a persuasive essay, a formal letter and a report. Section 32 in this band held a reflective personal essay in one voice, and section 36 showed that under exam time the planning minutes must be fast. This section puts all seven side by side so the choice is quick and right. Section 38 asks the last question of any genre, is this piece finished, and section 40 asks you to choose a genre and run the whole process alone.

Here is a hard case, and it deserves care. Some tasks do not name a genre at all. "Write about a place that changed you" could be descriptive, narrative or reflective. The words write about decide nothing; the words that changed you decide almost everything. A pure description of the place would answer half the task, and a story of one visit a little more, but changed points to meaning, so the governing genre is reflective, with description and narrative working inside it. When a task is open, find the word that carries its purpose and let that word choose the frame. A second hard case: two genres pull against each other inside a hybrid. A report that ends with recommendations is partly persuasive, and a writer who cares about the outcome is tempted to argue in the findings. Do not. The findings are where a report earns trust; a recommendation is believed because the findings before it were neutral. Persuasion lives in one place, at the end, and it is still worded as a recommendation, not a plea. And a third: a convention that seems to fight the governing idea. A story about one quiet afternoon may have no dramatic turn. Is it then a failed narrative? Not necessarily. The turn in a quiet story can be small - a realisation, a word said or not said - but something must change between the first paragraph and the last, or the piece is a description told in the past tense. Name what changes. If nothing does, the piece may honestly belong to another genre, and the plan should say so.

Did you know that the word genre simply means kind? It came into English from French, where genre means a kind or a sort, and it goes back to the Latin genus, the same word biologists still use for a group of closely related species. To ask a piece's genre is to ask what kind of thing it is, before asking whether it is good. Some genre names carry their jobs inside them. Essay comes from the French essai, an attempt or a trial, from essayer, to try; the French writer Michel de Montaigne gave the title Essais to the collection of short prose pieces he published in 1580, pieces in which he tried out his thinking on a subject. Report goes back to the Latin reportare, to carry back: a report carries back to its readers what was found. Narrative comes from the Latin narrare, to tell. So each name is a small reminder of the outline. An essay tries out an idea, a report carries back its findings, a narrative tells what happened. Choose the kind first, and the name already tells you half the shape.

Genre shapes are conventions, not laws. A skilled writer may bend one on purpose, but only once the reader can tell which genre is being bent. Formats for letters and reports vary between schools, boards and workplaces. Follow the format your task or teacher gives; the outlines here are the common core, not any paper's rule. Many real pieces are hybrids. Naming the frame and the inner genre is more useful than forcing a piece into one box. Genre chooses the outline, not the quality. A piece can have exactly the right shape and still need its plan, its revision and its proofread.

  1. Underline the task word, the reader and the form before planning.
  2. Write the genre in the margin in one word, and the frame as well if the task is a hybrid.
  3. Choose the outline shape from the genre, not from the last piece you wrote.
  4. Name the governing idea in the genre's own terms: a change, an impression, a thesis, a claim, an insight, a request or a finding.
  5. Decide the voice in a phrase before the first sentence, and hold it (Chapter 28).
  6. In a letter or a report, follow the format the task gives and Chapter 22 teaches.
  7. In a persuasive piece, use Chapter 29's argument outline inside whatever frame the task sets.
  8. Keep an inner genre in proportion to the frame's purpose.
  9. In the revise stage, read once for genre alone: does each paragraph do what this reader expects?
  10. Run all four stages of the Whole-Piece Check in every genre, without exception.
Genre decides the outline's shape and the voice; it never decides whether the four stages happen. 1. Read the task word, the reader and the form; name the genre in one word. 2. Choose the shape: a narrative by time to a change; a description by space around one impression; an explanation by logic under a thesis; an argument by claim, reasons, objection and close; a reflection from experience to meaning; a formal letter with its purpose first and its request last; a report under headings, findings before recommendations. 3. Fix the voice: neutral for a report, formal for a letter, honest first person for a reflection. 4. In a hybrid, let the frame hold the outside shape and the inner genre order the body. Siddharth used one essay template for a story, a letter and a report, and each one failed its reader. Maryam named the genre first, and each piece did its job. Next: is this piece finished?

