Relational Justice Analysis (RJA)

Relational Justice Analysis (RJA) is a modular, rule-based qualitative analysis framework designed to model relational dynamics, narrative structure, and justice-oriented meaning-making in interview and discourse data. Unlike black-box AI systems, RJA operates without machine learning and emphasizes epistemic transparency, reproducibility, and theoretically grounded annotation logic.

The framework integrates four analytical layers: (1) narrative structure detection, (2) frame-based discourse analysis, (3) affect trajectory modeling, and (4) relational justice mapping. Through stable annotation identities and cross-level integration, RJA enables the reconstruction of escalation patterns, coherence breaks, tension constellations, recurring interpretative bundles, and normative positioning within social interactions.

RJA models justice not as an abstract norm but as an emergent relational configuration shaped by affective density, discursive framing, and interactional sequencing. By operationalizing interpretative decisions in explicit rule sets and graph-based representations, the framework offers an explainable alternative to probabilistic language models while maintaining analytical depth.

Designed for digital humanities, social anthropology, oral history, and socio-technical research contexts, RJA provides an extensible infrastructure for theory-driven qualitative data analysis under conditions of methodological transparency.