Cross-Ideological Trust Under Divide-and-Conquer

Question. Does working a hard problem together, and the disclosure it requires, build durable trust across an ideological divide where dialogue does not?

Analogy. It works like a two-person crosscut saw: the blade only cuts when both sawyers pull and release in rhythm, so each has to read the other's timing or the saw binds and stops, and a working trust forms in the cut between people who never speak across it and might agree on nothing off the log. The study asks whether that wordless interdependent work builds a more durable cross-divide trust than talking the difference out, whether the bond holds once the sawing stops, and whether it survives a message built to split the pair.

What's at stake. The proposal's structural counter to divide and conquer is not counter-messaging but a sparse lattice of bridge-builders: a few trusted people per community who hold a real cross-divide tie, so a divisive message meets someone the bridge-builder already knows and trusts on the other side. The resilience comes from those few ties, not from mass goodwill, and the proposal leaves the load-bearing question open: whether the disclosure of working a hard problem together, not dialogue about the divide, is what forges them. If it does not, the counter is rhetorical.

The two answers it decides between. Either cross-divide trust needs the divide worked through directly, the way cohesion programmes are run and what makes them expensive; or it forms sideways from the paired work itself, as each person discloses their real situation and engages the other's thinking, the divide never the topic. The study reads variation the platform already makes: far-apart pairs against same-side pairs doing identical work, and whether trust toward the other side grows with the amount of paired engagement. If it does, the costly engage-the-difference step is not what builds the tie, and the platform produces it for free.

What a null result would mean. If far-apart pairs build no more durable cross-divide trust than same-side pairs, or trust does not rise with the amount of paired engagement, the structural counter to divide and conquer needs revision: the tie would require the engagement the technique was meant to make unnecessary. That is a finding about the theory, not a sign the platform was built wrong.

Why this matters to defence. Divide and conquer turns Canada's multicultural structure into its largest attack surface, and the Foreign Interference Commission named social cohesion as a target (DRDC Objective 6; DRDC Objective 5). The result decides whether cohesion programmes are built around paired problem-work or around dialogue and counter-messaging, and, since the paired work costs almost nothing once the platform exists, whether the cheaper route is the one that holds under a real divisive campaign.

How we would run it. At intake, measure each operator's ideological position, baseline trust toward the other side, and prior openness (the covariate that guards against self-selection). Pairs form across the full range of distance, and content-blind telemetry already logs how much and how deeply each pair works together, so nothing is staged. Re-measure trust later for rise, durability once practice stops, and resistance to a divide-stoking message. Same-side pairs are the free internal comparison, and the effect is read against the cohesion-programme effect sizes already published, so no dialogue arm is funded. The naturalistic cross-divide pairing is what a managed cohort cannot reproduce. The cross-community bridges study carries the network-scale version; this is the dyad it depends on.

Earliest start. Stage 2: the study needs guide-operator pairs working real problems together, which the Guide Protocol delivers. Funded.