Does fuzzy logic solve the sorites paradox?

It offers a treatment: as you remove grains from a heap, the sentence 'this is a heap' gradually drops in degree of truth rather than flipping sharply from true to false. Critics object that this just relocates the problem, since fuzzy logic still requires precise numerical degrees and faces higher-order vagueness about where those degrees lie.

Many-Valued and Fuzzy Logics

Many-valued and fuzzy logics replace the two classical truth values with three, finitely many, or a continuum of degrees, chiefly to model vagueness and borderline cases.

Definition

A many-valued logic admits more than two truth values; fuzzy logic in particular assigns sentences a degree of truth in the real interval from 0 to 1, with the connectives computed by functions over those degrees.

Scope

This topic covers logics that abandon bivalence in favour of additional or continuum-many truth values. It treats Lukasiewicz and Kleene three-valued systems, Zadeh's fuzzy sets and degree-theoretic logic, the application of these tools to the sorites paradox and vagueness, and rival treatments of vagueness — supervaluationism (truth-value gaps) and epistemicism (sharp but unknown boundaries) — that bear on whether degrees of truth are the right response.

Core questions

Should vagueness be modelled by extra truth values, truth-value gaps, or neither?
How are the classical connectives generalized to many or continuum-many values?
Does fuzzy logic resolve the sorites paradox or merely relocate it as higher-order vagueness?
Is there a fact of the matter about borderline cases (epistemicism) or not?

Key concepts

bivalence and its rejection
three-valued logics
degrees of truth
fuzzy sets
sorites paradox
higher-order vagueness

Key theories

Fuzzy (degree-theoretic) logic: Building on Zadeh's fuzzy sets, vague predicates are assigned degrees of truth in [0,1], with conjunction, disjunction, and negation given by min, max, and complementation, so borderline cases take intermediate values.
Supervaluationism: Fine treats a vague sentence as super-true iff it comes out true on every admissible way of making the language precise, preserving classical logic while allowing truth-value gaps for borderline cases without adopting degrees of truth.

History

Lukasiewicz introduced three-valued logic in the 1920s to handle future contingents, and Kleene gave a three-valued logic for partial functions. Zadeh's 1965 fuzzy sets generalized this to a continuum of degrees, which was applied to vagueness; Fine's 1975 supervaluationism and Williamson's 1994 epistemicism offered influential alternatives.

Debates

How to model vagueness: Whether vagueness calls for degrees of truth (fuzzy logic), truth-value gaps with classical logic preserved (supervaluationism), or sharp-but-unknowable boundaries with bivalence retained (epistemicism), and which best handles the sorites and higher-order vagueness.

Key figures

Jan Lukasiewicz
Stephen Kleene
Lotfi Zadeh
Kit Fine
Timothy Williamson

Seminal works

zadeh1965
fine1975
williamson1994

Frequently asked questions

Does fuzzy logic solve the sorites paradox?: It offers a treatment: as you remove grains from a heap, the sentence 'this is a heap' gradually drops in degree of truth rather than flipping sharply from true to false. Critics object that this just relocates the problem, since fuzzy logic still requires precise numerical degrees and faces higher-order vagueness about where those degrees lie.