Necessary vs Contingent Truths

How the distinction works

Take any true sentence and ask: could it have been false? If the answer is “no, the alternative is incoherent or impossible,” the truth is necessary. If the answer is “yes, things could have gone differently,” the truth is contingent.

• “2 + 2 = 4.” Could it have been 5? No — there's no consistent way to make addition work otherwise. Necessary.
• “Paris is the capital of France.” Could it have been Lyon? Yes — the National Assembly could have voted differently in 1789. Contingent.
• “Bachelors are unmarried.” Could a bachelor be married? No — you'd no longer be a bachelor. Necessary (also analytic).
• “Water boils at 100°C at sea level.” Could the boiling point have been 90°C? Yes — different intermolecular forces. Contingent (though “water is H₂O” is necessary).

Logicians formalise the distinction with two operators: □p means “p is necessary,” and ♢p means “p is possible.” The two are duals: □p ≡ ¬♢¬p (necessarily p just means it's not possible that not-p). A truth is contingent if both p and ¬p are possible: ♢p ∧ ♢¬p.

The possible-worlds picture

Leibniz introduced the metaphor: God surveyed all possible worlds before actualising the best. A possible world is a complete, consistent way things could have been — a maximal consistent set of propositions. The actual world is one of these.

With possible worlds in hand, the modal definitions become geometric.

Modal status	Holds in...	Example
Necessary	Every possible world	2 + 2 = 4
Possible	At least one world	I become a baker
Contingent	Some worlds, not all	Paris is France's capital
Impossible	No world	A round square
Actual	This world (whatever its modal status)	You're reading this

How seriously should we take the worlds? Modal realists like David Lewis (On the Plurality of Worlds, 1986) say they exist as concrete entities, just causally isolated from ours. Ersatzers say they're abstract representations — sets of sentences, propositions, or properties. Fictionalists treat them as a useful pretence. The semantics of necessity is the same; the metaphysics differs.

Necessary vs contingent — at a glance

	Necessary	Contingent
Could be false?	No	Yes
Possible-worlds gloss	True in every world	True in some, false in others
Modal symbol	□p	♢p ∧ ♢¬p
Paradigm examples	Math, logic, identities	Empirical facts, history, choices
Knowledge route	Often a priori, sometimes a posteriori	Usually a posteriori
Coined / sharpened by	Leibniz; modal logic Kripke 1959	Same
Modal scepticism target?	Yes — van Inwagen, Hawthorne	Less commonly

Worked example: Hesperus and Phosphorus

Ancient astronomers tracked two bright bodies — Phosphorus rising before dawn, Hesperus shining after dusk. They thought these were two different stars. In the 6th century BC, Pythagoreans realised they were the same planet (Venus). The proposition “Hesperus is Phosphorus” turned out true.

Now ask: is it necessarily true? Pre-Kripke, philosophers tended to say no — we discovered it empirically, so surely it could have been false. Kripke replied: proper names are rigid designators, picking out the same object in every world where the object exists. “Hesperus” rigidly refers to Venus; so does “Phosphorus.” The identity Hesperus = Phosphorus is therefore identical-with-itself in every world. It's metaphysically necessary, even though only knowable a posteriori.

Same shape: “water is H₂O,” “heat is mean molecular kinetic energy,” “gold has atomic number 79.” Each is a necessary truth that took empirical work to find.

The Kripkean grid

Necessary/contingent is metaphysical (about reality); a priori/a posteriori is epistemic (about how we know). Pre-Kripke, philosophers tended to assume they line up. Kripke argued they don't.

	A priori	A posteriori
Necessary	2 + 2 = 4 — uncontroversial	Water is H₂O — Kripke's signature
Contingent	Standard metre = 1m — Kripke's signature	Snow is white — uncontroversial

Kinds of necessity

• Logical necessity. True under every interpretation of the non-logical vocabulary: p ∨ ¬p.
• Mathematical necessity. True given the axioms of mathematics.
• Conceptual / analytic necessity. True in virtue of meaning: “all bachelors are unmarried.”
• Metaphysical necessity (Kripke). True in every metaphysically possible world: “water is H₂O,” identity statements with rigid names.
• Nomological necessity. True given the laws of nature: nothing exceeds the speed of light. Weaker than metaphysical — laws could have been different.
• Practical / moral necessity. “You must keep your promises.” Different again.

The strength ordering is roughly: logical > mathematical > metaphysical > nomological > practical. Higher kinds entail lower ones (logical necessities are also metaphysical), not vice versa.

Objections and rivals

• Quine's modal scepticism. In “Reference and Modality” (1953), Quine argued necessity makes no sense de re — applied to objects rather than sentences. Mathematicians necessarily count, but Cyrano necessarily has a long nose only relative to a description. Kripke's later work persuaded most philosophers that de re necessity was coherent, but Quineans remain.
• Lewis's modal realism. Concrete possible worlds let us reduce modal talk to quantification, but at heavy ontological cost — uncountably many universes nobody can interact with. Most philosophers prefer ersatz worlds, even at the cost of an unanalysed primitive.
• Modal fictionalism (Rosen). Treat possible-worlds talk as a useful fiction. Avoids ontological excess, but struggles to explain why the fiction is so reliable.
• Two-dimensionalism (Chalmers, Jackson). Distinguishes “primary” intensions (a priori, conceptual) from “secondary” intensions (Kripkean, metaphysical). Aims to recover an a-priori-knowable necessity behind Kripke's necessary a posteriori.
• Modal scepticism (van Inwagen). We have reliable modal intuitions only for cases close to the actual — talking dogs, no; counterfactual histories of the past century, yes. Far-flung claims (panpsychism is metaphysically possible) outrun our evidence.

Why the distinction matters

• Free will. If determinism is true, every action is nomologically necessary. Whether that excludes “could have done otherwise” depends on which kind of necessity matters.
• Counterfactuals. “If the match had been struck, it would have lit” quantifies over close possible worlds where the match was struck.
• Essentialism. Some properties an object has necessarily (Aristotle is necessarily human), others contingently (Aristotle was contingently a teacher). Modern essentialism trades on Kripkean necessity.
• Philosophy of mind. The conceivability of zombies (creatures functionally identical to us but lacking consciousness) trades on metaphysical possibility — if zombies are possible, physicalism faces a problem.
• Theology. Anselm's ontological argument tries to derive God's actual existence from God's necessary existence. Modal arguments are still active in philosophy of religion.

Common confusions

• Necessity ≠ certainty. Mathematicians have been certain of false claims (Frege's Basic Law V). Necessity is metaphysical; certainty is psychological.
• Necessity ≠ a priori. Kripke's main lesson — “water is H₂O” is necessary but a posteriori; “the standard metre is one metre long” is contingent but a priori.
• Possible ≠ conceivable. Some things we can imagine are impossible (water that isn't H₂O); some things we can't easily picture are possible (eleven-dimensional spacetime).
• Determinism ≠ necessity. Determinism says given the laws and past, only one future. That future is nomologically necessary, not metaphysically necessary; the laws could have been different.
• “Necessary” in plain English. Often means “essential” or “mandatory.” Philosophical “necessary” means “true in every possible world” — a much stronger claim.