From so simple a beginning…

Balancing Life has an insightful post on why a PhD., that seemingly venerated degree takes so long. Among the reasons he cites are the nature of the advisor, and also the whole framework of the PhD. in US graduate schools in America. All points well taken.

But one question I really have is, why is a PhD. necessary? My advisor is surely going to call me a back stabber when I am saying this, but after ‘n’ years in graduate school (where n is an arbitrary number greater than zero), I have come to one conclusion; the PhD. is surely overrated. And I am not saying this because it’s easy. In fact, one truth about almost any PhD. is that it builds character, and can be a sobering, highly educational in many respects, and most importantly, highly frustrating experience. That always teaches you a lot. If there’s one drink you had in your life that was justified, it may be that graduation champagne bottle.

But the problem is this. Think about the name of the degree- “Doctor of Philosophy”. Now, one of the definitions of Philosophy pertains to how well you have a general “feel” for your subject. This feel does not come with simple skill, although that is certainly a prerequisite. Just like you cannot be a good cook simply because you know how to grate, mash, sautee, and sprinkle exceedingly well, so can you not really get a feel for a discipline simply by learning how to do routine manipulations well. I am not underestimating the importance of routine, menial necessary work here; it’s an essential part of what physicist Freeman Dyson calls the “craftsmanship of science”. But craftsmen don’t necessarily become creators. What is important for becoming a true creator is real insight into the subject.

Every subject has its critical elements of insight and its own unique philosophy. For example, if you want to be a historian, you have to realise among other things that testimonies, memoirs, and documentation of events don’t constitute history themselves. They are the raw material that history is made up of, and it is your job to weave them into a coherent history. If you aspire to be a mathematician, you must have a real inner grasp of what the whole idea of “proof” is. As I once read in the preface of a book on differential equations, understanding a proof is not simply understanding the logic of the steps, but actually internalising the logic.

In any case, becoming a PhD. is surely something much more than learning how to skillfully learn the manipulations of the laboratory, or the application of a theoretical algorithm to a problem. And yet, most PhD.s, especially in the applied sciences, step out of graduate school after four or five years doing exactly that. They are excellent technicians, but often not much more. Do you really need a PhD. to become an excellent technician? In my opinion, you can take a good masters student or even a bachelors degree student, and train him to become an excellent technician in one year. If you do this, after five years, that technician will be an even better technician that the freshly minted PhD. Now consider how many resources have been spent for five years in moulding this technician. The graduate school has spent a stipend on him or her. The advisor has spent resources on him. But most importantly, consider the student himself; after five years, he is half as good as a masters degree student who already trained on the job for the past five years. Of course this does not happen all the time, but there is a good number of students who would face this situation.

Another problem with the PhD. is that it essentially weds you to one advisor for five years. But this wedding can be a nightmare compared to the usual not-so-predictable one. It might be a blessing (as it is in my case) or a woeful curse. Being stuck with a slave driving advisor is likely going to kill your enthusiasm and inspiration, sometimes for good. Students commiting suicide may be an extreme case, but working for a slave driving advisor leaves you exhausted by the time you graduate, a time when you should actually be raring to launch your career.

Tying you to one advisor has other problems of its own, especially for bright students. Often, you need to work on the project your advisor is working on, even if you don’t want to. For especially creative students, their advisors may sometimes give them freedom to choose a project, but since the advisor himself is constrained for funds, he himself might be forced to work on certain projects. Naturally, the student can pick what advisor he wants beforehand, but I have heard of advisors’ and students’ fortunes changing overnight in the middle of projects. Some have been asked to wrap up and leave in a few months. Others at the least have been consigned to work as TAs until their graduate, something that sucks away precious time from research. One way or another, there is a risk working on a PhD. for five years. Even though the majority of students make it through, bad things do happen to good people, and hearing about such things is not as rare as one might think.

Just as a PhD. schedule may hamper a bright student, it also may hamper a bright advisor. Academic professors are quite aware of the fact that they would need to supervise PhD. students, but some of them don’t want to keep on feeding a student for five years. They want to work on a project for one or two years, then move on. A PhD. which essentially would usually involve solving a single problem, may not interest every academic scientist as much. On the other hand, and this is significant, many academic professors especially in the experimental sciences actually are hungry for PhD. students, because they want cheap and hard-working labour for doing technician type work in the lab, and this again goes back to the previous point about technician-PhD.s I was making. The advisors want to publish, and part of successful publishing is throwing the largest number of PhD. students and postdocs at a problem. At the end of the whole story, the advisor has five papers, the PhD. students have their names on it, and there they are again, a bunch of skilled technicians, but not necessarily original problem solvers.

Now, one can say that that’s just fine. After all, society needs more technicians than original problem solvers, and there’s of course no ignominy in being a technician. As I mentioned before, there would be no science without technicians. But the question is, do those technicians really need a PhD.? If a company team leader hires five good masters degree students to do such kind of work, he can train them very well in five years. They would still have their names on those papers, but would not have spent five years getting a PhD. in good technician work.

And that’s probably the major problem, that goes beyond mere cost-benefit analyses of advisor benefits and student costs. We live largely in a system where the PhD. is highly prized, and where advisors need to publish or perish. Many jobs in industry need a PhD., or even if they would allow a masters degree student in, the pay would be quite a deterrent. Think of how many good students would lead more technically and financially successful lives after ten years if they had been picked up straight after their masters degree. But the system does not largely allow that, and because of the way the entire system is set up, the students themselves end up thinking that they would become original scientists with a PhD. It’s only after they have spent ten years in industry that they realise their possible folly, when they see the odd masters or bachelors degree technician doing the same type of work much better than they can, because by the time they started their job, he had already been on it for four or five years. I have heard this many times, about PhD.s constantly finding themselves working with technicians who are better than they are. Of course, they are supposed to be better than those technicians because they are supposed to be the thinkers, but that’s the point, that it unfortunately does not often work that way.

What I have said is valid for applied science and industry, not to academic jobs. But in case of academic jobs, this is really a more trivial issue. That’s because I am not arguing against a PhD. for the few truly original thinkers who are going to blaze trails in academic science. Their PhD. is justified. The real argument is that they are few, and many of the students who do their PhD. are learning how to becoming good technicians, who are not going to go into academics (or rather, should not). They are important, no doubt about it, but they don’t need a PhD. for doing what they do. Also, a Phd. is a character building and rewarding experience, there’s also no doubting that. But when it comes to applying science for practical benefits, the system needs to do a good analysis of the inputs and the gains. The students, their advisors, employers, and society, all need to understand this. Instead of rating the degree, rate the requirement, and the human beings.