The easiest dragon inconsistency to notice is a changed eye color. The damaging inconsistencies are usually deeper.

A dragon is described as rare but every village has anti-dragon architecture. It is sacred, yet nobody objects when merchants sell its body parts. It controls rivers, but the setting’s farming, shipping, flood politics and calendar look exactly like a world with no river dragons. It cannot cross iron, until the plot needs it to.

Those are not trivia errors. They are broken consequences.

A useful dragon audit therefore begins with a counterintuitive rule: do not start by checking the dragon sheet. Start by checking the world around the dragon.

The checklist below is designed for fantasy creators adapting historical motifs or building entirely invented species. When historical examples are used, keep them labeled as sources rather than pretending that “the East Asian dragon” is one uniform creature.

Audit 1: can you write the dragon’s rule in one sentence?

Write the rule that matters most.

Examples:

  • “River dragons can redirect water only inside watersheds where they have an established shrine.”
  • “Sky dragons cannot create storms; they can only amplify weather already forming.”
  • “Imperial dragons are not a species but a legal category granted to certain lineages.”

A rule becomes auditable when it contains a boundary. “Dragons control water” is too loose. What water? At what cost? Under what conditions? Can the rule fail?

If the writer cannot state the rule without adding five exceptions, the setting may already be using the dragon as a plot permission slip.

Audit 2: does the cultural meaning match the behavior?

Historical dragon imagery gives a good lesson in consistency: symbols acquire meaning through repeated relationships, not through a generic parts list.

The Metropolitan Museum of Art describes a Qing vase with a dragon amid clouds as drawing on several associations, including water, protection, happiness, fertility and imperial rule. A late-Ming panel in the same collection describes dragon-like long as symbols of imperial power and benevolence. A Japanese scroll by Maruyama Ōkyo uses a dragon-and-tiger pairing inherited from Chinese pictorial traditions and notes the dragon’s association with water and thunderstorms.

These examples do not prove that every Chinese or Japanese dragon always means the same thing. They show why creators should connect symbol and use.

If your invented society calls dragons “protectors of harvests,” where is that belief visible?

  • irrigation rituals;
  • tax exemptions for shrine land;
  • weather proverbs;
  • legal penalties for harming nesting grounds;
  • political arguments after a drought.

If none of those consequences exist, “protector of harvests” may be decorative lore rather than functioning culture.

Audit 3: is rarity visible in infrastructure?

Write down the population estimate you imply, even if characters do not know the number.

Then inspect infrastructure.

If dragons are genuinely rare, highly specialized anti-dragon towers in every small town may be wasteful unless the threat is catastrophic enough to justify them. If dragons are common, ordinary people should have mundane expertise: roof materials that resist claws, standardized warning flags, insurance exclusions, migration maps, trained livestock behavior.

Rarity should change how society spends money.

A useful test: pick one medium-sized town and list three things it would not build because dragons are too rare, and three things it would maintain because the risk is still worth preparing for.

That forces the setting to choose.

Audit 4: does the ecology balance?

Even supernatural creatures occupy relationships.

Ask:

  • what does the dragon consume;
  • what competes with it;
  • what habitat feature limits its range;
  • what season changes its behavior;
  • what traces does it leave;
  • what happens after it dies.

A mountain-sized carnivore cannot eat livestock every day without leaving an enormous economic and ecological footprint. If you want the visual scale without that footprint, change the biology or metaphysics: perhaps it feeds rarely, metabolizes mineral heat, spends decades dormant, or is partly nonmaterial.

The point is not realism for its own sake. The point is consequence accounting.

Magic is allowed. Free consequences are what break trust.

Audit 5: can institutions disagree without contradicting facts?

A believable setting can contain conflicting interpretations of the same dragon.

Priests may call a flood a judgment. Engineers may describe a damaged levee. Farmers may blame an upstream dragon. The court may suppress all three explanations because admitting loss of control is politically costly.

Those viewpoints can disagree while the underlying event stays consistent.

The audit question is: are factions disputing interpretation, or does each chapter quietly invent a different physics?

Create a table:

Group What it believes What evidence it sees What it wants
River temple dragon withdrew protection shrine pool dried restore offerings
engineers channel silt raised flood risk surveys, flow records fund dredging
governor panic is the main threat market flight control public message

The table does not tell you which side is right. It prevents accidental authorial contradiction from masquerading as cultural disagreement.

Audit 6: trace one dragon rule through five systems

Choose the highest-leverage rule and test it against:

  1. household life;
  2. economy;
  3. law;
  4. religion or symbolism;
  5. conflict.

Suppose dragons can recognize sworn promises and violently react to deliberate oath-breaking.

Households may use dragon motifs at weddings. Merchants may prefer contracts witnessed near a shrine. Courts may debate whether dragon reaction counts as evidence. Religious specialists may control access to oath sites. Criminals may develop loopholes based on technically true wording.

