To encode facts in a story, pin each fact to its own plot event, keep exact figures out of the narrative and in a fixed refrain, and then test the story by passing it through three successive retellings and counting which facts survive. The facts that drop out mark the weak events, and the facts that are altered mark the places where the story has become more plausible than true.
Part of Narrative Embedding
Encoding facts in a story is a design-and-test procedure: assign each required fact a unique plot slot, then measure how the story degrades under retelling.
A Fact Needs Its Own Slot
A story is a chain of events. A fact survives retelling if it is the cause or consequence of an event that the plot cannot do without. If a fact is only mentioned in passing, a teller can drop it and the story still works, so it will be dropped. The design goal is that removing any fact breaks the plot.
What You Need
- The fact list — each fact in one line, marked FIXED (must not change) or FREE
- A cast and a place — small and consistent (Character-Fact Association)
- A refrain — a short, rhymed or rhythmic line for the FIXED figures
- Three testers — people who have not heard it
Method: Peg, Refrain, Chain Test
Step 1: Order the facts into a causal chain. Each fact should either cause the next event or be needed to understand it. If two facts have no causal link, tell them in two stories or add a plot reason.
Step 2: Make each step an irreversible event. A character who skips the step must fail visibly (the water is still foul, the seed rots). Failure makes the step necessary, and a necessary step is remembered.
Step 3: Move figures to the refrain. The plot says “he boiled it as the rhyme says”; the rhyme holds the number. Use the substitution test from Rhythm, Rhyme & Song.
Step 4: Run the chain test. Teller A tells the story to B, B to C, C to D, with D writing down every fact they recall. Compare D’s list with the original and note, for each fact, whether it was kept, dropped or altered.
Step 5: Repair. A dropped fact needs a stronger event. An altered fact needs a figure moved to the refrain or a more distinctive hook. Run the chain again with new people.
Worked Example
A story must carry six facts, two FIXED and four FREE. After one chain test, D’s list holds five of the six.
- Facts kept: 5 of 6, so the survival rate for this chain is 5 ÷ 6 = 83%.
- The lost fact is a FIXED one, so the story fails the test even though most of it survived.
- Repair: give the lost fact a failure event and move its figure to the refrain, then retest.
The 83% is a result of this single small test and says nothing general about stories. The method’s value is the list of exactly which facts fell out.
Diagnosing Failure
- Dropped in the second hop but not the first — the fact is an aside. Make it a cause.
- Altered but kept — the story supports a believable wrong version. Give the right version a vivid consequence.
- Different in every chain — the fact has no anchor in the plot, and tellers are filling the gap differently.
- Everything survives but the numbers — the usual pattern; move all numbers to the refrain.
Common Mistakes
- Testing with people who already know the content — they fill gaps from memory. Use fresh listeners.
- Testing once — one chain is a sample, not a verdict.
- Encoding a hazard identification in a story — see the warning in Warning Tales.
What’s Next
- Learn next: Character-Fact Association — assigning facts to named characters
- Combine with: Correction Protocols — fixing what the chain test finds