skills/novel-story-telling/references/time-travel-and-loop.md

# Time Travel And Time-Loop Story Design

## Contents

1. Temporal ontology
2. Rule contract
3. Timeline and knowledge state
4. Loop progression
5. Escalation and reveals
6. Resolution patterns
7. Deterministic audits
8. Failure modes
9. Research basis

## Select One Temporal Ontology

### Fixed single history

The past cannot be changed; the traveler always participated in it. Interventions complete known history. Use for predestination, irony, tragedy, and causal-knot mysteries. Avoid chains of arbitrary accidents that exist only to force consistency.

### Mutable single history

Intervention overwrites or propagates through one history. Define propagation speed, memory immunity, duplicate handling, and whether erased causes can leave effects. Use for repair narratives and unintended consequences. Audit downstream state strictly.

### Branching histories

Intervention creates or enters a different history. Define whether the origin continues, whether return is possible, and how branches are identified. Saving a branch does not automatically repair the origin. Use for replacement, abandonment, proliferating selves, and moral responsibility.

### Closed causal loop

Events, objects, or information can cause themselves. Tag every bootstrap object or fact. Decide whether absence of external origin is the answer, the theme, or an unresolved anomaly.

### Reset loop

A bounded interval restarts after a trigger or deadline. Define what persists: memory, injury, age, objects, entropy, relationships, marks, or nothing. Use for mastery, investigation, moral iteration, and transformation through repetition.

Do not mix ontologies unless the transition itself is seeded, modeled, and central to the climax.

## Temporal Rule Contract

```yaml
ontology:
anchor_time:
trigger:
destination_rule:
reset_condition:
exit_condition:
memory_retainers: []
physical_state_retained: []
objects_retained: []
subjective_aging:
duplicate_policy:
causal_propagation:
branch_creation:
branch_return:
paradox_policy:
bootstrap_policy:
entropy_or_degradation:
known_exceptions: []
```

Demonstrate rules before using them to resolve a major conflict. Treat an unknown rule as unknown consistently across character action and narration.

## Timeline And Knowledge State

Maintain two clocks:

- **world time:** timestamp within a history or loop;
- **personal time:** the traveler's subjective sequence and age.

### Event row

```yaml
event_id:
timeline_id:
loop_index:
world_time:
personal_time:
location:
participants: []
cause_ids: []
effects: []
world_state_before: {}
world_state_after: {}
observers: []
rememberers_after_reset: []
information_gained: []
information_used: []
persistence_class:
```

### Knowledge row

```yaml
character:
fact_id:
learned_event:
source:
confidence:
survives_reset:
shareable:
invalidated_by_branch:
```

A character may act only on facts available in their personal sequence.

## Loop Progression

### Phase 1: ignorance

Establish the baseline interval, trigger the first reset, confirm repetition through stable anchors, and define the immediate survival or escape goal.

### Phase 2: experimentation and mastery

Begin each dramatized loop with a hypothesis. Change one meaningful variable. Show feedback, downstream effect, and a gain in knowledge, cost, or relationship state. Compress repeated passages.

### Phase 3: disillusionment

Show that optimization cannot solve the deepest conflict. Add memory burden, subjective aging, moral injury, loop degradation, adaptive opposition, or alienation. Reframe the exit condition or original objective.

### Phase 4: irreversible choice

Remove unlimited experimentation. Force a final attempt, permanent branch, sacrifice, memory loss, acceptance, or choice to remain. Resolve through an established temporal rule and changed value judgment.

### Per-loop delta test

Dramatize a loop only if it changes at least one of:

- hypothesis;
- tactic;
- information;
- emotional meaning;
- relationship;
- cost;
- antagonist response;
- rule understanding;
- exit probability.

Otherwise summarize or omit it. Repeat only enough anchor detail to orient the reader, then move to the delta.

## Escalation And Reveals

Possible escalation:

1. recognition;
2. low-risk tests;
3. competence or exploitation;
4. failed optimization;
5. hidden participant or rememberer;
6. opponent adaptation;
7. accumulated trauma or subjective age;
8. loop instability, shortening, leakage, or memory loss;
9. conflict between escape and protecting someone who will forget;
10. final irreversible attempt.

Distribute reveals across mechanism, scope, memory, history, agency, exit, persistent cost, and meaning. Answer one question while sharpening another. Do not withhold facts a viewpoint character already knows.

## Resolution Patterns

- **Fixed-point fulfillment:** prevention attempts cause the known event, but change its meaning or cost.
- **Rule exploit:** a demonstrated persistence exception carries a needed person, object, or signal.
- **Knowledge sacrifice:** exit requires loss of accumulated memory.
- **Self-sacrifice or substitution:** save others at the cost of remaining, dying, or being erased.
- **Branch acceptance:** save one branch while acknowledging the origin remains lost.
- **Causal-knot closure:** the final action supplies an earlier unexplained cause.
- **Objective transformation:** exit follows when the protagonist stops optimizing the wrong goal.
- **Negotiated release:** another rememberer or system ends the loop after incentives or relationship change.
- **Containment:** the loop remains but its scope or harm is bounded.

## Deterministic Audits

- **Rule-use audit:** flag climax mechanisms not introduced or demonstrated.
- **Knowledge-provenance audit:** trace every action-relevant fact through personal time.
- **Persistence audit:** compare pre-reset and post-reset state with the contract.
- **Counterfactual audit:** update dependent events after an intervention while preserving independent facts.
- **Causal-cycle audit:** tag cycles as intentional bootstrap, fixed point, or error.
- **Branch audit:** prevent silent movement between timeline IDs.
- **Duplicate audit:** enforce the duplicate policy at shared world times.
- **Delta audit:** flag consecutive dramatized loops without meaningful change.
- **Subjective-age audit:** calculate age and trauma exposure from personal time.
- **Reveal audit:** require reader-facing evidence before a rule resolves the climax.

## Failure Modes

- The ontology is never defined.
- Persistence changes for convenience.
- Characters use information they never learned.
- Repeated scenes add no causal or emotional change.
- Unlimited attempts remove urgency without cost or deadline.
- Fixed history relies on conspicuous coincidence.
- A new branch is treated as repairing an abandoned origin.
- A bootstrap object is called explained despite lacking external provenance.
- Subjective age, trauma, injury, or duplicate identity is ignored.
- The final chapter introduces a temporal exception.
- A hidden rememberer lacks a credible reason for earlier silence.
- Changed causes leave dependent consequences untouched.

## Research Basis

- Stanford Encyclopedia of Philosophy, [Time Travel](https://plato.stanford.edu/archives/fall2025/entries/time-travel/).
- Stanford Encyclopedia of Philosophy, [Time Travel and Modern Physics](https://seop.illc.uva.nl/entries/time-travel-phys/).
- Barak Shoshany and Jacob Hauser, [Time Travel Paradoxes and Multiple Histories](https://arxiv.org/abs/1911.11590).
- Lianhui Qin et al., [Counterfactual Story Reasoning and Generation](https://arxiv.org/abs/1909.04076).

The narrative schemas and audit methods are practical synthesis from causal and storytelling concerns.