Richard Feynman makes an interesting point when an interviewer asks “Why do magnets repel or attract?” that since laws of physics are based on observations, the question of “Why does X happen?” sets you up for an infinite series of questions until eventually the person honestly says “I don’t know.”

http://www.youtube.com/watch?v=wMFPe-DwULM

Models are imperfect, that’s why we refine them. And by “imperfect” I mean as a Mathematics instructor would count points off an exam for a student writing an answer in decimal approximation rather than keeping it as a (reduced) exact expression. Models of the universe are made arbitrarily accurate through application of study and research so these imperfect models can make useful predictions within an allowed measure of error.

And, seeing as how a student may be very naive to a subject, one might find advantage to teaching about George Washington’s proverbial cherry tree before moving on to a general discussion of civics and the social contract.

Bowen: … the book’s answer key should only list the odd-numbered problems.

Rookie error right there. Good catch.

(Seriously though, this is a horrible context and I hope that if I were writing about linear models, I would choose something a lot more appropriate. Besides, my own toaster actually DOES have a linear model for its 7 settings!)