FID-058 · Open question
Tradition-Specific Constraint Formalization
How can tradition-specific boundaries, source hierarchies, doctrinal constraints, and disagreement patterns be translated into machine-checkable specifications without flattening differences across faith traditions?
Why the question remains open
Faith-facing AI systems often answer as if there is one generic religious perspective. In reality, traditions differ on authoritative sources, doctrine, practice, institutional authority, and acceptable forms of counsel. Formalizing these boundaries could help systems avoid giving tradition-confused or overgeneralized answers.
Working hypothesis
A proposition to test, not a finding.
Some tradition-specific constraints can be expressed as bounded specifications: which sources count as authoritative, which claims require qualifiers, which topics are contested, which roles carry authority, and when an answer should name disagreement rather than synthesize it away.
Proposed method
How the question could be tested
- 01Select a small number of traditions, denominations, schools, or institutional contexts with willing reviewers.
- 02Define source hierarchies, required qualifiers, forbidden conflations, authority boundaries, and known disagreement areas.
- 03Convert these into structured constraints that can be applied to model and RAG outputs.
- 04Measure whether constraints reduce tradition-confusion errors without suppressing legitimate diversity or uncertainty.
Needed controls
What must constrain the study
- 01Do not present one group's formal specification as representative of all members of a broader faith tradition.
- 02Require named scope, source basis, reviewer basis, and expiration or revision process for each specification.
- 03Include disagreement and minority-view handling.
- 04Keep descriptive constraints separate from product policies or institutional approval decisions.
Expected outputs
Artifacts the work should produce
- 01Tradition-constraint specification template.
- 02Source hierarchy and qualifier schema.
- 03Evaluation set for tradition-confusion and overgeneralization failures.
- 04Reviewer workflow for maintaining and revising constraints.
- 05Public guidance on using formal constraints without overstating authority.
Open questions
Uncertainties the protocol must resolve
- 01Who is authorized to define a tradition-specific specification?
- 02How should conflicting reviewers be represented in the spec?
- 03Which constraints are stable enough to encode, and which should remain human review notes?
- 04How can this support Christian use cases while remaining useful across faith traditions?
Related calls
Continue through this research area
FID-077
Independent Agent Incident Investigation and Evidence Sufficiency
What operational evidence lets independent investigators reconstruct an agent incident, distinguish competing explanations, and identify which interventions could have changed the outcome?
FID-005
Scripture, Tradition, and Moral-Framing Interventions
Do Scripture, sacred tradition, religious identity, familial embeddedness, or other morally thick framings measurably change model behavior in faith-facing tasks, and can those effects be separated from style, length, familiarity, and response-bias artifacts?
FID-006
Faith-Facing Retrieval Grounding and Citation Reliability
How reliably do faith-facing AI systems retrieve, cite, and represent religious sources when users ask theological, historical, pastoral, or institution-specific questions?
Open question
Open work
Primary need: formal specification, theology, pluralism, evaluator design
- Contribute tradition-specific source hierarchies and boundary cases.
- Translate expert review rubrics into formal constraints.
- Test constraint-checking against model and RAG outputs.
- Review whether specifications preserve disagreement rather than erase it.