Internal consistency `¬□⊥` at a world is exactly seriality there, and it does
ProvedTangledSoundness.sat_con_iffaether-cataloglogic
Internal consistency ¬□⊥ at a world is exactly seriality there, and it does
not depend on the valuation.
theorem TangledSoundness.sat_con_iff(F : KFrame) (V : α → F.W → Prop) (w : F.W) :
sat F V w (MFormula.con (α := α)) ↔ ∃ v, F.R w v := by sorry
Formalization Note Transplanted verbatim from the Aether Catalog source Logic/ProvabilityLogic/IteratedReflection.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.
Preamble
-- Thm stub generated from Logic/ProvabilityLogic/IteratedReflection.lean
import Mathlib
import Definitions.Def_Logic_ProvabilityLogic_GLPFrames
import Definitions.Def_Logic_ProvabilityLogic_IteratedReflection
import Definitions.Def_Logic_ProvabilityLogic_SoundnessTopology
import Definitions.Def_Logic_ProvabilityLogic_TangledSoundness
/-
# Cycle 3: The Spectrum of Tangles — Iterated Reflection and Cycle Length
Cycles 1–2 showed that internalising the soundness schema `□φ → φ` at a world is
*equivalent* to a self-loop, and traced the consequences (no rank, no Löb fixed point,
no interior semantics, one loop per stratification step). This cycle calibrates the
phenomenon: how *weak* can an internal soundness principle be before the tangle
disappears?
Two sharp answers:
* **From below the schema cannot be weakened.** `atomicSound_iff_uniformlySound`:
reflection for *propositional variables only* already forces the self-loop, hence
the full reflection schema for all formulas. There is no non-trivial fragment of
soundness that a well-founded system can afford.
* **From above the tangle can be stretched, not removed.** `iterSound_iff_cycle`:
the `n`-fold reflection principle `□ⁿφ → φ` holds at `w` (uniformly) **iff** `w`
lies on a cycle of length `n`. The cycle frames `ZMod n` realise every point of
this spectrum (`cycleFrame_iterSound_self`, `cycleFrame_not_iterSound_lt`), so
internal soundness comes in a strictly increasing hierarchy of tangle lengths —
but every one of them tangles the transitive closure
(`iterSound_transGen_isTangled`) and none is available on a GL frame
(`glFrame_no_iterSound`).
* **Where the boundary really is.** Consistency (`¬□⊥`) is *not* on this spectrum:
`twoChain` is converse well-founded, loop-free, and internally consistent
(`twoChain_consistent_true`). Consistency costs nothing; reflection costs a loop.
## Relationship to catalog
Extends `Logic.ProvabilityLogic.TangledSoundness` (Cycle 1) and
`Logic.ProvabilityLogic.SoundnessTopology` (Cycle 2); uses `GLFrame`, `MFormula` from
`Logic.ProvabilityLogic.GLPFrames` and `IsTangled` from `Logic.TangledHierarchies`.
-/
open TangledSoundness
open GLPLogic
universe u
variable {α : Type*}
/-! ## Part A — Atomic reflection already forces the loop -/
/-! ## Part B — `n`-step accessibility and `n`-fold reflection -/
/-! ## Part C — Realising the spectrum: cycle frames -/
/-! ## Part D — The boundary: consistency is free -/Formal statement
theorem TangledSoundness.sat_con_iff(F : KFrame) (V : α → F.W → Prop) (w : F.W) :
sat F V w (MFormula.con (α := α)) ↔ ∃ v, F.R w v := by sorrySource