Galileo, Day 1: OK, now it gets good.

On page 63 things get interesting.  Salviati has spent a few pages talking about sun spots.  The purpose of this was that after dozens of pages discussing the notion of a perfect universe or perfect celestial objects of...whatever it was that Aristotle was smoking, Salviati has introduced sun spots as evidence that the heavens are changeable, asymmetric, and altogether imperfect.  Simplicio counters by suggesting that the sun spots are planets that are really, really close to the sun, and Salviati swats down that theory with observational evidence.

Finally, Salviati and Sagredo back Simplicio into a corner, reminding him of Aristotle's insistence that the evidence of the senses should take precedence over eloquent argument.  I have two thoughts on this:
1) It is quite ironic that Aristotle's teachings were treated as dogma by subsequent generations who tried to suppress or disregard evidence of the senses in favor of Aristotle's eloquent arguments.  This reminds me a bit of Hofstadter's final word on Dewey, whose disciples tried to institutionalize anti-institutional teaching methods.

2) If we told our students to trust their senses over even the most seemingly logical argument, we might persuade them that Newtonian mechanics is wrong (and not in the way that relativity and quantum mechanics disagree with Newtonian mechanics).  A great many physicists, from a great many pedagogical schools, have noted that when you grow up in a world with friction, air resistance, and gravity, and where most physical objects have substantial moments of inertia (i.e. they can't be treated as gravity), the abstractions of Newtonian mechanics make precious little sense.  We don't have point objects.  We don't have frictionless surfaces.  We can't really conceive of motion "in the absence of an applied force" because real objects either slow down while moving on the ground if you don't push them, or fall if you let go of them.

Moreover, the known inadequacies of freshman lab equipment once led the great teacher and author David Griffiths to make this statement:

I can explain the conservation of momentum in 15 minutes, but three hours in the lab would only convince an honest student that the law is false.

If the question is what we should do about this in the classroom, the answer is that we should do Newtonian mechanics experiments on air tables with digital timers and photogates.  However, the bigger point here is that many fallacies are rooted in real observations, and it isn't always obvious how un-systematic those observations are.  Short of a really, really good experimental investigation of Newtonian mechanics, one will almost certainly be solving mechanics problems on the basis of logical argument rather than the observations of one's own senses.  Ironic, no?  And that's not just true of Newtonian mechanics (where, in principle, a university could get some really, really good freshman lab equipment and really take the time to....HA HA HA HA HA, I made a good joke there, didn't I?), it's even more true of other branches of science.  None of us will ever personally observe, let alone conduct, more than a tiny fraction of the experiments needed to establish the foundations of our branches of science.