Quick answer: when a child tells a lively story aloud but produces a flat, short or confusing written version, the problem may not be imagination. The bottleneck may be oral-to-written transfer: selecting the useful parts of spoken thought, ordering them, making references explicit, compressing repetition and encoding causal relationships for a reader who cannot interrupt to ask questions.
This historical Simei Creative Writing page now owns that specific craft problem. It is not a claim of a current Simei eduKateSG centre. Its job is to show how speech can be used as a diagnostic and rehearsal layer without turning writing into dictation or letting the tutor become the real author.
The one-sentence answer
Let the student tell the event naturally, mark the meaningful state changes, select the causal spine, compress spoken repetition, make missing links explicit, and then write for a silent reader rather than transcribing speech word for word.
Speaking and writing share language but impose different constraints
Conversation is interactive. A listener can interrupt with “Who?”, “Why?”, “Then what happened?” or “Do you mean before or after?” The speaker can repair gaps immediately with gesture, tone, repetition or clarification.
A page cannot ask. The writer must anticipate the reader’s needs.
| Speech can rely on… | Writing must often supply… |
|---|---|
| gesture and facial expression | clear verbal reference |
| listener questions | anticipated explanation |
| restarts and self-correction | stable sequence |
| repetition for emphasis | selection and compression |
| shared physical context | enough orientation for an absent reader |
| tone of voice | word choice, syntax and punctuation |
A child can therefore possess strong narrative intelligence while still struggling to encode it on the page.
The first diagnostic: can the child tell the story?
Before assigning another full composition, ask for a sixty- to ninety-second oral version. Do not polish it. Listen for the architecture underneath the speech.
- Is there a stable starting situation?
- Does something create pressure?
- Can the child explain why the character acts?
- Are important objects or information tracked?
- Does the decision produce a consequence?
- Does the ending grow from what happened?
If these elements exist orally but disappear in writing, the intervention should target transfer. If they are absent orally too, the student may need story architecture or prompt interpretation before compression work.
Oral rehearsal is not dictation
The student should not simply transcribe the spoken version. Spoken language contains fillers, repeated subjects, vague pronouns and side-tracks that are natural in conversation but costly on the page.
Spoken:
So I was going to the library and then I saw this boy, like he was looking everywhere and he looked really worried, and then I saw this wallet under the chair.
Written:
On my way to the library, I noticed a boy searching beneath the tables. A wallet lay under the chair beside me.
The written version does not copy every word. It preserves the useful distinctions and removes spoken noise.
Build a spoken causal spine first
A useful rehearsal can be reduced to six moves:
- At first…
- But then…
- Because…
- So I decided…
- That caused…
- By the end…
If the child cannot complete the sequence coherently, the story may not yet be structurally ready. If the child can, the written draft has a backbone.
Mark state changes, not every spoken event
Children often tell stories with many small actions. Writing should preserve the actions that change knowledge, goal, relationship, possession, risk or consequence.
For a lost-wallet story, the state changes might be:
wallet belongs to stranger → wallet found → temptation appears → decision made → wallet returned → relationship or self-understanding changes.
The walk to the teacher’s desk may need one sentence. The decision to return the wallet may need a paragraph. Compression follows importance.
Compression is selection, not shortening for its own sake
A shorter sentence is not automatically better. The question is whether the sentence preserves the distinction the reader needs.
Useful information includes:
- cause;
- change;
- motive;
- evidence;
- constraint;
- consequence;
- relationship.
Spoken repetition may be removed; missing causal links may need to be added. Compression can therefore make a text slightly longer in one place and much shorter in another.
Pronouns expose oral-to-written transfer problems
In speech, a child can point or rely on shared context: “Then he gave it to him.” On the page, the reader may not know which person is which.
A useful revision check is to circle every he, she, it, they, this and that. Ask whether a silent reader can identify the reference immediately. If not, restore the noun or restructure the sentence.
“Then” often hides missing relationships
Young writers may transfer spoken sequencing directly into writing:
Then I ran there. Then I saw her. Then she cried. Then I helped her.
