Tuesday, March 07, 2006


It's been about 3 years since I started off as a PhD. student, and by this time, I have become as they say, a sadder but wiser man. Even the perpetual reading about science and scientists that I indulged in ever since a kid is no match for what the flavour of scientific research actually tastes like. So here is a distillation of some common lessons that I learnt through rather uncommon and meandering, not to mention excruciating, ways. More will be forthcoming, of course, as I suffer more:

1. Research involves ideas, not answers: Probably the hardest thing to come to terms with. Unfortunately, the way we are taught science in school and college is as a set of results, theorems and laws. Start doing actual research, and one quickly realises that about ninety percent (and I am being highly optimistic here) of research is clerical work, donkey work, monkey work, whatever you would like to call it. It is the last ten percent that scientists usually can die for. But most of what is routinely done is a far cry from the world of cut and dried facts that are encountered in textbooks. The average scientist or student can well spend spend most of his or her time in going down blind alleys, solving mundane but intractable problems that seem to have sprung up out of the blue, troubleshooting errors that don't even seem related to what you are doing, and most excruciatingly, getting unexpectedly stranded at the very beginning of a project for ages so that it seems that you are never ever going to progress to the juicy, creative part. In fact, this is the single most important situation that drove me to tears; getting stuck up with something that I least expected, that is the most boring yet essential part of the project, and that by itself is anything but creative work as such. You tend to lose all hope if the beginning is where you seem to be stuck till the very end. And yet, you have to endure.

In a nutshell then, if you don't enjoy the doing much more than the fruits, which would be rare if they exist in the first place, research is not for you. If you think that you want to do research to build a better washing machine or mouse trap, or to mix more nutritious cereal, or to find quick fixes for practical problems plaguing humanity, think again, at least if you are thinking about academic research. That might never happen possibly till you are halfway through your career, and possibly till your grandson is halfway through his. Moreover, since science has become a highly collaborative effort these days, it is very rarely that one gets to sample a Eureka Moment, when he can revel in the ecstasy of an idea that is his and his alone. Enjoy the labours more than the fruit then, and don't expect practical results unfolding daily before your eyes. Throw in an infinite reservoir of patience and tolerance, and you could be well on your way to becoming a scientist. Otherwise, GE, IBM (Google?) and Lucent Technologies always beckon you, but most of what the public perceives of these companies is definitely not the kind of research I am talking about here. So that's a different ball game you want to be involved in then.

On a side note, there was a time when high quality academic style research was being pursued in these corporations. Note the bonanza of physics Nobel prizes gathered by IBM in the 80s for example. But predictably, as companies became more enamoured with stock holders than with the fruits of pure research, so have the standards of pure research in these companies declined over the years, and in fact that's a rift in American science that is being vigorously discussed today.

One key fact that today's politicians and administrators should keep on reminding themselves about, is that it is a hard and proven fact that most research that is hazy and improbable in its time usually turns out to have practical consequences, and sometimes enormous ones, in the future. The atomic bomb, microwave devices, semiconductors, medicines, MRI, lasers, and genetic engineering are but a few examples of the kind of research that started off as mere curiosity in the ways of the natural world, and led to multibillion dollar practical technologies.

2. It's hard to know what's important: James Bryant Conant holds the distinction of being one of America's top notch organic chemists, president of Harvard, and one of the leaders of the Manhattan Project...all this being possible in the 1930s of course. When his student Frank Westheimer, again a Nobel calibre chemist, went to him with an idea for a research project he had, he was told that if he was successful in the project, he would be a "footnote to a footnote" in the history of science. While Westheimer did turn out to be much more than a footnote in his career, the message here is clear and has been enumerated by a number of successful researchers- 'Work on important stuff'.
While it's easy to inculcate this Olympian ideal in your mind, it's only when you start doing research that you realise with a lot of consternation and ask the question- 'What's the important stuff??'. Usually, your advisor would have an idea about what's important. But more often than not, you may also land up with some speculative project that, IF successful, MAY turn out to be quite important. But science and progress being what they are, it's naive to expect that all these ifs and mays would materialize and especially till the end of your PhD. The compensation for this uncertainty, as I can testify to a minor extent, is the joy of discovery, no matter how trivial, and the excitement about the future that is compunded with the uncertainty. Also is the compensatory feeling that what you are doing is a part, no matter how small, of a grand enterprise that will bear frution someday. Moreover, echoing good old Tom's words, if you survive the tears and the frustration, maybe you too can beam with triumph someday and say "I have not failed...I have found ten thousand ways that do not work". But try telling that to your PhD. committee.

