Re: The Mind of a Mnemonist

From: Harnad, Stevan (
Date: Thu Nov 09 1995 - 16:32:05 GMT

ws> From: "Smith, Wendy" <>
ws> Date: Mon, 6 Nov 1995 09:07:42 GMT
ws> Investigations couldn't establish limits to S's recall. Lists of 70
ws> numbers could be recalled even sixteen years later.

Great summary, Wendy!

Notice that S's rote memory violates Miller's Magical Number 7 +/- 2
limits on identification and memory. Is this an advantage or a

Did anyone manage to read Borges's "Funes the Memorious" (In
"Labyrinths", a collection of translated Borges short stories edited
by Jim Irby)?

ws> How did S remember so much, so clearly? Two features of his memory are
ws> apparent. First, synesthesia was present to a more than normal extent.

But it's not easy to say whether synaesthesia was cause or effect: He
had such a full and faithful memory that any little accompaniment of an
event, in any sense modality, stuck to it unforgettably.

ws> The second feature was his remarkable visual imagery. When he heard a
ws> word or number, or whatever, he would conjure up a whole story.
ws> (NB - the story may have helped him remember the
ws> formula, but how did he remember the story? Shades of homunculi here?)

Just the right question to ask. It seems as if the primary thing is
again the limitless memory power; it's not clear that the story is the
"means" or merely another symptom, like synaesthesia and vivid memory
images, of his unerring rote memory.

They DO use story telling and imagery to try to enhance memories, both
in "how-to" courses and in working with amnesic patients, and some say
these work, so what is cause and what is effect is not entirely clear
in all this.

ws> his imagery could interfere with reality in a more prosaic way.
ws> Sometimes the sounds of the word led to a visual image which was in
ws> stark contrast to the meaning. For example, the Russian word svinya
ws> conjured up elegance and fineness for him; and yet meant a pig.

And he would have intrusions of visual images from other places and
times when he was trying to read and image a novel. You are right to ask
whether this is all advantage or handicap.

ws> Also, he didn't appear to be taking in the meaning. He could be given a
ws> list of words, and recite them forwards, backwards, or give words
ws> before or after a probe word. However, on one occasion he was given a
ws> list with the names of birds embedded in it. He was unable to extract
ws> the names of the birds from the list, but had to work his way
ws> painstakingly through the whole list to find them.
ws> He was also given the following list:
ws> 1 2 3 4
ws> 2 3 4 5
ws> 3 4 5 6 etc
ws> He noticed no order in these numbers, and proceeded to memorise them as
ws> usual. He could have memorised Luria's book, word for word, in the time
ws> it took me to read it; but he would have been totally incapable of
ws> summarising it, as I am here.

Very good analysis, and extrapolation. Now what was his problem?
What was the trade-off between the "gift" he had in superabundance (we
would all like better memories) and the price he seemed to be paying for

You mentioned meaning, and also obvious simplifying patterns embedded in
the rote memories, which he found harder than we do to notice, extract
and use.

Could the price of limitless rote memory be the inability to abstract?
the inability to IGNORE what is irrelevant, and select out only what is
irrelevant? Could understanding be, to some extent, the ability to forget
and ignore too, and not just to remember everything?

ws> Poetry bewildered him, he found it hard to move from the figurative to
ws> abstract thinking. This could help him solve some problems - he didn't
ws> get bogged down in verbal manipulations - but in others he seemed
ws> almost child-like in his grasp of the ideas. If he couldn't "see" it,
ws> he didn't understand it.

It's not clear whether it was figurative-to-abstract that was his
problem. He had trouble moving from the CONCRETE and the SENSORY to the
abstract. Images, unanalysed sensory wholes, were fine, but abstract
features and patterns embedded in them were not noticed. Yet abstraction
and generalisation depend on being able to ignore the concrete, sensory
details and focus on a few salient features, the ones that matter.

ws> First was the nature of his experiences - his strong synesthesia. How
ws> well does this fit in with the ideas of rehearsal or echoes?

During the seminar, Denise (I think) brought up the idea that the extra
dimensions that synaesthesia would add to the stimulus might be a way of
getting around the magical Miller limit. But surely S's memory went far
beyond the amount that extra dimensions would give you: It seemed
practically limitless, whereas each dimension adds about 3-4 items more.

It would have been interesting to test S on other Miller limits: Can he
remember more along one stimulus dimension? Can he remember every jnd of
difference absolutely? Control experiments in which normal subjects were
given artificial synesthesia (accompanying sounds that vary only in
loudness with a lot of other uniquely associated stimuli, in different
sense modalities: the question is, how many?).

So the question remains: is synesthesia -- the extra sights and sounds
and smells and states that seem to be stuck to everything -- a means of
perceiving, hence remembering more? or just another symptom of being
unable to ignore or to remember less?

ws> Second was his strong visual imagery, and how it related to his
ws> reasoning and memory. Language appeared to hinder him - how does it
ws> help most people?

Language uses abstractions. Whenever you tell someone something instead
of showing it, you are abstracting (a picture is worth not just a
thousand, but an infinite number of words: the words just capture a
selected PART of it). Vivid imagery is part of a limitless concrete rote
memory, but abstraction is at odds with it, requiring you to forget or
ignore most of the concrete and to abstract from it only selected bits.

So abstracting is at odds with concrete imaging and remembering. And
verbalising is really a string of abstractions.

ws> Third was whether his memory was an advantage or disadvantage. Does it
ws> help to retain everything in total, but with little meaning? And how
ws> does "normal" memory extract the important stuff, and lose the waste?
ws> And is it related to either the experience or the imagery?

Not that much is known about this aspect of how normal memory works,
perhaps because one cannot entirely abstract remembering something from
perceiving it: The evidence you gave about the number patterns above
suggests that S may have problems not only with overabundant memories,
but from the absence of some initial filtering: Would he notice the
pattern while viewing it in the first place? Could he learn to abstract
it, form many instances, if he were learning to categorise? Could the
interference that was spoiling his novel reading also affect his
understanding of the world?

Today we would have wanted to do a lot more cognitive neuropsychological
testing of S...

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