Trust Repair
The core repair sequence after breach: acknowledgment, accountability, restitution where needed, changed behavior, and testable new terms.
Full Practice - Bond - Repairing After Rupture
Mechanism
Trust Repair is the practice of rebuilding warranted trust after a breach. It begins with a hard constraint: trust is not repaired by wanting the relationship back. It is repaired when behavior, accountability, and new terms make renewed exposure reasonable.
A breach changes the trust object. Before the breach, the question may have been "Can I trust this person to tell me when something is wrong?" After the breach, the question becomes sharper: "Can I trust this person to tell me the truth after failing to tell me the truth before, and what evidence would make that trust warranted?" The old relationship cannot simply be resumed. The rupture has added evidence, and the repair has to answer that evidence.
Repair is not the restoration of a feeling. It is the reconstruction of warranted exposure.The Control failure is defensive repair. The person or institution narrows the breach, disputes language, manages liability, demands forgiveness, or treats repair as reputational containment. Control wants the record closed before the harmed party can finish naming what happened.
The Decay failure is performative repair. Apologies multiply, processing expands, everyone says the right words, but no behavior changes and no new terms become testable. Decay can also appear as permanent rupture, where no repair attempt is allowed to become evidence because the breach has become the whole identity of the relationship.
The Range form is disciplined repair: the breach is named specifically, responsibility matches the violation, restitution addresses what can be repaired, future behavior is constrained by new terms, and trust increases only as those terms are honored over time.
Practice
The diagnostic question is: "What broke, what must change, and how will we know?"
Use the tool after a missed commitment, betrayal, misrepresentation, safety failure, institutional cover-up, or repeated small breach that has crossed into trust damage.
Name the breach in behavior terms. Do not begin with character labels. State what happened, what commitment or expectation it violated, who was exposed to harm, and what changed because of it. "You lied" may be true, but "you told the team the data had been checked when it had not" gives repair something to act on.
Accept the right responsibility. Match the response to the failure. Competence failures require correction, support, training, changed process, or reduced exposure. Integrity failures require truth-telling, accountability, and evidence that the person's relationship to truth or obligation has changed. Do not use apology to cover a competence gap, and do not use process improvement to avoid an integrity failure.
Make repair observable. Restitution is not always possible, but repair must become visible somewhere: a corrected record, returned resources, changed access, a public correction, a new review step, a witnessable commitment, a boundary honored without resentment. If no one can observe what changed, the repair remains atmospheric.
Rebuild under new terms. The next relationship needs terms that answer the breach. Lower exposure may be appropriate at first. Monitoring may be appropriate. A narrower domain of trust may be appropriate. If trust is rebuilt, it should be because the new terms were honored, not because enough time passed for the discomfort to fade.
Repair can end in reconciliation, narrowed cooperation, or separation. All three can be honest outcomes. The false outcome is pretending repair happened because people stopped talking about the breach.
In the Wild
A team lead promises that concerns raised in a review will be addressed before launch. The launch happens anyway. The repair is not "sorry, we were under pressure." The repair names the behavior: the team was told one thing and the decision process did another. The repair changes the process so unresolved objections have visible status before launch, and it gives the people who raised the concerns a clear way to verify whether their signal changed the decision.
A partner shares private information in a moment of frustration. The first repair step is not demanding forgiveness. It is naming the breach, accepting that the harmed person now has evidence about confidentiality under pressure, making whatever amends are possible, and offering narrower terms until behavior under future pressure can be observed.
An institution fails to enforce its own rule when a powerful member violates it. The repair cannot be only a statement of values. It has to correct the record, answer why the rule failed, change the authority structure or incentive that produced the failure, and accept external or independent verification. The breach was institutional, so the repair has to be institutional too.
If you caused the breach, do not ask first how to be trusted again. Ask what trusting you now requires.
If you were harmed, do not let anyone turn your willingness to repair into evidence that the breach was small. Repair begins with truth about the rupture. It continues only where responsibility, restitution, and changed terms make future exposure warranted.
Lineage
Roy Lewicki and Chad Brinsfield's review of trust repair is the broad research background. It distinguishes short-term repair strategies, such as apology, excuse, denial, and compensation, from longer-term strategies such as structural changes, monitoring, reframing, forgiveness, and silence. The Codex inherits the distinction between repair speech and repair structure: what is said after a breach matters, but the repair does not stop there.
Peter Kim, Donald Ferrin, Cecily Cooper, and Kurt Dirks provide one of the sharper empirical cautions. Their work on competence and integrity violations shows that the right response depends on the kind of alleged failure and on later evidence of guilt or innocence. That matters because repair advice is often too simple. "Apologize" is not a universal repair mechanism. The responsible move is to match the response to what actually happened.
Nicole Gillespie and Graham Dietz extend the problem to organization-level trust repair. Once the breach is systemic, repair has to be systemic. An organization that caused harm through incentives, culture, reporting structures, or authority design cannot repair trust only through leadership statements.
Restorative justice, apology research, and mediation practice sit nearby. The Codex does not collapse these fields into one method. Trust Repair is the Workshop's practical synthesis for one narrow job: rebuilding warranted exposure after trust has broken.
Cross-references
Within the category. Common Knowledge Generation becomes necessary when the repair terms need to be known by more than the two people speaking. Peacemaking Circles offers a structured process when the breach affects a wider community.
Across the Workshop. Trust Thermocline often diagnoses the accumulated breach before repair begins. Trust Diagnostics helps decide what level of renewed trust is warranted after repair has begun. The Exclusion Problem treats failed repair as a precondition before exclusion can be considered.
Limitations. Trust Repair cannot require reconciliation from the harmed party. It cannot make unsafe contact safe by procedure. It cannot repair a breach while the behavior that produced the breach is continuing. It can only name the work required and let future behavior decide whether renewed trust is warranted.