Re: Miller: Magical Number 7 +/- 2

From: Harnad, Stevan (harnad@cogsci.soton.ac.uk)
Date: Mon Nov 20 1995 - 21:12:26 GMT


> From: "Matsers, Kate" <CMM93PY@psy.soton.ac.uk>
> Date: Thu, 16 Nov 1995 09:00:21 GMT
>
> GM> "Absolute judgement is limited by the
> GM> amount of information, immediate memory
> GM> is limited by the number of items."

But there may be a connection: Recoding many bits into one chunk occurs
not only in symbolic codes that are remembered sequentially. It can also
occur in parallel, as we learn to see many parts as one whole. Or even
when we learn to see many things as all one KIND (category) of thing.
This surely involves changes in dimensionality too.

km> This relates back to something I once said in a seminar about the fact
km> that we can group information and could therefore retain seven, times
km> seven, times seven and so on, so long as no single list was longer than
km> seven. This falls apart under test though, when two groups of seven are
km> given to a participant for memorising, they will lose some of the
km> information and are likely to only recall approximately ten items.

Unless you overlearn the contents of each list of seven, so you know
them forever and all that's needed is a mention of the list-name to
bring them back: In that case. you should be able to remember 7 such
lists of 7. The trick, of course, is in the overlearned chunk. But in
general Miller says nothing about this process of learning and
overlearning.

km> Miller brings us back to the ongoing discussion of the possibility that
km> all cognition is hinged upon or driven by language. Miller describes
km> language as the ultimate process of recoding which enables us to chunk
km> information into retainable portions in order to overcome the limit of
km> sevens.

It's certainly one way. But perceptual learning does a lot of rechunking
too.

km> This reminds me of the discursive psychology we were introduced to last
km> year. Although Harre went about it in a rather annoying manner, I could
km> always see his point then, and I would like to pursue this thought
km> further by reading a little more of the discussion that have attempted
km> to explain our thinking and behaving in terms of the linguistic
km> constructions we find ourselves contained within.

If you want to pursue Harre further, I will soon have an electronic
paper by him that we can all discuss.



This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:56 GMT