A physical object is representation and control at once

What it is

The fader labeled “track 3” both shows where the level sits and is the handle that changes it—one object doing representation and control. Tangible interaction makes digital objects into things you can grasp, array, and stack: pose, heading, and presence on the table are read and rewritten by the same stuff. It is not a touchscreen in a shell, and it is not a surface that changes shape by itself. The core is grasp affordance: the object’s geometry is the data’s avatar. Desync with the digital side, and the difficulty of undo and copy, are the next two problems.

Why it happens

Grasp locks the object into the hand’s frame: once pinched, rotate, translate, and stack need no cursor aim. Geometry speaks—taller blocks read as “more,” a line as “sequence,” a token on a map as “here”—so reading and writing share one spatial channel, skipping “see an icon, then hunt a menu.” The constraint is the same binding: objects have inertia, footprint, and a unique present pose; they cannot duplicate instantly or shrink into a tray like a window. To change the representation you must move the thing; a glance is not an operation, but an operation always rewrites that representation. Around a table, the object also makes “who is changing it” visible as a hand’s place, not as an anonymous pointer.

Studying it

Compare the same task on tangibles, mouse, and touch, scoring spatial understanding and collaboration, not only time.

Independent variables: whether objects are graspable, whether representation and control are the same object, table versus personal display, number of people. Dependent measures: layout comprehension, verbal reference in collaboration (“this” versus coordinates), contention for the same object, completion time.

Single-person sort times will convict tangibles as “slow blocks.” More informative: after legends are removed, can people still read state from pose and stacking; and when two people change two objects, do gaze and hands collide.

Where it stops holding

When the object set churns with every task (a different file tree each session), you cannot mint a physical avatar for every digital object; the tangible channel covers only a few chosen roles. Pockets have no table; grasp becomes a dangling charm with too little display area. Pixel-level drawing is coarser than blocks. Purely personal, efficiency-only form filling gains almost nothing from spatial avatars, while footprint and object-finding costs arrive first. If the object is merely a stylus substitute (tap an icon), representation and control have not fused—it is a different pointer.

Applying it

• Mint graspable objects only for spatially stable things that must be both read and moved (tracks, roles, sample points). Do not pour a whole filesystem into a box of blocks.
• Let geometry carry the primary reading of that datum: height, stack, pose on a base map. Legends are secondary, not the only representation.
• For tabletop collaboration, give each person objects they can grasp at once; a unique relic becomes a contention point.
• Verify: hide printed legends and ask someone new to say what each object is and what state it is in. If they cannot, representation still lives in type, not in graspable geometry. Add a second operator and watch whether they fight over the same piece.

Related

• Same group: C10.05.2 Physical and digital state can fall out of sync · C10.05.3 Tangible interfaces are hard to undo and copy
• Adjacent: C10.06 Shape-Changing Interfaces · C10.07 Tradeoffs Between Physical and On-Screen Controls
• Search: tangible interaction · representation and control · graspable user interface