Robin Edgard Ulrik Mertens

How institutions stay aligned without central control. How can large organisations remain coherent without becoming rigid? This paper introduces Distributed Coherence: alignment as an emergent property maintained across decision systems rather than enforced from the top. By connecting architecture, measurement, and governance, it explains how institutions preserve intent while enabling distributed decision-making. Coherence becomes a system property—not a command. About the Coherence Programme The Coherence Programme studies why institutions drift despite appearing aligned. It shows that decisions are made not on intent itself, but on how intent is translated into criteria, metrics, and allocation rules. Using the Operating Spine, the programme traces how purpose becomes action across governance layers, making drift and coherence directly observable within decision systems. The research applies to public institutions, capital allocation, and AI-mediated environments, where the durability of decision rules determines long-term institutional reliability. Programme citation: Mertens, R. E. U. (2026). The Coherence Programme: A Conceptual Overview and Entry Point to the Research Programme. Resources: Coherence Programme OSF repository and https://thecoherenceprogramme.org Version 1.00: First public release of the Capstone synthesis essay. This manuscript develops the programme’s theoretical integration and research agenda. Empirical methods, measurement approaches, and design frameworks are presented in detail in the preceding programme papers. Version 1.01: Programme Consolidation Update: This version consolidates the manuscript within the unified Coherence Programme structure.Titles, terminology, and internal cross-references have been harmonised across the series to stabilise the programme’s core constructs: Translation Drift (mechanism) Translation Coherence (metric) Interpretive Maintenance (governance function) Distributed Coherence (theoretical integration) No changes have been made to the formal architecture, boundary conditions, methodological logic, or theoretical claims. The update improves cross-paper traceability, indexing consistency, and conceptual coherence across the programme. Version 1.02: Terminology harmonisation and minor structural refinements to improve consistency across the Coherence Programme. No changes to the theoretical framework, constructs, or research design. Version 1.03: Minor conceptual clarifications, consistency improvements, reviewer-informed refinements, and editorial updates across the programme.

How institutions forget what they are for. Modern institutions have become extraordinarily capable of coordination, measurement, optimization, and control. Yet many people increasingly experience organizations as strangely disconnected from the realities they were created to serve. This paper explores how institutions gradually drift from purpose—not through collapse or bad intent, but through layers of abstraction, translation, metrics, and representation that slowly become operational realities in their own right. Across governance, organizations, AI systems, and public institutions, decision systems often become highly effective at managing representations of alignment while losing visibility into the human and operational conditions beneath them. The paper examines how drift of purpose emerges structurally as institutional intent is translated into criteria, dashboards, allocation systems, and operational proxies that eventually begin governing decisions in place of the realities they were meant to reflect. What begins as coordination gradually becomes substitution: organizations optimize representations of alignment while losing relation to their underlying purpose. Bringing together the conceptual frameworks, empirical observations, and governance architectures developed across the Coherence Programme, the paper argues that resilient institutions require more than efficiency or innovation alone. They require the capacity to remain reflectively connected to the realities they claim to steward—even as complexity, scale, and technological mediation continue to increase. Coherence is therefore not control. It is the ongoing ability of an institution to remain connected to its purpose as decisions, systems, and interpretations evolve over time. At its core, this is a paper about whether institutions can still remember what they are for. This paper is part of the Coherence Programme.

How do formal representations displace institutional intent? This paper examines how institutions can appear aligned and high-performing while drifting from their stated purpose. It argues that this condition arises because intent does not enter decision-making directly: to guide action, it must be translated into criteria, metrics, and allocation rules. In this process, what begins as a representation of intent can come to replace it. The paper identifies the Double Translation Trap: a condition in which a proxy both guides allocation and serves as the standard by which success is judged. As a result, the system can no longer detect whether it is diverging from its original purpose. The paper introduces Translation Drift as the condition in which decision systems optimise for these representations rather than for the intent they were meant to serve. Once embedded through allocation, the institution learns to succeed on its own representations rather than on its intended purpose. The paper contributes a diagnostic framework that makes translation drift observable, actionable, and correctable at the point of allocation, enabling institutions to maintain alignment between intent and what they fund. This paper is part of the Coherence Programme.

