Question — partition predicate, Setoid bridge, QUD bridge #
@cite{groenendijk-stokhof-1984} @cite{roberts-2012} @cite{ciardelli-groenendijk-roelofsen-2018}
Bridge from Question W (downward-closed nonempty families of information
states) to partition-style inquiry in mathlib's Setoid.IsPartition
sense, plus a legacy bridge to the Bool-based QUD W.
Main definitions #
IsPartition P—Setoid.IsPartition (alt P). Mention-some, intermediate-exhaustive, and conditional-question alternative sets are not partitions and so fail this predicate.toSetoid h— the equivalence relation derived from a partition issue (two worlds equivalent iff they share an alternative cell).toQUD h— bridge to legacy Bool-basedQUD W.noncomputable, uses classical decidability of the underlying equivalence.
Main theorems #
info_eq_univ,pairwiseDisjoint,nonempty_of_mem_alt,sUnion_alt_eq_univ— partition issues are non-informative and their alternative sets are partitions in the standard sense.isPartition_polar,isPartition_ofList— the basic Hamblin constructions yield partition issues under nontriviality / disjointness / cover hypotheses.toSetoid_fromSetoid—Setoid → Question → Setoidis the identity (round-trip withfromSetoidfromCore/Mood/PartitionAsInquiry.lean).
IsPartition predicate #
An issue is a partition iff its alternative set is a partition
of W in mathlib's Setoid.IsPartition sense: no empty
alternative, and every world lies in a unique alternative. The
Groenendijk–Stokhof partition shape.
Equations
- P.IsPartition = Setoid.IsPartition P.alt
Instances For
Partition issues are non-informative: every world is covered by
some alternative, hence by some resolving proposition, so
info P = Set.univ.
Alternatives of a partition issue are pairwise disjoint. Direct
inheritance from mathlib's Setoid.IsPartition.pairwiseDisjoint.
Alternatives of a partition issue are nonempty. Mathlib's
Setoid.nonempty_of_mem_partition specialized to Question.
The union of a partition issue's alternatives is the whole universe.
The setoid derived from a partition issue: two worlds are
equivalent iff they lie in the same alternative cell. Built from
mathlib's Setoid.mkClasses on the alternative set.
Equations
- Core.Question.toSetoid h = Setoid.mkClasses P.alt ⋯
Instances For
The equivalence classes of toSetoid h are exactly the
alternatives of P. The defining roundtrip property of
Setoid.mkClasses.
Two worlds are related under toSetoid h iff they share an
alternative cell — the linguistic reading of the underlying
equivalence relation.
Bridge to legacy Bool-based QUD W: two worlds get the same
answer iff they share an alternative cell. noncomputable —
decidability of the underlying equivalence is supplied
classically.
Equations
- Core.Question.toQUD h = { toSetoid := Core.Question.toSetoid h, decR := fun (w v : W) => Classical.dec ((Core.Question.toSetoid h) w v) }
Instances For
toQUD's relation holds iff the two worlds share an alternative cell.
toQUD's sameAnswer is true iff the two worlds share a cell.
The cell of a world under toQUD h is in alt P.
A world lies in its own toQUD cell.
The toQUD cell of any world in an alternative p is exactly p.
Round-trip with fromSetoid #
The Setoid → Question embedding (fromSetoid, in
Core/Mood/PartitionAsInquiry.lean) and Question → Setoid projection
(toSetoid) compose to the identity on partition-shaped issues.
Under Nonempty W, fromSetoid r is a partition issue. The
nonemptiness assumption rules out the degenerate ∅-alternative
that would otherwise appear when W is empty.
Round-trip: the setoid derived from a partition issue built
from a setoid recovers the original setoid. The Setoid → Question → Setoid direction of the partition–setoid correspondence.
IsPartition for the basic Hamblin constructions #
Question.polar p is a partition issue when p is non-trivial: the
two alternatives p and pᶜ partition W by LEM.
Question.ofList L is a partition issue when L is a nonempty list of
pairwise-disjoint nonempty cells that exhaust W.