Status: Phase 1 stub. Matures in Phase 5 per §8.1.

Statement

Animation exists to clarify cause and effect, to spend the user's attention deliberately, and to mark transitions of state that would otherwise feel discontinuous. Animation that does not serve one of those three purposes is removed.

Why

The audit found six surfaces with motion that a calmer designer would not have shipped: a bouncy cubic-out on every tab change, a 600ms fade on the messaging composer, a permanent shimmer on a button that wasn't loading. The user's attention is finite and motion consumes it. A 600ms fade on every interaction is not "delight" — it is a tax the user pays for nothing.

Conversely, the audit found four surfaces that needed motion and didn't have it: the route transition between an inbox conversation and the message detail (felt like a hard page-replace), the moment a message is sent (no acknowledgment that the bubble flew anywhere), the keyboard appearing (the input jumped). These are exactly the cause-and-effect clarifications motion was invented for.

How

Motion vocabulary (F-6) names motion by meaning: motion.enter, motion.exit, motion.shift, motion.attention, motion.handoff. A primitive that does not have a clear narrative reason for one of these tokens does not animate. Reduced-motion preference (A-2) maps every named token to a reduced equivalent.

The bar for adding motion is: "what would be confusing if this didn't move?" If the answer is nothing, it doesn't move.

Phase 5 deepens this stub with per-surface motion citations, the audit counter-examples, and the cause-and-effect taxonomy.