We've got an interesting discussion going here,
spread across several threads.
To summarize:
1) What should we NOT do when somebody asks a
question that embodies a fallacy?
a) We should not perpetuate the fallacy.
b) We should not berate and abuse the questioner.
2) What SHOULD we do when somebody asks such a
question?
a) Figure out the intended meaning, retate the
question, and answer the restated question.
Note: The foregoing applies not just to students
and not just to classrooms. It applies in the real
world to bosses who ask questions, customers who
ask questions, etc. etc. If you don't want to
answer the question, trying to "get off on a
technicality" is a losing strategy.
Note: It is worth teaching this to students
explicitly: "You should expect me to not
nitpick your questions. And I expect you to not
nitpick my questions. This is because ...."
To continue the summary:
3) So what are some examples of fallacies?
a) The Pathetic Fallacy. http://www.m-w.com/cgi-bin/dictionary?va=pathetic+fallacy
If one were to say, "the electron didn't follow
the predicted path because it didn't feel like
it" ... well, that would be a pretty pathetic
example of the Pathetic Fallacy. [sorry]
b) ?Anthropomorphism? It has been alleged that this
is a fallacy of the worst sort. But I don't think
that is necessarily true. Let me discuss this
point. (I think there is consensus on all the
preceeding points.)
There have been several questions lately of the general
form
-- How does the electron know .....
-- How does the constraint know .....
When we speak of an inanimate object "knowing" something,
we are using colorful language, and anthropomorphic
language -- but I'm not at all convinced it's fallacious
or even risky. What the electron "knows" is very
different from what it "feels". In many cases, "knowing"
is a colorful shorthand for _information_ which absolutely
is a proper subject for technical discussion.
A good feel for information goes hand-in-hand with a
good feel for entropy.
Information and entropy are often unduly neglected in
the syllabus -- but we ought to fix that. We ought to
encourage (indeed demand) that people think about such
things.
=======
Sometimes, questions that ask about "knowing" are really
asking about stability-and-control issues. (This is
neither entirely the same nor entirely different from
asking about the information budget.)
This is another topic that is often neglected in the
curriculum. But it shouldn't be.
Example: How did James Watt's steam engine know what
speed to maintain? Answer: Flyball governor.
Feedback control occurs in more-complicated systems
(computers, corporations, ...) and also in systems that
are less complex (or at least look less complex) than a
steam engine:
Example: In a dime-store balsa-wood glider, what
makes it maintain a definite angle of attack? How
does it know what angle of attack to maintain?
Answer: decalage, mostly. http://www.monmouth.com/~jsd/how/htm/aoastab.html
Example: How does the wall know how much force to exert
on the bouncing ball? Answer: Hooke's law. The wall
deforms just enough to develop the required force. Under
ordinary conditions, this deformation plays no role in
the energy budget or the momentum budget, so we usually
don't pay much attention to it.