Is this piece finished?

A piece is finished when the Whole-Piece Check has actually been run on it and it still serves its one governing idea. A piece has merely stopped when something else - the clock, the word count, tiredness - made the decision.

Finished, or stopped? Four questions.

1. Has every stage been run on this version - Plan, Draft, Revise and Proofread?
2. Does every paragraph still serve the governing idea from the plan?
3. Does the ending close what the opening opened, with no loose ends left?
4. Would the next change make the piece better for its reader, or only different?

Yes, yes, yes and only different: it is finished. Leave it.
The signWhat it tells youFinished or stopped?
The time is upa fact about the clockneither, by itself
The word count is reacheda fact about lengthneither, by itself
I am tired of looking at ita fact about the writerneither, by itself
The proofread was skippeda stage was never runstopped
The thesis promises three reasons; the body gives twoa loose endstopped
Every stage run; every paragraph serves the idea; the ending answers the openingthe check has been passedfinished

This is the last question of the Whole-Piece Check, this chapter's four stages - Plan, Draft, Revise and Proofread: does the piece still serve the one governing idea it began with? If yes, it is finished. The proofread that comes just before it is Chapter 23's Four-Pass Check, and it must be run again on anything you change afterwards.

Past the finishing point, revision changes direction. It stops repairing the piece and starts wearing it away. Learn the commonest late changes and what they cost.

The late changeWhat it does to a finished pieceThe question that catches it
A specific detail replaced by a general phrasethe paragraph loses the evidence that made it realWhat does the reader now see that they saw before?
A new idea added to the conclusionthe piece now serves two ideas, and argues one of them in a single sentenceIs this in the plan's governing idea?
A plain voice made grander or more formalthe voice the whole piece was built in wavers or stiffensWould the writer of paragraph 1 say this?
Words swapped and swapped backtime is spent and nothing is gainedIs this better, or only different?
A reviewer's taste followed against the piece's purposethe piece drifts toward someone else's essayWhat problem does this change solve?

Notice three things. First, every one of these changes feels like work, and that is why it is tempting. A writer who is anxious about a piece wants to keep touching it. The questions in the table turn that feeling into a test.

Second, keeping a finished choice is an active decision, not laziness. Editors have a mark for it: stet, written beside a crossed-out phrase to mean let it stand. A writer who can say why the plain ending, the short paragraph or the specific detail stays is still in control of the piece.

Third, the one change that is never over-revision is a real repair: a slip the proofread finds, a loose end the check exposes, a sentence that contradicts the thesis. Finished does not forbid change. It asks every change to solve a problem the reader would actually have.

Picture a potter at the wheel. Shaping the clay is the work, and stopping too early leaves a lopsided bowl. But a potter who keeps pressing after the bowl is true does not make it truer; the walls thin, the rim wobbles, and the bowl collapses back into clay. A skilled potter knows the moment the bowl is finished, lifts their hands away, and lets it dry. Finishing a piece of writing is that lifting of the hands: a decision made on purpose, at the right moment, for a reason.

Discover's section 9 asked a simple question: is it finished? It gave a checklist of a beginning, a middle, an end, a title and one read-through. Section 31 in this band ran the full Whole-Piece Check on a demanding piece, section 33 revised a full draft and section 36 kept every stage alive under an exam clock. This section asks the Discover question again at Master level: finished is now a judgement you make and defend, and over-revision is a fault as real as stopping too soon. Section 39 reads one complete published piece start to finish, and section 40 asks you to finish one of your own without help.

