Making Retroduction Transparent, Not Algorithmic: An Integrated Danermark–RRREIC Protocol for Mechanism-Based Explanation in Critical Realist Qualitative Research

Critical Realist (CR) empirical research routinely invokes retroduction and generative mechanisms without specifying how these operations are performed on data. Published CR case studies rarely distinguish retrodiction from retroduction as separate inferential operations (Fletcher, 2017; Mingers and Standing, 2017). This article integrates Danermark et al.'s (2002) seven-stage explanatory protocol with the RRREIC inferential framework, separating retrodictive reconstruction from retroductive mechanism conjecture as distinct, documented stages. Taking seriously Bhaskar's warning against reducing applied CR to 'a formula or recipe or algorithm', the article treats transparency as reconstructability and contestability rather than procedural convergence: mechanism claims need not converge across analysts, only remain traceable, scrutinisable, and contestable. A seven-case hypothetical demonstration in a knowledge-intensive acquisition context illustrates the protocol's operation; no empirical data were collected or analysed, and the demonstration provides no empirical validation of the protocol itself.