"Should master's programs train their students for the number of students they would have in an ideal world or the number of students they will have in this era of expanding class sizes and bulging enrollments at community colleges?"
The latter.
This piece about the programs that grant master's degrees in composition raised a crucial, if underappreciated, issue. And it's certainly not limited to writing programs or English departments.
I've seen any number of programs/initiatives/interventions over the years that accomplish wonderful results when a tremendous amount of resources are spent on a single, small group of students. If we could get every class size down to 15 or less, with supplemental instructors and dedicated advisers and ample money for extras and plenty of course releases to keep the faculty fresh, then yes, we could improve student success rates markedly. For reasons of their own, certain foundations (cough Gates cough) love this kind of thing. Pouring improbable amounts of money into tiny pilot programs can get you some impressive percentage gains.
But you can't scale them up. They're not sustainable. While they may make everyone involved feel good for a while, they're ultimately beside the point. Any intervention that requires tripling our costs per student is not to be taken seriously, even if it works on a pilot basis. That's especially true when the grant expires, and the project either moves to internal funding -- the budget that just keeps getting cut -- or dies. If you're serious about sustainable collegewide change, projects like these are just irrelevant.
One of the comments to the IHE story used a nice metaphor. Instead of thinking of community colleges as just like every other college but a little bit worse, think of them as creatures unto themselves. Just as a doctor in an ER must work differently than a doctor in a surgical practice, so too must a professor in a community college work differently than a professor at a tony SLAC. That's not because ER doctors are worse; it's because the demands of the setting are different. As with the ER, community colleges need people who work well in this distinctive setting.
In the context of teaching writing, it means learning how to grade productively when you have four sections of twentysomething each. (I personally think that any college that packs forty students into a composition class should have its accreditation revoked, but that's me.) In the context of teaching, say, math, it involves learning how to teach around phobias, language issues, and rustiness. In the context of teaching almost anything we offer, it involves projecting a certain comfort level with students whose personal styles will be wildly different from your own, and whose roughness around the edges can be bracing.
It isn't just about stamina. The stamina metaphor implies doing the same thing you did elsewhere, just a little more slowly. That's missing the point. It's about studying in a systematic and serious way the most productive ways to prioritize. If you can only spend 15 minutes per paper when you grade, what's the best use of that 15 minutes? Many people recoil at the question, but failing to ask it is failing to come to terms with the reality of the college. Worse, personal observation suggests that people who don't think seriously about the adjustment eventually just give up, and revert to multiple-choice testing. The "how best to use 15 minutes" question strikes me as empirically testable, and the (potential?) results strike me as obviously useful.
I say all of this while freely admitting that I am not up to date on the latest research in the teaching of composition. If it's already there, then great. But I don't get that impression, and composition is only one example.
At least the composition programs are asking the question; full credit for that. I'd love to see other disciplines do something similar. In my scholarly discipline, the attention paid to teaching is minimal, and often openly contemptuous. My dissertation advisor was blunt in telling me to spend as little energy as possible on teaching, since as he memorably put it, teaching "doesn't count for shit out there." I wanted him to be wrong, but he wasn't. And even reducing it to "teaching" is misleading, since teaching two classes at a selective liberal arts college is a very different undertaking than teaching five classes at a community college. You can't just copy and paste. What "teaching" means varies drastically by institutional context; it would be nice if new grads were prepared for that.
I understand the argument from 'appeasement,' the one that says it's wrong to capitulate to harried conditions. But I think the 'capitulation' view takes too much for granted. Given that reality is not optimal, and is not likely to be anytime soon, can we do better than we're doing now? It seems like a fair question to me.
It's a hell of a lot more useful than yet another boutique project that won't scale.
