OK, I have to disagree, simply because I think it's that's a wonderful argument and HN bans empty ++ posts.
Companies that do basic research are a different kettle of fish. They make money on breakthroughs and innovation. Java schools make money on students, so the thinking is an entirely different kind - formalizing existing knowledge (so the students can study something more "academic"). Sure, you can formalize code as category theory, algebra, sets, and so on, but it's not exactly an out-of-box breakthrough. Also, the type of researchers you get are excellent students, not engineers who made innovations in the field and got kicked up to the lab.
Java schools don't make money from the students by "formalizing existing knowledge", though. It's more of a prisoners' dilemma: they make the most money by relying on other schools to formalize knowledge, and then delivering it to their own students as cheaply, but with as much implied authoritativeness, as possible.
But I meant to indict here more than just what you normally think of as "Java schools"—virtually all Universities are receptacles for students who go to them to get a high-paying (and high-status!) job, and thus virtually all Universities cater to those students. Java schools are just more upfront about it.
How Universities actually make their money is a combination of convincing students to study there, and receiving grants. Both of these rely on the status of the researchers in the employ of the University—even though the researchers (the Ph.Ds) are very infrequently the students' lecturers. Students pay not to learn from high-status people, but simply to coexist in an environment with them, and have the "status" of the institution rub off on them.
In order to reverse this process, we have to remove "has studied at high-status University X" as a judging criterion for high-status jobs. That may well be impossible, though, given how embedded the idea has become in our culture.
"even though the researchers (the Ph.Ds) are very infrequently the students' lecturers"
That's not true at MIT; are there any other exceptions to this sad pattern in research universities?
One of the advantages of the lowest tier schols that don't do (much) research is that their faculty are there to teach first and foremost. When the system works, when the instructor is sufficiently in command of his material, you get a better result than in the mid-tier schools.
though the researchers (the Ph.Ds) are very infrequently the students' lecturers.
Who does teach at these Universities? Almost all my lecturers where either current researchers or former researchers currently focusing on writing and teaching. I can only think of one or two courses where the lecturer didn't at least have (or was working towards) a relevant PhD.
I'm pretty sure derefr meant the ones who already have their Ph.Ds, not the archetypal grad student who's learning how to teach at the same time he's teaching an entire class (instead of a recitation, section or tutorial).
Companies that do basic research are a different kettle of fish. They make money on breakthroughs and innovation. Java schools make money on students, so the thinking is an entirely different kind - formalizing existing knowledge (so the students can study something more "academic"). Sure, you can formalize code as category theory, algebra, sets, and so on, but it's not exactly an out-of-box breakthrough. Also, the type of researchers you get are excellent students, not engineers who made innovations in the field and got kicked up to the lab.