Seven Plus Or Minus Two

There seems to be a strictly limited amount of stuff people can hold in their short-term memories. It is probably wise to bear that in mind, when designing or communicating.

GeorgeMiller wrote an article in the 50s called "The Magic Number Seven Plus or Minus Two." [On-line at http://www.musanim.com/miller1956/] It talks about the number of "chunks" that people can remember in short-term memory.

Chunks are not very well defined. The experiments Miller described used random numbers and words. However, it also works with sentences, pictures, and anything else where the items concerned make a whole. [TODO: Double-check this against the article, per the author's request.]

Think of them as seven "pointers" that can point to things that are already in your long-term memory

Apparently, people can be trained to make more use of long-term memory so that they can extend their short-term memory almost indefinitely (up to about a thousand items has been achieved I think). The fact that these long lists are remembered for a long time shows that they are using long-term memory for these amazing feats.

See also SevenPlusOrMinusTwoDiscussion.

Contributors: Anonymous, ChrisBooth, EdwardKiser

Thanks for finding the paper. I've been referring to it for years but never read it. It is QUITE different from what I expected. The author mentions the bit about short-term memory, but states clearly that everybody knows this already. Most of the paper is about categorizing things, and showing studies in which 7 categories are all that are meaningful. It is a very interesting paper! -- RalphJohnson

ForthLanguage programmers note the parallel with the ideal length of a definition: about seven words.

I think I could turn this into a self-defense move. See, if someone comes up to you on the street and asks for your wallet, you first begin to take out your wallet to placate him. Then, quickly speak a sentence with seven random unrelated concepts in them, in order to displace the robbery in his mind. Then, offer to buy his knife for five dollars, thus replacing his confusion with a context in which you don't have to relinquish your wallet. -- NickBensema

(For complete security, you'll have to say nine random unrelated concepts.)

Hilarious! GeekFu at its best! -- PeteHardie

The best movie of the year -- Newsweek ;)

I've seen something like this in TheLongDarkTeatimeOfTheSoul by DouglasAdams...

What if one of your random unrelated concepts drops out of his mind?"

They never drop; they are always pushed.

How is it they always say we only use 10% of our brain? Does expanding medium and long-term memory "stretch" the short-term also?

"They" always say that because it is a popular folk myth. It was based on cutting-edge medical research circa 1880; doctors noticed that patients who lost chunks of their brain still seemed "functional" except for the 10% or so that controls the heart etc. The problem is that they were not very particular about their definition of "functional"; in famous cases the patients were unable to continue working in their old jobs, suffered severe deleterious personality change (e.g. becoming completely apathetic or antisocial), etc, but this was glossed over.

That 10% is a misconception. For years, I was on a mission to train myself to use more, but then I found this is not true :) It's that you USE 10% of your brain AT A TIME. You use your whole brain. I guess proof could be findings from that paragraph above me... -Erik (the dead) Swanson

Discussion about BrainCapacity moved there.

It's a bit of research that seems to suffer a continual curse of misinterpretation.

People can, and do, process more than 7+-2 items of information every day. See this letter from Miller http://members.shaw.ca/philip.sharman/miller.txt. To quote Miller, the 7+-2 number only applies to "unidimensional stimuli (pitches, loudness, brightness, etc.) and also a limit for immediate recall".

For those who are interested - Miller's original paper can be read at http://www.well.com/user/smalin/miller.html. Research on short-term memory has obviously moved on a fair bit since 1956. Miller's "immediate memory" concept has mutated several times, and other experiments have shown that that the "magic" number is more likely to be lower than 7+-2 (see http://www.bbsonline.org/documents/a/00/00/04/46/bbs00000446-00/bbs.cowan.html for more recent info).

Even ignoring the more recent research, the original paper doesn't warrant the conclusions that many people draw from it. It's very specifically *not* about limits to general information comprehension.

Some discussion on the various ways it has been misinterpreted can be found at:

Also see:

J.-L. Doumont "Magical Numbers: The Seven-Plus-or-Minus Two Myth," (Interface) IEEE Trans. Prof. Comm., vol. 45, pp. 123-127, 2002.

I am quite sure that BobbyFisher? or PaulMorphy? routinely held a lot more than this in his short-term memory :)

Not necessarily! It is likely that their "chunks" look a lot different than those of a novice, however. And it could be that they pop back and forth between contexts more quickly as well (the basis for look-ahead). DeGroot? did a lot of research on the differences between the mental models of chess experts compared to chess novices.