The timeline is visible, but the relationships are not. Revision asks which connector the meaning actually requires:
- because for cause;
- although for contrast;
- when for time relationship;
- so for consequence;
- but for changed direction.
The goal is not to ban “then”. It is to stop using sequence as a substitute for logic.
Paragraph boundaries turn spoken flow into readable units
Speech can run continuously while listeners use intonation and pauses to organise it. Writing needs visible units. A new paragraph may mark a shift in action, speaker, time, location, focus or story state.
After oral rehearsal, students can label the main moves and let those moves guide paragraph boundaries. The page becomes easier to follow because the structure is externalised.
Oral rehearsal can expose missing mechanism
A student says, “Then the teacher knew the wallet belonged to the boy.” Ask one small question: how?
The child may realise that the wallet contained a student card, the boy had already reported it missing, or the name matched a label. The missing mechanism must enter the written story.
This is why speaking can be diagnostic. The mouth often reveals a logical jump before the red pen does.
Oral rehearsal can expose implausible motive
If a student cannot explain aloud why a character suddenly lies, runs away, confesses or helps, the written version will probably feel forced. Ask the student to explain the choice conversationally. If the explanation requires a fact the story never established, add the fact or change the decision.
Do not let the tutor become the hidden author
Oral discussion can easily turn into adult plot construction. Twenty leading questions may produce an excellent story that the student did not truly generate.
Use a few discriminating prompts:
- What changed?
- Why did the character do that?
- What would the reader not know yet?
- What consequence follows?
The child supplies content. The adult helps expose structure.
Worked example: oral story to paragraph plan
A student tells a lively story about finding a wallet and returning it. The oral version includes a long account of walking to the library, a funny aside about lunch, repeated explanations of the boy’s worried face and a good decision moment near the end.
The transfer process might become:
- Paragraph 1: notice worried boy and find wallet.
- Paragraph 2: identify what is inside and recognise the temptation/responsibility.
- Paragraph 3: make decision and act.
- Paragraph 4: return wallet and reveal ownership.
- Paragraph 5: consequence and changed relationship or understanding.
The lunch aside disappears because it does not change the causal spine. The worried behaviour remains because it helps the reader connect the owner to the missing object.
Worked example: the child has ideas but writes too little
A Primary 4 student can orally explain that two friends quarrel because one thinks the other lost a key card. On paper, the child writes: “We argued. I was angry. Later we found it and became friends again.”
The missing work is not more vocabulary. The writer has compressed away the evidence, accusation, decision and discovery.
Return to the spoken version and mark the load-bearing distinctions:
- what evidence caused the accusation;
- what each friend believed;
- what action made the argument worse;
- what discovery disproved the assumption;
- what changed after the apology.
The student now has reasons to write more because each added sentence restores a distinction the story needs.
P3: say the whole event before writing it
Primary 3 students benefit from short retells that stabilise beginning, problem, action and outcome. The goal is sequence, not polished speech.
P4: compress repetition and clarify reference
Primary 4 students can notice repeated “then”, ambiguous pronouns and spoken loops, then convert them into clearer written relationships.
P5: rehearse motive, evidence and consequence
Primary 5 students can use oral rehearsal to test whether decisions are believable and whether the story contains enough evidence for later inference.
P6: compress the rehearsal itself
By Primary 6, the rehearsal may become a rapid internal causal check rather than a long discussion. The purpose is to reduce mid-draft repair and preserve writing time. Current examination details should always be verified against the latest SEAB materials for the relevant cohort.
A three-student oral-to-written exercise
In a 3-pax group, one student tells, one listens for missing causal links and one listens for repetition or unclear reference. Roles rotate. Then all three compress the same oral story into one sentence.
Different one-sentence versions reveal what each learner believes is load-bearing. The tutor can ask which distinctions must survive into the written draft.