Here is a hard one, and it is worth taking slowly. How do you tell the last useful revision from the first harmful one? The changes look alike: both swap words, move sentences, cut and add. The difference is what they answer to. A useful revision answers a problem a reader would have - a paragraph that does not serve the thesis, a link that is missing, a sentence that cannot be followed. A harmful revision answers the writer's restlessness. When you cannot name the reader's problem in one sentence, you are usually past finished. A practical sign: your changes start undoing each other, or each new version is a little less specific than the last. A second hard case: the deadline arrives before the piece is finished. Then honesty matters more than the label. Use the remaining minutes on the parts that decide whether the piece works - the ending, a promised paragraph in short form - and do not pretend the piece is finished to yourself. Knowing exactly what is missing is what lets you finish it properly later, if the work allows, and stops you calling the same gap finished next time. And a third: is a piece ever finished if it is not perfect? Yes. Finished means complete for its reader and its purpose, not flawless for every possible reader. A paragraph that is good rather than brilliant, a word that is plain rather than striking, an example a reviewer would not have chosen - none of these reopens a finished piece. A contradiction, a missing part or a slip does.

Did you know that the word finish comes from the Latin word for a boundary? It reached English through French from the Latin finire, to bring to an end, from finis, an end or a limit. The same root sits inside define, which once meant to set the limits of something, and definite, which describes something whose limits are clear. To finish a piece, in the oldest sense, is to draw its boundary: to decide where it ends. And here is a second surprise. Editors and proofreaders have long used a mark for undoing a change: the Latin word stet, which means let it stand. When a correction has been marked on a page and someone decides the original was right after all, they write stet beside it, often with a row of dots beneath the crossed-out words, and the original returns. So the profession that exists to change writing keeps a special word for not changing it. Knowing when to write stet is as much a part of finishing a piece as knowing what to cut.

Finished is judged against the piece's own reader and purpose. A piece can be finished and still not be the best thing its writer will ever write. Some tasks really do end at a deadline. Then the honest description is what matters: say, at least to yourself, what is still missing. Not every late change is over-revision. A slip the proofread finds, or a sentence that contradicts the thesis, should always be fixed, and then proofread again in place. The four questions in this section summarise the Whole-Piece Check; they do not replace it. If a stage was never run, run it before asking whether the piece is finished.

  1. List the four stages and tick only the ones you actually ran on this version.
  2. Reread the governing idea from your plan before judging the whole piece.
  3. Read the opening and the ending one after the other.
  4. Search the draft for placeholders, half-sentences and promises the plan made.
  5. Treat the time, the word count and tiredness as facts, not verdicts.
  6. For every late change, name the reader's problem it solves in one sentence.
  7. If a change makes the piece only different, do not make it.
  8. Keep a finished choice for a stated reason: stet.
  9. Proofread again anything you change after the proofread.
  10. When a deadline wins, protect the ending and say honestly what is missing.
Finished is a judgement, not a moment on the clock. 1. Has every stage been run on this version - Plan, Draft, Revise, Proofread? 2. Does every paragraph still serve the governing idea? 3. Does the ending close what the opening opened, with no loose ends? 4. Would the next change make the piece better for its reader, or only different? Yes, yes, yes and only different: it is finished. Leave it. Stopped is not finished. Ritvik's essay reached 600 words with a paragraph broken off; the clock decided, not the check. Over-revision is a fault too. Shalini's Monday essay was finished; by Thursday its specific detail had become a generic phrase and its ending a slogan. Finished means complete for its reader and purpose, not perfect. Keep a finished choice with a reason, and proofread again anything you change. Next: close reading: one complete piece, start to finish.

Close reading: one complete piece, start to finish

Close reading a complete piece means reading it the way its writer built it: one governing idea, served by parts in order, each linked to the one before, and closed on purpose. The Gettysburg Address is short enough to hold in mind whole, and complete enough to show every stage.

Reading a whole piece, start to finish.

1. The governing idea: say what the whole piece is for, in one sentence of your own.
2. The shape: name its parts and the order they come in.
3. The links: find the words that carry each part into the next.
4. The sound: hear what repeats and where the weight lands.
5. The close: ask what the ending answers, and whether it adds anything new.

