AI-enabled scholarly workflow

Agentic Inquiry, Human Authority

AI expands how far scholarly inquiry can travel — across evidence, interpretations, methodological questions and competing explanations. Human judgement determines what is ultimately trusted.

This workflow combines agentic AI capability with two complementary safeguards. Independent audit challenges reasoning before it hardens into conclusion, testing whether important claims remain defensible when examined from another angle. Human adjudication and verification preserve researcher authority at consequential moments — deciding what should be accepted, revised, qualified or left unresolved.

Together, they create a model of scholarly work in which AI can explore ambitiously and reason extensively, while audit strengthens confidence and human judgement retains authority. The inquiry is then advanced through two complementary human-led perspectives. One researcher develops the Three Clocks inquiry, tracing the paper through historical position, theory, construct genealogy, methodological reconstruction and later evidence. A second researcher develops the Six Drivers inquiry, testing the study through design, sampling, measurement, unit of analysis, analytical technique and rival explanations. In both streams, AI supports exploration, challenge and checking, but the researcher retains responsibility for interpretation and adjudication. The two independently developed judgements are then harmonised into a single scholarly judgement and translated into one shared Three Clocks, Six Drivers simulation.

Durham–emlyon Global DBAInteractive scholarly guardrail mapv0.1
Clock 1 — Before fieldworkTheory, constructs and evidence available before the 2008 fieldwork.
Clock 2 — Fieldwork to publicationEvidence fixed through fielding, analysis and the 2016 publication.
Clock 3 — After publicationLater evidence, contemporary reinterpretation and translation into new artefacts.

Three Clocks, Six Drivers. One governed inquiry.

Read the workflow from bottom to top. The frozen left rail keeps every workflow stage visible while the historical evidence plot scrolls horizontally.
Harmonised endpoint: two human-led inquiry streams — Three Clocks and Six Drivers — converge in one shared scholarly judgement before being translated into the simulation; the workflow itself closes the record as an accountability artefact.
Purple — human authority Red — AI audit Purple + red — agentic–human analysis
Select any row to reveal its role, evidence window and guardrail function.