Some musings on academic research

Anindya Chatterjee

IIT Kanpur

Comments: (1) These are merely opinions; my colleagues may disagree with me. (2) It is a longish piece. Apologies. (3) I have tinkered with it a bit. This version: September 27, 2015.

Many people are writing many research papers these days. An article I read somewhere, but cannot find now, said the number of papers published annually worldwide has been growing exponentially for 200 years. A mind boggling trend! Perhaps never before in the history of humanity have so many had so much to say with so few to hear them.

Incidentally, for academics like me, papers are counted. Variations across fields are large. On September 9, 2015, a Scopus search for documents in Physical Sciences with the word steel in the title, abstract, or keywords, gave 554076 documents, of which the three most prolific authors have written 414, 360 and 322 documents respectively. A similar search with the words fluid and turbulence in place of steel showed only 41810 documents, wherein the three most prolific authors have written 83, 79 and 69 respectively. Is 83 in fluid turbulence the same as 414 in steel (a factor of 5)? More dangerously for young academics, is 10 in turbulence the same as 50 in steel? Does anybody know? These days, many institute administrations think they do.

Time to introspect.

Question 1.

It is irrelevant that many of us like writing research papers. The puzzle is that society pays us to write them. Why?

The naive view is that these papers supply new knowledge that society uses to make profits or improve the human condition. I find this view naive because I have read many papers. I think the vast majority of academic research leads to no directly visible immediate economic value. If a car company wants to improve its vehicles, I think its money is better spent on in-house research efforts than on funding university research. (Please do not tell the car companies.)

It is common for a PhD student to write complex computer code for, say, nonlinear solid mechanics, only to join industry and use a commercial package, never to use his own code again. Or for a PhD student to make challenging one-of-a-kind experimental measurements, only to then work in a lab where all the test equipment is bought readymade, for rather less unique R&D efforts. The thesis work is not directly used in either case. Who gained what?

The benefit to society is mainly indirect. The training imparted has value for society, although the specific research activity used to train the student is almost irrelevant. The transferable qualities are relevant: general ability, maturity, discipline, persistence, clarity in reporting, etc.

Clearly, to the student, the problem must be new and unfamiliar. The problem could in principle be known to others. But if problems are repeated, copying of PhD theses from other universities would begin in earnest. This is a practical reason for requiring novelty in PhD work.

I suspect that research publications mostly serve as a means of statistical quality control as follows.

The PhD student finishes some work. Then she writes a research paper and sends it to a reputed journal that serves her specialization. The editor of the journal is a respected scholar in the general area, and selects some anonymous reviewers who have the needed expertise. They tell the editor whether the paper is acceptable. This process both maintains the standards of PhD theses and provides community-level awareness within each field so that subsequent PhD students can select defensible problems with better chances of success. Professors guiding these students get a perk: they develop communities where they know each other, meet for conferences, exchange ideas and stories, and are able to better guide new PhD theses.

In this prosaic view it seems unnecessary to write papers unless you either have a research report that needs screening (run of the mill) or a great idea (rare). Nobody really needs to count the total number of papers by a given professor, nor compare the non-comparable like papers on steel versus fluid turbulence.

Unfortunately, the number of papers a professor writes is now a part of her assessment for promotion. She is rewarded for writing many papers and punished for writing too few; the original goal of student training is obscured; there is a direct reward for rushing into new areas where not fully mature work can still be published; and a few professors have even been caught doing startlingly cheap things (like using fake identities to review and accept their own papers).

The older, more idealistic view was that a professor was throwing a ball as high as she could, and there was just an informal wall somewhere in there to check if the throws were sort-of, kind-of, high enough. Now there is a highly nonuniform wall and a ticking clock, and a professor is encouraged to walk along the wall looking for the lowest possible spot, and quickly chuck as many little balls over as possible until somebody comes along and raises the wall there.

Imagine the conversation this whole process implies:

Humble Assistant Professor: I have published the results of my research.

Exalted Promotion Committee: In how many separate documents?

I may laugh, but only because I am a full professor already.

Question 2.

How to choose a research area? An unavoidable question.

Many follow what is hot. Hot means an area where many people are publishing many papers. It seems practical to join the crowd. It is easy to publish, to find reviewers, to get cited, to get research funds, and to seem trendy.

Our national academic research agenda is apparently set by what is popular internationally. Here, internationally mostly means the US, followed by Europe.

But how does the US choose its research areas? Here is my theory.

The main fact is that in the US (and other wealthy societies), a huge population now wants a college education. US universities have beautiful heated and air-conditioned buildings, lovely manicured lawns and so on, all expensive to maintain; and so the cost of education is astronomical. Yet, even wealthy societies are unable to extract immediate visible direct economic value from so many people with such educations. They are willing to pay salaries for only so many engineers and doctors and musicians and literary critics and psychologists and cultural anthropologists, and the rest must take lower paying jobs they are overeducated for.

Do not misunderstand me. Of course widespread education is a good thing! It leads to more orderly people who pay taxes in response to politely worded automated emails (as opposed to much of the Indian public, which only starts moving when a stick-wielding policeman gets within striking range), better hygiene in homes and schools and hospitals, more interesting conversations in millions of living rooms, and a finer appreciation among worried parents of all they may have done wrong while trying to produce well adjusted yet high achieving children.

