Friday, September 25, 2009

Looking back at Perimeter Institute

As you probably read already on Peter Woit's blog, the current issue of Nature has an article about Perimeter Institute, titled "The edge of physics," as well as what is supposed to be a review on Howard Burton's book "First Principles" by Joao Magueijo. You can read my review here.

Unfortunately, both articles are access restricted so the majority of our readers won't be able to look at them. Let me briefly summarize that "The edge of physics" is a well-written brief description of how PI came to be, and what it is today. It brings forward a healthy skepticism on some issues. For example that "Lazaridis has had an unusually strong hand in the management of Perimeter," to which the new director Neil Turok comments that "the make up of the board helps give the place a risk-taking spirit that is more in keeping with a Silicon Valley start-up than an academic venture."

There is some truth to both. Focusing power usually helps things run faster and doesn't make it necessary to please everybody. On the other hand, it's likely to piss off people who believe that their opinion is worth listening to. Unfortunately, most people with a degree fall into this category. As you can guess, the Power-of-Mike issue is frequently discussed among newcomers, and a point usually not addressed in public. It is thus interesting to find it mentioned in this article. In practice however, that power structure has no consequences unless you're involved into the highest level of administration, or you're a person of principle who cares about the basics of the system. Idealistically, I find it an uncomfortable position, in particular if the mission statements babble something about a "flat hierarchy." Practically, it works pretty well, probably because Lazaridis is a smart man.

The article also tells an anecdote in which Nima Arkani-Hamed and Freddy Cachazo work till 4am before they collapse, which, probably unintentionally cynically, is summarized as "the kind of effort that Turok wants: undirected, unconventional, ambitious." It's not that I have never worked till 4am, but people who do this frequently don't display ambition but a miserable time management combined with an unhealthy dose of masochism. You can blame that on my European upbringing, but I believe that time to relax is an essential ingredient to sustainable creative work.

To move on to the second article, Joao Magueijo's review of Howard Burtons' book isn't much of a review, but rather an expression of his opinion about the change PI has undergone since its startup. And it is not a positive opinion. Something went wrong along the way, is what Joao writes "the sought utopia had become a dystopia." He criticizes Howard for being an "insecure country cousin awed by the sophistication of established scientists and their fancy dinner parties."

Well. I haven't been at PI in its early days, but I've had the opportunity to follow its development during the past three years. It is of course true that PI has changed, and it is still changing. You can't run an Institute with 100 people as you run a place with 10, and PI is supposed to grow to more than twice its current size, both in facilities as well as researchers.

Part of the change growth brings is in administration. You'll need some sort of policies and procedures. You'll need some way to efficiently get information to where it needs to be, to coordinate efforts. You'll set up meetings, and committees, circulate drafts of guidelines, and discuss for several hours who is supposed to cleans the coffee mugs.

Another part of the change is social. The more people are at a place, the more difficult it becomes to encourage interdisciplinary exchange. At some point, researchers will begin to cluster into groups according to their prime interests. This has at PI reflected in the formal introduction of research groups like you find at any other physics department. I personally find this very disappointing and an (unnecessary) step into the wrong direction. In contrast, the Santa Fe Institute for example has chosen to keep its research staff deliberately small to avoid exactly this falling apart. When I arrived in Waterloo in 2006, PI had just about reached the size where it was basically impossible for everybody to know everybody. This represents quite a dramatic change from what must have been one large family to an increasingly larger group of researchers who just happen to share the same employer.

But the most serious part of the change is one in spirit. And that is what I think Joao is mostly criticizing. There is an inherent and unsolvable tension between trying to create an institute that is "different," and trying to create an institute that "competes with other top-institutes." You either play according to everybody else's rules and adjust, or you give them the finger, make your own rules, and accept that you will be regarded with skepticism. Unfortunately, I have heard that tension being denied repeatedly by various people at PI. I think it is possible to find a balance between both, the risky quirky and weird stuff with a high nonsense factor, and the established mainstream research programs that are promising to pursue. In most places, the emphasis is on the latter. PI was meant to have the emphasis on the former, but the trend I have witnessed is one towards adjustment with the heatbath. That is the natural thing to happen if new people come in who bring their expectations, their opinions, and their strategies from elsewhere. Especially if these people sit on temporary contracts and know they will be thrown back into the heatbath. Working against this trend requires conscious effort. And that effort hasn't been made.

