Only world-locked content can be jointly pointed at by several people

What it is

Two people looking at “that one” can only get finger or ray to land on the same place if the thing sits in a shared world frame, not in someone’s body or view. A body-locked strip travels with its owner; a bystander sees a plate at the owner’s hip that means nothing to them. Head-locked status exists only on the owner’s retina. What can be pointed at together, and named as “the third on the left”, is the world-locked layer alone.

Why it happens

Deictic language (this, that, the one on the left) depends on a shared spatial frame. World-lock puts the object in the room, so two gazes and two gestures intersect in one Euclidean geometry and reference can converge. Body-lock puts the object in one person’s egocentric coordinates; the second person has no readout of that frame, and the gesture points at empty space or at their own different carry-on UI. Head-lock is more extreme: the content occupies no place in the room. Collaboration failures are often filed as “communication problems” when the frames simply do not intersect. Pointing tools themselves — avatar, hand, ray — must also be visible to the other person and must live in the world; a ray drawn only on the emitter’s head-locked layer is invisible as pointing.

Studying it

A dyad: one person indicates a target, the other selects. Compare the target world-locked, body-locked to the indicator, and body-locked to the operator.

Independent variables: target’s reference frame, indication channel (hand / ray / speech), whether both occupy the same physical room. Dependent variables: time from indication to selection, wrong selections, count of verbal repairs (“not yours, the one on the wall”).

Hindmarsh-style analyses of collaborative pointing and deixis experiments in shared AR sit here. Remote joiners whose world origins are not aligned will make world-lock fail for a different reason — anchors, not lock choice.

Where it stops holding

Solo use has no addressee; this is not a constraint. A trained pair can replace gesture with language entirely (“third item in the menu”) and still collaborate on a body-locked list, but that is no longer spatial deixis. Mirroring “your strip is my strip” can fake sharing with body-lock, but one turn misaligns both sides and it only barely works face-to-face with little movement. Pointing at a real chair does not depend on a virtual lock; the virtual layer only needs not to hide the chair.

Applying it

• Content that two people must see, edit, and call “this” goes world-locked. Carry-on tools and private status must not impersonate the shared layer.
• Draw indicating hands, avatars, and rays in the world and sync them to the other person. Do not draw them only on the indicator’s head-locked layer.
• Pin an object from a private body-locked panel into the world before talking about it, rather than speaking across two carry-on menus into empty space.
• How to check: two people wear the devices; one may only point, not name. What the other can hit stably is the shared layer. What aligns only after speech is still in someone’s private frame.

Related

• Same group: N3.08.1 Different parts of one interface can use different locking modes · N3.08.2 A change of locking mode needs a visible transition · N3.08.3 Lazy follow lets body-lock keep up without shaking
• Nearby: N3.01 Locking Modes · N5.05 Shared Space · N5.10 Shared Space and Multi-user Collaboration
• Search terms: shared deixis · joint pointing · world-locked collaboration