Quote only what you have checked, exactly as the copy prints it.
The partFrom the Bliss copyWhat it does
The past: the founding"Four score and seven years ago our fathers brought forth on this continent, a new nation, conceived in Liberty, and dedicated to the proposition that all men are created equal."names the nation and the idea it was founded on: the ground the whole speech stands on
The present: the war"Now we are engaged in a great civil war, testing whether that nation, or any nation so conceived and so dedicated, can long endure."Now turns the time; that nation, conceived and dedicated pick up the opening; it asks the question the speech must answer
The present: the place"We are met on a great battle-field of that war."narrows from the war to the field, and that war points back
The present: the occasion"It is altogether fitting and proper that we should do this."affirms the dedication the crowd came for, just before the speech turns away from it
The future: the taskthe last two sentences, which turn to the unfinished work and close on "government of the people, by the people, for the people, shall not perish from the earth"answers the second sentence: the nation can endure, if the living take up the task

Close reading is the Whole-Piece Check, this chapter's four stages of Plan, Draft, Revise and Proofread, run backwards by a reader: you find the plan inside the finished piece, trace how each part serves it and links to the one before, test whether anything fails to earn its place, and quote it as exactly as a proofreader would check it. The links are the work of Chapter 21, Transitions and Flow; the sound belongs to Chapter 24, Sentence Rhythm, and Chapter 26, Figures of Speech. This section only shows where each one appears in a whole piece.

One word comes back again and again in the Address: dedicate. Follow it, and you are following the governing idea.

WhereThe wordsWhat is being dedicated, and by whom
the first sentence"dedicated to the proposition that all men are created equal"the nation, to an idea, by the founders
the second sentence"so conceived and so dedicated"the same nation, now being tested by war
the fourth sentence"We have come to dedicate a portion of that field"the ground, by the crowd that has gathered
the sixth sentence"But, in a larger sense, we can not dedicate", followed by two more verbs of the same kind, consecrate and hallowthe ground again, which the living say they cannot truly dedicate
the ninth sentence"It is for us the living, rather, to be dedicated here to the unfinished work"the living themselves, to the task

Notice three things. First, the repeated word carries the argument. Dedication moves from the nation, to the ground, to the people standing on it. The speech never needs to say that its purpose has changed; the word's journey says it.

Second, the close returns to the opening. The first sentence speaks of a nation brought forth and conceived: the language of birth. Near the end the speech asks "that this nation, under God, shall have a new birth of freedom". The ending does not add a new idea. It completes the first one.

Third, each part is linked twice: once by a linker that names the relation, and once by content carried forward. Now names a turn in time; But names a contrast; rather names a correction. At the same time, that nation, that war and that field each repeat something from the sentence before. That is the Bridge Test from Chapter 21, working across a whole piece.

Picture a bridge seen from a riverbank. A visitor admires the carved lamps along its rail. An engineer looks at the spans: where each one rests, how the load passes from one span to the next, and why the last one lands exactly on the far bank. A close reader looks at a piece the way the engineer looks at the bridge. The lamps are real and worth admiring, but they are bolted to the spans, and it is the spans that carry the reader across.

Section 38 asked whether a piece of your own is finished or merely stopped. This section asks the same question of a piece by someone else, and answers it by reading the whole. Section 31 ran the Whole-Piece Check forwards, from plan to proofread; here you run it backwards, from the finished words to the plan they serve. Section 40, the Academy's capstone, asks you to run it forwards once more, alone, on a piece of your own that a reader could study this closely.

Here is a hard one, and it rewards patience. Where is the thesis? An essay usually states its governing idea near the start. The Address does not. Its first sentence gives the founding; its idea arrives in full only in the last two sentences. Is it, then, badly planned? No. Genre decides the outline's shape. A speech for an occasion can open with the ground it stands on and arrive at its purpose at the close, as long as every earlier part prepares for that close. Test it: read the last two sentences first, then each earlier one, and ask whether it leads there. Each one does. Here is a second. The eighth sentence says "The world will little note, nor long remember what we say here". A reader today may smile, since the speech is widely remembered. Is it a false claim that a careful writer should have cut? Read the sentence for its job in the piece. It sets what we say against "what they did here", and so moves the weight from the speakers to the dead, and from the dead to the task. It is judged by what it does in the whole, not as a forecast. And a third. Is the repetition wordy? Apply the Cut Test from Chapter 20. Cut or any nation so conceived and so dedicated, and the stakes shrink from every such nation to one. Cut by the people from the last phrase, and both the rhythm and the meaning drop. The Cut Test asks whether meaning, tone or emphasis is lost. Here emphasis is lost every time, so the repetition earns its place.

