At 11:29 AM 2/3/01 -0500, Chuck Britton wrote:
>Taxonomy helps us communicate,
Right.
For that matter, names are useful even when they're not taxonomic names.
> but, as Feynmann was want to say, naming something
> doesn't denote understanding.
Feynman was indeed very fond of saying that, and there's a germ of truth in
it, but I think he went waaaay to far with it.
At a formative age Feynman was walking through the woods with his father,
who made the point that if you know the name of a bird, you don't know
anything about its song, or its feeding habits, or its other behavior.
Well, that's misleading or worse. Suppose you tell me such-and-such a bird
is called a "wren". Although the name by itself tells me nothing, I can
remember the name and go home and look it up, or ask somebody. For that
matter, maybe I've _already_ read about the behavior of wrens.
In some cases the names are mnemonic all by themselves:
-- what is the song of a chickadee?
-- what is the behavior of a woodpecker?
-- what is the color of a bluebird?
In some cases the name is AFAICT completely arbitrary:
-- jay
-- wren
which are un-mnemonic except for the obvious taxonomic implications, e.g.
the idea that gray jays are probably related to blue jays.
And there are intermediate cases where the etymology of the name is so
obscure as to be useless for typical English-speaking users (eagle, junco).
========================================
There are things that names can do for you, and things they can't
do. Changing the name of a bird won't change the bird's behavior.
On the other hand, names are important. They play a central role in
communication, and they even help frame your thoughts when you are thinking
by yourself.
A mnemonic name is nice, but an arbitrary name will do.
OTOH a bad name can be a significant impediment to clear thinking and clear
communication.
1) One type of bad naming occurs when something appears to be
mnemonic, but isn't. For instance, in airplanes there is a device which is
often mis-called the yaw-string but which ought to be called the
slip-string because it measures slip angle not yaw angle. Ironically there
is another device which is often mis-called the slip/skid ball, even though
it doesn't measure the slip angle. It is properly called the inclinometer
ball, because it measures inclination.
2) Another common type of bad naming is when two different concepts
are called by the same name.
A perfect example of this involves the concept of "charge on a
capacitor". Imagine the simplest possible capacitor, consisting of two
parallel plates. In most cases, when people ask how much "charge" is on
the capacitor, they are talking about a balanced _displacement_ of
elementary charges from one plate to the other, of the sort that arises
when you run a current through the capacitor in accordance with Kirchhoff's
laws. However, occasionally somebody wants to put some static electricity
on the capacitor, i.e. an unbalanced net number of elementary charges,
which is perfectly permissible from a physical point of view but which
violates Kirchhoff's laws. Alas this is sometimes called by the same name,
"charge on the capacitor". Yuuuckkk!
Suggestion: If ever you detect a situation of this kind, coin a new name
for one of the concepts.
In many cases the worst thing you can do is to reuse a name. If you have a
new concept, give it a new name. If you can't come up with something
original, at least use adjectives to distinguish one form from the other
(e.g. the American robin is unrelated to the Old World robin). Don't
simply recycle the name of an old concept unless you are absolutely sure
that the old concept is useless and completely forgotten.
3) Standard suggestion: Define your terms, unless you know that you
and your audience share a clear common understanding of the terms. This is
especially important when speaking to students; using a term in a
"standard" way doesn't suffice; students weren't born knowing the standard
definitions.
If it is not possible to give a crisp, declaratory, dictionary-style
definition (and it often isn't), at least give some examples and/or rules
of thumb that roughly delimit the concept.