Properness of `S4.2`.
ProvedMultiverseModalForcing.B_failsaether-catalogapplications
Properness of S4.2. The S5 axiom B : p → □◇p fails in the
(reflexive, transitive, directed) sink forcing frame: the atom is true at
wT, yet from the accessible sink world wF no forcing extension recovers
its truth. Hence forcing does not validate S5.
theorem MultiverseModalForcing.B_fails:
¬ MValid sinkFrame.R sinkFrame.M
(.imp (.atom true) (.box (MSentence.dia (.atom true)))) := by sorry
Formalization Note Transplanted verbatim from the Aether Catalog source Applications/MultiverseModalForcing.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.
Preamble
-- Thm stub generated from Applications/ProofTheoryAndLogic/MultiverseModalForcing.lean import Mathlib import Definitions.Def_Applications_ProofTheoryAndLogic_MultiverseModalForcing /- # The Modal Logic of Forcing — a Kripke-semantic deepening of Multiverse Set Theory This file equips the combinatorial core of the set-theoretic *multiverse* with a **modal structure of forcing**, following the Hamkins–Löwe programme: * **possibility** `◇p` — `p` holds in *some* forcing (generic) extension; * **necessity** `□p` — `p` holds in *every* forcing extension. Forcing extensions form a Kripke *frame*: an accessibility relation `R` on worlds (models of set theory). We give the standard Kripke semantics for a language of modal set-theoretic sentences and prove that the forcing frame is **sound for the modal system `S4.2`**, the theorem at the heart of the modal logic of forcing: * `sound_K` — the distribution axiom `□(p → q) → (□p → □q)` (any frame); * `sound_Nec` — the necessitation rule (any frame); * `sound_T` — `□p → p` from **reflexivity** of forcing (you force over yourself); * `sound_Four` — `□p → □□p` from **transitivity** (iterated forcing); * `sound_Two` — `◇□p → □◇p` from **directedness** (any two extensions have a common further extension — the amalgamation/product-forcing property). Together `K, T, 4, .2` axiomatise `S4.2`. We also prove the **duality** `◇p ↔ ¬□¬p`, monotonicity and distribution laws, package the result as `forcing_sound_S42`, instantiate the abstract frame with a concrete *flip-reachability* forcing frame over the multiverse, and finally show the logic is **properly weaker than `S5`**: the characteristic axiom `B : p → □◇p` **fails** in a directed reflexive-transitive forcing frame (`B_fails`). Thus forcing is genuinely `S4.2` and not `S5` — you cannot, in general, force your way back. Everything is proved from first principles over `Mathlib`, self-contained. -/ open MultiverseModalForcing open Classical /-! ## Worlds and modal sentences -/ open MSentence /-! ## Kripke semantics for forcing `R w v` reads "`v` is a forcing extension of `w`". `M` is the multiverse of admissible worlds. `meval R M w p` is the truth of `p` at world `w`. -/ /-! ## Duality between necessity and possibility -/ /-! ## Soundness of the modal system S4.2 for forcing frames We isolate the three frame conditions that forcing satisfies and show each validates the corresponding modal axiom. -/ /-! ## Derived modal principles -/ /-! ## Packaging the S4.2 soundness theorem -/ /-! ## A concrete forcing frame: flip-reachability We model a generic extension deciding an atom the other way by *flipping* its truth value. Two worlds are mutually accessible when they differ on only finitely many atoms — the reachable class under finitely many forcing steps. -/ /-! ## Properness: forcing is `S4.2`, not `S5` The characteristic `S5` axiom `B : p → □◇p` fails in a directed reflexive transitive forcing frame. Hence the modal logic of forcing is *properly* weaker than `S5`: from a generic extension you cannot in general force *back* to the ground model. -/
Formal statement
theorem MultiverseModalForcing.B_fails:
¬ MValid sinkFrame.R sinkFrame.M
(.imp (.atom true) (.box (MSentence.dia (.atom true)))) := by sorrySource