Joao Magueijo btw left PI briefly after I arrived. I recall being disappointed because it was good to have somebody around who had worked on the phenomenology of quantum gravity. Not to mention that he's a cool guy. Meanwhile, PI has a new director, and though I didn't have much overlap with Neil Turok, my impression is that he is full of energy and eager to lead PI into a new phase, the recently launched PSI program is part of that. Yet where PI's change will lead, I don't know.

My experience with PI has been very good. I found it to be a welcoming and openminded place where, best of all, I could just do what I wanted. German as I am, the disorganization and maladministration I encountered was utterly frustrating, but then this sort of dysfunction isn't specific to PI. And since the limousine transfers from the airport are mentioned in the Nature article, unless it falls under moving expenses you have to pay them from your travel grant.

In any case, I wish PI the best luck. And I hope that the place remains truthful to its original goals.

19 comments:

Tkk said...

Managing a research institution, especially one with PI mission, is difficult. It is not like a business organization where there are lots of knowledge from business schools and lots of experienced people to consult.

The whole point of business is focused effort under tight management to produce a highly structured success from products and services.

But the whole point of PI is the exact opposite. PI's output is 'black swans', the exact opposite of what just about all other organizations want.

Offering constructive advises is far more responsible than offering criticisms.

Matt Leifer said...

You think that PI had bad organization and maladministration whilst you were there? Ha, you should have been there in 2004!

Bee said...

Hi Tkk,

I completely agree. You can't one the one hand ask for people who are independent, bring a rebellious spirit, think outside the box, and constantly try to improve things - and on the other hand expect them to accept a top down management and keep their mouth shut.

Constructive advice has actually two requirements. For one, you need somebody who isn't afraid to voice it. (Is what I referred as the asshole in my little group theory.) But as important as that, you need an organization that embraces this criticism and works with it. It is of course much more convenient to reach decisions only with people who are likely to agree. But that way you don't reach very good decisions. Surowiecki btw also addressed these questions of intelligent decision making in his book "The Wisdom of Crowds". Best,

B.

Bee said...

Hi Matt,

Good to see you around :-) What is funny about maladministration is if it gets really bad, it becomes less frustrating because you stop expecting it to work. There are whole countries that are good examples for that. The problem is if you think it makes some sense and rely on it to work but then the outcome is just nonsense and a waste of time. Best,

B.

Tkk said...

My read is Joao has an axe to grind and he used his critique of Burton book to swing the axe on PI under Turok. He arrived at PI during the early phase. The place was small, unstructured, exciting, everybody does things their way. This fits very well with his personality. He is, after all, a man who makes a living selling the snake oil of Faster than the Speed of Light! Trouble is, the publishing side is decidedly underwhelming. Instead of being creative and scientific, he became a bit of pain in the ass. His visitor status permits a face-saving exit.

Back to management. As any organization grows, and Lazardis did say Grow, one faces the classic organization challenge of freedom and control. When you reach a hundred people, planning and directing must grow or the free-wheeling become useless chaos. PI has now entered this stage, which will only get more challenging.

With size comes alienation of of the original free-wheeling members, who won't help things because they usually feel 'superior' as 'founders'. Many soon leave. This is unavoidable. The task of managing growth is to replace early mavericks with a larger number of disciplined people who are just as creative and determined. So some works to 4am - but that's their problem.

It is important for executive to clearly spell out the 'smell' of the place especially to potential newcomers. So many postdocs from all over the world, differing in culture, language, expectations and personality. Many will find it hard to fit. They waste their first year. Even a world-traveling German Bee got bent a little out of shape.

New arrivals should understand either they learn to fit and participate else try another place. Management must create the culture - the most difficult part, that's why they got paid the big dollars. Then clearly set directions, goals, expectations. That culture should avoid any sense and reality of micro-managing. Provide excellent technical and administrative support - but make their use mostly optional. The faculty is more than knowledge and experience; they are the day-to-day face of the research culture.

