Why seven?

Miller's 1956 paper, "The Magical Number Seven, Plus or Minus Two: Some Limits on Our Capacity for Processing Information," is a classic in cognitive psychology. In it, George A. Miller humorously framed his observations around being "persecuted by an integer" (the number 7), while reviewing evidence from information theory applied to human performance. He distinguished two related but distinct limits:

 -Span of absolute judgment (unidimensional stimuli, like judging pitch, loudness, or position without comparisons): People can reliably distinguish about 7 categories (roughly 2.5–3 bits of information), beyond which confusions rise sharply. Examples included pitch discrimination (~6 tones), loudness (~5 intensities), or visual position (~10–15 points in some setups), with a narrow range across studies (mean ~2.6 bits).

 -Span of immediate memory (e.g., digit span or word recall): Roughly 7 ± 2 "chunks" (e.g., ~9 binary digits, ~7 decimal digits, ~5 monosyllabic words). This is measured in chunks (units), not raw bits of information—the key insight.

 Miller emphasized that humans overcome these bottlenecks through recoding or chunking (grouping information into larger, meaningful units, like turning binary sequences into octal or letters/words), multidimensional stimuli (adding cues like size + color), and successive processing. Chunking dramatically expands effective capacity (e.g., experts recoding binary into phrases). He noted the "magical" coincidence of ~7 appearing in both domains but treated it cautiously as possibly serendipitous, not a universal law. The paper ends by suggesting these limits shape thought and language but can be stretched, calling for more research on recoding.

It was never meant as a rigid "law" for short-term memory (STM) in all contexts—Miller himself later reflected on its tongue-in-cheek tone.

So, why seven? That our limit...or at least it was in 1956...

You can subscribe if you'd like to stay up to date and receive emails when new content is published!