Chapter 3 The Peace‑DAG
In the previous chapter, we learned that even once war has begun, escalation is not inevitable. By intervening on key causal links, we can weaken the forces that pull actors toward joint attack and instead open pathways toward restraint. Chapter 3 takes this logic one step further. Here we construct the full Peace‑DAG—a causal architecture designed not merely to stop escalation, but to make de‑escalation the default outcome of a crisis. Using the same actors and the same triggering event, we reorient the arrows so that the system channels conflict into mediation, monitoring, incentives, and ultimately a structured peace process. This chapter shows how peace can be designed into the system itself, transforming a crisis from a spark for war into a trigger for diplomacy.
The Peace‑DAG keeps the original crisis node—Z attacks I—but rewires everything that follows. Instead of feeding into alliance pressure and joint attack, the crisis activates a sequence of mechanisms that lead toward ceasefire and negotiation.
The Peace‑DAG is not a list of new mechanisms. It is a redirection of causal flow using the glossary’s tools.
3.1 Redirect the Crisis Node
Instead of:
\[ B \rightarrow K \rightarrow C \rightarrow Y \]
We design:
\[ B \rightarrow M \]
Mediation becomes the new mediator, replacing alliance pressure.
3.2 Block Harmful Paths
In the Peace‑DAG, Q is the monitoring and verification node:
B → M → Q
Monitoring Q blocks:
- Misperception (S, Pᵤ, P_z) — confounders that distort threat assessment
- Proxy misinterpretation — e.g., mistaking mobilization for aggression
- Co‑incident escalation — avoiding conditioning on joint responses
3.3 Introduce Positive Mediators
Incentives (D) become mediators of restraint:
\[ M \rightarrow D \rightarrow C' \]
3.4 Remove Effect Modifiers of Escalation
Security guarantees reduce the modifying effect of fear and vulnerability.
3.5 Avoid Co‑Incident Nodes
Instead of “joint attack,” the co‑incident node becomes:
\[ C', Z' \rightarrow CF \]
A joint ceasefire, not a joint attack.
This chapter demonstrated that peace is not merely the absence of war but the product of a deliberate causal design. By redirecting the crisis node into mediation, monitoring, incentives, and de‑escalatory responses, the Peace‑DAG transforms the same initial shock that once led to joint attack into a pathway that culminates in joint restraint and structured settlement. The architecture of escalation becomes the architecture of peace.
Yet the Peace‑DAG is built around a specific scenario: one attacker, one ally, one alliance structure. Real conflicts vary widely in scale, actors, and political context. Some involve internal factions rather than states; others unfold within regional systems or under international oversight. To make the causal tools of this book broadly useful, we must generalize the logic of the Peace‑DAG into a model that applies across conflicts.
Chapter 4 takes this step. It distills the lessons of the war‑DAG and Peace‑DAG into a universal framework for conflict mediation and resolution. By abstracting from the specific scenario, we identify the essential causal elements that appear in most conflicts—triggers, confounders, mediators, co‑incident nodes—and show how they can be systematically redirected toward negotiation and settlement. The result is a general DAG model that policymakers, mediators, and analysts can use to diagnose where a conflict stands and design interventions that reliably shift the system toward peace.