Cooperating Under Bad Faith — Architecture
The Bond's discipline of keeping cooperation functioning when the framework itself is being exploited — and the lens that makes predation, capture, and the weaponization of cooperative vocabulary legible from inside the cooperation.
Bond
What This Category Holds
The discipline of the Bond is cooperation. Most of its categories teach how cooperation is built and maintained under the structural pressures every cooperative system faces — incentive misalignment, evolved biases, entropy. This category teaches something different. It teaches what cooperation requires when one party has come prepared to take from the framework. The pressure has changed shape, and the discipline has to change with it.
The category equips a practitioner to keep cooperation functioning when its own vocabulary is being turned into a weapon. Good faith, the steelman requirement, connection before correction — these are strengths when both parties are accountable to them. They become attack surfaces when one party invokes them while exempting themselves. Without the work this category names, the rest of the Bond runs in an idealization. With it, the Bond's commitment to cooperate becomes operational under the conditions it has to survive, not just the conditions in which it works smoothly.
Cooperative language is genuine when both parties are accountable to it. It is weaponized when one party uses it to constrain the other while exempting themselves. The line is the diagnostic the category runs on.
Cooperating Under Bad Faith holds the Range against two specific pulls. The pull toward Control is suspicion as default — the cooperative framework destroyed in order to defend it, every disagreement read as potential sabotage, the practitioner becoming the kind of actor the category was built to recognize. Bond §06 names this directly: it is Severance wearing armor. The pull toward Decay is naive openness — the refusal to acknowledge that bad faith exists because acknowledging it feels uncooperative, exploitation allowed to operate unimpeded under the cover of charitable interpretation. §06 names this too: it is Fusion without self-awareness. The discipline is neither. It is calibrated trust carrying a specific new capacity inside it — the ability to recognize, in real time, when the cooperative framework itself is being used as a weapon.
The Adversarial Threshold
A cooperative framework has to distinguish adversarial exploitation from structural drift. This category begins when the evidence supports a strategic or asymmetric pattern: another party studies, games, or invokes the cooperative system in ways that constrain others while leaving themselves unbound.
The adversary in this model is rarely the caricature. They are usually intelligent, often more fluent in the cooperative framework's language than its average member, and frequently convinced they serve some higher purpose. Bad faith here does not require private malice. It requires observable asymmetry: the actor learns the system's norms, accumulates authority through visible alignment, and then uses the norms or authority strategically while resisting the accountability they impose on others. A cooperative framework that does not account for this reads, to anyone who has watched exploitation happen in real time, as a framework for sincere people in a world that is not only sincere.
Environments can also reward defection, erode monitoring, and produce exploitative outcomes without any actor adopting that strategy. Those conditions matter, but they do not constitute bad faith by themselves. They belong first to systems diagnosis: read the incentives, rules, pressures, and feedback loops before attributing an adversarial operating mode to a person or group.
The boundary protects both directions. If every bad outcome becomes evidence of bad faith, the framework becomes paranoid: every disagreement is read for positioning, every dissent becomes potential sabotage, and suspicion-as-default becomes the Control move the category was built to refuse. If every asymmetry is explained away as incentive drift, engineered exploitation remains invisible. The category therefore asks for evidence of pattern, reciprocation failure, strategic use, and asymmetric accountability before it changes operating mode.
The Tools Inside
The category's tools share a structure: one lens, four mechanisms. The lens makes adversarial dynamics legible inside cooperative systems. The four mechanisms are what the lens brings into focus.
Adversarial Dynamics. The lens. Names the adversarial threshold and the diagnostic question the category runs on — whether the other party is accountable to the same norms they are invoking. Three practices convert the diagnostic into something usable in a live situation: the Reciprocation Test, the Pattern Assessment, and the Structural Check. Sources: Axelrod's tournaments, Olson and Ostrom on institutional cooperation, Stigler on regulatory capture, and Popper and Rawls on the limits of tolerance.
Trust Mining. The two-phase accumulation-then-extraction pattern. Trust capital built up specifically in order to be spent. The accumulation phase is indistinguishable from an ordinary career inside a cooperative system: calibration tests passed, alignment demonstrated, authority earned, patiently, sometimes over years. Then the exploitation phase begins. Institutional capture is trust mining run at scale, through legitimate channels, over time. The specific vulnerability the mechanism reveals: calibrated trust, the Bond's own prescription, can be gamed by an actor patient enough to pass the calibration.
The Cooperative Vulnerability. The framework's own vocabulary turned into an attack surface. Steelmanning becomes a legitimation tool. Good faith becomes a shield demanded asymmetrically. Connection before correction becomes a guarantee that a manipulative position will be received sympathetically before it can be challenged. The answer is not to abandon the cooperative practices — that would surrender, becoming the thing being resisted — but to make them conditional on reciprocation and to build the diagnostic capacity to see when reciprocation is being performed rather than practiced.
Sabotage Diagnostics. The discipline of telling genuine dissent from strategic sabotage. From outside, the two can be indistinguishable: both oppose prevailing positions, both name problems, both are uncomfortable for the institution on the receiving end. The difference shows up only as a pattern over time, across five behavioral signatures. No single signature is conclusive; the pattern is the diagnostic. The risk of false positives is severe — labeling dissent as sabotage is among the oldest Control moves in institutional history — which is why the diagnostic is built around structural evidence rather than any one observation.
The Exclusion Problem. The hardest operational question the category raises: when does a cooperative setting have to exclude a participant in order to survive? Four conditions together justify exclusion from that setting — a sustained pattern of bad faith rather than a single incident, evidence of strategic exploitation distinct from honest disagreement, failure of the full repair sequence, and an exclusion that serves the Range rather than the excluder's comfort. The paradox of tolerance, adapted from political philosophy to cooperative settings.
Chapter Note
The Bond chapter carries this category as named chapter material at §06 (When Cooperation Is Attacked). The §06 treatment is substantive: it names the engineered-failure pattern as distinct from the structural drift the rest of the Bond's failure modes describe; it names trust mining as the accumulation-then-extraction mechanism; it names the cooperative vulnerability — the Bond's own practices becoming attack surfaces when one party is not practicing in good faith; and it frames the response posture as calibrated trust carrying weaponization-recognition capacity, with the Range pulls articulated as Severance wearing armor on the Control side and Fusion without self-awareness on the Decay side. §06 also forwards the operational detail to the Inheritance — the diagnostic criteria, the mechanics of capture, the conditions for exclusion — which this category now holds.
The Inheritance sharpens one boundary that the chapter leaves implicit: structural pressure can produce harmful outcomes without supplying evidence of bad faith. Diagnosing that pressure matters, but this category changes the cooperative operating mode only when the record supports strategic use or asymmetric accountability. That threshold keeps the Inheritance's adversarial tools available without making adverse outcomes proof of an adversary.