Graduated Reciprocity
Building cooperation through small, observable, reciprocated moves, then expanding trust only as behavior warrants the next level of exposure.
Full Practice · Bond · Calibrating Trust to Behavior
Mechanism
Graduated Reciprocity is how trust becomes safer without becoming small. The practice begins with a cooperative move that is real enough to matter and limited enough that a failure will not destroy the relationship. Then it watches what comes back. If the other party reciprocates, trust can expand. If they defect, ignore, exploit, or respond asymmetrically, the next move changes.
The tool exists because the two common trust strategies both fail. Total withholding protects you from betrayal by preventing cooperation from reaching any meaningful scale. Total extension creates intimacy, authority, or dependence before the other party has shown what they do with exposure. Graduated Reciprocity gives the Bond a third option: make the first cooperative move, but make it small enough that behavior can teach you what comes next.
Trust scales by demonstrated reciprocity, not by the wish for closeness or the fear of betrayal.Reciprocity here does not mean score-keeping. The practice is not "I gave, now you owe." It is a way of letting behavior reveal the cooperative field. A party who can receive a small trust extension, respond proportionately, repair small misses, and respect the same constraints they ask of you is showing the pattern larger trust requires. A party who consumes the extension without reciprocating is also showing you something. The point is to learn before the stakes become too high.
The Control failure is never making the first move. The Decay failure is moving too far too quickly because early warmth feels like evidence. The Range form is incremental extension with live feedback: real cooperation, bounded exposure, and revision after each round.
Practice
The diagnostic question is: "What is the next smallest trust extension that would let behavior reveal whether the next level is warranted?"
Use it when building a partnership, delegating responsibility, repairing after a breach, collaborating across factions, or testing whether a strained relationship can return to cooperation.
Make a small first move. Offer a cooperative move with real content and bounded risk. Share a draft, not the whole strategy. Delegate a defined task, not the whole domain. Make one concrete concession, not a blank check. The move has to be meaningful enough that reciprocity can be observed.
Read the reciprocation. Watch whether the other party responds in kind, ignores the exposure, exploits it, or demands a larger extension before reciprocating the smaller one. Do not over-read one interaction. Do read patterns across repeated small moves.
Scale, hold, or step back. If reciprocity appears, increase exposure one level. If the pattern is unclear, hold the level and gather more behavior. If the other party exploits or refuses reciprocity, step back without turning the single failure into a global verdict on the person. The next move should answer the pattern, not your hope or your fear.
Graduated Reciprocity works best when the steps are visible. If the other party does not know what trust extension is being made, they cannot respond to it deliberately. This does not require ceremonial process. It can be as plain as: "I am comfortable sharing this part first. If the collaboration works cleanly, we can open the next layer."
In the Wild
Two teams have been competing for resources. A joint project is proposed, and everyone says they want collaboration. The Control move is to protect everything until the other side proves itself. The Decay move is to merge the work immediately and hope goodwill carries the rest. Graduated Reciprocity starts with one shared deliverable, clearly scoped ownership, and a review point after the first exchange. If both teams honor the exchange, the next project can carry more exposure.
A manager wants to rebuild trust after mishandling feedback. Apology matters, but the trust cannot be restored by declaration. The graduated move is a small, specific commitment: "In the next critique meeting, I will summarize the criticism before responding, and you can stop me if I defend before understanding." The team watches the behavior. If it holds, the next trust extension can be larger. If it fails, the failure is visible at low cost and repair can begin there.
A friend asks for access to a sensitive professional contact. The relationship is warm, and you want to help. Graduated Reciprocity asks for a smaller first move: let them send a short note for review, or introduce them in a low-stakes context first. The point is not suspicion. It is respect for the fact that trust in friendship and trust with professional reputation are different objects.
When trust is uncertain, do not choose between closed doors and full exposure. Find the next move that is cooperative, bounded, and observable.
Offer enough trust for cooperation to begin. Watch what the other person does with it. Let the next level answer the behavior you have seen.
Lineage
Robert Axelrod's The Evolution of Cooperation is the central game-theoretic lineage. Axelrod's tournaments showed how reciprocal strategies can sustain cooperation in repeated interaction when future encounters matter, reputation can form, and defection can be answered. Anatol Rapoport's Tit for Tat strategy is the clean reference point: begin cooperatively, then respond to the other party's prior move.
Graduated Reciprocity is not identical to Tit for Tat. Tit for Tat is a strategy in a formal repeated game. The Workshop tool is a practical trust-building discipline for messy human cooperation. It inherits the simple insight that the first move can be cooperative without making cooperation unconditional.
Charles Osgood's GRIT strategy, Graduated and Reciprocated Initiatives in Tension-reduction, is the closest diplomatic and conflict-reduction neighbor. GRIT proposed small unilateral initiatives designed to invite reciprocation in hostile settings. The Codex inherits the graduated structure and the reciprocation read, while keeping the tool broader than international tension reduction.
Elinor Ostrom's work on commons governance sits nearby through the institutional idea that cooperation needs monitoring, graduated response, and conditions for repair. Ostrom's "graduated sanctions" are not this tool, but they share a practical understanding: cooperative systems should respond proportionately over time rather than swing between permissiveness and expulsion.
Cross-references
Within the category. Graduated Reciprocity uses Good Faith as Default as the opening posture and Trust Diagnostics as the assessment layer. Skin in the Game changes how quickly trust should scale: shared downside supports larger steps; offloaded downside calls for smaller ones.
Across the Workshop. Prisoner's Dilemma carries the strategic structure this tool practices in relational form. Repairing After Rupture will carry the formal repair sequence when trust has already been breached. Cooperating Under Bad Faith is the boundary condition: if the other party uses small trust extensions as extraction opportunities, graduated reciprocity should slow or stop rather than keep offering material for exploitation.
Limitations. Graduated Reciprocity can become transactional if every move is treated as a debt. The practice is not bookkeeping. It is learning. It can also under-serve relationships where immediate large trust is required, such as emergency response, caregiving, or high-tempo operations. In those cases, the trust calibration shifts to role, training, institutional accountability, and after-action review because the situation does not allow slow scaling.