This paper advances a philosophical account of institutional agency grounded in the structure of translation. It argues that institutions do not act directly on intention, purpose, or meaning as such, but on representations produced as these are rendered decision-relevant within structured systems of evaluation and allocation. From this perspective, institutional action is mediated by forms—criteria, metrics, models, and signals—that both enable and constrain what can be seen, compared, and selected. These forms are not neutral carriers of meaning. They are transformations of it. In becoming actionable, intention is selectively encoded, simplified, and redefined. What remains is not the original intent, but a representation that substitutes for it within the decision system. Institutional reality is therefore not merely interpreted but constructed through processes of translation. What counts as value, performance, or relevance is determined not only by normative commitments or actor beliefs, but by the representational forms through which decisions are made. These forms delimit the space of the possible: what cannot be represented cannot be selected. Within this framework, institutional drift is neither an anomaly nor a failure of rationality or alignment. It is a structural consequence of translation. As representations stabilize and become authoritative, they progressively redefine the meaning of the intentions they were meant to realize. Institutions may thus remain coherent in their operations while becoming increasingly detached from their originating purpose. The philosophical implication is a shift in the locus of agency. Agency is not located solely in actors, intentions, or decisions, but in the translation architectures that determine how meaning becomes actionable. To understand institutional behaviour is therefore to analyse how intention is transformed into representation—and how, through that transformation, the conditions of action are themselves constituted.

How can a system remain internally coherent while drifting away from its intended purpose? This paper examines the certification of the Boeing 737 MAX to show how institutional intent—flight safety—can be translated into an operational proxy that comes to guide both action and evaluation. As this proxy becomes decision-authoritative, contradictory evidence may no longer enter the decision process. The paper introduces the Double Translation Trap: a structural condition in which an operational proxy, initially used to implement intent, becomes the standard for judging alignment. Under this condition, systems can remain procedurally coherent while progressively losing correspondence with the conditions they are meant to represent. Using a reproducible artefact-tracing protocol, the analysis reconstructs how certification criteria could render internal test data non-decision-relevant within the formal decision chain. This allows misalignment to be identified within the formal record prior to outcome failure, as a mismatch between governing proxies and a system’s own evidence. The argument contributes a structural account of how decision systems can become closed to relevant information while continuing to function as designed, with implications for institutional governance, safety-critical systems, and AI-mediated decision environments. The paper contributes a conceptual and methodological framework for analysing institutional decision-making as a translation process in which meaning is progressively stabilised, reduced, and reconstituted across governance layers. By locating where translation becomes binding, it shifts the analysis of alignment from outcomes to the structure of decision formation itself. This paper is part of the Coherence Programme.

Institutional decision systems do not act directly on their stated purpose; they act on the representations into which that purpose is translated. This paper develops a structural account of how strategic intent becomes decision-relevant within complex governance systems. Drawing on insights from institutional theory, commensuration, and sensemaking, it argues that the transformation of meaning is not merely interpretive but becomes binding at the point where concepts are encoded into rules, criteria, and signals that govern allocation decisions. Using evidence from three domains of European Union governance—fiscal coordination, industrial policy, and research funding—the paper traces how concepts such as cohesion, strategic autonomy, and scientific excellence are translated into operational forms that enable comparison, verification, and coordination. It shows that, under these conditions, decision-making operates on these representations rather than on the originating concepts, and that aspects of intent that cannot be expressed in such forms are systematically excluded. The analysis identifies a structural condition—termed the Translation Trap—in which alignment is maintained at the level of rules and metrics while diverging from underlying purpose. This condition does not arise from implementation failure but from the requirements of decision systems operating under constraints of commensurability, auditability, and coordinated allocation. The paper contributes a conceptual and methodological framework for analysing institutional decision-making as a translation process in which meaning is progressively stabilised, reduced, and reconstituted across governance layers. By locating where translation becomes binding, it shifts the analysis of alignment from outcomes to the structure of decision formation itself. This paper is part of the Coherence Programme.

