Linguistic Synonymy
Source: Cruse, Lexical Semantics, 1986; Edmonds & Hirst, Computational Linguistics 28, 2002; Edelman & Gally, PNAS 98, 2001
Finding
Natural languages contain extensive synonymy — multiple words conveying the same or nearly the same meaning. English has “big,” “large,” “enormous,” “vast,” “immense,” “huge,” “gigantic,” each with slightly different connotations and registers but all pointing to great size. Cruse distinguished absolute synonymy (interchangeable in all contexts — extremely rare) from near-synonymy (overlapping but not identical — pervasive). This is not inefficiency but structural robustness: the redundancy provides resistance to ambiguity, registers for social context, and expressive range. The parallel to biological degeneracy is structural: 64 codons encode 20 amino acids; multiple antibodies recognize the same antigen; multiple neural configurations produce the same behavior; multiple words express the same meaning.
Pattern Mapping
Proportion — Synonymy proportions expression to context. “Deceased” in legal documents, “passed away” in condolence cards, “died” in news reports, “kicked the bucket” among friends — same meaning, register proportioned to situation.
Humility — No single word has total authority over a meaning. Each synonym captures a facet; the full meaning emerges from the set.
Alignment — When the right synonym is chosen, expression aligns with context. When the wrong synonym is chosen (“the deceased kicked the bucket”), the misalignment is immediately felt.
Connections
- Sapir-Whorf Hypothesis — if language shapes thought, then synonymy gives multiple cognitive entry points to the same concept
- Universal Grammar — deep structure may be shared; synonymy exists at the surface level of lexical choice
- DNA Error Correction — codon degeneracy protects against mutation; linguistic synonymy protects against ambiguity (→ Meta-Pattern 10 - Degeneracy)
- Convergent Evolution — multiple evolutionary paths to the same function; multiple words for the same meaning
- Distributed Systems and Consensus — both achieve robustness through redundancy: no single node or word is a single point of failure
Status
Cruse (1986) standard reference. Edmonds & Hirst (2002) formalized computationally. Biological degeneracy parallel from Edelman & Gally (2001). Connection to degeneracy is this project’s interpretation.
The mapping to the five properties is this project’s structural interpretation.