Contingency Logic

Origin. Montgomery and Routley (1966). Contingency as primitive. Δφ: φ is contingent. Non-interdefinability with necessity. Foundation for modal analysis.

Models. Kripke frames. Δφ = ◇φ ∧ ◇¬φ. Alternative: Δ primitive, □ derived. Different expressiveness.

Formalism.

Contingency operator: Δφ: φ is contingent. True at some worlds, false at others.

Standard definition: Δφ ≡ ◇φ ∧ ◇¬φ Contingent = possibly true and possibly false.

Non-contingency: ∇φ ≡ ¬Δφ ≡ □φ ∨ □¬φ Necessarily true or necessarily false.

Δ as primitive: Can we define □ from Δ? No! Δ has less discriminating power. Same Δ-theory, different □-theory possible.

Frame definability: ∇φ → ∇(∇φ ∨ ψ): valid on transitive frames (Kuhn's K4 axiom). Transitivity itself is not ∇-definable (Zolin). Different axioms than standard modal.

Axioms (over K): Δφ → Δ¬φ (symmetric) ¬Δ⊤, ¬Δ⊥ Δ(φ ∨ ψ) → Δφ ∨ Δψ ∨ Δ(φ ∧ ψ)

Equivalences: □φ ≡ φ ∧ ∇φ (with reflexivity) ◇φ ≡ ¬□¬φ (standard)

Higher contingency: ΔΔφ: contingently contingent. Iterate for complex properties.

Model classes: Different from standard: {reflexive} not definable by Δ alone. Restricted expressiveness.

Symbols.

Metatheory. Montgomery and Routley (1966) axiomatized the non-contingency fragments of T, S4 and S5; Humberstone (1995) and Zolin (1999) settled the harder minimal case, giving a complete axiomatization of the ∇-logic of all Kripke frames and establishing the finite model property, hence decidability. The expressive gap is exact rather than approximate. Over reflexive frames □ is recovered as □φ ≡ φ ∧ ∇φ, so the two languages coincide there; over arbitrary frames □ is not ∇-definable, because the one-point reflexive frame and the one-point irreflexive frame validate exactly the same ∇-formulas — every ∇ψ is true at the single world in both — while they disagree on □p → p. Zolin gave a general criterion for which modally definable frame classes remain definable in the non-contingency language. ∇ is congruential but not monotone: ∇⊥ is valid, since □⊤ is, and ⊥ → p is valid, yet ∇p is not, so ∇ is a non-normal operator belonging to the classical rather than the normal hierarchy.

Applies to. Philosophy of modality. Free will. Determinism. Modal analysis.

Limitations. The language records that a matter is settled but not which way: ∇φ and ∇¬φ are equivalent, so necessary truth and necessary falsity are indistinguishable, and any application to determinism or to knowledge of the future has to reintroduce □ to recover the direction. Reflexivity and seriality are not ∇-definable, so the fragment cannot state the axioms that make the alethic reading intelligible in the first place. Because ∇ is not monotone, the inference from the validity of φ → ψ to ∇φ → ∇ψ fails, which blocks normal-form arguments, filtration in its usual form and Sahlqvist-style correspondence; each frame class needs its own completeness proof rather than inheriting one. Axiomatizations do not compose either — the ∇-fragment of a normal modal logic is generally not axiomatized by the ∇-translations of that logic's axioms, and no transfer theorem covers the general case.

© 2026 Lingenic LLC