Documentation

Linglib.Semantics.Quantification.Properties

Properties of generalized quantifiers #

This file proves the general theorems about the properties of generalized quantifiers defined in Quantification/Defs.lean. The negations and the dual reverse monotonicity and preserve quantity invariance; conservativity and monotonicity are closed under the Boolean operations and under adjectival restriction, as Keenan and Stavi show; and conservativity together with symmetry or with the left monotonicities yields the characterizations of van Benthem and of Peters and Westerståhl. Barwise and Cooper's theorems on type ⟨1⟩ quantifiers close the file.

References #

Duality Theorems #

Outer negation turns a scope-upward monotone quantifier into a scope-downward one.

Outer negation turns a scope-downward monotone quantifier into a scope-upward one.

Inner negation turns a scope-upward monotone quantifier into a scope-downward one.

Inner negation turns a scope-downward monotone quantifier into a scope-upward one.

Outer negation turns a restrictor-upward monotone quantifier into a restrictor-downward one.

Outer negation turns a restrictor-downward monotone quantifier into a restrictor-upward one.

theorem Quantifier.GQ.innerNeg_innerNeg {α : Type u_1} (q : GQ α) :

Inner negation is an involution.

theorem Quantifier.GQ.dual_dual {α : Type u_1} (q : GQ α) :
q.dual.dual = q

The dual is an involution.

QuantityInvariant closure #

Outer negation preserves quantity invariance.

Inner negation preserves quantity invariance.

Dual preserves QuantityInvariant.

theorem Quantifier.GQ.QuantityInvariant.inf {α : Type u_1} (f g : GQ α) (hf : f.QuantityInvariant) (hg : g.QuantityInvariant) :

Meet preserves QuantityInvariant.

theorem Quantifier.GQ.QuantityInvariant.sup {α : Type u_1} (f g : GQ α) (hf : f.QuantityInvariant) (hg : g.QuantityInvariant) :

Join preserves QuantityInvariant.

theorem Quantifier.GQ.intersection_conservative_symmetric {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hInt : q.IntersectionCondition) :
Std.Symm q

A conservative quantifier satisfying the intersection condition is symmetric (Barwise and Cooper's Theorem C5).

Scope-downward monotonicity is equivalent to scope-upward monotonicity of the inner negation (co-property characterization, P&W §3.2.4).

Conservativity, Symmetry, and Strength #

theorem Quantifier.GQ.Conservative.congr_scope {α : Type u_1} {q : GQ α} (h : q.Conservative) {X S S' : α → Prop} (hSS : ∀ (x : α), X x → (S x ↔ S' x)) :
q X S ↔ q X S'

A conservative quantifier reads its scope only on the restrictor.

theorem Quantifier.GQ.conserv_symm_iff_int {α : Type u_1} (q : GQ α) (hCons : q.Conservative) :
Std.Symm q ↔ q.IntersectionCondition

Under conservativity, symmetric ↔ intersective (P&W Ch.6 Fact 1). This is the single most important bridge theorem — it explains why weak determiners allow there-insertion.

theorem Quantifier.GQ.symm_not_positive_strong {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hSym : Std.Symm q) (hNontrivF : ∃ (R : α → Prop) (S : α → Prop), ¬q R S) :

Non-trivial symmetric quantifiers are not positive strong (P&W Ch.6 Fact 7).

theorem Quantifier.GQ.conservative_iff_livesOn {α : Type u_1} (q : GQ α) :
q.Conservative ↔ ∀ (A : α → Prop), NP.LivesOn (q A) A

Conservativity of a GQ is equivalent to its restricted quantifiers living on their restrictors (P&W §3.2.2).

theorem Quantifier.GQ.Conservative.iff_of_inf_eq {α : Type u_1} {q : GQ α} (hq : q.Conservative) {R S S' : α → Prop} (h : R ⊓ S = R ⊓ S') :
q R S ↔ q R S'

A conservative quantifier's value at a restrictor depends only on the scope's meet with it.

