Tolerant Exhaustification (Chierchia 2013) #
@cite{chierchia-2013}
A second Excluder: negate every alternative not entailed by the
prejacent, even if the result is contradictory.
An alternative a is non-weaker than φ iff φ ⊄ a (some φ-world
is outside a). Innocent exclusion further requires that the negation
of a be jointly compatible with the negations of other choices; the
tolerant operator drops that requirement and excludes every non-weaker
alternative unconditionally. The result tolerant.exh may therefore be
empty (a contradiction).
Tolerance is empirically motivated by EFCIs (@cite{alonso-ovalle-moghiseh-2025}):
applying tolerant to an unembedded existential free-choice item produces
the contradiction that drives modal insertion or partial exhaustification.
Relationship to innocent #
Tolerance always excludes at least as much as innocent exclusion (every
innocently-excludable alternative is non-weaker), so
tolerant.exh ⊆ innocent.exh. The converse can fail when symmetric
alternatives leave both choices outside IE while still being non-weaker.
Chierchia 2013's contradiction-tolerating excluder: pick every alternative not entailed by the prejacent.
Equations
- Exhaustification.tolerant = { excluded := fun (ALT : Finset (Finset W)) (φ : Finset W) => {a ∈ ALT | ¬φ ⊆ a}, excluded_subset := ⋯ }
Instances For
A world w survives tolerant.exh iff it satisfies the prejacent
and falsifies every non-entailed alternative.
Alternatives entailed by the prejacent are kept, never negated.