A Lawvere-Tierney topology on a topos is an endomorphism j: Ω → Ω of the subobject classifier satisfying three axioms:
- j(true) = true (the maximal sieve is closed)
- j ∘ j = j (idempotence — applying j twice is the same as applying it once)
- j(p ∧ q) = j(p) ∧ j(q) (j preserves meets — closure of a conjunction is the conjunction of closures)
These three conditions make j a closure operator on the internal logic of the topos. The “topology” is not a collection of open sets but a way of deciding which propositions count as “locally true” — which truth values survive closure.
A Lawvere-Tierney topology determines which presheaves are sheaves: a presheaf F is a j-sheaf if it satisfies the sheaf condition with respect to the sieves that j classifies as dense. The category of j-sheaves forms a subtopos — a full subcategory that is itself a topos, with its own internal logic modified by j.
The connection to classical topology: on the topos of sheaves over a topological space X, each Lawvere-Tierney topology corresponds to a classical topology on X. But the concept generalizes beyond spatial topoi to any elementary topos, where it provides a purely algebraic way to define “local truth.”
In the semiotic universe, the modal closure operator j on the Heyting algebra H is a Lawvere-Tierney topology. It determines which semiotic structures count as stable (j-closed) versus transient (not j-closed). The j-sheaves are the semiotically coherent objects: those whose local interpretations glue together consistently under the closure discipline that j imposes.