This post at Mama Ph.D. raised a number of worthwhile issues. It's basically about people making distinctions between the 'essential' subjects -- the ones at which your performance really matters -- and the 'frills.'
I've heard students talk about this more times than I care to remember. But I have to admit that there's something to it.
My personal sense of it is that the distinction between core and periphery is largely a function of purpose. If your goal in life is to be an exhibited artist, then you might well decide that art is essential and history a frill. If your goal is to be an engineer, I could understand valuing a math class over a psych class. Since different students have different purposes, I wouldn't be surprised to discover that one student's frill is another student's priority.
But the questions go deeper than that. How do we decide what it's okay for people to suck at? And when do they get to start?
In high school, I remember feeling very strongly the pressure to be "well-rounded" for the sake of my college application. I felt like I didn't have the cultural permission to suck at anything. That didn't stop me, of course -- I'm still convinced that geometry is an elaborate prank, and I can't carry a tune in a bucket -- but it didn't feel okay to admit those.
Even in college, 'distribution requirements' came across as yet another postponement of the time when I'd finally get to write off certain things. I was eager to get down to the stuff I considered important, but first I had to do penance by fulfilling various curricular requirements (including phys ed). While I understood some level of abstract argument in favor of well-roundedness, the simple fact of the matter is that I never enjoyed lab sciences. I just didn't. Call it a character flaw if you want, but it's true, and I resented having to slog through some cookbook chemistry in order to get to the stuff I actually cared about. (And yes, if science is your thing, you could flip the variables; the structure of the argument still stands.) It felt more like hazing than like education.
I worry, too, about the effects on motivation of making college students slog through courses about which they just don't care. Patience is a virtue, but real achievement comes with passion. And in an age in which the big safe structures of the past are falling down left and right, it seems to me that we should encourage idiosyncrasy at least as much as well-roundedness. Well-rounded people can do well in established structures, but they aren't the ones who'll spend fifteen hours a day on the next big breakthrough. Innovators tend not to be well-rounded; if anything, they seem generally to be a bit unbalanced. At this historical moment, we need innovators.
All of that said, though, I recognize that students may not know what they need, and that there's a valid educational purpose served in exposing them to subjects they wouldn't have sought out on their own. (Pragmatically, I'm also aware of the requirements for transfer.) Given the choice, plenty of students would probably bypass composition classes, even though most students need them pretty badly. In my teaching days, I always counted it a personal triumph when a student admitted that she took my class to fulfill a requirement, only to find that it was far more interesting than she imagined it would be. If not for the requirement, that 'aha!' moment wouldn't have happened.
My own sense is that it's reasonable to make distinctions between serious subjects and frills, as long as we recognize that the difference is more about the student than about the subject. The trick is in recognizing that difference without sliding into 'anything goes.' On the ground, we use distribution requirements to do that, which is a second-best approach that annoys nearly everybody without mortally offending anybody.
The harder part is the cultural part. At what point do we start telling students that it's okay to be awful at certain things, and why? And how do we choose what's okay to write off, as opposed to the stuff everybody should do at least passably well?