> In the case of testing it’s because you choose to focus on the obviously bad thing (current state of testing) rather than the very complex and difficult question behind it: HOW do you measure knowledge? And when you decide how, how do you scale it?

I would actually focus on the question of "Why do I need to quantify everybody's knowledge at a high resolution?"

When I was TAing, I held the position - never accepted I should say - that we should make more courses pass/fail; and instead of investing effort in the numeric grading keys, try to give more meaningful feedback on assignments.

Some alternative suggestions I brought up:

* I suggested that the final grade be a combination of the assessment a roll of 1D6 points - to hammer it in that the grading is to a great extent artificial. Somehow this was even less popular of a suggestion...

* I once proposed we offer people a perfect passing grade if they just never show up to class nor submit anything, and only people who want to learn would risk an imperfect grade. I really liked that proposal, because it put the two motivations - learning and making the grade - which are often conflated, at direct odds with each other.

Of course none of this was taken seriously - even though I was serious. Kind of.