Trust and Network Patterns — Redundancy and Chokepoints 1.2.6-1

What Network Trust Reveals

The article admits network trust and chokepoint analysis as a source-pattern family for bounded interdependence. This sidebar expands the vocabulary — not the argument — so that readers can apply the structural concepts to cooperative arrangements they encounter.

A network, in the structural sense relevant here, is any system in which participants depend on connections to other participants — for information, for value exchange, for access, for credentialing. The network need not be digital. A professional association, a supply chain, a referral community, and a platform marketplace are all networks. What makes network analysis load-bearing for bounded interdependence is a single insight: the structure of connections determines the distribution of power, and power concentrates at chokepoints whether anyone intends it to or not.

Three Structural Concepts

Chokepoints are nodes or links through which a disproportionate share of network traffic must pass. A credentialing body that controls access to a profession is a chokepoint. A platform that hosts the primary marketplace for a type of service is a chokepoint. A data custodian that holds records no other entity can replicate is a chokepoint. The structural problem is not that chokepoints exist — some degree of concentration is inevitable in any network. The problem is that chokepoint controllers acquire dependency power over every participant who must traverse the chokepoint, and that power accumulates silently, without any single act crossing into coercion. It is a design liability that grows on its own — a structural condition worth naming and monitoring, not a rights violation waiting to be discovered.

Redundancy is the presence of multiple independent paths through the network. Redundancy is the primary structural defense against chokepoint power because it provides alternatives. If there is only one credentialing body, every professional depends on it. If there are three, each operating independently with mutual recognition, no single body can hold the profession hostage. Redundancy is not duplication — it is not the same service replicated identically. It is the availability of functionally equivalent alternatives that a participant can switch to without prohibitive cost.

Contestability is the structural capacity to challenge, route around, or replace a node that abuses its position. Redundancy provides the alternatives; contestability provides the mechanism for switching to them. A network with multiple paths but prohibitive switching costs has redundancy without contestability — the alternatives exist in principle but cannot be reached in practice. Contestability requires that switching costs be bounded, that interoperability standards exist between alternative paths, and that the act of switching does not trigger retaliation from the node being abandoned.

Fault Assumptions

Network trust analysis builds on a design discipline called fault assumptions — the practice of designing systems under the assumption that any node can fail, defect, or be captured at any time. The discipline originated in distributed computing, where the question was not will a node fail but when — and whether the system can continue functioning when it does.

The relevance to bounded interdependence is direct. A cooperative framework that assumes every participant will act in good faith has made a trust assumption that network analysis treats as structurally naive. The design question is not whether participants will defect, extract value, or capture chokepoints. The question is whether the cooperative framework can detect those conditions, route around them, and recover — without requiring that every participant be trustworthy at all times. This is the network-level expression of the Rule of Admission's third test: scales without saints.

The restraint principle (1.2.3) provides the design logic: the system must fail safely, be auditable, and have replaceable components. Fault assumptions apply that logic to network structure. A cooperative network designed under honest fault assumptions has no single point of failure, no irreplaceable node, and no chokepoint whose capture would bring the network down. Whether such a network can be achieved in practice is a design challenge. That it must be the design target is a structural requirement.