The three load types draw on one shared, limited processing capacity

What it is

Intrinsic, extraneous, and germane load aren't three independently measured axes — they draw on one shared, limited pool of working-memory capacity. That's the central claim of cognitive load theory (Sweller) about how the three relate to each other. Once their sum exceeds that shared capacity, the task or the learning fails; the three trade off against each other, so the more extraneous load takes up, the less room is left for germane load, even when intrinsic load itself hasn't changed at all. This theory answers a question about a single task: how the three load types split a fixed working-memory budget within that one task — a different level of analysis from how attention resources get allocated across multiple parallel tasks and channels, which is what multiple resource theory addresses. The two frameworks both talk about "resource allocation," but they aren't answering the same question: one operates inside the working memory of a single task, the other across several tasks running at once.

Why it happens

Working-memory capacity is a general-purpose bottleneck that's limited and can't be directly expanded through practice. Modeling the three load types as sharing one budget explains a pattern that's easy to miss: reducing extraneous load alone can look like it "solved" an intrinsic-load problem — but intrinsic load hasn't moved at all. What changed is the capacity extraneous load used to occupy: a task that used to fail because total demand exceeded capacity now becomes doable because total load has dropped back under the ceiling, and the freed-up margin can even go toward the schema-building work of germane load.

Studying it

The main research approach manipulates extraneous load (presentation) and the task's intrinsic load (complexity) together, and observes how learning or task outcomes shift across their combinations — this tests the shared-capacity assumption directly: if the three types really do share one budget, then reducing extraneous load should produce a bigger benefit on high-intrinsic-load tasks than on low-intrinsic-load ones, because high-intrinsic-load tasks already sit closer to the capacity ceiling, so any freed-up margin matters more for whether the task can be completed at all. That interaction pattern is itself the evidence for the shared-capacity claim. Methodological note: none of the three load types has an independent, directly additive measurement unit yet — most evidence comes from this kind of indirect inference (manipulate presentation and difficulty, watch what happens to outcomes), not from reading off three numbers and summing them.

Where it stops holding

Shared capacity is a modeling assumption internal to cognitive load theory, not something every researcher accepts — whether germane load is genuinely the same kind of, and directly additive with, intrinsic and extraneous load is itself disputed. Total capacity also varies clearly across individuals: whether the same task exceeds the "shared budget" depends on a user's working-memory capacity and level of expertise, so no single conclusion generalizes across users.

Applying it

When design resources are limited and a screen-by-screen extraneous-load audit of the whole product isn't feasible, prioritize the audit and redesign effort on tasks or screens where intrinsic load is already high — complex configuration flows, decision interfaces with many variables, first-time onboarding — rather than spreading effort evenly across every page. The shared-capacity model predicts that the same amount of extraneous-load reduction frees up more usable margin, and has a bigger effect on whether the task succeeds, in these high-intrinsic-load settings, giving a better return on the same design investment. Verification: apply an equal amount of extraneous-load reduction to a high-intrinsic-load task and a low-intrinsic-load task, then compare the gain in task success rate; if the high-intrinsic-load task shows a clearly larger gain, the current prioritization is well-founded.

Related

• Same group: A9.06.2 intrinsic load can be reduced through segmenting or pre-training, not just accepted passively · A9.06.3 the three-way split originates in schema theory, and schema-building is exactly what germane load does · A9.06.4 whether germane load is an independent resource, rather than a facet of intrinsic load, remains disputed · A9.06.5 classifying a given load requires checking whether it changes with task expertise
• Nearby: A9.01 the three load types (definitions and design leverage) · A9.10 multiple resource theory (resource allocation across tasks and channels — a different level of analysis from this group)
• Search terms: cognitive load theory · working memory capacity · Sweller