Fortolbarhed

Fortolbarhed is a concept used in Danish-language scholarly discourse to describe the degree to which the meaning, structure, and implications of information, data, or computational models can be understood and explained by humans and, where appropriate, by automated systems. The term derives from fortolkning, meaning interpretation, and barhed, a suffix indicating capability or attainability.

In practice, fortolbarhed encompasses clarity of representation, transparency of processes, provenance, and the availability of accessible explanations for outputs. It is applied across fields such as data science, artificial intelligence, linguistics, and policy analysis, where stakeholders must assess how conclusions are reached and what assumptions underlie them.

Assessment methods combine qualitative judgments with quantitative metrics. Qualitative approaches may include expert reviews and scenario-based testing, while quantitative measures attempt to gauge cognitive load, time to understand, and the clarity of model explanations or data schemas. A system with high fortolbarhed typically offers interpretable representations, well-documented data lineage, and user-facing explanations of decisions.

Relation to other terms: fortolbarhed is closely related to interpretability and explainability but emphasizes overall understandability and communicability, including nontechnical audiences. High fortolbarhed can involve trade-offs with performance or complexity.

Limitations: Interpretability is context-dependent and subjective. Cultural and linguistic factors influence what is considered understandable, and achieving universal fortolbarhed may be impractical.

See also: interpretability, explainability, transparency, data documentation.