The only way to get a bunch of cats with PhDs to yield black swans is to leave them alone but come down hard once a year.

Bee said...

Hi Tkk,

I basically agree with you, just that I don't think it will work if "executives" or "the management" provides what I called spirit and what you call the "smell." Generally my impression is that scientists won't let anybody who is not of their kind tell them what to do (even though this would sometimes be a good idea). I thus think this can only be done by those members of the place who can provide some continuity, ie faculty. I am not sure how much of a difference the director actually makes to that end, especially if he is a new arrival too. Best,

B.

Bee said...

PS: Regarding the world-travelling Bee, the more I get around the more difficult I find it to fit in anywhere, even my home country.

Phil Warnell said...

Hi Bee,

“PS: Regarding the world-travelling Bee, the more I get around the more difficult I find it to fit in anywhere, even my home country.”

Perhaps this is because in the process you are developing the perspective of a globalist and as of yet there exists no place that represents being your home. I find myself much in the same position, only in this case I attribute it more to my journey through time, rather than space. So perhaps they not being discernable holds more relevance then simply in respect to the physical, which I suggest with no pun intended.

Best,

Phil

Phil Warnell said...

Hi Bee,

I think there exists a paradox in the PI vision, being that on one hand they would like to have all of their people be mavericks in the sense of their research direction and yet collaborative in their social perspective. Hence you have two different models for success. the forner based on individual strengths and the latter reliant on those of a collective. To tell you the truth, if they can ever make it work there may be some hope for humanity after all; again with no pun intended.

Best,

Phil

Bee said...

Hi Phil,

I don't think there's a paradox. A monoculture of either type of research style is something to avoid. There's no reason to think "mavericks" are more able scientists than "sociable people," but you don't want to exclude either, for it's not a criterion relevant for their scientific abilities. Frankly, in my experience literally everybody who succeeds in science possesses both qualities: individual strength and the ability to draw on colleagues knowledge. Best,

B.

Phil Warnell said...

Hi Bee,

I’m envisioning Einstein and Godel with the first as being socially astute and the latter the opposite. Einstein said of his later years that the thing he enjoyed most about being at the Princeton Institute for Advanced Studies was no longer his research, yet the conversations he had with Godel as they walked to and back from work. It would be nice to imagine that PI become a place that could facilitate such an experience for all.

Best,

Phil

Giotis said...

It is hard to find a safe criterion to say that such an institute is successful or not. In my opinion, with so many physics institutes around, such a criterion would be to create a school, a new way of thinking and doing physics that will produce innovative work. But you can't do that from top to bottom i.e. you shouldn't try to make a fuss and build a reputation by attracting VIPs with big salaries. VIPs came and go. Instead it should be the other way around; people should distinguish themselves and make a name through their work in the institute. You should try to think out of the box and invest on young, ambitious scientists and give them enough room for their creativity to flourish.

Bee said...

Hi Giotis,

only time will tell whether PI is successful or not. And I'm thinking 50-100 years.

Regarding the VIP's and the young ambitious scientists. I think a mix of different levels of experience is essential. Otherwise the young ambitious people will just repeat mistakes of the past. The issue with the VIPs is more tricky. While I sympathize with what you say, you have to keep in mind that the young ambitious scientists you want to attract will want to go where they have good chances of being on the front of research and where they can learn from the brightest of the brightest. If your place doesn't have any people with names, you're not likely to grab the best applicants out there. I tend to think the people who go for the big names are in the majority lost causes anyway, but that's how you have to think if you play the game by other people's rules.

Best,

B.

PS: And then there are the big names you put on the website which makes everybody throw up who has to turn around three times every dollar in their travel grant.

Tkk said...

Bee:
By 'executive' I mean those prominent scientists who run the place. I can't imagine otherwise.