I suspect that the real cost of a typical college education in the US is not covered by tuition fees, although they may seem high. As a loose example, when you pay 100 dollars for a US motel room, that 100 is about what the room costs (or maybe 85, but surely not 50). With that in mind, if you wanted a big lovely campus to study in, with sports facilities and cafeterias and lawns and everything, perhaps you might agree to pay 100 dollars a day just for infrastructure costs. That is 36500 for the year, and the actual education part has not been paid for yet. So it seems to me that if you are in a beautiful big campus of a private university, being taught by people with excellent academic credentials, and paying 40000 dollars a year in tuition, then you may actually be ahead of the game in terms of costs.

If in fact the real cost of a typical US college education is not covered by tuition, there is a serious problem. If universities raise tuition to cover costs, people will protest that it takes 20 years to repay college loans. Then who will pay? The government, obviously. But big government subsidies to universities cannot be allowed by a strong capitalist ethos. The answer is (drum roll!) research grants.

The naive view is that research grants pay for research. But look at the facts. They actually seem to mostly defray the costs of running the university (call it DCRU). Research grants pay for university overheads off the top (DCRU), summer salaries for professors (DCRU), student research assistant tuition fees (DCRU), stipends for the same (DCRU, if you think of near-insolvent graduate student food bills as a part of their educational expenses), equipment which is only partially used for the research project and then stays in the university for general use (DCRU), and for conference travel for professors and graduate students (a part of their perks and educations, respectively, hence DCRU). Thus, it seems that the cost of so many Americans going to college is indirectly subsidized by their government in the notional form of research grants. No direct subsidies, because we are not socialists!

Unfortunately, every professor now faces pressure to write grant proposals, and the competition is brutal. Of many excellent proposals, only a few get funded. Much faculty time is wasted on unfunded proposals. The government, unwilling to make case by case decisions on so many unrelated proposals, develops a list of focus areas. These focus areas are identified by people who are senior and respected and clever and all but not, I imagine, clairvoyant. Yet, they predict what will be hot; the government funds those topics; funding-hungry professors jump on those topics; followers in other countries start work on those topics some time later, when they notice where the action has moved; and, as with a self-fulfilling prophecy, these areas do get hot for a while.

And so it was biotech some time back, then MEMS, then nanoscience, then multiply hyphenated things like nano-bio-something; and here, in India, with its crumbling roads and dismal primary education and leaky (when not dry) faucets and unmanageable urban trash, we have young bright people, trained in western methods (perhaps even trained in the west), working long, long hours on problems that were recently hot in the US, writing papers without knowing who exactly will read them, constantly watching the scoreboard and getting all stressed out about it. Seems to me like running really hard without knowing where or why.

Imagine asking a young faculty member, recently returned to India after a PhD in the US, what he is interested in. Energy storage devices, or the mechanics of cells, or haptics for surgery, he says, with impressive self-knowledge. That is what he is going to focus on, that is where his passion lies. Off he goes, writes really competent grant proposals, gets money, buys equipment, puts two students on it, and gets really interested in this topic. Enough to really establish his career in that area. Invest-a-decade type of interest.

But the whole thing seems accidental to me, based on random events that led to him getting funding to work on one topic instead of another. In another throw of the dice this same person, with the same self-knowledge, might be as interested in something quite different. Yet, dare I suggest to a young faculty member that his present strong interest is irrelevant, and he should work on something more practical, or something that directly attracts better students? He may point out that I am infringing on his academic freedom (which seems to be his right to be deeply interested in his accidental thesis topic), or ask angrily if I am suggesting he should compromise his academic passion for the mundane goal of pleasing students.

I am slightly troubled by an alternative thought, though. The apparent orchestration of many individual research agendas, through deciding on what areas get funded, may have another purpose as well. Very large crowds without some kind of common agenda or external discipline tend to descend into chaos. Perhaps the business of funding one research area, then another, and then another, achieves some level of harmony in place of chaos. But it seems somehow similar to the slash and burn type of agriculture practiced by some primitive nomadic tribal communities. To the extent the analogy holds, what does that say about an academic who refuses to follow the trend and walks in the forest as per his own whims and wishes? What is a solitary nomad without a tribe?

Academic freedom is important. I certainly do not want to give up mine. But what is the true basis for our academic choices?

Question 3.

Is happiness possible in academia?

I believe modern Indian academia offers excellent opportunities for happiness. The internet is a leveler. Our salaries seem decent to me. Students are available in good numbers. Travel is easier. Research project funding is there to be had (though I am not sure how long that will last).

All I really need is motivated students (the department has kindly given me a bit of space, of course: that is essential).

A couple of mundane thoughts:

It helps if you put keywords and headings and pictures into each thesis so that HR people on interview panels can see that the student is somewhat oriented toward practical things. I knew a person in the US who studied a frictional oscillator, but early in his thesis he explained why it was relevant to trains; he got three jobs, all with railroad companies. In India, I had a student who wrote equations for the dynamics of a sliding box that has an unbalanced rotor in it, wherein he stated that it was relevant to washing machines; he got a job with a washing machine manufacturer, and another with a company that builds equipment containing rotors. Finally, I had yet another student who worked on a really interesting (to me) academic problem, relevant to vibration damping; but we did not clearly state how it was relevant to damping. This student found it hard to get a job. Nobody wanted to talk to him after they heard his thesis topic: it meant nothing to them. That is the world of employment outside our campuses, and any student or professor who ignores it is impractical.

We must pay attention to the employability of our students. As our students do well in their professions, so do we in ours. As we build their careers, they build ours. If students are interested and working hard, enough papers get published. I fear that our preoccupation with our individual scoreboards (papers, projects, whatever) is blinding us to this basic issue. We may pay dearly for this. We may think we exist to do research, but to our students we really sell career outcomes. Students are our customers. No customer, no business.

 << back