Basic Left Monotonicity and Smoothness ([PW06] §5.5-5.6) #

Persistence → ↑_SE Mon.

Persistence → ↑_SW Mon.

↑_SW Mon ∧ ↑_SE Mon → Persistence ([PW06] Prop 6).

Persistence ↔ ↑_SW Mon ∧ ↑_SE Mon ([PW06] Prop 6).

Anti-persistence → ↓_NW Mon.

Anti-persistence → ↓_NE Mon.

↓_NW Mon ∧ ↓_NE Mon → Anti-persistence.

Anti-persistence ↔ ↓_NW Mon ∧ ↓_NE Mon.

theorem Quantifier.GQ.UpSEMon.compl {α : Type u_1} (q : GQ α) (h : q.UpSEMon) :

Outer negation reverses ↑_SE to ↓_NW ([PW06] Prop 8a).

theorem Quantifier.GQ.DownNWMon.compl {α : Type u_1} (q : GQ α) (h : q.DownNWMon) :

Outer negation reverses ↓_NW to ↑_SE.

theorem Quantifier.GQ.UpSWMon.compl {α : Type u_1} (q : GQ α) (h : q.UpSWMon) :

Outer negation reverses ↑_SW to ↓_NE.

theorem Quantifier.GQ.DownNEMon.compl {α : Type u_1} (q : GQ α) (h : q.DownNEMon) :

Outer negation reverses ↓_NE to ↑_SW.

theorem Quantifier.GQ.DownNEMon.innerNeg {α : Type u_1} (q : GQ α) (h : q.DownNEMon) :

Inner negation switches ↓_NE ↔ ↓_NW ([PW06] Prop 8b).

theorem Quantifier.GQ.DownNWMon.innerNeg {α : Type u_1} (q : GQ α) (h : q.DownNWMon) :

Inner negation switches ↓_NW ↔ ↓_NE.

theorem Quantifier.GQ.UpSEMon.innerNeg {α : Type u_1} (q : GQ α) (h : q.UpSEMon) :

Inner negation switches ↑_SE ↔ ↑_SW.

theorem Quantifier.GQ.UpSWMon.innerNeg {α : Type u_1} (q : GQ α) (h : q.UpSWMon) :

Inner negation switches ↑_SW ↔ ↑_SE.

Smooth ↔ outer negation is co-smooth ([PW06] Prop 8a).

Smooth ↔ inner negation is co-smooth ([PW06] Prop 8b).

theorem Quantifier.GQ.scopeMonotone_of_smooth {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hSmooth : q.Smooth) :

CONSERV ∧ Smooth → Mon↑ ([PW06] Prop 9).

theorem Quantifier.GQ.upSW_and_downNE_of_symm {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hSym : Std.Symm q) :

CONSERV ∧ Std.Symm → ↑_SW Mon ∧ ↓_NE Mon ([PW06] Prop 7).

theorem Quantifier.GQ.symm_of_upSW_of_downNE {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hUpSW : q.UpSWMon) (hDownNE : q.DownNEMon) :
Std.Symm q

↑_SW Mon ∧ ↓_NE Mon → Std.Symm (under CONSERV).

theorem Quantifier.GQ.symm_iff_upSW_and_downNE {α : Type u_1} (q : GQ α) (hCons : q.Conservative) :
Std.Symm q ↔ q.UpSWMon ∧ q.DownNEMon

A conservative quantifier is symmetric iff it satisfies ↑SW Mon and ↓NE Mon ([PW06]'s Prop 7).