Giotis:
Don't dismiss 'VIP'. The 'VIP', essentially the faculty and chairs of some prominence, are vital. Because young scientists are no more than young people engaged in science. And all young people can benefit from mentorship and visionary guidance. In additional to technical guidance of course. How many postdoc publish a significant paper all by themselves, without assistance and review by faculty? How many love to publish with a 'VIP' name tagged along?

Pope Maledict XVI said...

Tkk asked:

"How many postdoc publish a significant paper all by themselves, without assistance and review by faculty? "

Only the good ones, the ones who will one day do something important.

"How many love to publish with a 'VIP' name tagged along?"

All the bad ones, the people who sink without a trace.

One apparently trivial but actually all-important aspect of this whole business has been neglected: the actual *location* of the institute. It is *absolutely essential* that such a place be located in a town or region which is extremely beautiful/interesting. I'm convinced that most of the advantages derived from working in such institutions comes from the pleasant lifestyle, or [agreeable] shock to the system, associated with their locales. I'm sure that Bee is finding that living in Stockholm is very beneficial to her work. Anyone who has visited the ICTP in Trieste will tell you that the charms and challenges of living in Trieste are an essential part of the experience. Conversely, you can have the best staff and equipment in the world but to no avail if your institute is in Dallas or Kalamazoo or Los Angeles.

Bee said...

Hi Tkk,

I see. I think we agree on the question of who is responsible for keeping up the spirit then.

Reg your reply to Giotis: You don't need VIPs for mentorship and guidance. In my experience people who think themselves VIPs are rather annoying to deal with. If they aren't about to appear on radio, talk with a journalist, jump on a plane to somewhere, or a busy writing and promoting their recent book anyway. You know what I mean. That isn't to say they are all like this.

How many love to publish with a 'VIP' name tagged along?

Indeed. And that are the people you should forget about. The reason for want of prominent coauthors is that they count on it being helpful for them to find another job. And it will be helpful because places don't only hire a person, they also hire that person's network. We all know that. Just that these games have nothing to do with actually gaining scientific insights. "Prominent coauthors" is one of the factors I have repeatedly referred to as "secondary criteria," that are irrelevant for the actual question of whether or not research is promising.

Best,

B.

Bee said...

Hi Pope,

Incidentally, I found Kalamazoo to be a very nice place, in particular during the summer, whereas I found Trieste noisy and generally annoying. That is just to say people may differ in what preferences they have for location.

What I find frankly much more important is that an institute be located in vicinity of a major international airport. The one thing that annoyed me most about Waterloo wasn't its absence of culture or nightlife, but that the Toronto airport is 100 km away, and that 100 km you have to go on a highway that especially in the Winter is permanently jammed, such that on a bad day it can take you up to 3 hours. (There is an airport in Kitchener, but it's a joke, so let's not mention that. Last time I looked the only place you could get from there was Detroit, and who wants to go to Detroit?)

But yeah, I do appreciate a nice scenery and the Waterloo skyline isn't exactly thrilling.

In any case, while I agree on your assessment of people who want a VIP name on their paper (see my reply to Tkk above), I think you underestimate the value of getting feedback from somebody with experience. It can be tremendously helpful and save a lot of time. You don't need a Nobel Prize winner for that though (unless you want to hear that you better don't take a shower at 6am if you're suspecting a call from Stockholm.) Best,

B.

Anonymous said...

Observer:I just hope that the New director N.Turok brings to P.I. The neccesary strong hand to keep everybody focused. I was amazed to see several "new arrivals" come with "Prima Donna" attitudes. Intellectual freedom is right and good but come on... you have to adjust yourself to the rules and regulations.

Bruce Bartlett said...

Neil Turok has, of course, been the driving force behind the African Institute of Mathematical Sciences (AIMS), which he essentially founded six years ago. AIMS is situated just outside Cape Town in South Africa and basically offers a nine-month postgraduate training programme for students from all over the African continent in the math sciences (that includes physics!). Nowadays there are also opportunities for postdoctoral researchers. It's been a big success, so kudos to Neil Turok. So: if you feel like coming to sunny Cape Town and delivering a three week postgrad lecture course, or spending some research time at the institute, visit the website!