Unsupported arm-raise fatigue sets in within tens of seconds

What it is

Interactions that require the arm to stay lifted and unsupported toward a vertical or tilted surface — large touch walls, mid-air gesture zones, floating panels in front of a headset — push shoulder muscles into a costly load state that goes largely unnoticed until it's already advanced. Human factors and interaction design call this the gorilla arm effect, a term Jef Raskin used when discussing early touchscreen interfaces, named for the posture's resemblance to an ape's suspended forelimb. The defining feature isn't a binary "can hold it / can't hold it" — it's how long the posture can be sustained. Users don't need to reach exhaustion: merely hovering at a height consistent with the interaction task produces measurable and felt fatigue on a timescale far shorter than a typical task — usually tens of seconds to a couple of minutes, not the "several minutes to tens of minutes" that intuition suggests.

Why it happens

Holding an arm aloft toward a target requires the shoulder flexors and abductors to continuously work against the moment created by the weight of the whole arm — forearm, hand, and possibly a held device. This is a static load: the muscle barely changes length, it just maintains tension, and static loads fatigue faster than dynamic movement at comparable intensity — a general property of static loading, not something unique to raising an arm. What makes arm-raising land so squarely in the steep part of that fatigue curve is that the shoulder flexors are inherently weak relative to the moment they must produce to hold the whole arm in the vertical plane: even just holding the arm horizontal or slightly above requires a relative muscular effort (percentage of maximum voluntary contraction) far higher than intuition suggests, and the higher that relative load, the shorter the sustainable duration.

Studying it

The standard measurement has participants hold the arm at a set angle or perform a simulated task and records the time from onset to a predefined stopping criterion — self-reported fatigue reaching a set level, inability to maintain posture, or compensatory tremor — a duration called endurance time. Surface EMG (sEMG) is commonly recorded alongside it, watching the median frequency of the signal shift downward over time as objective evidence of local metabolite buildup, cross-validated against the subjective scale.

A methodological caution: endurance time is highly sensitive to angle, added weight (holding a device or not), and exactly how the posture is defined. Absolute numbers don't transfer across studies — they need to be remeasured under the actual posture the target device imposes.

Where it stops holding

This timescale applies only to unsupported, airborne arm-raise postures. Once the forearm or elbow gets even partial support — an armrest, a desk edge, the device itself bearing some weight — the load structure changes entirely and endurance time extends substantially; the conclusion no longer applies. It also only describes scenarios where the shoulder is the primary load-bearing joint — if the main effort is at the wrist or fingers (say, with the forearm resting and only the fingers active), this timescale does not hold.

Applying it

• For any interaction requiring the arm raised to shoulder height or above with no physical support, cap continuous hover time at well under a minute; beyond that, provide a gap to lower the arm or a lower alternative input.
• Place frequently-used hit targets on interactive walls or mid-air gesture zones at or below the user's standing shoulder height, reducing default reliance on high reaches for common actions.
• Verification: have test users operate continuously at the device's actual angle, recording a subjective fatigue rating (e.g., Borg CR-10) every 10–15 seconds and noting any postural compensation (shoulder hiking, leaning forward). Use the point where the rating curve turns sharply upward or compensation appears as this design's actual endurance threshold — not a number borrowed from a study on a different device.

Related

• Same group: A8.19.2 fatigue accelerates with both angle and duration · A8.19.3 perceived fatigue lags physiological fatigue · A8.19.4 short usability tests miss this problem · A8.19.5 high-frequency actions must not require overhead reach
• Nearby: A8.20 static vs. dynamic load · A8.14 hand support and precision · A8.25 reach envelope and range of motion
• Search terms: gorilla arm effect · endurance time · isometric shoulder load · sEMG median frequency