Documentation

Linglib.Theories.Semantics.Events.SpatialTrace

Spatial Trace Function σ #

@cite{gawron-2009} @cite{krifka-1998} @cite{talmy-2000} @cite{zwarts-2005} @cite{zwarts-winter-2000}

The spatial dimension of event structure: σ maps events to their spatial trajectories (paths). This parallels τ (temporal trace, EventCEM.τ_hom) and θ (thematic role, RoleHom) as the third Krifka/Zwarts dimension.

Three-Dimension Picture #

Temporal: Events →τ Intervals →dur ℚ (temporal dimension)
Spatial: Events →σ Paths →dist ℚ (spatial dimension)
Object: Events →θ Entities →μ ℚ (object dimension)

All three use the same QUA/CUM pullback mechanism via MereoDim.

Key Results #

Sections #

  1. SpatialTrace Class
  2. IsSumHom Instance for σ
  3. MereoDim for σ
  4. Telicity Transfer Through σ
  5. PathShape → Telicity Bridge
  6. Motion Verb Path Annotations
class Semantics.Events.SpatialTrace (Loc : Type u_1) (Time : Type u_2) [LinearOrder Time] [cem : Mereology.EventCEM Time] [SemilatticeSup (Core.Path.Path Loc)] :
Type (max u_1 u_2)

Spatial trace: maps events to their spatial trajectories. @cite{zwarts-2005}, @cite{gawron-2009}: σ(e) is the path traversed in event e. Parallels τ (temporal trace) from EventCEM.

σ is required to be a sum homomorphism: σ(e₁ ⊕ e₂) = σ(e₁) ⊕ σ(e₂). This ensures CUM pulls back through σ (atelic paths → atelic VPs). For QUA pullback, injectivity must be assumed separately (via σ_mereoDim), just as for τ.

  • σ : Ev TimeCore.Path.Path Loc

    Extract the spatial path of an event.

  • σ_map_sup (e₁ e₂ : Ev Time) : σ (SemilatticeSup.sup e₁ e₂) = σ e₁σ e₂

    σ preserves sums: σ(e₁ ⊕ e₂) = σ(e₁) ⊕ σ(e₂).

Instances

    σ as an IsSumHom instance, derived from SpatialTrace.σ_map_sup. Enables cum_pullback to work automatically for σ. Parallels instIsSumHomRuntime for τ.

    σ with injectivity is a MereoDim: the spatial dimension is a mereological morphism, enabling QUA pullback through spatial paths. Parallels the pattern for τ (injective τ → MereoDim).

    Equations
    • =
    Instances For

      Bounded path predicate → telic VP. "Walk to the store" is telic because "to the store" is QUA (no proper subpath of a to-the-store path is also to-the-store) and QUA pulls back through σ. @cite{zwarts-2005}: bounded PPs yield telic VPs.

      theorem Semantics.Events.SpatialTrace.unbounded_path_atelic {Loc : Type u_1} {Time : Type u_2} [LinearOrder Time] [cem : Mereology.EventCEM Time] [SemilatticeSup (Core.Path.Path Loc)] [st : SpatialTrace Loc Time] {P : Core.Path.Path LocProp} (hP : Mereology.CUM P) :

      Unbounded path predicate → atelic VP. "Walk towards the store" is atelic because "towards the store" is CUM and CUM pulls back through the σ sum homomorphism. @cite{zwarts-2005}: unbounded PPs yield atelic VPs.

      Map PathShape to Telicity: bounded/source → telic, unbounded → atelic. @cite{zwarts-2005}: the boundedness of a directional PP determines whether the VP it creates is telic or atelic.

      This is the spatial analog of the QUA/CUM ↔ telic/atelic correspondence from vendlerClass_telic_implies_qua_intent / vendlerClass_atelic_implies_cum_intent in Events/Mereology.lean.

      Equations
      Instances For

        PathShape telicity agrees with PathShape boundedness licensing: telic ↔ licensed (closed scale), atelic ↔ blocked (open scale). This connects the spatial classification to the scale-theoretic one.

        Whether a verb class inherently specifies a path shape. Inherently directed motion verbs (Levin 51.1: arrive, come, go) lexicalize a bounded path. Manner-of-motion verbs (51.3: run, walk) are path-neutral — the path comes from a PP complement. @cite{talmy-2000}: verb-framed vs. satellite-framed distinction.

        Equations
        Instances For