Did you know that the opening words are a sum? A score is twenty, so four score and seven is eighty-seven. Count eighty-seven years back from 1863 and you reach 1776, the year the Declaration of Independence was adopted. The founding the speech means is the one that declaration announced, and its first sentence borrows the declaration's own idea that all men are created equal. Lincoln could have written eighty-seven years ago. Readers often hear the older count as more solemn, and the same counting word appears in the King James Bible's "threescore years and ten". And here is a second surprise. The speech never names the town, the battle, a general, a state or a side. It says "a great civil war", "a great battle-field", "that nation". Leaving out every particular keeps each sentence on the one governing idea, and it lets the speech speak to any reader, in any place, who asks whether such a nation can last.

This section quotes only the Bliss copy, the version Lincoln wrote out in 1864. Other copies in his hand exist, and a reader checking another copy may find small differences in wording. The Bliss copy uses dashes in its later sentences. Those spans are described here, not quoted, so that no quotation is altered. The shape named here is what the text shows. It makes no claim about how Lincoln drafted, revised or delivered the speech. Close reading reads the words of a piece. It is not a judgement on the war, its causes or its politics, and every student and every other passage here is invented.

  1. Read the whole piece aloud once, straight through, before marking anything.
  2. State its governing idea in one sentence of your own words.
  3. Name the parts in order, and say where each one begins.
  4. Find the words that carry each part into the next, and name the relation they show.
  5. Follow any word that keeps coming back, and ask what changes each time.
  6. Ask what the close answers, and check that it adds nothing new.
  7. Hear where the weight lands; name Chapter 24 or Chapter 26 when you describe it.
  8. Try the Cut Test on anything that seems repeated, and say what would be lost.
  9. Quote exactly as the copy you name prints it: never correct a quotation.
  10. Write your own sentences about the piece in standard English, and proofread them.
Close reading a complete piece means reading it the way its writer built it. 1. Read the whole piece aloud once, without marking. 2. State the governing idea in one sentence of your own: the living honour the dead by taking up the unfinished work, so that the nation endures. 3. Map the shape: the past (the founding), the present (the war, the field, the dedication), the future (the task remaining). 4. Find the links: Now, that nation, that war, that field, But, rather. 5. Follow the repeated word: dedicate moves from the nation, to the ground, to the living. 6. Read the close: it answers whether the nation can long endure, and it returns to the image of birth. 7. Quote exactly from one named copy. The key contrasts: the topic is not the governing idea; a list of lines is not a reading; a close that answers is not a close that adds; repetition that carries the idea is not wordiness. Next: the Academy's capstone.

The Academy's capstone

Independent means that nobody hands you the next step. You name each stage yourself, reach for each tool yourself, and decide for yourself when the piece is finished.

The capstone, alone.

1. Read the prompt for its reader, purpose, genre and limit.
2. Plan: one governing idea, and the paragraphs sketched in order.
3. Draft: each paragraph to its plan, linked to the one before.
4. Revise: voice, rhythm, punctuation and length, then cut what does not earn its place; run the Argument Check if it argues.
5. Proofread: the Four-Pass Check on every line, the last paragraph too.
6. Ask the last question, and answer it.

