Foundations.

Coordination Geometry and Binding Architecture in Hybrid Organisational Systems

Conference paper · in preparation · 2026

Distributed execution systems separate what gets codified from what gets interpreted. Procedures, protocols, and decision rules can scale independently of the contextual judgment that grounds them. This paper argues that the resulting divergence — not authority clarity alone — is the primary structural driver of coordination instability.

The model introduces two orthogonal dimensions: operational formalisation (the degree to which activity is rendered executable) and interpretive orientation (the degree to which execution remains grounded in contextual judgment). These dimensions vary independently. Their divergence produces systemic strain. The coordination space they define precedes and conditions governance.

Within this geometry, binding architecture governs where commitment transfers: from situated interpretation through encoded authority to system execution. Decision maps, delegation structures, and review mechanisms are derived from this coordination logic rather than from practitioner intuition.

Hybrid and automated systems intensify divergence by accelerating formalisation growth. But the underlying geometry is not technology-specific. It applies wherever execution can be codified independently of contextual interpretation.

The contribution is layered. To organisation theory: an intra-organisational alignment condition independent of environmental contingency. To design theory: implementable governance artefacts derived from a minimal explanatory geometry. The claim is that organisational instability in distributed systems is less often a product of unclear authority and more often a product of structural disproportionality.

PDF available upon submission.

The diagnostic instrument

The method I use assesses how a workflow behaves under delegation: where decisions sit, how review expands, how escalation behaves, and where ownership becomes unclear.

The instrument operates on two axes. Three communication channels — how teams coordinate with each other, how people interact with the systems they delegate to, and where automated handoffs require human oversight. Four operational layers — whether shared understanding is established before work begins, whether there is structured iteration during work, whether authority boundaries are defined, and whether verification improves the process or merely catches errors.

Their intersection produces twelve diagnostic dimensions. Each is scored on a five-point maturity scale, from absent to integrated. The gap between current state and designed practice defines the scope of intervention.

Decision Architecture Framework

Structural framework for preserving decision continuity under delegation

The Decision Architecture Framework specifies the structural conditions under which decisions remain coherent as they propagate across distributed actors, systems, and time. It formalises the topology, propagation dynamics, and structural integrity of delegated decision environments.

Two independent properties determine whether delegated decisions remain stable: clarity of definition at origin, and continuity under propagation. Their intersection produces four structural conditions.

Decision architecture precedes governance.

Framework operational since 2025. Applied in organisational and AI-mediated decision environments.

Traditions

The coordination geometry is derived from several established research traditions, each contributing a specific structural element to the model.

Distributed cognition (Hutchins) relocates cognitive performance from the individual to the system — the move that makes proportional divergence between formalisation and interpretation legible as an organisational condition rather than an individual failing. Grounding theory (Clark) establishes that shared understanding is not a precondition but an active, ongoing accomplishment — which is why interpretive orientation requires continuous calibration rather than one-time specification. Situated action (Suchman) demonstrates that plans do not fully determine behaviour; action unfolds in context — the theoretical ground for treating interpretation as structurally irreducible to codification.

Actor-network theory (Latour) insists that all participants in a workflow, human and non-human, function as actants requiring explicit accounting — the basis for treating hybrid execution systems as coordination problems rather than technology problems. Joint intentionality (Tomasello) specifies the conditions under which collaborative action achieves functional alignment — structuring the binding architecture that governs commitment transfer. Organisational learning (Argyris & Schön) distinguishes single-loop correction from double-loop recalibration — the architecture of review that detects and corrects proportional divergence.

Origin

Where Decision Architecture began

Decision Architecture did not begin as a governance theory. It began as a constraint.

The early work concerned human-to-AI and system-to-system communication — environments in which ambiguity cannot be socially repaired. A system cannot "roughly understand." It must bind interpretation to execution. An instruction must resolve into a state transition.

At that boundary, something became visible. The decision itself does not move. Only its representation moves. The receiving actor — human or system — reconstructs under new conditions. The reconstruction is competent. It is also structurally different.

Initially, this appeared to be a machine problem. It was not. The same pattern existed in purely human environments: executive to team, team to team, specification to implementation, escalation to resolution. In every case, instability originated not in competence, but in transfer.

Authority shifted. Interpretation narrowed. Commitment formed. And continuity could not be assumed.

The hinge — the delegation boundary where interpretation becomes commitment and responsibility transfers — proved to be a structural property of coordination itself.

From this recognition followed instrumentation: a dimensional representation of delegation stability, the Decision Architecture Framework, programmes to develop structural perception, and formal articulation of Decision Architecture as a field.

The hinge was not invented. It was exposed.

Context

Why delegation architecture is now a board-level issue

Enterprise systems are scaling delegation faster than they are scaling clarity. AI agents act. Distributed teams execute across platforms. Regulatory regimes mandate oversight and traceable accountability. The "accountability gap" is now a named problem.

At its core, this is not a technology issue. It is a delegation issue. When authority transfers, only a representation of the decision moves. The receiving actor reconstructs it under new constraints. Unless that transfer is deliberately specified, continuity cannot be assumed.

Tooling can log decisions. Governance can assign responsibility. Compliance can enforce documentation. None of these stabilise delegation boundaries.

Decision Architecture operates beneath governance and tooling. It addresses the structural integrity of the hinge — the point where interpretation becomes commitment and responsibility transfers.

As delegation scales, instability scales with it. The current moment did not create this condition. It exposed it.

The question is no longer whether decisions are well made. It is whether they survive propagation intact. Structural questions require architectural answers.

The Coordination Layers Model

These programmes are grounded in a single structural observation: what organisations call "culture" is the visible surface of deeper stabilisation layers.

Architecture → Constraints → Reinforcement Signals → Behaviour

Causality flows upward. Visibility flows downward. Participants see behaviour. Architecture generates it.

The model explains why behavioural patterns persist — not because people believe in them, but because the system makes them rational. It identifies the leverage layer (constraints) where structural intervention produces durable change. And it predicts where intervention will fail: when it targets behaviour or values while leaving reinforcement conditions intact.

Each programme corresponds to a layer of the model. Alignment makes stabilised behaviour visible. Decision-Making exposes the reinforcement signals that maintain it. Decision Clarity and Ownership modifies the constraint structures that produce it. The direct workflow work operates at the architecture layer.

The instrument and the framework sit behind both the programmes and the workflow intervention.