The number of items working memory can hold at once is limited

What it is

When a task requires holding several things "in mind" at once — mentally computing a string of digits, comparing several values across two screens, tapping through a sequence of prompted buttons — the number of independent, immediately usable information units a person can keep active simultaneously is small. This ceiling is working memory capacity. It has nothing to do with long-term memory's near-unlimited storage: long-term memory is a question of whether something sticks; working memory capacity is a question of how much can be online at the same moment. It also isn't the same as "poor memory" in the everyday sense — the limit is a shared feature of cognitive architecture, not a deficit in some people, and practice does not train it away.

Why it happens

Working memory isn't a passive store; it's closer to a mechanism that must keep spending resources to hold content "active" and ready for immediate retrieval. Only a limited number of items can be maintained this way at once. When the number of independently presented items exceeds that limit, newly arriving content competes for the same resources being used to maintain the earlier content, and the earlier content gets displaced rather than quietly filed away for later. This is why working-memory failures typically look like "the first few items I was tracking dropped out" or "the later ones bumped the earlier ones," rather than an even, gradual fuzziness across everything — because the constraint is really about how many slots are online at once, not an overall storage ceiling.

Studying it

The basic measurement paradigm is a memory-span task (digit span, letter span, spatial span): a sequence of independent items is presented and the participant recalls it immediately afterward, with span defined as the longest sequence correctly reproduced. A second paradigm is change detection: several items are shown briefly, then masked, and the participant judges whether one location changed; detection accuracy is used to back out how many items were held simultaneously. Both paradigms share a requirement — blocking strategic tricks like rehearsal or grouping as much as possible, otherwise what gets measured is the effectiveness of the strategy rather than raw capacity itself.

Where it stops holding

Performance near the capacity ceiling degrades gradually rather than collapsing abruptly, and the exact number reported varies with material, presentation speed, and test method — which is why different studies don't converge on one identical ceiling. This boundary isn't about which specific estimate is "the right one"; it only establishes that the capacity limit is a real, stable architectural constraint, while treating it as one fixed, precise constant is a misreading of the methodology behind it.

Applying it

• Audit the interface for every point where a user must hold independent information in mind with nothing recorded on screen to check against — if the count runs past single digits at any step, externalize part of it (keep it visible, write it into a summary panel, or make it available as a checkable history).
• Avoid asking users to compare several independent values across screens without a side-by-side aid — for example, requiring them to remember three parameters from the previous page while filling in the next one.
• How to check: walk the task flow and mark, at each step, how many independent items the user must hold unaided in mind; tally the peak simultaneous count. Any step whose number is conspicuously higher than the rest is the first candidate for redesigning how that information is presented.

Related

• Same group: A6.02.2 Retention is brief and vulnerable to interference · A6.02.3 Requiring users to remember content across pages is itself a design flaw · A6.02.4 The classic capacity estimate is closer to four chunks than the earlier figure of seven · A6.02.5 Without rehearsal, working memory content decays naturally within roughly ten to twenty seconds · A6.02.6 Rehearsal extends retention but draws on the same limited processing resource · A6.02.7 Capacity varies with item complexity, so simple and complex items can't be compared on one number · A6.02.8 Capacity limits and time-based decay are two independent mechanisms, and easing one doesn't substitute for the other
• Nearby: A6.01 Sensory memory · A6.03 Chunking · A5.02 Divided attention and dual-tasking
• Search terms: working memory capacity · memory span · change detection task