One rule now produces a world.

If a rule touches none of the five systems, it may be too weak. If it changes all five in exactly the same way, it may be too simple.

Audit 7: separate historical inspiration from invented anatomy

This is essential when using living cultural traditions.

Create two columns.

Documented inspiration

  • museum object or text;
  • date/period;
  • culture;
  • what the source actually says.

Invented adaptation

  • your species biology;
  • added powers;
  • political role in the fictional setting;
  • visual combination from multiple sources.

For example, The Met’s object records support specific claims about particular Chinese and Japanese artworks. They do not authorize a claim that all Asian traditions share one dragon theology. The Chinese Text Project can help creators locate premodern Chinese texts, but a search result still needs context and interpretation.

This source/adaptation split improves both respect and creativity. Once you know what you are borrowing, you can change it deliberately.

Audit 8: test the creature at three scales

Scene scale

Can the dragon’s abilities and limits remain consistent over ten minutes of action?

Story scale

Do consequences persist after the scene? If a dragon burns a district, does the next chapter acknowledge displacement, repairs, smoke, politics or fear?

Generation scale

Would the dragon’s existence change institutions over decades? Architecture, roads, songs, law and trade should accumulate responses to recurring forces.

Many worlds pass scene scale and fail generation scale. The dragon works as a monster encounter but leaves no historical footprint.

Audit 9: run the “offstage dragon” test

Remove every on-screen dragon from one chapter.

Can a reader still feel that dragons exist?

Look for indirect evidence:

  • a river gate designed around a migration season;
  • a child’s rhyme about cloud color;
  • a tax category for damaged orchards;
  • a taboo on bells near nesting cliffs;
  • a guild that repairs clawed roofs;
  • a political insult based on imperial dragon imagery.

Do not add all six. Two precise traces are more convincing than a paragraph of encyclopedia exposition.

If nothing remains when the creature leaves the page, the world may still be waiting for the dragon to become part of it.

Audit 10: search for exception inflation

List every exception to the dragon’s main rule.

Then label each one:

  • priced — the exception has a cost;
  • foreshadowed — the audience could have known it was possible;
  • institutionalized — the world has adapted to it;
  • convenient — it appears only to solve a current plot problem.

One convenient exception may be survivable. Five convenient exceptions mean the rule is no longer a rule.

The fix is often not deleting the exception. Give it a price, a history and a consequence.

Audit 11: check names, categories and translations

If different cultures interact with the same creature, terminology should have reasons.

A scholar’s category, a local nickname and a religious title may all coexist. But do not add synonyms simply to make the glossary look ancient.

For every name, ask:

  • who uses it;
  • when;
  • what distinction it implies;
  • whether another group considers it wrong.

Translation can create productive ambiguity, but accidental inconsistency is not ambiguity. Maintain a small term table and note deliberate variants.

Audit 12: let readers discover rules through evidence

The final audit is narrative.

Take one rule and delete its explanatory paragraph. Can readers infer it from events, objects and choices?

A law, a damaged roof, a ritual, a merchant’s refusal and a child’s warning can establish the presence of dragons without a lecturer saying “in this world, dragons are important.”

If the rule survives without exposition, it is probably integrated.

If it disappears entirely, you may have written information rather than worldbuilding.

A compact dragon consistency scorecard

Before publication, score each item 0–2:

  • core rule has a visible boundary;
  • symbolism matches cultural behavior;
  • rarity matches infrastructure;
  • ecology or metaphysics pays its costs;
  • institutions disagree without changing physics;
  • high-leverage rules propagate into society;
  • historical sources are separated from invention;
  • scene/story/generation scales agree;
  • offstage footprint exists;
  • exceptions are priced and foreshadowed;
  • terminology is controlled;
  • rules can be inferred from evidence.

A perfect score is not required. The scorecard tells you where the world is leaning on author explanation rather than consequences.

The goal is not to make dragons realistic animals. It is to make them reliably unreal: creatures whose impossible qualities still produce stable, traceable effects.

Run one contradiction drill with a hostile reader

Before locking the article or story bible, ask someone who did not design the dragon to attack the rules.

Give them three prompts:

  • “How could a clever character exploit this rule?”
  • “What ordinary job or business would exist because of this creature?”
  • “Which earlier scene becomes suspicious if this new fact is true?”

Do not defend the setting during the first pass. Write down every answer. An outsider is especially good at noticing assumptions the designer has stopped seeing.

Then sort the objections. Some expose a real contradiction. Some reveal an interesting loophole worth keeping. Some depend on information characters do not have. The point is not to seal every loophole; a world with no exploitable edge can feel sterile. The point is to know which gaps are deliberate.

A productive audit often ends with fewer rules, not more. Combining two overlapping exceptions into one priced mechanism can make the dragon easier to write and the world easier to remember.

Sources & Further Reading

Related Reading