Causal Contribution Timeline¶
Evidence map — instantiates Outcome Responsibility Attribution Calibration
Reconstructs the outcome as a time-ordered chain of actions, omissions, warnings, and conditions, so causal contribution is read from the actual sequence rather than from whoever is most visible at the end.
Causal Contribution Timeline reconstructs an outcome as a time-ordered chain: each action, omission, ignored warning, decision, and background condition placed in sequence, with the agent behind each step named. Its defining move is to establish causation from the sequence before anyone assigns blame. By laying every contributing step in order, it surfaces the upstream and the invisible — the omission three weeks earlier, the alert nobody escalated — that the eye skips in favor of the last, most visible actor. It is the shared evidentiary base the rest of the appraisal reasons over: it shows how each step contributed, but stops short of judging fault, control, or share.
Example¶
A payment service is down for hours, and the obvious culprit is the engineer who ran the deploy that triggered it. The timeline reconstructs the chain instead: a config change merged without review two days earlier, a monitoring alert that fired but routed to an unstaffed channel, the deploy that finally exposed the latent fault, and a rollback delayed because the runbook was stale. Now four contributing steps and several agents sit on the record, in order — and the deploying engineer is one link in a chain, not "the cause." That ordered map is exactly what the control tests and the apportionment step then reason over; without it, blame lands on the last hand on the keyboard.
How it works¶
- Record the outcome and reconstruct the sequence of contributing steps — actions, omissions, warnings, decisions, conditions.
- Attach each step to the agent(s) behind it, so the cast of everyone who touched the outcome is on the page.
- Separate contributing causes from mere background, and log omissions beside commissions — without yet weighting any of them.
- Freeze the chain as a shared record before the blame conversation begins.
Tuning parameters¶
- Lookback horizon — how far upstream the chain reaches; longer horizons surface root conditions but risk sprawling toward infinite regress.
- Inclusion threshold — how causally relevant a step must be to make the timeline; low catches enablers, high keeps it legible.
- Omission sensitivity — how aggressively things that didn't happen (an unescalated warning) are logged as contributions; raising it counters the bias toward visible action.
- Granularity — event-by-event sequencing versus a phase-level summary.
When it helps, and when it misleads¶
Its strength is that it is the single best guard against fixating on the last visible actor, because it physically puts the upstream contributors on the same page. It counters the fundamental attribution error[n1] — the pull to explain an outcome by the salient person's disposition while ignoring the situational chain — by making that chain an explicit, ordered record. Its failure modes: a timeline can be curated to foreground or bury particular actors (where the chain starts is itself a choice, and starting it at the convenient actor is the classic manipulation), and sequence is not causation — mere adjacency in time can imply a link that isn't there. It also cannot say how much each step mattered; reading a share off a timeline is a category error. The discipline is to fix the lookback and inclusion rules before knowing whom they will implicate, and to keep the artifact descriptive — handing weight to the apportionment step.
How it implements the components¶
outcome_event_record— it records the outcome and its contributing events as the base artifact everything else reads from.causal_contribution_map— the ordered chain is the map of how each step causally contributed.
It is deliberately descriptive: it does not test whether any agent could have acted otherwise (agency_control_boundary — Counterfactual Control Test), sort culpability (exculpation_and_mitigation_record — Just Culture Review), weigh the steps into shares (attribution_weighting_rule — Blame–Credit Apportionment Table), or confer legitimacy on the account (appeal_or_contestation_path — Outcome Responsibility Review Panel).
Related¶
- Instantiates: Outcome Responsibility Attribution Calibration — it is the shared causal record the later steps reason over.
- Sibling mechanisms: Counterfactual Control Test · Just Culture Review · Omission–Commission Parity Check · Blame–Credit Apportionment Table · Responsibility Attribution Matrix · Attribution Uncertainty Label · Role–Duty Mapping · Scapegoat Screening Review · Responsibility Diffusion Check · Outcome Responsibility Review Panel · Credit Contribution Register
Editorial Notes¶
Form Classification¶
Form family: Record, Log & Register
Rationale: The mechanism reconstructs and freezes the actual time-ordered actions, omissions, warnings, decisions, conditions, and responsible agents behind an outcome, so its operative form is a causal event record.
Nearest alternative: Representation, Specification & Plan — The timeline visually represents sequence, but its value depends on preserving the actual historical trace before attribution begins.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Engineering & Design
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Safety and reliability engineering established event-and-causal-factor timelines that reconstruct actions, inactions, conditions, failed barriers, and contributing causes in sequence before corrective responsibility is assigned.
Related originating lineages:
- History & Historiography — Chronological source reconstruction contributes disciplined ordering and separation of background conditions from claimed causes.
- Law & Governance — Legal causation and responsibility analysis contribute distinctions among action, omission, authority, foreseeability, and contribution.
- Organizational & Management Science — Incident-learning and just-culture practice contribute accountable participation, omissions, management conditions, and a non-scapegoating review frame.
- Psychology — Attribution research motivates making upstream conditions visible rather than assigning the outcome to the most salient final actor.
Review resolution: Both reviewers offered plausible but only medium-confidence lineages. NASA's mishap procedure directly specifies chronological event-and-causal-factor charting, including actions, inactions, management practices, and contributing causes. That exact method makes engineering safety primary, with organizational just culture, law, attribution psychology, and historiography retained as independently formative contributors.
Attribution caveat: Legal and historical reconstruction use similar ordered evidence, but official mishap-investigation methods document the closest complete ancestor of this artifact; the Encyclopedia version adds an explicit responsibility-calibration and anti-scapegoating purpose.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
- NASA: Procedures for Mishap Reporting, Investigating, and Recordkeeping
- NASA: Spaceflight Mishap Investigation Handbook
- AHRQ: System-Focused Event Investigation and Analysis Guide
Notes¶
A timeline that logs only what was done and never what was culpably left undone silently exonerates the passive — the enabler who ignored a warning drops off the record while the last actor absorbs everything. Whether omissions are captured as rigorously as commissions is exactly what Omission–Commission Parity Check audits; set the omission sensitivity high enough that it has something to find.
[n1] The fundamental attribution error (correspondence bias) — the tendency, named by Lee Ross, to over-attribute others' behavior to their disposition and under-weight the situation. The timeline's remedy is to make the situational chain explicit rather than leave it to inference. ↩