From Fragmented Research to Cumulative Strategic Knowledge: A Directed Research Architecture for Federating Scientifically Autonomous Projects

Abstract

CONCEPT NOTE Directed Research Architecture (DRA): From Fragmented Research to Cumulative Strategic Knowledge Complex strategic problems increasingly generate research across multiple disciplines, institutions and methodological traditions. However, the accumulation of individual studies does not necessarily produce cumulative strategic knowledge. Research may remain fragmented across projects, organisations, disciplines and methods, limiting the capacity to compare findings, preserve contradictions, identify evidence gaps and progressively reconfigure future research activity. The Directed Research Architecture (DRA) is proposed as a research-governance architecture designed to federate independently governed scientific projects around a shared strategic problem space while preserving substantive scientific autonomy. The DRA does not seek to impose a unified theory, methodology, interpretation or publication process on participating researchers. Instead, it provides a common evidence architecture through which autonomous projects can remain scientifically independent while contributing to a structured and cumulative body of knowledge. The architecture is organised around five core functions: 1. Shared Strategic Problem SpaceParticipating projects are connected through a clearly defined strategic problem space. The shared problem establishes a common field of inquiry without predetermining the theoretical approaches, methods, findings or conclusions of individual projects. 2. Shared Evidence ArchitectureA common evidence architecture provides the structures required for traceability, comparability and cumulative analysis across projects. It may include shared metadata, evidence records, definitions, indicators, data dictionaries, reporting structures and other scientifically justified mechanisms that facilitate cross-project interpretation without eliminating methodological diversity. 3. Independently Governed Research PortfolioThe research portfolio consists of projects that retain substantive autonomy over theoretical, methodological, interpretative and publication decisions. Independent scientific governance therefore refers to four essential dimensions: theoretical autonomy, methodological autonomy, interpretative autonomy and publication autonomy. Federation does not imply central scientific control. 4. Contradiction-Preserving Cross-Project SynthesisEvidence generated by participating projects is compared and synthesised across project boundaries. Convergent findings are examined alongside divergent, null, negative and context-dependent results. Contradictions are not treated as failures to be eliminated but as scientifically relevant evidence requiring explanation, contextualisation and further investigation. 5. Evidence-Driven Federated Portfolio RenewalThe synthesis produced at portfolio level is used to identify unresolved questions, emerging evidence gaps, new uncertainties and areas requiring further investigation. These findings then inform the questions, evidentiary priorities or composition of a subsequent independently governed research portfolio. The process is therefore recursive rather than linear. The central scientific outcome proposed by the DRA is Cumulative Strategic Knowledge. This refers to a progressively structured body of evidence generated by critically connecting autonomous research contributions while explicitly accounting for convergence, contradiction, contextual variation, methodological limitation and remaining uncertainty around a shared strategic problem space. Research federation is understood here as the structured connection of independently governed scientific projects through a shared evidence architecture that enables cross-project comparison, contradiction-preserving synthesis and recursive portfolio renewal without requiring unified scientific governance. Federation is therefore analytically distinct from integration. Integration concerns how heterogeneous knowledge is brought together; federation concerns how independently governed knowledge-producing units are connected without requiring their scientific governance to be unified. A central governance principle of the DRA is interaction without epistemic subordination. Scientific and institutional actors may interact, exchange information and coordinate activities while maintaining explicit responsibility boundaries between strategic agenda-setting, scientific production, cross-project synthesis and institutional decision-making. Strategic priorities may help determine which problems are important to investigate, but they must not predetermine scientific methods, findings or conclusions. The DRA is conceived as a configurational contribution rather than an elemental one. It does not claim novelty for problem-oriented research, interdisciplinary collaboration, research portfolio management, knowledge integration, evidence-gap mapping, continuous evidence synthesis, adaptive learning, programme evaluation or research-priority setting considered separately. These mechanisms are already represented in existing literatures. The proposed contribution lies instead in their configuration around a federated portfolio of independently governed research projects. The framework is therefore presented as a theoretically specified and empirically testable research-governance architecture, not as a validated or superior model. Its scientific value remains an empirical question. Future development should proceed through external conceptual review, operational measurement, bounded pilot implementation, mechanism testing, comparative research and international replication. A bounded pilot could involve a clearly delimited strategic problem space, several autonomous research projects, multiple disciplinary and institutional settings, a shared baseline evidence architecture, formal cross-project synthesis points and at least one documented portfolio-renewal decision. Initial evaluation should focus on feasibility, traceability, preservation of scientific autonomy, cross-project synthesis and the development of cumulative strategic knowledge rather than on claims of policy effectiveness. The Directed Research Architecture ultimately seeks to address a fundamental scientific challenge: how independently governed research projects can remain genuinely autonomous while contributing to a structured, cumulative and strategically relevant body of knowledge. Author: Jamel DouraJournal: DIM Review – International Review of Knowledge Management and Strategic PlanningISSN: 3125-9308Year: 2026Article type: Conceptual and Methodological Research Article