Sunday, January 28, 2007

Notes from a Paris maths department

Before I came to Paris, I was very, very excited. It wasn't just the thought of living in a great city like Paris, it was also the chance to talk to and work with some of the best mathematicians in the world. Paris is one of the great centres of the mathematics world - along with Boston and Cambridge/London and Tokyo/Kyoto. Historically, Germany was the place but post WWII, their influence in the maths world has declined.

The average American university is always very busy. The professors and postdocs have heavy teaching loads, grading and administrative tasks along with the research they do. While most professors are very informal and receptive to talking, you can see the stress on their faces and somehow that attitude trickles down to the postdocs and phd students. Professors in the US also move constantly till they get tenure and it's normal for a tenured professor to have been in 3-4 different universities till they settle down in one place. Even after that it's pretty common for them to move around. In between juggling teaching, grading, administration, travelling for conferences and research it's nowhere close to the cushy, comfortable life non-academics consider it to be - even though mathematicians don't need anything more than a pen and paper to do their work. My experience in Germany wasn't too different, though considering how obsessed Germans are with rules, even a beach bum would look busy.

When I came here I wasn't too sure of what I was supposed to do. My host had said I should come in 2-3 times a week, try talking to some other people and give a talk while I'm here. I thought he was trying to understate things and make things sound laidback. My office window faces the physics department and I can look into a lab and it's a bit scary how hard they work. They're all in before I come in and there till I leave. Apart from a brief lunch break they all seem to sit there plugging away with their machines and computers.

The maths dept is on a different wavelength though.

It's now the end of January, and I still hear people saying "Bonne Annee (Happy new year)" in the corridors. That says a lot about how often they see each other. Lunch is an incredibly relaxed affair. There's no regular time or group that goes for lunch. There are people who arrive daily around lunch time and then proceed for a relaxed, leisurely lunch which takes at least an hour. There's very little stress I see, and after 5 there aren't too many people around in the dept. It's very common to see a bunch of professors, sitting in the common room sipping coffee and talking to each other - not just about maths. There was a postdoc I was trying to meet but was told that he hasn't been spotted for the last 2-3 months, and nobody seemed to think it unusual. One postdoc comes in each day only around lunchtime and pushes off around 5 to play with his string quartet.

Seminar talks are interesting. In the US, there's room for spillover time, but talks generally tend to have a fixed length, and every talk I gave over the last year tended to be of the same duration. In Germany, the faculty looked suprised if I didn't send an email on time about the exact time (5:00 or 5:10), location and length of the lecture. Out here, there's a regular time but no time limit. Talks range from 30 mins to almost 3 hours and when I asked about how long I should speak for next month, they all looked surprised.

The department in terms of people is huge for a maths department - about 40-50 fulltime tenured faculty members. Most of have done outstanding research and when I came here, I expected to see an incredibly busy and tense place. I couldn't figure out how these guys could do research of the highest quality, have all the time in the world to do whatever they feel like and also keep travelling for all kinds of conferences. The answer is simple.


CNRS is a national organisation which funds research in the basic sciences. At a young age (you have to be under 30 I think), you can apply to CNRS for a position. If you get it, for the rest of your life you're a salaried government employee with all possible benefits and have no obligations to teach. Assuming you do good research you keep getting promoted, but it's a permanent position. It doesn't pay well, but after 2 months here I've realised you don't need too much to have a comfortable life here. It's quite amazing, how happy and relaxed the whole CNRS lot is. It's obviously not easy to get but an equivalent position in the US would involve competition of a different magnitude. The salaries in a comparable position in the US are also of a different magnitude though.

The French though have a nice system. Rather than funding only superstars, they fund a much larger bunch of potential superstars. For the rest of their life, these people are free to do whatever they want. They can work from home, cafes or not work at all. But the quality of research they churn out is exceptional. Mathematics is one area where capitalism and free markets don't have the edge.

Last week I went to the Institut Henri Poincare to attend some lectures given by people who'd been awarded the National Medal of Science. The speaker I was interested in listening was sitting in the audience with 4-5 of his friends. I recognized all of them as they'd come to Boston last year for a special program, and they're all bigshots. While in Boston, they all sat in a group and behaved like kids, talking and giggling, irrespective of who was lecturing. Their behaviour wasn't too different in Paris, except one of them got up, accepted a prestigious award and went on to give a great lecture. That was followed by some of the best food and wine that I've ever had, and they all sat around again just chatting away and sipping their wine and nibbling on their food.

