Change a relationship or constraint with known authority.
Capability 04 · Causal Engineering
Change the relationship that keeps reproducing the problem.
Causal Engineering designs structural interventions in incentives, dependencies, constraints, feedback, timing and buffers—then carries their liabilities into execution and learning.
Symptom versus structure
Visible action can leave the governing relationship untouched.
Removing a node may trigger replacement. Adding capacity may increase demand. Accelerating a process may transfer risk downstream. Causal Engineering asks what must change so the system stops reproducing the unwanted outcome.
Action on the event
Responds to what has already appeared. It may be necessary, but it rarely changes the entire generating architecture.
Intervention in the structure
Targets a relationship, rule, bottleneck, incentive, buffer or conversion path and anticipates adaptation.
Intervention anatomy
A credible intervention carries its causal argument with it.
The institution should be able to state what changes, why that change matters, who can authorize it, what resistance is expected and how failure will be detected.
Target relation→Mechanism→Authority→Sequence→Adaptation→Learning
Portfolio, not magic lever
Leverage is conditional on timing, legitimacy and system state.
A robust portfolio combines moves with different time horizons and failure modes. It can include direct action, designed non-intervention, experiments, buffers, option protection and institutional redesign.
Absorb propagation while deeper change becomes viable.
Test a causal proposition without committing the full system.
Preserve an alternative before path dependence closes it.
Withhold action when it would reinforce the harmful structure.
Change incentives, mandates, ownership or coordination.
Second-order discipline
The system gets a move.
Actors reroute, narratives reframe, resources shift and thresholds move. Monitoring therefore tests the intervention’s causal argument, not only whether a delivery milestone was completed.
Decision contract
Objective, mechanism, owner, constraints and failure conditions.
Propagation signals
Early movement, adaptation, displacement and delayed liabilities.
Review & abort
Explicit thresholds for continue, revise, pause or reverse.
Structural memory
What was known, assumed, authorized, observed and learned.
What it produces
A governed intervention portfolio—not a black-box prescription.
Leverage Map
Bottlenecks, couplings, delays, dependencies and viable openings.
Intervention Portfolio
Sequenced actions, buffers, experiments and protected options.
Decision Contract
Authority, boundaries, preconditions, ownership and review gates.
Monitoring & Memory
Signals, model revisions, outcomes, liabilities and institutional learning.
Govern the consequence
Intervention is the beginning of governance.
Dynamic Causal Governance connects the portfolio to legitimate authority, execution, adaptation and structural learning over time.