One line in the margin for every decision: the stage, what you did, and why.
StageRhea's decisionHer one-line defence
Planreader: adult residents; purpose: to persuade; genre: argumentative essaythe prompt names the reader and asks should, so the piece argues and takes the essay outline from Chapter 29
Planthesis: restore the step-well as a working part of the town, not as a monumentone claim every paragraph can serve; a fact such as the well is old could not govern an essay
DraftParagraph 3 opens: That stored water is also the answer to the cost.a content link: it picks up Paragraph 2's water and names the relation, the Bridge Test from Chapter 21
Revisecut the paragraph on the builder's familytrue and well written, but it serves no part of the thesis: the Cut Test from Chapter 20
Revisekept the two-word sentence Water stays.the Read-Aloud Test from Chapter 24: the weight lands where the argument turns
Proofreadit's for its in Paragraph 4, and the steps was for the steps werethe Four-Pass Check from Chapter 23: the mark pass and the verb pass
Then askfinishedevery paragraph serves the thesis, and the conclusion answers the prompt's question without a new idea

The capstone is the Whole-Piece Check, this chapter's four stages of Plan, Draft, Revise and Proofread, run end to end by you alone, with one habit added: a line in the margin for every decision. The tools stay where the other chapters left them: the Paragraph Check in Chapter 27, the Bridge Test in Chapter 21, the Read-Aloud Test in Chapter 24, the Effect Test in Chapter 25, the Cut Test in Chapter 20, the Voice Check in Chapter 28, the Reader Test in Chapter 22, the Four-Pass Check in Chapter 23 and the Argument Check in Chapter 29.

What changes when nobody helps? Every stage is still there. What disappears is the person who used to start it for you.

StageWith a scaffoldAlone
Planan outline is handed to youyou write the reader, purpose, thesis and paragraph sketch yourself, from the prompt
Drafta model paragraph shows the wayyou hold each paragraph against your own plan as you write it
Revisea reviewer points to what to changeyou read the whole piece for voice, rhythm, punctuation and length, and choose what to cut
Proofreadslips are underlined for youyou run the Four-Pass Check on every line, the last paragraph included
Finishedthe teacher says stopyou ask the last question and answer it

Notice three things. First, a defence names the stage, the reason and the home chapter. It sounds better is not a defence; the Read-Aloud Test puts the weight where the argument turns (Chapter 24) is.

Second, keeping is a decision too. Rhea tested the two-word sentence Water stays. against the Read-Aloud Test and kept it, and that keep gets a line in the log just as a cut does. A log with only changes in it hides half the judgement.

Third, when the last question is answered no, go back to the stage that failed, not to the commas. A conclusion that adds a new idea is a plan problem: return to the thesis. A link that names the wrong relation is a drafting problem: return to the Bridge Test. A slip in the last paragraph is a proofreading problem: run the passes again, from the top.

Picture a pilot flying alone, working through a written checklist before take-off, item by item: fuel, checked; flaps, set; instruments, checked. Nobody is watching. The checklist is not there because the pilot is weak. It is there because even a skilled pilot knows that memory skips things under pressure. Your decision log is that checklist, written as you go: each stage named, each answer recorded, so that nothing is skipped and every choice can be explained afterwards.

This section uses everything before it and adds nothing new. Section 31 ran the Whole-Piece Check in full, with guidance; sections 33 and 34 revised and proofread a full draft; section 35 added a reader; section 36 put the process under a clock; section 37 fitted it to a genre; section 38 judged when a piece is finished; section 39 read one finished piece closely. Here you do all of it at once, alone, on a piece of your own, with a chapter behind every stage.

Here is a hard one, and no checklist settles it for you. The thesis changes while you draft. Halfway through her third paragraph, Rhea realised her real argument was not restore the step-well for its history but restore it as a working part of the town, because it still stores rain. Should she force the draft back to the old plan, or change the plan? Change the plan, on purpose. She rewrote the thesis, then held every paragraph against the new one. Her history paragraph now had to earn its place, so she cut it to two sentences that explain why the well was built to hold water. Her log says: Plan, rewritten mid-draft; Paragraph 2 trimmed to serve the new thesis. Here is a second. The best paragraph is off the thesis. Rhea's paragraph on the family who built the well was the finest writing in her draft. It served no part of her argument. She cut it, and kept it in her notebook for another piece. Many writers keep a file of cut passages for exactly this reason: cutting from one piece is not throwing away. And a third. When do you stop? With ten minutes left, Rhea could have gone on swapping one good word for another. But each change now altered her taste, not whether the piece served its thesis. Finished is not perfect. It means the check has been run and passed. When another pass would only trade one good word for another, ask the last question, answer it, and stop.

