Being a harmless science blogger, I've stayed out of the whole Harvard/Summers/women-versus-men tar pit. (Proof that I don't spend all my time fishing for traffic, as if posts on patent law weren't enough evidence already.) If you want, you can find more discussion of that controversy than you could want on any of the current-affairs blogs. But, still, I was struck by a comment from Virginia Postrel. She's discussing what might be done to increase the female presence in the sciences, given that biological clocks for reproduction work very differently for women and men (i.e., fathering a child at 45 is a lot easier than getting pregnant at that age. Neither Virginia nor I make any claims about the wisdom of doing either one; we're just talking biological feasibility):
"If, however, you spend six years in grad school and another two as a postdoc, you'll be 30 when you get your first tenure-track post--and that's assuming you don't work between college and grad school. I don't have the numbers, but science training is notorious for stretching out the doctoral/postdoc process, in part because the researchers heading labs benefit from having all that cheap, talented help. Female scientists who want kids are in trouble, even assuming they have husbands who'll take on the bulk of family responsibilities."
Fortunately, that long a stint in academia is unusual by chemistry standards, but molecular biology is notorious in just the way she's talking about. I've seen biology postdoctoral positions break up marriages, because the other partner eventually just wanted to finally, finally move on with life. Her suggested remedies?
"So, if a university like Harvard wants to foster the careers of female scientists, this is my advice: Speed up the training process so people get their first professorial jobs as early as possible--ideally, by 25 or 26. Accelerate undergraduate and graduate education; summer breaks are great for students who want to travel or take professional internships, but maybe science students should spend them in school. Penalize senior researchers whose grad students take forever to finish their Ph.D.s. Spend more of those huge endowments on reducing (or eliminating) teaching assistant loads and other distractions from a grad student's own research and training."
I got my first real PhD-level job at 27, after a year's post-doc, but that's a year or so younger than average for organic chemistry. I spent my undergraduate summer breaks doing research internships (of greater and lesser value), but I should make clear to those outside the field that graduate students in the sciences already work all through the summer. When I was in grad school, we watched the law students across the street pack up and leave in the spring while we cranked away in the lab days, nights, weekends, and holidays. I treasure a memo in my files from the chemistry department head, pointing out that the university vacation calendar did not apply to grad students - and he wasn't just talking about summers, of course. Do not, the memo warned, attempt to take all these holidays, things with names like "spring break", even though you may hear people talking about them.
As for Virginia's other prescriptions, I think penalizing slowpoke professors is a great idea. I know that some schools talk about doing this, but I've never seen any of them follow through. I think that the inverse idea, rewarding those research groups with a high percentage of students finishing on time, would be worth looking into as well. There are plenty of groups that could use a better work ethic - not in terms of the number of hours put in, but in terms of making sure that everything the students do is devoted to the great and holy cause of getting the hell out of graduate school. That's something you should do on general principles, man or woman, whether you plan to start a family or not. Grad school is for getting through, not for lingering.
Reducing TA assignments would also help. I know that many professors, if they have enough grant money, try to get their students out of teaching assistant positions as early as the university will let them (I did one year of it, the minimum.) But if you work for someone without as much of the ready cash, you can be TA-ing until your last year, and in an increasingly bitter mood about it, too.
Speeding up graduate education can be done. You don't want to turn out a bunch of unprepared losers, but as far as I can see, the system we have now does that anyway, but often too slowly. It's true that real research projects take time - you're never going to get well-trained chemistry PhDs out the door in two and a half years. But you shouldn't be expecting five and six years out of people as the norm.