> From: "Holloway, Stephen" <firstname.lastname@example.org>
> Date: Mon, 4 Mar 1996 10:27:29 +0000 (GMT)
> In the last lecture, Mr Harnad said that, with reference to the word
> game where you were given say ten words, then have them taken away
> again and asked if you remember a certain word, that we just knew if
> that word was there but could not work out how we knew because it
> happened so quickly. Is it possible to slow down the mind or brain to
> such an extent that the thought mechanism can be observed?
There are some things you do instantly, like seeing or hearing
something. Some things you sometimes do more slowly, like recognising
someone, or remembering a name. But in neither case do you know HOW you
do it just by introspecting on what is going on in your mind. If you
look up the name in a book, you do know how, but in that case it wasn't
your mind that did he work; it was the book.
In the case of remembering whether a "probe" word was or was not in a
list of words you have just heard (the Sternberg task), the reason it
is noteworthy that the answer comes so quickly is that the experimental
evidence shows that the longer the prior list is, the longer it takes
you to decide whether or not the probe word was on that list. We are
tempted to think that the reason for this is that we think of each word
on the list, one after the other, so if the list is longer, we have to
do that more times, so it takes longer. But in reality, it all happens
so quickly that we are not aware of thinking of each word on the list,
one at a time, and comparing it to the probe word.
Besides, even if it happened slowly enough for us to do it
deliberately, that would not explain HOW we remembered what was on the
list! And no matter how slowly we remembered, we still wouldn't know
how we remembered.
In general, you can't slow down the mind for the convenience of reaction
time experiments; but even if you could, it is not clear that you would
be any better at saying HOW you were doing whatever you were doing, be
it ever so slowly and consciously.
Notice that if instead of unconsciously searching a list of four words
in your mind, you were consciously looking under four bowls on a table,
one after the other, then the question would be: How do you know that
you should look under the bowls? How do you look under the bowls? How do
you know they match when they match? So the deliberate, conscious
actions would not explain themselves; they themselves would need to be
Well, if you do it consciously in your head, that still needs an
explanation -- because, remember that assuming that it's done by a little
man in the head (a homunculus), looking at the four mental
words and comparing them to the probe word, is no explanation at all:
You would just have to explain what's going on in the head of the little
man in your head. In the end, it all has to be cashed out in unconscious
processes that just DO it, rather than relying on an unexplained
homunculus to do it.
That's why, in reverse engineering, the test of whether you really have
an explanation of HOW you do something is whether you can get a mchine
to do it. The way the machine does it (unconsciously) is the explanation
-- or AN explanation, at least -- of how YOU do it.
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:38 GMT