OT: Miller’s 7+-2 “law”

[Dmitry Olshansky]

On Saturday, 3 March 2018 at 16:59:56 UTC, Russel Winder wrote:
> On Sat, 2018-03-03 at 16:06 +0000, Dmitry Olshansky via
  DuckDuckGo search.
> It's a 1956 paper by Miller that claims 7 is the magic number 
> for short term memory, the number of chunks of stuff you can 
> keep for a certain period. A chunk is not a defined thing such 
> as characters or words, but they are examples.  I am not sure 
> what the experimental status is of this "theory", but I suspect 
> no-one has disproved it as yet.

I know people who indirectly proved that theory to be correct in 
many unexpected ways. In particular when people are asked to 
define “distant” or “hot” as a set of classes they usually settle 
for around 7 states and cannot distinguish finer ones. Same 
problem with colors, as in defining shades of the same color.


All that said, the trick is that ~7 applies to any “thing” and 
thusly your capacity increases if you can “merge” things to a 
single entity or otherwise establish relations or laws, doing 
reduction on a number of entities. Likely composition is a 
sideeffect of this tendency and 7 is not exact number in any wat.