[carlob]

I also didn't much like the tone of the article (and I'm a physicist). However there is something to be said about working with data and not being aware of the most fundamental approximations you've done, how long can they hold, under what assumptions they will break.

As a familiar example in physics, take Hooke's law: a spring will try to retract with a force that is proportional to its extension. I think everyone will agree that this is only an approximation, and as soon as you start pulling too much you'll get out of the linear regime, or if you pull long and hard enough you'll start seeing plastic effects and the spring will change shape permanently.

This is to say that having a linear model of springs or an exponential model of economic growth is very useful, but it is important to train scholars to know the limits of their approximations so that they can recognize the need to use a more complex theory.