The Cooperative Vulnerability
How the Bond's own practices — good faith, steelmanning, connection before correction, tolerance, repair — become attack surfaces when one party invokes them asymmetrically while exempting themselves from the constraint.
Full Practice · Bond · Cooperating Under Bad Faith
Mechanism
Cooperative practices were designed for symmetric application. Good faith as default works because both parties extend it. Steelmanning works because both parties construct the strongest version of the other's argument before responding. Connection before correction works because both parties take the time to recognize what is worth preserving in the other's position before challenging what is not. The practices are not unilateral commitments — and the moment you notice that, the vulnerability becomes legible. They are reciprocal disciplines, and their effectiveness depends on the reciprocity.
The vulnerability is what happens when one party invokes the practice as a demand on the other while exempting themselves from the same constraint. Steelmanning becomes a legitimation tool: the bad-faith actor demands that you engage the strongest version of their position while engaging yours in caricature. Good faith becomes a shield demanded asymmetrically: any challenge to the actor must be presented in maximally charitable form, while their challenges to others are routinely sharp. Connection before correction becomes a guarantee of sympathetic reception: the actor's positions arrive wrapped in warmth that must be reciprocated, while their corrections of others arrive without it. Tolerance becomes a one-way obligation: the framework must tolerate the actor's positions on tolerance grounds, while the actor refuses to extend tolerance to anyone outside their faction.
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 mechanism works because the cooperative framework's own language obscures the asymmetry. The framework was built to extend trust, to engage charitably, to receive disagreement with care, to repair after rupture. The language of those practices carries moral weight. When one party invokes the language, the receiving party's default reading is that a cooperative practice is being requested, and the cooperative response is to extend it. Recognizing that the language is being deployed unilaterally — that the constraint is being applied in one direction only — requires the receiving party to step out of the practice they are being asked to perform and audit whether the audit is itself a violation of the practice. Inside the practice, the audit looks like bad faith. From outside, the practice without the audit looks like surrender.
What the framework needs is not a rejection of the cooperative practices. Rejection would surrender to the adversary by becoming what the adversary already is: an actor who refuses to extend cooperation. What the framework needs is a way to keep the practices operational while distinguishing reciprocal invocation from unilateral exploitation. The distinction has to be made inside the framework, on its own terms, without requiring the framework to abandon what makes it work when it works.
Practice
The diagnostic question is: "Does the cooperative practice the other party is invoking bind them as it binds me?"
The question is about the practice, not the person. It is auditable, public, and grounded in observable behavior. It does not require the practitioner to read the other party's mind or impute strategic intent. It requires noticing whether the practice is operating reciprocally or unilaterally.
The Reciprocation Audit. Before extending the cooperative practice in response to its invocation, audit whether the other party is themselves accountable to it. The audit is concrete: when someone has invoked good faith as a constraint on the practitioner's challenge, look at whether the same actor extends good faith to challenges they receive. When someone has demanded steelmanning of their position, look at whether they steelman positions that contradict their own. The audit is not adversarial. It is the practice operating with its own architecture intact. Reciprocity is the condition the practice runs on; auditing it is part of practicing it.
The Asymmetry Test. When the audit surfaces asymmetry, name it specifically. The asymmetry is the diagnostic, not the verdict. Some asymmetries are accidental — different parties with different defaults, mismatched calibration, situations where the practice has not yet had a chance to operate in both directions. Other asymmetries are structural and sustained. The test is what the asymmetry does when surfaced. A party operating in good faith who hears the asymmetry named typically adjusts. A party operating the cooperative vocabulary asymmetrically as strategy typically argues that the surfacing itself is a violation of the practice, which is the structural signature the test is built to catch.
Conditional Practice. Adjust the practitioner's own practice to be reciprocation-conditional rather than unconditional. The shift is small in form and large in operation. Unconditional good faith is "I extend good faith because the practice asks for it." Conditional good faith is "I extend good faith with parties demonstrably accountable to good faith." The conditional form does not weaken the commitment. It strengthens it, because it makes the commitment durable against the failure mode that would otherwise hollow it out. A framework that practices unconditional cooperation against parties who treat the framework as an exploitation surface collapses into either capture or burnout. A framework that practices conditional cooperation can extend the practices fully where they are reciprocated and withdraw them precisely where they are not.
The three practices operate together. The Reciprocation Audit produces the data. The Asymmetry Test surfaces what the data shows. Conditional Practice converts the finding into a sustainable operating mode. Each step is part of the practice itself rather than a departure from it.
In the Wild
A founder built a company on the value of radical candor — direct feedback exchanged in both directions, the conviction that uncomfortable truths spoken plainly were the foundation of high-trust collaboration. The norm was named in onboarding, in performance reviews, in the company's public account of its culture. When the founder critiqued employees, the critiques arrived plain and direct, often hard. When employees critiqued the founder, the directness was reframed as misalignment with the company's values, as a failure to extend the charitable interpretation the founder's complex decisions required. The cooperative vocabulary was the same. The constraint it imposed ran in one direction only. Over time, employees who absorbed the message that their critiques required a different register began producing critiques in that register — softened, qualified, oriented toward the founder's preferred frame. The asymmetric invocation of the radical-candor norm had quietly produced the absence of candor the norm was supposed to produce.
