Symmetric Adoption in Negotiation

Question. When both sides of a real dispute bring strategic empathy, does it converge to win-win or escalate into an arms race of cleverness?

Analogy. It works like two convoys forced onto one road, each able to read the other's timetable: read each other right and they interleave so both columns move faster than either could alone, or each grabs for the road first and both jam it solid, the harder they out-time each other the worse the stall. The study asks whether two-sided strategic empathy interleaves the traffic, converging to a settlement both sides value, or jams it, escalating into an arms race of cleverness where each grows more deviously evasive.

What's at stake. The proposal makes a large dual-use claim about strategic empathy: the same capability that lets an operator defeat an adversary, finding the assumption their plan depends on and breaking it, lets two civilian parties find a deal neither could see from inside their own frame. The whole civilian half of the case, cross-divide negotiation, cohesion as a counter to divide-and-conquer, the strategic bridges that bind the network, rests on strategic empathy being cooperative when both sides have it. But there is an obvious darker reading: strategic empathy is a weapon that does not care which way it points, so two trained sides could simply out-deceive each other faster, an arms race rather than a handshake. The proposal names both poles and bets on the first; no study has tested which one actually happens.

The two answers it decides between. Either, when each side can accurately model the other, each does creative thinking on the other's behalf and surfaces options that expand the deal, so two-sided strategic empathy produces better joint outcomes and faster settlements than one-sided or neither (the integrative-bargaining view); or strategic empathy is an offensive capability, so two sophisticated modellers escalate deception and manoeuvring, and two-sided training produces more deadlock and mutual evasion, not less (the security-dilemma view). The shape of real two-sided negotiations tells them apart.

What a null result would mean. If two-sided strategic empathy produces no more win-win outcomes, and no faster convergence, than one-sided or untrained, or produces worse, the proposal's negotiation-empathy and cross-divide-cohesion claims need revision: strategic empathy would be an adversarial weapon, not the cooperative bridge the civilian case relies on. That is a finding about the theory, not the platform, and one of the more consequential nulls, because the civilian and cohesion half of the proposal leans on it.

Why this matters to defence. Whether trained strategic empathy de-escalates or escalates is a real planning question at population scale (DRDC Objective 6, the cognitive level; DRDC Objective 5 and cohesion against divide-and-conquer; DRDC Objective 3). It changes a concrete call: whether to treat population-scale strategic-empathy training as a cohesion and de-escalation asset or to flag it as an escalation risk, and whether cross-divide cohesion and negotiation programmes can rely on mutual strategic empathy to bridge fault lines.

How we would run it. Set up real or realistic negotiations and disputes and vary how many sides are trained: neither, one, or both. Measure two outcomes: how good the final settlement is for both parties together, the size of the joint pie and whether it is win-win or win-lose, and how long it takes to reach it. Vary the conditions likely to be decisive: the power balance between the sides, the time pressure, and whether a trusted person bridges them. Track whether two-sided training pushes outcomes toward win-win and faster settlement, or toward deadlock and mutual deception, measured as parties increasingly misleading versus increasingly disclosing. The only-us part is real stakes with trained parties on both sides of a genuine dispute, which a managed cohort cannot stage and which only the trust network produces once adoption reaches both sides of real conflicts.

Earliest start. Stage 13: the study needs adoption to have reached both sides of real disputes, so trained parties face each other across a genuine conflict.