Did you know that the names of the stages carry their own instructions? Draft is the same word as draught, and both go back to the verb draw: a draft is something drawn up, as a plan or a sketch is drawn. Revise comes from the Latin revisere, to look at again. And a proof, in printing, was a trial copy taken from the type so that it could be checked before the full run was printed; proofreading first meant reading that trial copy. Draw it up, look at it again, test it before it goes out: the words have held the process for centuries. And here is a second surprise. Professional writers rarely work alone at the end, even when they wrote alone. Editors, sub-editors and proofreaders in newspapers and publishing exist because no writer, however skilled, sees every slip in their own work. Unassisted is a test of whether you can run the process yourself. It is not a claim that good writers never need a reader.

Rhea's sixty minutes, with eight for the plan and twenty-five for the draft, is one sensible split, not a rule. A prompt with a different limit needs a different split, but every stage still happens, as section 36 showed. A decision log is a habit for practice and for real pieces. In many exam halls there is no room for it, and the decisions are made in your head; make them anyway. Unassisted means no help while you write. It does not mean a finished piece should never meet a reader afterwards. Rhea, Arnav, the town and its step-well are invented. Step-wells are a real kind of structure found in many parts of India, but no real well, town or newsletter is described here.

  1. Read the prompt twice: underline the reader, the purpose, the genre and the limit.
  2. Write the Whole-Piece Check from memory at the top of the page before you plan.
  3. Write a one-sentence thesis that every paragraph can serve, and sketch the paragraphs in order.
  4. Split your time so every stage fits, and keep minutes back for revising and proofreading.
  5. Draft each paragraph to its plan, and link it to the one before by content as well as by a linker.
  6. If the thesis changes while you draft, rewrite it on purpose and recheck every paragraph.
  7. Revise the whole piece before the sentences: unity and order first, then voice, rhythm, punctuation and length.
  8. Run the Four-Pass Check on every line, the last paragraph included.
  9. Log every decision, keeps as well as changes, with the stage, the reason and the home chapter.
  10. Ask the last question and answer it; if the answer is no, go back to the stage that failed.
The Academy's capstone is one prompt, one complete piece, and no help. 1. Read the prompt for its reader, purpose, genre and limit. 2. Plan: one governing idea, and the paragraphs sketched in the genre's order. 3. Draft: each paragraph to its plan, linked to the one before. 4. Revise: the whole piece first, then voice, rhythm, punctuation and length; cut what does not earn its place; run the Argument Check if it argues. 5. Proofread: the Four-Pass Check on every line. 6. Ask the last question. If yes, it is finished. If no, go back to the stage that failed. The key contrasts: a strong draft is not a finished piece; changing the thesis on purpose is not drifting from it; a kept sentence is a decision too; a defence names a stage, a reason and a chapter, never a feeling; finished means checked, not perfect, and not merely out of time. Arnav wrote well and stopped. Rhea planned, drafted, revised, proofread, and finished. Next: the Grammar Academy.

