# Provenance-Sensitive Multiplicity
**Domain:** Explainable AI / Computational Linguistics / Evidence Architecture
**Doc Type:** Canonical Project-Derived Principle
**Maturity:** Developed
**Related:** [[wiki/Provenance|Provenance]], [[wiki/Multi-Index Language Representation|Multi-Index Language Representation]], [[wiki/Neuro-Symbolic AI|Neuro-Symbolic AI]], [[wiki/Construction Transparency|Construction Transparency]], [[wiki/Symbolic Language Engine|Symbolic Language Engine]]
---
## Definition
**Provenance-Sensitive Multiplicity is the principle that a system may offer several kinds of linguistic suggestion while preserving what kind of relation and evidence produced each one.** Multiplicity is valuable only when heterogeneous outputs are not flattened into one undifferentiated confidence score.
## Relation Types
A reconstructed [[wiki/VersePerfect|VersePerfect]] could explicitly identify:
- a phonological match derived from a pronunciation representation;
- a semantic or taxonomic relation from [[wiki/WordNet|WordNet]] or another ontology;
- an association observed in a corpus;
- a candidate satisfying a selected verse-form constraint;
- a vector-neighborhood result from an embedding model; or
- a generative suggestion produced by a language model.
These outputs may converge, disagree or carry different evidentiary weight. The interface should make those differences legible to the writer.
## Wider Corpus Context
The principle connects creative tooling to [[wiki/Provenance|Provenance]], [[wiki/Construction Transparency|Construction Transparency]] and the larger rights architecture. Systems acquire power when model inference, retrieved evidence and administrative classification become indistinguishable. Preserving relation type is therefore both an epistemic and constitutional safeguard.
## Key Insight
**Do not ask one confidence score to erase the difference between evidence, relation, constraint and conjecture.**
## See Also
[[wiki/Provenance|Provenance]], [[wiki/Multi-Index Language Representation|Multi-Index Language Representation]], [[wiki/Knowledge Representation|Knowledge Representation]], [[wiki/Neuro-Symbolic AI|Neuro-Symbolic AI]], [[wiki/Candidate Generation and Ranking|Candidate Generation and Ranking]], [[wiki/Construction Transparency|Construction Transparency]]
## Sources / Provenance
- Principle articulated in [[projects/Ten Years Building a Symbolic Language Engine|Ten Years Building a Symbolic Language Engine]].