Vision and hearing are separate perceptual resources; speech and manual response are separate response resources

What it is

Among the resource dimensions multiple resource theory lays out, two pairings get used directly in interface design more than any others: on the perceptual side, vision and hearing are two independent perceptual resources (input); on the response side, speech and manual input (keys, touch) are two independent response resources (output). These two pairings are the direct theoretical basis behind most "spread it across multiple channels" design advice.

Why it happens

The independence on the perceptual side comes from vision and hearing being processed on separate neural pathways from early stages onward — an overload on one channel doesn't directly seize the other channel's low-level processing resources, except once both need to converge on shared central attention for interpretation. The independence on the response side is similar: the motor pathways behind speech (larynx, articulation) and behind manual action (arm, fingers) are largely independent at the execution level, and can be activated simultaneously without queuing behind one another. That's exactly why "voice interaction paired with simultaneous manual operation" is such a common solution: it draws on two independent dimension pairs at once — vision and hearing separated on the perceptual side, speech and manual response separated on the response side.

Studying it

Evidence for this comes from the same dual-task paradigm, run as a crossed design over perceptual channel (visual vs. auditory stimulus) and response channel (keypress vs. spoken reply), comparing dual-task cost across the four resulting combinations: same-channel combinations (visual input paired with a keypress, for instance) carry the highest cost, and cross-channel combinations (visual input paired with a keypress, auditory input paired with a spoken reply) carry the lowest. This crossover pattern supports the claim that the two pairs of dimensions are genuinely independent, rather than a single difficulty factor explaining everything.

Where it stops holding

Independence on the perceptual and response sides isn't absolute isolation. If visual and auditory information both have to be converted into the same verbal code to be understood — listening to a spoken instruction and reading a written one both ultimately require verbal comprehension — the two still compete on the coding dimension even though their perceptual channels differ. Manual and verbal response channels are independent at the level of motor execution, but if both have to pass through the same central decision about what to say or press, that decision stage may still share resources. These two pairings are also simply the dimension combinations tested most often in research — they don't mean these are the only two dimensions worth caring about in the theory.

Applying it

When a user needs to receive two information streams at once, assign one to vision and one to hearing rather than putting both on vision; when a user needs to give two responses at once, assign one to speech and one to manual input rather than requiring two manual responses. After making the assignment, still check whether the coding level is genuinely separated too — if both streams ultimately have to be converted into language to be understood, separating perceptual channels alone won't fully avoid competition. Verification: compare a "two visual streams" design against a "one visual, one auditory stream" design on dual-task performance; confirming the latter shows clearly lower interference is what verifies that this assignment rule actually paid off in the specific product context.

Related

• Same group: A9.10.1 attention resources aren't one pool — they split into pools by channel, stage, and response · A9.10.2 tasks sharing a resource pool interfere more than tasks using different pools · A9.10.4 multiple resource theory explains why dual-task interference can't be predicted from task difficulty alone · A9.10.5 designing parallel tasks should deliberately assign them to different resource pools to reduce interference
• Nearby: A5.02 divided attention and dual-tasking (the observed cost and behavior of dual-tasking)
• Search terms: perceptual modality · response modality · visual-auditory dual task