Common questions about putting it all together

What are the stages of the writing process?
Four stages, every time. Plan: decide the reader, the purpose and one governing idea, and sketch the paragraphs. Draft: write each paragraph to its plan and link it to the one before. Revise: read the whole piece for voice, rhythm, punctuation and length. Proofread: check every sentence for slips. Then ask whether it still serves its governing idea.
What is a thesis statement?
A thesis statement is the one sentence that says what a whole piece will show. "Our town should keep its library because it helps people learn, meet and read for free" governs every paragraph after it. It is different from a topic sentence, which governs only one paragraph. A bare fact, such as "Libraries have books", is not a thesis.
How do I make an outline before writing an essay?
Write one short line for each paragraph, in order: an opening line, one line for each main idea of the body, and a closing line. Then check three things. Does every line serve the main topic? Do any two lines repeat an idea? Does the order make sense for this kind of piece? Fix the outline before you draft.
How do I write a good introduction paragraph?
An introduction opens the whole piece. It names the topic, gives the reader a reason to read on, and shows what the body paragraphs will cover. In an essay it usually ends with the thesis. "Every morning our street fills with traffic, and three changes could make it safer" opens a piece and signals its plan.
What should a conclusion paragraph do?
A conclusion closes the whole piece, not just the last paragraph. It returns to the main idea, sums up the main points and gives a final thought. It should not add a new argument. "In conclusion, it is good" closes nothing; a conclusion that answers the opening and gathers every point does.
How do I know when my writing is finished?
A piece is finished when it has been checked, not when you stop. Ask five things. Does the beginning say what it is about? Does the middle stay on one topic? Does the end answer or close the beginning? Does the title fit? Have you read it aloud and fixed every bump? If all five are true, it is finished.
What is a beginning, a middle and an end in a story?
The beginning starts the story and says who and where. The middle tells the most: the problem and what happens, in order. The end tells how it turned out and answers the problem. Writing "The End" is a label, not an ending; the ending is the sentence that says what happened at last.
How should I divide my time in a timed writing task?
Give every stage some of the time. For one piece, a sensible start is about a tenth to plan, most of the time to draft, and a few minutes each to revise and proofread. Write the finishing time for each stage in the margin. If time runs short, protect the ending and the thesis before any polish.
How do I link paragraphs in an essay?
Begin each paragraph with a link that names how it relates to the one before: adding, contrast, cause, result, time or example. "However" signals a turn; "As a result" signals a consequence. The link should also point back to the thesis. Chapter 21, Transitions and Flow, owns the linking words; this chapter uses them across a whole piece.
How do I give useful feedback on someone else's writing?
Read the whole draft first. Then give two or three comments, starting with the largest: the governing idea, the structure, anything that wanders, then style, and proofreading last. Make each comment specific, kind and usable: where it is, what you experienced as a reader, and a suggestion or question. Say what works as precisely as what does not.
What is the difference between revising and proofreading?
Revising changes the piece: its thesis, its structure, the order of paragraphs, what is cut and how it sounds. Proofreading checks the finished piece for slips in grammar, spelling and punctuation. Revise first, from the largest level down, then proofread, because a sentence you correct today may be cut tomorrow.
How do I come up with ideas before I write?
Brainstorm in two halves. First collect many ideas quickly without judging them, using a list, an idea web or the question words who, what, where, when, why and how. Then choose: keep the idea that fits the task and the reader, gives enough to say, and is small enough to finish.
What should I decide before I start writing?
Decide two things before the first sentence: who the piece is for, and what it is about. Write them as two short lines, For and About. The About line decides which ideas belong; the For line decides how you say them. A small topic, such as "our cat's first bath", keeps a short piece whole.
How do I stop my writing from going off topic?
Check every sentence against your About line, not just against the sentence before it. A piece can drift one step at a time: a kite, the sky, aeroplanes, a trip abroad. Each step follows the last, but the piece has left its topic. Cut back to the first sentence that drifted.
Why should I read my writing aloud?
Reading aloud makes your voice meet every word, so it catches slips your eyes skip: a doubled word, a missing word, a sentence that does not make sense, or a jump where a step is missing. Point to each word as you read, mark each place you stumble, then fix it and read the whole piece again.

Keep exploring

▶ Practise putting it all together in the interactive lesson
Other guides: NounsVerbsAdjectivesPronounsArticles & DeterminersAdverbsPrepositionsSentence StructurePhrasesClauses & ConjunctionsSentence TypesSubject-Verb AgreementTense ConsistencyPronoun AgreementParallel StructureCommas & SemicolonsApostrophes & QuotesCapitalizationHyphens & DashesConcisenessTransitions & FlowTone & RegisterError SpottingSentence RhythmPunctuation & Editing for EffectFigures of SpeechParagraph ArchitectureVoice & StyleArgument & Persuasion · All 30 chapters
Copyright © Pawan Nayar · LLOS.ai · 2026 — Original pedagogy, voice, and design — all rights reserved. This is the readable guide; the interactive lesson lives at putting-it-together.html.