Schlenker, Lamberton & Lamberton (2026) #
@cite{schlenker-lamberton-lamberton-2026}
Traveling Shots in Language: Towards an Analysis of Dynamic Viewpoints in ASL. To appear in Linguistic Inquiry.
Core Contribution #
Extends Iconological Semantics (@cite{schlenker-lamberton-2024}) by showing that viewpoint variables may denote dynamic (traveling) viewpoints — functions from time-world pairs to static viewpoints — not just static observation points.
In ASL, a classifier denoting a static object (e.g., TREE-cl) can move in signing space to represent the object's apparent motion from a moving character's perspective. This is the linguistic analogue of a traveling camera shot in film.
Two Analyses #
Analysis I (§7): All viewpoint variables can be dynamic. Relative motion arises whenever a viewpoint variable takes a non-constant value.
Analysis II (§8): Only viewpoints introduced by Role Shift (as overt context shift) can be dynamic. Standard viewpoint variables denote static viewpoints. Under Role Shift, a distinguished variable π* reads the character's dynamic viewpoint from the shifted context.
The paper leaves both options open, pending a consensus on the definition of Role Shift.
Data #
Paradigms were elicited from two Deaf native ASL signers. Acceptability is on a 7-point scale (7 = best, 1 = worst). Classifier direction (left vs. right) systematically determines the character's inferred path.
The static viewpoint analysis is a special case of the dynamic one. When a viewpoint is constant, dynamic projection reduces to time-invariant projection. The dynamic framework is a conservative extension.
The traveling shot condition: there exist two times at which the object's projection differs, producing apparent motion. This is only possible when the viewpoint is dynamic.
Equations
- One or more equations did not get rendered due to their size.
Instances For
A static viewpoint never produces a traveling shot: projection is constant across time, so no apparent motion is possible.
A viewpoint assignment: maps viewpoint variable indices to dynamic viewpoints. The iconic analogue of an assignment function for pronouns.
Equations
Instances For
Resolve a viewpoint variable against a viewpoint assignment and an agent. Free variables read from the assignment; the context-bound variable π* reads the dynamic viewpoint associated with the agent.
Under Role Shift (which changes the context's agent to the character), π* yields the character's dynamic viewpoint.
Equations
- Phenomena.Iconicity.Studies.SchlenkerEtAl2026.resolveViewpoint (Semantics.Iconic.ViewpointVar.free i) assign agentVP agent = assign i
- Phenomena.Iconicity.Studies.SchlenkerEtAl2026.resolveViewpoint Semantics.Iconic.ViewpointVar.contextBound assign agentVP agent = agentVP agent
Instances For
Role Shift as context shift. Structurally identical to
attitudeShift but labeled .roleShift — it changes the agent to
the character (and the world to a Role-Shift-accessible world).
The viewpoint consequence is indirect: since π* reads the agent's
viewpoint via resolveViewpoint, changing the agent changes what
π* denotes.
Equations
- Phenomena.Iconicity.Studies.SchlenkerEtAl2026.roleShiftCtx character rsWorld = { apply := (Core.Context.attitudeShift character rsWorld).apply, label := Core.Context.ShiftLabel.roleShift }
Instances For
Role Shift is a monster (non-identity context shift), connecting
to the Kaplan/Schlenker monster debate in Monsters.lean.
Under Role Shift, π* resolves to the character's viewpoint.
Free viewpoint variables are unaffected by who the agent is — they read from the assignment function regardless.
The restrictive theory: free viewpoint variables always denote static viewpoints. Only context-bound π* (introduced by Role Shift) can be dynamic.
Equations
- Phenomena.Iconicity.Studies.SchlenkerEtAl2026.RestrictiveTheory assign = ∀ (i : ℕ), (assign i).isStatic
Instances For
Under the restrictive theory, traveling shots are impossible for classifiers with free viewpoint variables.
The complete argumentation chain:
- Role Shift changes the context's agent from signer to character
- π* reads the agent's viewpoint, so it now reads the character's
- The character's viewpoint is dynamic (they are moving)
- Dynamic viewpoint + static object → traveling shot is possible
- Static viewpoint would make traveling shot impossible
This theorem captures steps 2-5: given that the character's viewpoint is genuinely dynamic (non-constant), the traveling shot condition is satisfiable — there exist times where projection differs.
An acceptability judgment on a 7-point scale.
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
A paradigm condition: classifier direction + Role Shift status.
- classifierDirection : ClassifierDirection
- roleShiftStatus : RoleShiftStatus
- judgment : Judgment
Instances For
Equations
- One or more equations did not get rendered due to their size.
Instances For
Paradigm (7): POLE-cl moving past signer. Ann driving drunk, nearly hits pole. Direction determines which side she nearly hits. No strict Role Shift (no body rotation).
Equations
- One or more equations did not get rendered due to their size.
Instances For
Paradigm (13): TREE-cl moving past signer. Ann runs past a tree. Direction determines which side the tree passes on. No strict Role Shift.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Paradigm (19): Role-shifted version of (7). Strict Role Shift (body rotation). Same interpretive effect.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Paradigm (7) instantiates the traveling shot: pole is static, classifier moves, motion type is relative.
Equations
- One or more equations did not get rendered due to their size.
Instances For
Paradigm (13) instantiates the traveling shot: tree is static, classifier moves, motion type is relative.
Equations
- One or more equations did not get rendered due to their size.
Instances For
The traveling shot arises both with and without strict Role Shift.
The classifier used in paradigm (7) is POLE-cl from the ASL fragment.
The classifier used in paradigm (13) is TREE-cl from the ASL fragment.