The next time I see someone sitting in a cafe staring into blank space, I'd be tempted to ask him if he's a CNRS person. It's possible he could be a bum, but it's also possible he could be the next Sartre or Serre.


Tabula Rasa said...

sounds yummy. the only question i had was -- don't the physicists get cnrs support?

bandafbab said...

I'm sure they do, but I suppose the experimental ones need to be in their labs to do their research. The theoretical ones probably live the same way as the math guys do.

Bea said...

oh man that sounds amazing! Vive la France!

Anonymous said...

loved reading this post of yours. have one addition - free markets and fundamental research in any field do not mix well. if it were upto the corportations, we all would have only learnt the tools to do our jobs well. In US almost all universities, including private ones, get massive grants from the government and private donors for their research...

Anonymous said...

Dude, what else would people do after a PhD in Math?

bandafbab said...

The US universites get massive grants for fundamental research, but there is a link between the funding and the ranking of the univ or dept. Getting a grant, a superstar prof, good students is based on competition like any free market. A place like MIT is a great place for fundamental research, but if you ever meet a prof, postdoc or student from there, you'll realise how much of a corporation it actually resembles - in fact they refer to it as a corporation.

Rahul said...

nice post.

tr - yes the CNRS supports all fields, obviously. And theoretical physicists may not be terribly different from mathematicians, but experimental physicists have more constraints.

It's pretty normal for people to come in close to lunchtime, but that's not a measure of the intensity of the research. Students there take 3 years flat to finish a thesis, compared to 5+ in the US, but they (at least, the ones I knew) do a phenomenal amount of work in those three years, easily the equal of any 5-year thesis I've known. The discussions with their advisors are very intense, and can go on for 3-5 hours (lunch often consisted of a sandwich in the office). And they find the time to have fun or pursue other interests.

But it has to be said that CNRS is a huge organisation and not all labs may be of the same high quality.

bandafbab said...


Phd's in the US take longer because those 5 years include about 2 years of coursework equivalent to a master's or mphil or the diploma that French or German students do. The actual amount of time one spends on a thesis (out of say a 5 yr program) is about 3 years.

I knew a few European students who finished their phd in the US in 2-3 years, because they'd already done the coursework requirements in their undergraduate/diploma studies in Europe.

The US system is very flexible though and because of that flexibility, almost anyone can join a phd program somewhere or the other. A lot of people also drop out or take inordinately long. In France and Germany, you won't be allowed to join a phd program unless you've done the right kind of degree or diploma prior to that. That would have meant someone like me wouldn't have had a hope of doing a Phd.

Rahul said...

Well, another measure is how old are they when they finish their Ph.D. Typically it's about 25 in France (and Germany and UK, I believe) and it's rarely less than 27-28 in the US (often 30+). Which also accounts for why there's the age limit for 30 for CNRS jobs -- such a limit would be very difficult for most people in the US (or India) when you add the postdoc years.

I remember looking at grad school brochures when I was an undergrad, and the median time for experimental particle physics at Caltech was 9 years.

But yes, the US is more flexible if you're going off the beaten track, so to speak.

bandafbab said...

The reason most people in Europe finish their phd at a younger age is because of their undergraduate degree or diploma. As in India, they have to choose their major the moment they finish high school, and then take courses only relevant to their major. So by the time they finish their phd, they're about 25.

In the US, the average undergrad doesn't do as much focussed coursework and I knew quite a few American students, who chose math as their major 2-3 years into their undergrad degree. But in math, it's very common to come across talented and motivated students, who do graduate coursework as undergrads, and finish around the time they're 25. The average student takes longer, but the motivated ones can finish pretty quickly.

Age is unfortunately a barrier in Europe, where you're assumed to be washed up if you haven't finished your phd while you're young. 2 friends of mine in Vienna can probably testify to that because they were around 30 when they finished their phds, and hence treated badly.

But then, CNRS is mainly for French students and researchers, so if you're from within the French system, they assume you haven't taken too long.

I guess there are benefits in both systems, but the European system lacks flexibility, and I'm happy to be here as a postdoc rather than as a student.