Friday, May 2, 2008

T(h)rashing Indian Science

Two articles within a few days of each other, have appeared, bemoaning the state of Indian science. The first is a vitriolic and polemical attack on Indian science (and scientists) in general, titled Science in the Sick Bay. It talks of the "prototypical Indian science institution (which) looks drab and is dingy, with sanitary facilities that rival those in our railway stations", except, curiously enough, those with "en suite" bathrooms! The other is more detailed, serious and dignified, as befitting the stature of the person writing it -- however it is no less gloomy and despairing, presenting a bleak picture of the direction in which Indian science is moving, Neither quality nor quantity. (Note that the latter is on Nature and may need registration, though not payment). At some future date, when I have managed to put some of my chaotic thoughts on paper, I might have some comments on these articles.


Anant said...

I have been wondering if it is really so that our science is doing as badly as all these articles say. Everyone will pull out statistics and say that our share has fallen from n% to m% over the last 2 decades. But were things really that hunky dory, say 3 decades ago? If our scientists were that great at that time, has their work stood the test of time? How much work done in India has entered textbooks? Are there really outstanding classics published based on work in India? In other words, things are perhaps really no worse now than in those days? On the other hand, do we now not have a large cadre of professional and highly trained scientists walking the halls of Institutes designed by outstanding architects, as well as in our less hallowed institutions? Do we not have persons who are as good as any in the world? While I have many reservations about way things are run, I don't really share the gloom. On that happy note maybe we should get back to thinking about writing that next paper...

Rahul Siddharthan said...

RahulB - I'm sure Anita Mehta will absolutely love your comment on "stature".

Anant - of course one cannot easily compare generations. But the question many ask is, when we had no research support to speak of, we produced a G N Ramachandran whose work has entered every undergraduate textbook in the field. Who do we have that is comparable today?

In physics, undergraduate textbooks have hardly changed for half a century, but even in graduate textbooks, the only Indian name I remember encountering was T V Ramakrishnan (and his co-author M Youssouf), for work done in the 1970s.

The only comparable name I can think of today is in computer science (Manindra Agrawal, IIT Kanpur).

I think it is a fair question. We have more institutions, more money, more international connectivity, than ever before. But the quality of the output does not reflect this. Meanwhile, the universities are dying. I don't think this is a coincidence. The universities have to be regarded as the bedrock on which elite institutions are built, but instead our elite institutions are cut off from the roots and have evolved into mutual back-scratching organisations.

Anant said...

Rahul (Basu and Siddharthan):

Yes, it is hard to compare generations. But we did not start the comparison. It has been started by assorted bean counters and science managers, using statistics and whatever else suits them. I agree that the work of GNR is in text books. It seems to be the exception that proves the rule. I am also aware that there are other pieces of work: I am told that the work of the Physical Research Laboratory group in Ahmedabad on the ionosphere or some other sphere is also very important. The old TIFR maths group and their work, etc.. I am not doubting these. In fact, it may seem that I am contradicting myself. But I am not: I am sure if there is enough work done to really probe what is going on, even with the present generation one would find bodies of work of individuals, of groups, of teams which is pretty good. No one has done this work. The bean counters simply start counting gross numbers of publications and then tie themselves into knots. Furthermore, the nature of the enterprise itself is different. But anyway, I am awaiting learned comments from the author of the post...not enough to simply say that some day he will put his chaotic thoughts down...why not put it down right now, this instant...?

Rahul Basu said...
This comment has been removed by the author.
Rahul Basu said...

Well, Anant, it's because while there is much that is probably wrong in Indian science, it's difficult to pinpoint what the factors that are responsible for it, are. I don't want to get into a detailed discussion without some homework. Also remember, we are all from so-called elite institutes with little connection with the universities and how they function. (I have collaborators and I hear only horror stories from them). So one needs to bone up a bit about the world of science beyond our ambit.

However, mind you (not mind it as Rajnikant/Shah Rukh/Dhoni would say) I don't subscribe to the extreme gloomy prognosis that both the articles try to sell.

Anant said...

