The non-gesture path serves recognition failure, fatigue, and bodily difference among other failure modes

What it is

A non-gesture path is not a taste switch for people who dislike gestures. It covers a cluster of failure modes: the recognizer misses, the arm is too tired to raise, hand shape or a prosthesis mismatches the model, something is being carried, the setting forbids waving. A button written only into settings for “advanced users” is not aimed at those modes. Fatigue, recognition, and environment each have their own mechanisms. What is required here is that the equivalent path serve this whole cluster, not act as a temporary patch for one of them.

Why it happens

The modes live on different time scales. Recognition failure can be this second’s lighting; fatigue shows up after twenty minutes; bodily difference is there on first use; social setting switches with place. A runtime fallback built only for recognition failure is already being called “sometimes broken” before fatigue arrives. Spare controls built only for an accessibility checklist, if they cannot be reached on the main surface where fatigue appears, do not exist. The cluster shares one equivalent path because users cannot diagnose which failure they are in—they need a method that works now. Design the path for “the hand cannot be input in this moment,” not as a branching menu of diagnostic names.

Studying it

Manufacture recognition failure, late-task fatigue, restricted hand shape (glove, one hand occupied), and a setting with bystanders, and see whether the same non-gesture path is reachable in all four. Code time to find it and whether a hint was needed. Do not measure only one mode and declare “an alternative exists.” In interviews ask “why did you switch to the button,” and see whether several reasons appear rather than only “I don’t know the gesture.”

Where it stops holding

Some failures should be repaired on site first (go turn on a light); the equivalent path is a parallel option, not the only answer. Users who cannot use any manual input need a third channel (voice, gaze); a button path does not cover all bodily difference. A game that designs “arm burn” as challenge would break its rule by offering a button mid-play; declare that at the start, and still offer an exit for people who cannot play. In an emergency, finding a button may take longer than trying the gesture once; put the path where the hands already are.

Applying it

• Accept on “the hand cannot input,” not on “recognition score is low” as the only switch that reveals the button.
• Put equivalent controls on the main surface people still look at when fatigue arrives, not only inside accessibility settings.
• Walk four scenes once each: recognition collapsed, after ten minutes of raised arms, one hand holding a cup, a stranger present. Missing the button in any one means the path is not yet aimed at this cluster.

Related

• Same group: C4.28.1 Any function carried by a mid-air gesture needs a non-gesture equivalent path · C4.28.3 The non-gesture path cannot be buried deeper than the gesture path, or it exists only in name · C4.28.4 The cost of providing the non-gesture path should be counted in the overall cost of the gesture design
• Adjacent: C4.12 Fatigue cost of mid-air gestures · C4.32 Social acceptability
• Search: failure mode · fatigue · body difference