How do institutions act on abstract purposes such as “improving health” or “saving the planet”? This paper argues that they do not act on such purposes directly, but only on the representations into which those purposes are translated. As institutional intent moves across governance layers, it is reformulated into criteria, categories, and signals that make alternatives comparable and decisions executable. The analysis develops a representational account of institutional decision-making. Institutions are conceptualized as systems that depend on commensuration: heterogeneous aims must be rendered into comparable forms in order to support evaluation and allocation. This transformation is structurally necessary, but it also alters the meaning of the concepts being operationalized. Decisions are therefore made on representations of purpose rather than on purpose itself. Using a minimal and replicable traceability method, the paper reconstructs how institutional purpose is translated across publicly available governance artefacts in three cases: Novo Nordisk, Eli Lilly, and Patagonia. The analysis shows that as representations stabilize within decision systems, they become the effective objects of decision-making. This gives rise to a systematic divergence between originating intent and enacted criteria, referred to as translation drift. The paper contributes to social ontology and the philosophy of social science by identifying a structural condition of institutional action: what can be represented in decision-relevant form determines what can be selected. Institutional coherence is therefore not a property of outcomes or stated intentions, but of the relation between originating concepts and the representations through which decisions are made. By demonstrating that these transformations are empirically observable in governance artefacts, the paper connects philosophical accounts of representation and commensuration with the operational structure of real-world decision systems.

This paper examines how funding decisions are shaped by the translation of institutional intent into evaluation criteria. It analyzes funding systems across major organizations, including the European Research Council, Novo Nordisk Foundation, and the Gates Foundation, to show how abstract goals become operational through rules, indicators, and review processes. The paper demonstrates that what gets funded depends not only on stated priorities but on how these priorities are represented within decision systems. It develops a framework for observing translation dynamics in funding contexts and shows how misalignment can emerge even under strong governance. The paper contributes to the Coherence Programme by extending translation analysis to resource allocation systems.

This paper introduces the concept of the margin of purpose to explain how institutions gradually lose alignment with what they explicitly claim to value. It examines how purpose is translated into decision criteria and how elements that are difficult to formalize are progressively excluded. The paper shows that this loss is not abrupt but accumulates at the margins of decision systems, where qualitative aspects of intent fail to be represented. Over time, these omissions reshape institutional priorities and outcomes. The paper argues that preserving coherence requires explicit attention to what is left out of formal representations. It contributes to the Coherence Programme by extending the analysis of translation dynamics to include systematic exclusion.

This paper examines why well-functioning organizations still experience strategic drift despite strong governance. It introduces the concept of the coherence advantage, showing how systems that are locally consistent and efficient can become globally misaligned over time. The paper argues that leaders often detect drift only after it has become embedded in decision criteria and operational signals. It reframes drift as a structural property of translation across governance layers rather than a failure of execution. By analyzing how meaning becomes operational and gradually diverges from intent, the paper explains why early warning signals are difficult to observe. It contributes to the Coherence Programme by linking coherence, drift, and detectability in institutional decision systems.

This paper develops the concept of the sovereign spine as a structural mechanism through which institutions maintain alignment with declared intent over time. It examines how coherence can be preserved without centralized control by ensuring that purpose is consistently translated across governance layers. The paper contrasts this with models of governance that rely on enforcement or coordination alone. It argues that alignment emerges from the integrity of translation processes linking purpose, strategy, and operational decisions. The sovereign spine provides a minimal architecture for sustaining coherence in distributed systems. The paper contributes to the Coherence Programme by extending the theory of institutional coherence into a dynamic account of agency and continuity.

This paper examines how institutional reliance on dashboards and performance indicators can lead to a systematic loss of alignment with declared intent. It introduces the concept of the “green dashboard trap,” in which decision systems signal success through internally consistent metrics while drifting from their original purpose. The paper shows how translation dynamics across governance layers transform intent into indicators that gradually become ends in themselves. As a result, institutions may appear well-governed while silently accumulating misalignment. The paper reframes governance failure not as a lack of control but as a structural consequence of how meaning is operationalized. It contributes to the Coherence Programme by identifying a recurrent failure mode in modern data-driven governance systems.

This paper introduces AI-augmented impact frames as a closed-loop architecture for aligning institutional decision-making with declared purpose. It examines how institutions translate intent into operational criteria and how these criteria are increasingly mediated by algorithmic systems. The paper develops a model in which AI systems do not merely optimize predefined metrics but participate in structuring the interpretive frames through which impact is defined and evaluated. By embedding feedback loops between purpose, criteria, and outcomes, the architecture enables continuous recalibration of decision systems. The paper argues that without such closed-loop structures, institutions risk reinforcing drift through automated optimization of misaligned signals. It contributes to the Coherence Programme by extending the concept of translation dynamics into AI-mediated governance contexts.