Rahul: thanks for your mail. The point I was trying to put across is simply is this: are things so much worse than before? Now, I am sure three decades ago there were plenty of horror stories as well. There has been a marked tendency to simply paint a glorious past and stories of how terrible things are now. I don't doubt that there is a lot of work to be done. The trouble is that the very forces responsible in most part of these terrible state of affairs are today decrying that state of affairs. There was once a statement that our Universities are the slums of our academia from someone very high up. I forget who it was. Now having run Universities into the ground, we are now crying about the terrible state of affairs. What is the solution? Build more isolated enclaves and gated communities? I am not sure these are the solutions. In any case, like you I do not subscribe to the gloomy scenario either.

Rahul Siddharthan said...

Anant -- I don't think anyone talks of a glorious past. What they say is, in spite of a wretched past where you had to wait hours for a long distance phone call, weeks for international letters and months for international journals, and where modern laboratory equipment was out of the question, some world-class work was done. Today we have world-class facilities, but where is the world-class work?

To some extent I agree we cannot judge world-class work in its time -- posterity will have to judge. But sometimes the importance of a piece of work is instantly apparent all over the world. As I said, the only Indian example I can think of in recent years is Agrawal's polynomial-time algorithm for determining primeness.

I agree, our science administrators are responsible for the mess today. Also, individual egos contributed to the proliferation of small research institutes and the destruction of the university system (and even the dilution of larger institutes like IISc).

Anant said...

Rahul (Siddharthan): points taken.

Now, the work on primality and polynomial time proof is an exceptional result. It is in the league of the proof of Fermat's last theorem, and therefore eyecatching, and also something that is accessible to the `common' (wo)man. It is obvious that work like that gets the limelight that it does.

But my point is not being taken: unless there has been stocktaking and inventory of the actual work, not in terms of gross publication data, how does one even know?

Finally, there has to be a whole new way about thinking about things. Now in Europe, no one talks about German science, Italian science, French science. It is simply science and it is intensely collaborative and transnational. If there is to be a future for us, we also have to reorient ourselves. One does at an individual level, but how about at the policy level?

Rahul Siddharthan said...

anant -
unless there has been stocktaking and inventory of the actual work, not in terms of gross publication data, how does one even know?

I'm not sure how best to quantify "quality of work", but one measure certainly is "impact factor" (citation count). It would be interesting to see who the most-cited Indians are. Recently I was at a meeting where a certain gentleman described himself as "India's Craig Venter" (he had just interrupted an unrelated meeting for the purpose of introducing the real Craig Venter, who was visiting). I found two things odd: one, Craig Venter is a decidedly anti-establishment character while "India's Craig Venter" is very much part of the establishment. Two, Craig Venter's top-cited papers have several thousand citations (according to Google Scholar), while "India's Craig Venter"'s most-cited papers have barely a hundred, and none of his recent papers (last 10 years) come close even to that number.

What our leading lights lack in impact, they certainly make up in pomposity.

Now in Europe, no one talks about German science, Italian science, French science.

They certainly talk of "European science" versus "American science", and they also talk of "British science" (recent example here). I suspect that talk of "German science", "French science" etc is conducted in those languages.

Rahul Basu said...

Anant/Rahul S

SPIRES has some interesting statistics for HEP particularly what they call the Top Cite Olympics. Take a look at the following

Its called the HEP statistics playgorund.

Anant said...

Dear Rahuls: I was being polemical. I do know that nation states that support science of an on will talk about their own science. [I typed all this earlier but lost it, but I will try again.] If I write a paper with a Frenchman does that make it desi-science of French-science? If you write papers with Americans, what does it make it? I am being pedantic no doubt, but I am trying to make a point that there cannot be simplistic conclusions on such a complex subject. I am aware of the hep olympics, but conclusions can be draw from it? If a post-doc or a student with a US postal address is on a highly cited paper, whether or not he is Algerian or a desi, it will show up as a US score. In other words, only a long term view point can give a worthwhile picture.

In any case, let this discussion go on with vim and vigour. And Rahul (Basu) can be happy with the number of comment here, although the number of commentators is still only three!


AMOK said...

As we know, talent+opportunity=productivity. There is no shortage of talent in India. The emphasis, however, is on technology and applied science for military and economic engines (missiles and autos). Tata has recently "invented" Jaguar and Land Rover. For a more statistical view of this subject and away from (Sir!)Naipaul-like observations see the link below.

The number of comments on this topic seem to draw it away from the ranks of the cranio-elitists. Regrets, Dr. Basu.

Sunil Mukhi said...

