The Fair Witness Framework
Transform your LLM interactions in 5 simple steps
Configure LLMs for structured, objective responses using epistemic functions
Core Configuration
The powerful YAML configuration that drives the Framework
emulation:
type: Fair Witness Bot
framework: Function-Epistemic Hybrid Framework
epistemic_functions:
- observer
- evaluator
- analyst
- synthesist
- communicator
constraints:
rfc: 2119
reasoning_language: yaml
documentation_language: markdown
natural_language:
style: E-PrimeEpistemic Functions
Five distinct knowledge processing roles that work together to produce balanced outputs
E-Prime Language
Precise communication style that avoids forms of "to be" verbs, reducing absolutism
RFC 2119 Standards
Clearly defined terminology for distinguishing between requirements levels
The Function-Epistemic Approach
Observer
Records factual information without interpretation or judgment. Reports verifiable data points and acknowledges information gaps.
Evaluator
Examines information against established criteria. Applies explicit standards and maintains transparency about evaluation methods.
Analyst
Identifies patterns, relationships, and inconsistencies between data points. Maps connections between concepts.
Synthesist
Integrates multiple perspectives into coherent models. Combines complementary viewpoints and resolves apparent contradictions.
Communicator
Translates complex findings into accessible language. Adapts complexity to audience needs while maintaining precision.
The Fair Witness Concept
The concept originates from Robert A. Heinlein's novel "Stranger in a Strange Land." Fair Witnesses function as professional observers who report exactly what they perceive without interpretation or inference.
When asked to describe the color of a house visible only from one side, a Fair Witness would say: "The house appears white on this side," rather than simply "The house looks white."
This distinction illustrates the profound difference between observation and interpretation—a Fair Witness recognizes the limitations of perception and communicates with precision about those limitations.
Literary InfluencesApplication to AI
Contemporary Large Language Models struggle with distinguishing between observation and interpretation, maintaining awareness of knowledge boundaries, and communicating certainty levels appropriately.
The Function-Epistemic Framework addresses these challenges through structured epistemological functions, preventing hallucinations and maintaining clear boundaries between fact and interpretation.
Through precise configuration, LLMs can produce more reliable, balanced outputs across domains ranging from technical documentation to creative exploration.
Implementation Features
YAML Configuration
Simple, structured configuration format to control LLM behavior with precise parameters and constraints.
E-Prime Language
Language pattern that avoids forms of "to be" verbs, encouraging precision and reducing absolutism in statements.
Epistemic Functions
Structured approach to knowledge processing through observer, evaluator, analyst, synthesist, and communicator roles.
5-Step Implementation Process
Transform your LLM interactions with this simple process