knowledge-base

knowledge-base IS a specific kind of tool — an artifact that stores and organizes information to extend human memory and reasoning capacity. Law 9a: a competent speaker would call a knowledge base 'a tool' in the broad sense of an artifact extending human capacity. Direction: specific (knowledge-base) → general (tool).

encyclopedia IS a specific kind of knowledge-base: a structured reference work organizing information across domains for retrieval and reference. Direction: specific (encyclopedia) → general (knowledge-base). Law 9 satisfied — a competent speaker calls an encyclopedia a knowledge base.

wiki IS a specific kind of knowledge-base: a collaborative hypertext system where multiple authors create, edit, and hyperlink content with version tracking. Direction: specific (wiki) → general (knowledge-base). Law 9 satisfied.

script-knowledge IS a specific kind of knowledge-base per Law 9. Script-knowledge organizes knowledge about typical event sequences and scenarios — it is a structured form of knowledge representation. A competent speaker would call script-knowledge a kind of knowledge-base.

A safety-knowledge-base is a specific kind of knowledge-base dedicated to safety information, standards, procedures, and lessons learned. It is the general knowledge-base construct specialized for the safety domain.

An inference-engine needs a knowledge-base to operate: inference is reasoning from stored knowledge. Remove the knowledge-base and the engine has nothing to infer from — it stops functioning. This is constitutive, not incidental: an inference engine IS the processing layer over stored knowledge.

RAG retrieves from external knowledge bases to augment generation. Per Law 8 removal test: remove knowledge bases, RAG stops operating. Genuine dependency.