I'm extremely surprised that Rahul Sid can only think of T.V. Ramakrishnan and Manindra Agarwal as world-class scientists in India.

If being a "world-class scientist" is measured against objective standards like being invited to speak at international conferences, having one's work well cited globally, being on editorial boards of leading journals etc, then (from whatever I know) Indian science is doing much better than before. Just to cite the example of TIFR, most particle physicists there failed to qualify on any of the above counts in the 1960's and 70's, but most do qualify on one or more counts today.

I know TIFR is special but I'm citing it to make the point that things - within the same institution - are definitely much better today, and not merely "no worse than before" as Anant puts it.

I think it's important we don't rely on anecdotal articles in the press (not to mention aimless, spaced-out articles like that of Anita Mehta) to form our opinions. As an example, the Indian press clearly doesn't know the standing of Ashoke Sen (nor evidently does Rahul Sid, which is less forgivable!). But Ashoke represents an incredibly high peak in Indian science, and this is not built on any past glory - to this day he writes dazzling papers literally on a weekly basis. I don't have the space to justify this statement here but please note that he became an FRS two years *before* T.V. Ramakrishnan despite being over a decade younger than him! (intriguingly Narlikar, another Indian science icon, is not an FRS despite his sterling British connections).

The fact that Ashoke (along with many lesser mortals) can so easily disappear off the map when statements are made like "we have world-class facilities, but where is the world-class work?" makes me feel it is the discussions that are not measuring up to expected standards.

I don't want to be seen as saying more than I am. My point is not that Indian science is doing very well on average, merely that we do ourselves the disfavour of forgetting about its successes - something that in no other country would be done so casually. Frankly the glee with which we rush into these "Indian science is dead" discussions strikes me as fratricidal in nature.

Rahul Siddharthan said...

Sunil -- either you did not read the comments or you deliberately misread them. The mentions of TVR and Agrawal were in the context of Anant's "How much work done in India has entered textbooks?" I mentioned that I saw only TVR's name (for 30-year-old work) in grad textbooks, and of the current crop of work, only Agrawal is an obvious candidate.

I obviously pushed some buttons in omitting Ashoke Sen's name, so let me push some more by clarifying. What I wrote above is "To some extent I agree we cannot judge world-class work in its time -- posterity will have to judge. But sometimes the importance of a piece of work is instantly apparent all over the world..." I think not only in Ashoke Sen's case, but in nearly all of string theory, "posterity will have to judge". Perhaps his name is already in string theory textbooks (I don't know), but will string theory textbooks be around in 50 years time, or will the field seem like those 19th-century theories of the aether?

Rahul Siddharthan said...

Sunil - p.s. yes the phrase "world class" was badly chosen and the comment needed to be read in the context of previous comments. I meant "sufficiently groundbreaking to get into not-too-esoteric textbooks". (I'm sure lots of Indian names are mentioned in sufficiently arcane textbooks.)

Anant said...

Sunil, Rahul(s): Happy to see that the discussion is going on vigorously. For the moment I do not have much to add, except to say that in my humble view science is much more of a profession in India than it ever was in the past. Perhaps the conditions of our seniors was different, and I respect them for the work that they did in their time, under their constraints (no email, no phones, no journals, no preprints, no what ever, walking up hill both ways and so on). But what the heck, I think those that are striving today are doing quite well, perhaps not as well as they could had they been elsewhere, probably better, depending on the cirucmstances. All I can say is that the show goes on. If the policy makers are not content with the state of affairs, they must first do their work thoroughly. As I have kept hammering away, it should not be just on basis of gross publication numbers. It really means nothing. To reiterate, I would be the last to suggest that things cannot be better. I think our brothers and sisters in less endowed Institutes deserve far more facilities than what they have. Do I think there are too many facilities concentrated in too few hands? Yes I do. But all that apart, the show does go on. I will not be writing much more as I plan to learn from what others have to say.

Sunil Mukhi said...

Rahul S., you wrote: "Perhaps his name is already in string theory textbooks (I don't know), but will string theory textbooks be around in 50 years time, or will the field seem like those 19th-century theories of the aether?"

I really wish some (actually many) of my Indian Condensed Matter Physics friends would save themselves the embarrassment of repeatedly peddling a line they read long ago and never bothered to update. It really detracts from my appreciation of their scientific standing.