Common transfer failure modes
| Symptom | Likely problem | Repair |
|---|---|---|
| Oral story lively, written story tiny | Over-compression | Restore state changes, evidence and consequence |
| Writing sounds like transcript | Insufficient compression | Remove fillers, repetition and spoken repair loops |
| Too many “then” links | Relationships not encoded | Specify cause, contrast, time or consequence |
| Pronouns become confusing | Shared context lost | Restore nouns or restructure reference |
| Plot jump appears on page | Listener supplied missing logic during speech | Add the mechanism explicitly |
| Tutor questions create most ideas | Support has become authorship | Reduce prompts and return content decisions to learner |
The silent-reader test
Give the paragraph to someone who did not hear the oral rehearsal. Can they reconstruct who did what, why it happened and what changed? If not, the writing is still relying on information that existed only in the conversation.
The compression test
Can two spoken sentences become one written sentence without losing a necessary distinction? If yes, compress. If one spoken sentence actually contains two important events or ideas, the written version may need expansion instead.
The transfer test
After practising with one story, give a new prompt and ask the student to complete the same process with less adult prompting. If the child can tell, mark, select, write and tighten independently, the bridge is becoming portable.
AI should not become the transcription and compression engine
AI can turn spoken notes into polished prose instantly. That may be convenient, but it removes the exact selection and encoding decisions this lesson is trying to teach.
A bounded use is:
student tells and writes independently → identifies one transfer problem → AI may compare two compressed versions or highlight ambiguity → student explains the difference → tool removed → new oral-to-written task completed independently.
The learner’s receipt is the next page, not the polished AI version.
What progress should look like
- Written stories retain more of the useful oral causality.
- Spoken repetition reduces during writing.
- Pronoun reference becomes clearer.
- Timeline jumps reduce.
- “Then” is replaced selectively by clearer logical relationships.
- Paragraphs correspond more closely to state changes.
- Students need fewer tutor questions to expose missing mechanism.
- New prompts can be rehearsed and encoded independently.
Questions parents and teachers can ask
- Can you tell me the story before writing it?
- What changed in your spoken version?
- Which part of the speech is repetition rather than story information?
- What would a silent reader not understand?
- Where does “then” hide a stronger relationship?
- Which sentence needs a clearer noun instead of “he”, “she” or “it”?
- What can be compressed without losing the causal spine?
Frequently asked questions
Is oral rehearsal suitable for every child?
It is particularly useful when spoken narrative is stronger than written output. Some students think more clearly through notes or diagrams; the diagnostic should follow the learner’s evidence rather than one universal routine.
Will speaking first make writing slower?
Long rehearsal can. The goal is to shrink the support over time. By upper Primary, the causal rehearsal should become quick enough to save drafting time.
Should parents write down the child’s spoken story?
Not as a default. That turns the adult into the encoder. Better to help the child identify the main state changes and let the child create the written representation.
Related eduKateSG writing routes
For composition architecture, continue to Build Composition Architecture Before Polishing Sentences. For controlling the first paragraph, see Scene Entry Control. For checking the actual written structure after drafting, see Reverse-Outline Verification.
Current eduKateSG class information
Selected Primary English and Creative Writing students are taught in focused 3-pax small groups, typically 1.5 hours. Current centres listed on eduKateSG are 83 Punggol Central, Singapore 828761 and 8 Fourth Avenue, Singapore 268674. This historical Simei URL is not a claim of a current Simei centre.
2017 photo archive
The original Simei page contained the photographs below. They are retained as historical archive material and do not establish a current Simei centre, current class size or current programme configuration.
The deeper principle
Speech is not a rough draft to be copied. It is a high-bandwidth diagnostic surface. Let the learner speak enough to reveal sequence, motive and causality; then require the writer to choose what a silent reader needs, compress what conversation can afford to repeat, restore the links speech left implicit, and carry the same story intelligence into controlled written form.
First published 31 August 2017. Rebuilt in 2026 as an extended Simei creative-writing guide to oral rehearsal, written compression and transfer. The original historical photo archive is retained in full.