Decision Log Update¶
Governance record amendment — instantiates Belief Revision Workflow
An amendment to the standing decision record that ties a revised belief to its governance consequences — which decision, condition, owner, and next-review point changed because the belief changed.
A belief can be revised on paper and still steer the organization exactly as before, because the decisions it justified were never revisited. A Decision Log Update closes that gap. It is an amendment to a standing decision record — a decision register, an architecture decision record, an investment memo — that carries a belief change into governance: it restates the decision's rationale in light of the revised belief, updates any conditions or triggers the decision now rests on, names who owns the change, and sets when it will next be reviewed. Its defining property is that it is the governance record, not the epistemic one: it does not preserve the belief itself, only the decision's altered dependence on it. It is what makes a revised belief consequential rather than ceremonial.
Example¶
An investment committee has held two positions on the premise that a mid-cap logistics firm's margins would expand as it scaled. A quarter of disappointing results and a channel check revise that premise: margins are structurally capped by a customer-concentration problem. The belief change is logged epistemically elsewhere — but that alone changes nothing about the portfolio. The Decision Log Update is what governs. Against the original "accumulate on weakness" decision, the committee records: rationale amended (the margin-expansion thesis no longer holds); the decision's condition changed from "add below $40" to "trim into strength, cap position at 2%"; the owner of the position reassigned to re-underwrite by next meeting; and a review trigger set for the following earnings report. This mirrors the discipline of a decision journal — recording not just what was decided but the reasoning and the conditions under which it should be revisited — so that when the belief moved, the standing decision moved with it.[n1]
How it works¶
- Locate the decisions the belief propped up. Trace which standing decisions rested on the now-revised belief; a belief change with no decision attached needs no governance entry.
- Amend the rationale and conditions. Restate why the decision stands (or falls) under the revised belief, and update the conditions or thresholds it now depends on.
- Assign the owner. Name who is accountable for carrying the change through — without an owner, the amendment is a note, not a governance act.
- Set the next review. Establish the trigger or date at which the amended decision is re-examined, so the change is provisional rather than frozen.
Tuning parameters¶
- Coupling breadth — how many downstream decisions one belief change is traced into. Broad tracing catches hidden dependencies but can cascade into re-opening everything.
- Owner specificity — a named individual versus a role or committee. Named owners drive follow-through; roles survive turnover but diffuse accountability.
- Review-trigger type — a fixed date, a metric threshold, or an event. Event triggers fire when they matter but require monitoring to notice.
- Amendment weight — whether the change is a light annotation or a formal re-decision requiring sign-off. Heavier process suits high-stakes reversals; lighter suits reversible ones.
- Traceability depth — how much of the superseded decision is preserved versus overwritten, trading a clean current record against an auditable history.
When it helps, and when it misleads¶
Its strength is defeating symbolic updating — the failure where language changes but behavior doesn't — by forcing a revised belief to land on a specific decision, owner, and review point. It is the component that makes "we changed our minds" verifiable in what the organization actually does next.
Its failure mode is governance theater: an amendment logged with a vague owner ("the team") and no real review trigger looks like accountability while producing none. A related misuse is coercive re-decision, where a powerful actor forces a decision change under the banner of "the belief was revised" without any genuine evidence behind the revision. The guarding discipline is to require a named owner and a concrete review trigger on every amendment, and to insist the governing belief change point back to a real, evidenced revision rather than a preferred narrative.
How it implements the components¶
action_implication— its core: it restates which decisions, conditions, and plans change because the belief changed, turning revision into governed action.revision_owner— it names who is accountable for carrying the amended decision through, so the change has a driver.revision_trigger— it sets the date or condition at which the amended decision is re-examined, keeping the change provisional and monitored.
It records the decision's altered dependence on a belief but not the belief's epistemic content — the prior belief, the conflicting evidence, and the before/after of the belief itself (current_belief, conflicting_evidence, revision_record) belong to its nearest twin, Belief Update Log; it also does not grade the confidence (confidence_update — Confidence Scale).
Related¶
- Instantiates: Belief Revision Workflow — it is the workflow's governance hook: the point where a revised belief changes what is actually decided.
- Consumes: Belief Update Log supplies the epistemic revision this amendment governs.
- Sibling mechanisms: Belief Update Log · Bayesian-Style Update Session · Confidence Scale · Dissonance-Safe Dialogue · Learning Retrospective · Prediction Error Review
Editorial Notes¶
Form Classification¶
Form family: Record, Log & Register
Rationale: Decision Log Update operates as a durable record, ledger, register, or trace whose value depends on preserving actual state or history because it an amendment to the standing decision record that ties a revised belief to its governance consequences — which decision, condition, owner, and next-review point changed because the belief changed.
Independent corroboration: The frozen evidence defines Decision Log Update as 'An amendment to the standing decision record that ties a revised belief to its governance consequences — which decision, condition, owner, and next-review point changed because the belief changed', so its operative form is Record, Log & Register.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Organizational & Management Science
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Decision-journal and governance practice cohered amendments that revisit rationale, conditions, owners, and review triggers when material assumptions change.
Related originating lineages:
- Computer Science & Software Engineering — Architecture decision records supplied versioned supersession rather than silent editing of earlier choices.
Review resolution: Decision-journal and governance practice cohered amendments that revisit rationale, conditions, owners, and review triggers when material assumptions change.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
The nearest twin, Belief Update Log, and this file split the record in two on purpose: the belief log answers "what did we believe, and what now?"; this decision log answers "what do we therefore do differently, who owns it, and when do we recheck?". A revised belief with no matching decision-log entry is exactly the symbolic-updating failure mode the workflow exists to prevent.
[n1] A decision journal is the practice — popularized in decision-making writing by, among others, Michael Mauboussin and Annie Duke — of recording a decision's rationale, expected outcome, and the conditions under which it should be revisited, so that later review is judged against the reasoning rather than the outcome alone. ↩