Let me start by saying that some of my work appears in a textbook on Conformal Field Theory. Will conformal field theory be around in 50 years? That's like asking if phase transitions will be around in 50 years. Why don't you tell me the answer.

Ashoke Sen's work includes a mathematical theorem on the existence of harmonic self-dual two forms on two-monopole moduli space. It was this work which inspired Atiyah to nominate him for the FRS, which he got. Will harmonic two-forms be around in 50 years? I expect so, unless mathematics itself dies an inexplicable death.

Ashoke Sen, as well as Atish Dabholkar, Sumit Das, Avinash Dhar, Gautam Mandal, Sandip Trivedi and Spenta Wadia have important work on the microscopic understanding of black hole entropy from a quantum theory of gravity. This work provides convincing evidence that black holes are a statistical system obeying the known laws of physics, and it was mentioned, for example, in a Scientific American article. Will their work be around in 50 years? Or will gravity, quantum theory and statistical mechanics all disappear, along with black holes and entropy?

My eminently world-class colleague Shiraz Minwalla at TIFR has recent work on the relationship between black hole physics and hydrodynamics. Among other things it suggests new ways to address old problems of hydrodynamics. Will hydrodynamics be around in 50 years? Perhaps not, once all the water on the planet dries up?

Finally - this one is not by Indians in India, but have you heard of the following paper in string theory: "Quantum critical transport, duality, and M-theory" by Herzog, Kovtun, Sachdev and Son? The abstract starts "We consider charge transport properties of 2+1 dimensional conformal field theories at non-zero temperature." Please read the paper, or at least the abstract, and then tell me if you want to stick by your statement about string theory and 19th century theories of aether.

Rahul Siddharthan said...

Sunil: one should distinguish a physical theory from the mathematical formalism (or fallouts thereof). Are you aware that much neoclassical economics owes its origin to the 19th century theories of the aether that I alluded to? (Not everyone has a positive opinion of that, of course.)

The formalism that arose in string theory could well be useful in other areas, and I plead ignorance of much of the work that you mention. As you retracted (on your blog) your claim that Ashoke Sen is ignored by the media, I'm willing to include him in my list. I already said the phrase "world class" was poorly chosen. Blogger, for good or bad, doesn't allow me to edit old comments.

Whether string theory's own claims to explain the universe are valid is still an open question, but precisely because of the useful "fallout" such as the examples you cite, I'm in favour of it continuing to be a topic of research, just as I favour pure mathematics. (In fact, Atiyah is a mathematician and a lot of string theory work has received more awards in mathematics than in physics...) Perhaps string theorists should do a better job of communicating what they're really up to.

Anant said...

Sunil: while you are at it, please also use this space to point out important work done in string theory in places other than TIFR. As you know, blogs are now a powerful medium, and it would be good have a balanced picture. Best regards, Anant

Sunil Mukhi said...

Anant: Aside from the minor fact that I've been talking about Ashoke Sen's work for days, and he's not from TIFR, I do want to make another point clear. What you call "a balanced picture" is exactly what - in the present context - I'm opposed to. It implies that there is limited space for recognition of work done in India and that highlighting one work is somehow at the expense of highlighting another.

I want us to free ourselves of this notion. Let us be additive, not competitive, in this particular context. Please feel free to add examples of world-class work (in any field) that you know about. Please ask anyone you know, and respect, to highlight anything they know about. Let's have string theorists other than from TIFR contribute to this thread, or more generally, to the discussion of what's good in Indian science. The more, the merrier.

Rahul Siddharthan said...

Anant, Sunil - for once I agree with Sunil. And let's not restrict ourselves to string theory please. Even within physics there is much more to talk about, plus there are the other sciences. I'd certainly call some of the biology work in the country world-class (as I said a few times, my use of that phrase above was ill-chosen). Whether it's future textbook stuff is hard to say, but who knows. Perhaps I'll write about it sometime (but not in this thread).

To reiterate something I said in Sunil's blog: I'm thinking of starting an online journal club where scientists can discuss published work in a form accessible to lay people. (The first goal is that it should be accessible to non-specialist scientists, of course: so a string theory paper should be presented in a way that a chemist or biologist or earth scientist will find interesting. And vice versa.) I'm hoping that, if we do a good job, it will be read by a few interested people including science journalists, and that will have a secondary effect. If someone wants to beat me to it in starting such a thing, please go ahead.