In an intellectual community committed to charitable interpretation, a senior figure routinely demanded that critics engage the strongest version of his work and rejected critiques that did not. His own engagements with opposing views, when audited, regularly engaged the weakest version, treated standard critiques as caricatures of more sophisticated objections, and dismissed the strongest forms entirely. The community took years to surface the asymmetry. The reason was structural: surfacing the asymmetry required a critic to engage in a meta-move that itself looked like bad faith inside the practice (you are not steelmanning the senior figure's recent post; you are auditing his pattern of steelmanning). The practice had been weaponized in a way that made surfacing the weaponization look like a violation of the practice.
A workplace adopted explicit norms around connection before correction. Direct feedback had to be preceded by recognition of what was working. A particular manager applied the norm strictly with reports who disagreed with her decisions — every challenge had to be wrapped in connection — and broke the norm freely when delivering corrections to those same reports. The team eventually noticed that the norm functioned as a buffer protecting the manager from challenge rather than as a relational discipline operating in both directions. The team's correction was not to abandon the connection-before-correction norm but to make it bidirectional and to make the bidirectionality auditable.
The answer the tool offers is not new practices, and not weakened practices. It is reciprocation-conditional practices. The same good faith, the same steelmanning, the same connection before correction, the same tolerance and the same commitment to repair — operating under the condition that the party invoking them is also accountable to them. The condition is what makes the practice survive contact with parties who would otherwise exploit it.
A framework that learns to extend its cooperative practices conditionally is not a framework that has lost faith in cooperation. It is a framework that has done the work to keep cooperation operational at the boundary where unilateral cooperation would dissolve it.
Lineage
Karl Popper's The Open Society and Its Enemies (1945) formalized the structural shape of the vulnerability. A society committed to tolerance has to be intolerant of intolerance, or the intolerant will use tolerance to destroy the society that extends it. The paradox is not rhetorical — it is the structural form of every cooperative practice's exposure to unilateral invocation. The defense Popper proposed (tolerance of the intolerant ends where the intolerant refuse to engage on rational terms) is the lineage ancestor of the reciprocation-conditional move on this page.
John Rawls' Political Liberalism (1993) developed the closely related concept of the limits of reasonable pluralism. A liberal framework owes reasonable disagreement its full engagement but does not owe unreasonable positions the same standing, because extending equal standing to positions that reject the framework's own conditions would unravel the framework. The Rawlsian distinction between reasonable and unreasonable disagreement is the political-philosophy form of the distinction this tool draws between reciprocal invocation and unilateral exploitation of cooperative practices.
The philosophical literature on tu quoque and procedural hypocrisy carries the argumentative form. The classical tu quoque fallacy holds that pointing out a critic's hypocrisy does not refute their argument — and that is correct as a matter of formal logic. But the cooperative vulnerability is not about logic. It is about the operation of practices that depend on reciprocity. When a practice requires reciprocity to function, asymmetric invocation is not a logical fallacy; it is a structural breakdown of the practice. The same act that would be a tu quoque in pure argumentation is a structural diagnostic in cooperative practice. The two literatures are not in conflict; they are operating on different layers.
Work on weaponized civility and procedural capture across institutional theory (regulatory capture, deliberative democracy critique, organizational behavior research on toxic norm enforcement) documents the asymmetric-invocation pattern empirically across many specific contexts. The patterns appear independently across the disciplines, which is the kind of evidence the Knowledge instruments-as-triangulation move treats as load-bearing.
Cross-references
Within the category. The Cooperative Vulnerability operates under the lens of Adversarial Dynamics, which carries the category's two-adversary architecture and the diagnostic question of whether the other party is accountable to the same norms they are invoking. The Cooperative Vulnerability and Trust Mining are structural siblings: Trust Mining names the temporal accumulation pattern; The Cooperative Vulnerability names the rhetorical-and-practice architecture that the accumulated authority operates through. In live cases the two run together — accumulated trust provides the platform from which the cooperative practices can be invoked asymmetrically with maximum effect.
Across the Workshop. Receiving Disagreement Well in the Bond is where Steelmanning's relational application lives, and Steelmanning is the single most-frequent cooperative practice the tool's mechanism operates on. The reframing the lens performs on Receiving Disagreement Well is that the Steelman requirement is reciprocation-conditional rather than unconditional — an asymmetric demand to steelman the other while the other engages in caricature is exactly the cooperative vulnerability operating at the level of a single conversation. The Reciprocation Test from Adversarial Dynamics' practice section is the operative diagnostic in that reframing.
Limitations. Three worth naming. First, the Reciprocation Audit can drift toward score-keeping that consumes more attention than the cooperative practice itself produces, which is a way of losing the cooperative posture in the act of defending it. The defense is to run the audit at the pattern level rather than the within-interaction level. Second, surfacing asymmetry can be costly even when the audit is correct, because the asymmetric actor often responds with framings that put the burden of cooperative practice back on the practitioner surfacing the asymmetry. The Bond's other categories carry the practices for receiving that pressure without absorbing it. Third, the tool is not a license for refusing cooperative practice generally. A practice that has been weaponized asymmetrically by one actor is not a practice that should be withdrawn from all parties; it is a practice whose extension should become reciprocation-conditional toward that actor specifically. Generalizing from one bad-faith encounter to a refusal to extend cooperative practice broadly is itself a known drift toward Severance.