This paper examines how institutional decision systems learn what is considered important over time. It develops a model of governance in which purpose is translated into capabilities, criteria, and signals that shape allocation and evaluation decisions. The paper argues that institutional learning is not neutral but structured by the representations embedded in decision systems. As these representations stabilize, they define what can be seen, measured, and acted upon. The paper shows how misalignment emerges when learning processes optimize locally coherent signals that diverge from original intent. It outlines design principles for building purpose-aligned governance systems that maintain coherence across decision layers. The paper contributes to the Coherence Programme by framing institutional learning as a translation

This paper introduces the coherence problem as a structural feature of institutional decision systems. It explains how misalignment emerges as meaning is translated across governance layers, producing persistent gaps between intent and action. The paper outlines how institutions can observe and respond to these dynamics, framing drift as an endogenous property of decision architectures. Companion papers examine how these dynamics manifest in practice (Why Strong Governance Still Drifts) and how they can be detected early (Detecting Strategic Drift Before Outcome Failure). This paper is part of the Coherence Programme.

This paper develops a minimal, rule-based method for observing how policy concepts are translated into decision criteria within institutional systems. Using three European Union programmes—the Recovery and Resilience Facility, Galileo, and Erasmus+—it traces how concepts move across governance artefacts into rules, indicators, and verification signals. The analysis demonstrates that translation dynamics are directly observable and that institutional decisions are shaped by translated representations rather than declared intent. This paper is part of the Coherence Programme.

This paper introduces the Coherence Programme, a research programme examining how institutional decision systems maintain or lose fidelity to declared intent. It models governance as a translation architecture in which purpose is progressively transformed into capabilities, strategy, and decision signals. The paper outlines the core concepts—translation dynamics, coherence, and drift—and positions the programme as a framework for analysing how meaning becomes operational in complex institutional settings. This paper is part of the Coherence Programme.

Tagline: The minimal architecture behind institutional learning. Institutions must make decisions in real time while learning under conditions of delayed feedback. This creates a persistent tension between operational performance and fidelity to declared purpose. This paper introduces the Operating Spine as a minimal causal architecture for analysing how institutional intent becomes operational decisions. By tracing how purpose is translated into capabilities, strategy, portfolio allocations, and performance signals, the framework identifies where interpretation can diverge from the meanings originally intended. Institutional drift is therefore treated not as an anomaly but as a structural possibility inherent in complex decision systems. The analysis provides a conceptual foundation for studying institutional learning as a problem of translation across decision architectures. This paper is part of the Coherence Programme.

Tagline: Governance as the maintenance of institutional meaning. In digitally mediated institutions, interpretation increasingly becomes embedded in technological artefacts such as dashboards, scoring systems, and algorithmic models. These artefacts structure how organisational actors interpret value, risk, and performance. This paper conceptualises such systems as a form of institutional Meaning Infrastructure. Rather than treating governance as episodic oversight, the paper proposes that institutions must actively maintain the interpretive structures through which meaning becomes operational decisions. The concept of Interpretive Maintenance captures this ongoing responsibility to monitor and recalibrate the translation interfaces that connect purpose to decision criteria. The analysis situates governance within the broader philosophical problem of how meaning remains stable within complex socio-technical systems. This paper is part of the Coherence Programme.

Tagline: Alignment without centralised control. How can complex institutions remain unified without suppressing distributed agency? Traditional theories of governance often assume that alignment requires centralised control or hierarchical enforcement. This paper proposes an alternative account in which coherence emerges as a structural property of institutional decision systems. The theory of Distributed Coherence developed here treats alignment as the outcome of how institutional intent is translated across layered decision architectures rather than imposed through authority. By linking interpretive stability to decision structures, the paper offers a framework for explaining how institutions can preserve fidelity to purpose while allowing distributed decision-making. The result is a conceptual account of institutional agency in which coherence is maintained through the architecture of translation rather than through centralised command. This paper is part of the Coherence Programme.