Boolean Closure ([KS86]) #

Conservativity is closed under complement.

theorem Quantifier.GQ.Conservative.inf {α : Type u_1} (f g : GQ α) (hf : f.Conservative) (hg : g.Conservative) :
(f ⊓ g).Conservative

Conservativity is closed under meet.

theorem Quantifier.GQ.Conservative.sup {α : Type u_1} (f g : GQ α) (hf : f.Conservative) (hg : g.Conservative) :
(f ⊔ g).Conservative

Conservativity is closed under join.

theorem Quantifier.GQ.ScopeMonotone.inf {α : Type u_1} (f g : GQ α) (hf : f.ScopeMonotone) (hg : g.ScopeMonotone) :
(f ⊓ g).ScopeMonotone

The meet of two scope-upward monotone quantifiers is scope-upward monotone.

theorem Quantifier.GQ.ScopeAntitone.inf {α : Type u_1} (f g : GQ α) (hf : f.ScopeAntitone) (hg : g.ScopeAntitone) :
(f ⊓ g).ScopeAntitone

The meet of two scope-downward monotone quantifiers is scope-downward monotone.

theorem Quantifier.GQ.ScopeMonotone.sup {α : Type u_1} (f g : GQ α) (hf : f.ScopeMonotone) (hg : g.ScopeMonotone) :
(f ⊔ g).ScopeMonotone

The join of two scope-upward monotone quantifiers is scope-upward monotone.

An existential quantifier is conservative, so E-Det lies inside DDet ([KS86]).

theorem Quantifier.GQ.Existential.adjRestrict {α : Type u_1} {q : GQ α} (h : q.Existential) (adj : α → Prop) :

The existential property is preserved under adjectival restriction, since an absolute adjective commutes with the meet of restrictor and scope ([KS86]'s PROP 16).

theorem Quantifier.GQ.Conservative.adjRestrict {α : Type u_1} (q : GQ α) (adj : α → Prop) (h : q.Conservative) :

Conservativity is preserved under adjectival restriction.

theorem Quantifier.GQ.ScopeMonotone.adjRestrict {α : Type u_1} (q : GQ α) (adj : α → Prop) (h : q.ScopeMonotone) :

Scope-upward monotonicity is preserved under adjectival restriction.

theorem Quantifier.GQ.ScopeAntitone.adjRestrict {α : Type u_1} (q : GQ α) (adj : α → Prop) (h : q.ScopeAntitone) :

Scope-downward monotonicity is preserved under adjectival restriction.

Type ⟨1⟩ Theorems (P&W Ch.2-3) #

theorem Quantifier.GQ.individual_upward_closed {α : Type u_1} (a : α) (P P' : α → Prop) (h : ∀ (x : α), P x → P' x) :

Montagovian individuals are upward closed (ultrafilter property).

theorem Quantifier.GQ.individual_meet_closed {α : Type u_1} (a : α) (P Q : α → Prop) :
NP.individual a P → NP.individual a Q → NP.individual a fun (x : α) => P x ∧ Q x

Montagovian individuals are closed under intersection.

[vB84] Characterization #

theorem Quantifier.GQ.vanBenthem_refl_antisym_is_inclusion {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hRefl : q.PositiveStrong) (hAnti : Std.Antisymm q) (A B : α → Prop) :
q A B ↔ ∀ (x : α), A x → B x

Under conservativity, inclusion is the only reflexive antisymmetric quantifier ([vB84]'s Theorem 3.1.1).

theorem Quantifier.GQ.zwarts_refl_trans_scopeUp {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hRefl : q.PositiveStrong) (hTrans : IsTrans (α → Prop) q) :

A conservative reflexive transitive quantifier is scope-upward monotone (Zwarts's theorem in [vB84]).

theorem Quantifier.GQ.zwarts_refl_trans_restrictorDown {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hRefl : q.PositiveStrong) (hTrans : IsTrans (α → Prop) q) :

A conservative reflexive transitive quantifier is restrictor-antitone (Zwarts's theorem in [vB84]).

theorem Quantifier.GQ.quasiReflexive_of_scopeUp {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hUp : q.ScopeMonotone) :

Under conservativity a scope-upward monotone quantifier is quasi-reflexive, the half of van Benthem's theorem on symmetric quantifiers that needs no symmetry.

theorem Quantifier.GQ.quasiUniversal_of_scopeDown {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hDown : q.ScopeAntitone) :

Under conservativity a scope-downward monotone quantifier is quasi-universal, the half of van Benthem's theorem on symmetric quantifiers that needs no symmetry.

