What is wrong in graduate education?

I spend a lot of time these days trying to find a way to talk about this. What is going on in physical sciences that gets in the way of learning, at the highest levels of academia?

You take a bunch of well-intentioned (this is my personal opinion, and I think the only place from which I want to study a group of people, otherwise I wind up on a witch-hunt for meanies), well-educated, motivated individuals, who graduate schools may be competing to admit to their programs, and by the end of 4, or 6 or 8 years, maybe half of them have graduated. What happened? Are we really going to stick to the story of "Oh, well, only those who had the fire in the belly made it, and those others weren't meant to be continue on"?

I was talking with a figure skating coach the other day - yeah, when is the last time you heard that sentence?! - picking her brain for how coaching works in skating. I have the sense that at least half the structures of graduate education (the weed out courses, and the fact that they are called that, the willingness to not do a self-check when lots of you most qualified candidates wind up leaving the field during your own grad program; the absence of a scientific writing and speaking course in the core of any graduate curriculum) comes from the belief that the most prominent scientists are born and not made. It gives the graduate program an excuse to not intentionally, carefully and fully mentor their students. And then the other half of the practices (saying students should be working 80-100 hours a week; saying that only the toughest survive) point to the belief that hard work makes a top scientist. In fact, it kind of follows the lines of which example is being given - when professors quote from their own lives and experiences, that impostor-syndrome argument for hard-work-and-luck comes out more. When the examples of others comes us, it is more of a "yeah, well she was clearly made to do this and didn't need to be trained in most of it" gets more airtime.

Regardless, there are a lot of very motivated, top-of-the-game students being accepted to grad schools and then many of them do not finish. Or do very well giving talks or writing papers. And a lot of things are not consistently, explicitly taught. You know what would happen to a coach who didn't teach his skaters how to do spins and only concentrated on jumps?

None of his skaters would win anything. The judges just wouldn't be able to award full points.

But perhaps that is part of the problem in academia - that we don't have an explicit enough scoring sheet for graduate students. What is it they will be judged on? Sure, sure, publication record. But what are departments actually looking for? How do hiring committees make their decisions? If a strong letter of support from a faculty member that knows the student is a big deal, then no wonder grad students are scared of sounding stupid in their home departments. Any verbal performance that impacts a letter writer's opinion of that student (especially about whether he or she is smart enough for the field) is a high stakes situation. All the more so if the professors writing these letters don't really know how to judge intelligence or potential any better than "if I knew that, it must be simple, so I can't believe that student didn't know it."