In my own experience then, most of the work I have done until now has involved mainly groping in the dark, massive amounts of clerical work that never end, and frustration heaped on top of more frustration. Why then, do we do this? Ask why people climb mountains, collect insects in the Brazilian rainforest, compose poetry, wait for eternity to meet their loved ones, arduously spend a Sunday in baking that exotic dish, and prove three hundred year old equations with no practical significance whatsoever, and we see the answer. It's the joy of the discovery that counts. But like other contrasting qualities, it's value is felt only when you go through its other wicked half- pain. And once in a while, yes, we do end up discovering a new kind of microwave oven too...


Anonymous Anonymous said...

"Try telling that to your thesis committee". Very apt. In general, try telling that to the world.
I guess some of the problem, as you rightly point out, exists because of the way academic teaching is structured. But even then, you can get a feeling for this when you solve problems (in a taught course) that are not based solely on the material taught in the class. Most of the times, most of us have no clue as to how to do those, because we think in terms of what we learn. There is a crucial difference though: in research, even a problem seemingly well defined on the basis of well-known methods may not yield results one would expect. But then many anomalies lead to interesting results as well. So there are probably as many exceptions as there are rules.
There are great debates here: Crick would have us believe that we should not bother with details too much but focus on the big problems. And so would Dick Hamming here:

But then, there are also those who devoted their lifetimes to detailed study of some phenomenon and understood it better than anyone else. And some of them were rewarded with Nobel prizes as well. There are other examples as well. People who follow traditional scientific methods (e.g. Helmholtz) vs those who rebel (e.g. Einstein) and then there are people who do many things in many ways. So, a posteriori, many complementary approaches to science work and produce "results".
So, in my experience, the question of what to believe in science is really tough to answer. One would be tempted to say, believe what your ideas tell you or what the data tells you. But then, there is no a priori or more commonly, a posteriori reason to believe that you are right or to assume that the data was collected according to an experiment that was optimally designed. I should probably sign off here, quoting Weinberg: "In science, you do not know the right question to ask until you are very close to the answer". Well, even that is just one opinion!

7:01 PM  
Blogger Ashutosh said...

Aniket: Yes, I have read Dick Hamming's views on this. I found them to be nice and engaging, but a little too cut and dried. As you say, it's not always possible to make a definite judgement about whether to keep on looking at the big picture or to get bogged down in details. Of course, most of the time even when you don't want to, you do get bogged down in details. Thomas Kuhn's model seems to work here too; sometimes, the time becomes ripe for enough details to accrue to make a quantum transition into a new paradigm. In general, the safe thing is to know WHICH details matter, which is again difficult to know. The Weinberg quote you mentioned does seem to be apt. Thanks for stopping by!

10:14 AM  
Blogger Hirak said...

Work on stuff that is important . Hah! That's like telling people to "Buy low and sell high".

* *
I rule that I have learnt in graduate school is:
"Avoid fishing experiments"

Which means that explore by all means, but every experiment has to have definite purpose and end in mind. According Whiteside, you should have the figures for your paper planned even before you even begin. This helps focus on the aims and design before too much time is spent meandering.
eg: Viagra was discovered while scientists at Pfizer were looking for a drug to cure heart trouble.
However, like all the "Golden Rules" in graduate school, it is easy in theory and hard in practice.

1:25 PM  
Blogger Kapilmuni said...

Looks like the corporatisation of academia is complete. Muahahahaha.

Much of what you said applies to most jobs on the planet - 90% drudgery 10% "what-I-thought-would-give-me-job-satisfaction-when-I-joined".

By the way how can "re-search", searching the same thing again and again lead to discovering something new?

3:05 AM  

Post a Comment

<< Home