Right-monotone quantifiers are right-continuous.

theorem Quantifier.GQ.irrefl_almostConn_scopeDown {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hIrrefl : q.NegativeStrong) (hAC : q.AlmostConnected) :

A conservative irreflexive almost-connected quantifier is scope-antitone ([vB84]'s Theorem 4.1.2).

theorem Quantifier.GQ.irrefl_almostConn_restrictorUp {α : Type u_1} (q : GQ α) (hCons : q.Conservative) (hIrrefl : q.NegativeStrong) (hAC : q.AlmostConnected) :

A conservative irreflexive almost-connected quantifier is restrictor-monotone ([vB84]'s Theorem 4.1.2).

Asymmetry and circularity #

[PW06] Ch 6.4

theorem Quantifier.GQ.asymmetric_negativeStrong {α : Type u_1} (q : GQ α) (hAsym : Std.Asymm q) :

Asymmetric quantifiers are negative-strong (irreflexive).

theorem Quantifier.GQ.asymmetric_antisymmetric {α : Type u_1} (q : GQ α) (hAsym : Std.Asymm q) :
Std.Antisymm q

Asymmetric implies antisymmetric (vacuously).

theorem Quantifier.GQ.circular_symmetric_quasiRefl {α : Type u_1} (q : GQ α) (hSym : Std.Symm q) (hCirc : q.Circular) :

Circular + symmetric → quasi-reflexive.

theorem Quantifier.GQ.circular_reflexive_symmetric {α : Type u_1} (q : GQ α) (hCirc : q.Circular) (hPS : q.PositiveStrong) :
Std.Symm q

Circularity + reflexivity → symmetry.

theorem Quantifier.GQ.isom_asymmetric_eq_diff {α : Type u_1} [Fintype α] [DecidableEq α] (q : GQ α) (hCons : q.Conservative) (hIsom : q.QuantityInvariant) (hAsym : Std.Asymm q) {A B : α → Prop} [DecidablePred A] [DecidablePred B] (hCard : Fintype.card { x : α // A x ∧ ¬B x } = Fintype.card { x : α // B x ∧ ¬A x }) :
¬q A B

Under conservativity, quantity invariance and asymmetry, Q(A, B) fails whenever A ∖ B and B ∖ A have the same size ([PW06]'s Prop 6.59, fixed-domain version).

"Aristotle reversed": square from inferential conditions #

[vB84] §3.3

theorem Quantifier.GQ.vanBenthem_symm_quasiRefl_is_overlap {α : Type u_1} [Fintype α] [DecidableEq α] (q : GQ α) (hCons : q.Conservative) (hSym : Std.Symm q) (hQR : q.QuasiReflexive) (hWitT : ∃ (x : α), q (fun (y : α) => y = x) fun (y : α) => y = x) (hWitF : ∃ (A : α → Prop), ¬q A A) (hIso : q.QuantityInvariant) (A B : α → Prop) :
q A B ↔ ∃ (x : α), A x ∧ B x

Under conservativity the only symmetric quasi-reflexive quantifier is overlap, that is some ([vB84]'s Corollary 3.3.2).

theorem Quantifier.GQ.vanBenthem_symm_quasiUniv_is_disjointness {α : Type u_1} [Fintype α] [DecidableEq α] (q : GQ α) (hCons : q.Conservative) (hSym : Std.Symm q) (hQU : q.QuasiUniversal) (hWitF : ∃ (x : α), ¬q (fun (y : α) => y = x) fun (y : α) => y = x) (hWitT : ∃ (A : α → Prop), q A A) (hIso : q.QuantityInvariant) (A B : α → Prop) :
q A B ↔ ∀ (x : α), ¬(A x ∧ B x)

Under conservativity the only symmetric quasi-universal quantifier is disjointness, that is no ([vB84]'s Corollary 3.3.3).