OT: losing memories (was Re: How many of you are Extreme Programmers?)

Stephen Horne intentionally at blank.co.uk
Thu Apr 24 22:49:48 CEST 2003


On Thu, 24 Apr 2003 11:52:07 -0400, Peter Hansen <peter at engcorp.com>
wrote:

>I think if pathways in the brain can be built up, through repetition of
>stimuli, dreaming, etc., then they can likewise be dissolved, whether
>through drug use, pyschological problems or chemical imbalances caused 
>by lack of sleep, and so on.  Perhaps simple disuse leads only to an
>asymptotic decay which never quite reaches zero, as your argument might
>suggest, but I think other processes could act in different ways.

There are those who claim that the human brain contains a record of
everything that ever happened to it, and that an inability to remember
results from the inability to bring that record back to conscious
recollection. Complete garbage. The brains capacity may be immense,
but it isn't that immense - and if it was, wasting resources in
remembering numerous memories that cannot be recalled or applied in
any useful way would be clearly against the principal of survival of
the fittest. Especially considering that the brain uses much more
energy per unit of weight than any other part of the body.

Some memories may still be present but difficult or impossible to
recall - due to the loss of the associative links between memories -
but the resources dedicated to such memories are prone to being
re-used. Memories don't stay in one part of the brain - at least, not
for several years at least (when they may finally make it into
permanant memory). They get shuffled back and forward between two
different parts of the brain during resorting processes. There are
plenty of chances to reallocate those resources to other memories if
they are unused.

Cognitive psychology and neurology have determined some very
interesting facts about human memory. One such fact is that all human
memory is subjected to extreme lossy compression. Basically, only
aspects of the memory which are considered particularly relevant and
different to expectations are stored. Everything else is factored out.
On recall, the full memory is reconstructed by filling in the gaps
with current knowledge, expectations and prejudices.

If the general state of knowledge, expectations and prejudices has
radically changed in the mean time, the reconstructed memory may not
resemble the original event very closely at all.

A surprising fact that courts love to ignore - a persons confidence in
the accuracy of their recall is no guide to the true accuracy of what
they recall. Confident people may speak confidently and may convince
people very effectively, but they also may not question the accuracy
of their own memory. A less confident person may report memories more
accurately, and with a more accurate sense of the limitations of those
memories, yet will tend to be disregarded and disbelieved due to that
lack of confidence.

There have been many cases where people convicted due to the evidence
given by victims or witnesses, who were absolutely confident in their
memories, have later been proved innocent. Fancy spending a few
decades in prison for rape, only to later be proven innocent by DNA
evidence, because of a victims over-confidence in an intense but
inaccurate memory? - It *has* happened.

Experts are only expert in particular fields. Just because Einstein
was a genius in the field of physics, that did not qualify him to
state that only 10% of the brain is unused. Assuming he claimed it at
all - I never checked, but it does get attributed to him now and again
- it was probably never intended as a statement of strict biological
fact. Benjamins view of Dijkstra's point can probably be seen in a
similar light.

Of course it is a basic fact of physics that any observer is affected
by everything it observes (as well as affecting the thing it
observes), and chaos theory states that the impact of small affects is
often unexpectedly large, but that isn't the same as a memory. You
can't reconstruct a recollection from those affects even if you could
accurately quantify them (which you can't). Its an information theory
thing.


Sorry - I have a broken brain (Asperger syndrome) myself, so I have an
obsessive interest in certain aspects of the brains inner workings.





More information about the Python-list mailing list