Parallel multi-limb input interferes with itself

What it is

Requiring two or more limbs to perform independent input actions at the same time — a hand doing one thing while a foot does something else with a different rhythm or direction at the same moment, or a hand gesture layered with a head gesture — doesn't produce the simple sum of the two channels' individual capabilities. Instead the channels drag each other down: accuracy and speed for each channel performed alone are both higher than when performed simultaneously. This cross-limb interference isn't a matter of either limb lacking capability on its own; it's an extra cost that comes from the act of running them in parallel.

Why it happens

The source of the interference is that independent limb movements, even when the task itself makes no demand that they relate to each other, tend to pull toward the same temporal rhythm or coupled pattern anyway — a consequence of motor control sharing scheduling resources at a central level. Planning and monitoring multiple limb movements is not a set of fully independent parallel processes; it competes for the same limited pool of central coordination resources. This resource competition has been repeatedly observed in dual-task psychology research (two near-simultaneous responses slow each other down), and what this shows is that the same bottleneck is not confined to successive actions within a single limb — parallel actions across limbs and body parts are bound by it too.

Studying it

A common way to assess cross-limb interference is to measure each limb's performance executing its task alone, then measure both performing simultaneously, and treat the gap between the two as the direct index of interference magnitude. How rhythm compatibility affects the degree of interference is also a commonly examined variable — having the two limbs execute in-phase same-frequency, anti-phase same-frequency, and different-frequency rhythm combinations, then comparing the interference magnitude across those combinations, is a standard way to test it.

Where it stops holding

Interference magnitude drops as each action becomes more practiced — a highly automatized action consumes far less central scheduling resource than one still in the cognitive or associative stage, so an experienced driver steering with the hands while working pedals with the feet shows markedly less interference than a novice learning both at once. Interference magnitude is also tied to the rhythm relationship demanded of the two limbs: the closer the required rhythms are to synchronous or a simple integer ratio, the smaller the interference tends to be; requiring the two to run fully independently with no simple ratio between them produces noticeably more interference.

Applying it

• Avoid designing fine operations for novice or occasional users that require two limbs to act independently and out of sync at the same time — this combination will almost certainly produce noticeable mutual drag for users without dedicated practice.
• If parallel multi-limb input is genuinely necessary, keep the rhythm relationship required of the two channels as simple and compatible as possible (synchronous or a simple integer ratio), rather than demanding each run on its own unrelated rhythm.
• When judging whether a parallel multi-limb operation is ready to ship, don't look only at experienced-user performance — budget enough of a practice period for interference to drop to an acceptable level as proficiency builds.
• Verification: measure the target users' accuracy and completion time executing each channel's task alone, then measure the same for both performed simultaneously, and compare the gap between the two — a larger gap means more severe interference. Re-measure this gap as the same users accumulate practice, and only treat the parallel design as viable for novices once the gap has clearly narrowed.

Related

• Same group: A8.26.1 Foot input suits binary switches and coarse adjustment · A8.26.2 Head and torso input trade precision for freeing the hands · A8.26.3 Non-dominant and novel limb input shows significantly lower precision and slower learning
• Nearby: A8.04 Psychological Refractory Period · A8.18 Symmetric Bimanual Cooperation · A8.23 Stages of Motor Learning
• Search terms: cross-limb interference · motor entrainment · bimanual coordination · psychological refractory period