Re: Difficult Explanations

From: Adrian Tymes (
Date: Mon May 14 2001 - 10:57:10 MDT

Lee Corbin wrote:
> At 08:58 PM 5/13/01 -0700, Adrian wrote:
> >"Eliezer S. Yudkowsky" wrote:
> >> Incidentally - this occurred to me afterwards - this is not me saying,
> >> "I'm a better explainer than you, nyah nyah." This is me saying "I get
> >> strongly annoyed when people say that things like this are hard to
> >> explain." It symbolizes a psychological disconnect. Our instinctive
> >> reactions, our automatic intuitions, carry all the necessary information.
> >> Trust your intuitions enough to verbalize them.
> >
> >Well said. I would add that it can be good to verbalize them not just
> >to others, but also to oneself. That way, one can know why one's
> >instincts lean a certain way, and know better when to trust them - and
> >when not to.
> But some things really are hard to explain! Isn't it true,
> for example, that some things are a lot harder to explain
> over the phone? Well, some things are hard to explain, period.

True. But if one practices explaining it to oneself as one would
explain it to a novice, or a child (if that metaphor is clearer), then
one can find ways to explain it to others easier...and, not
infrequently, come to understand it better oneself. At least, I have
used this technique myself when faced with confusing collections of
facts, and I feel that I understand those things better as a result,
though I am not certain how I could objectively test this. I have
found that for even the most difficult problems, repeating this
technique has allowed me to eventually boil anything I have tried it on
down to layman's terms. This happens most often through a combination
of finding commonly-used terms that adequately describe a concept, of
finding commonly known concepts with a small and clearly describable
set of differences from the concept being explained, and of determining
what large sections of complexity can be reduced to a few statements
while hiding the details from the basic explanation (with the
understanding that, if those details are desired, they can be requested
after the basics are digested).

This archive was generated by hypermail 2b30 : Mon May 28 2001 - 10:00:05 MDT