"Competition is Inherent to Science"

A conversation with Patrick Aebischer, EPF Lausanne
Interview: Florian Fisch, Labtimes 04/2014

Photo: Lonza Ltd

If you have heard of the Swiss Federal Institute of Technology in Lausanne or the Human Brain Project this is most probably thanks to Patrick Aebischer. The neuroscientist gives his view on governing a university and the role of big science.

When Patrick Aebischer became President of the Swiss Federal Institute of Technology in Lausanne (EPFL) in 1999, the Swiss university landscape changed forever. The professor of neurosciences started by upgrading the little sister to the grand ETH in Zurich, by opening a school for life sciences. The well-known and highly debated Human Brain Project became the flagship of the EPFL with the rock star scientist, Henry Markram, at the helm.

Under the presidency of Aebischer, the EPFL took over mathematics, physics and chemistry from the neighbouring university of Lausanne, and enlarged the campus by adding many new buildings, like the fancy library (Rolex Learning Center) and the SwissTech Convention Center, to name but only two prominent examples. Other rectors and presidents look grey and dull next to Aebischer, who started his academic career as a medical student in Geneva in the late 1970s. From the mid-80s to the early 90s, he worked at Brown University, USA, and returned to Switzerland in 1992. As a researcher, his studies revolve around understanding and treating neurodegenerative diseases, such as Parkinson's disease and Alzheimer's disease.

Most people agree that Aebischer has achieved a lot with the EPFL. Competitors from Zurich say that Aebischer would prefer to destroy a project rather than give way to others. Surprisingly, the president's office is on top of one of the old buildings. The grey-haired Aebischer gives a paternal impression and amidst presidential advertisements, he is also open for banter.

Lab Times: Being president of a university, is it more like managing a company or like managing a government office?

Aebischer: It's neither. You certainly cannot run a university like a company. My 380 professors can do and say whatever they want without having to fear any sanctions. Academic freedom is unique about a university. But as a president you have to set the ambition, the culture of the institution, which is what I did when I became president of the EPFL: I said that I wanted to transform this good engineering school into a world class technical university. This is different from government. You cannot be the best government.

How much influence does the president really have in such a process?

Aebischer: We have much more influence than we think. Most importantly, it is our responsibility to nominate professors. Up until now, I have nominated more than 80 percent of the faculty. This is key for the quality. Probably, the most important reform we undertook at the EPFL was setting up an assistant professor tenure track system for young scientists. This has attracted the brightest scientists and engineers because they know they can become independent early on in their career. The tenure track system is the strength of the USA compared to Europe. It was an experiment to adapt the US system to European conditions.

Did it not create resistance from older professors?

Aebischer: There was quite some resistance at the beginning. But with time, the faculty saw the advantage at being surrounded by the best, to be able to collaborate with them. That's the secret of good research institutions.

So you mainly determine the composition of your professorate?

Aebischer: You can also influence the areas of research. When I arrived at EPFL, I thought that it is important to have life sciences at a technical university. Apart from that, you also have the responsibility to ensure that the infrastructure is up to the level of the expectations and that you attract funds from both the public and the private sector. This allows you to modulate an institution. Being a university president is not like a conductor of an orchestra but rather like an organiser of a jam session. You put the best musicians together and you give them the beat.

You are said to have a strong hand, having made some enemies. Is it necessary to show aggressiveness as a president?

Aebischer: If you want to apply major changes, you need a strong hand, yes. At the beginning you have to break eggs to make an omelette. We, the presidents of the Swiss Federal Institutes of Technology in Lausanne and Zurich, are in a special position because we are not elected by the faculty but appointed by the federal counsel [government cabinet]. You could say parachuted. You still have to be accepted by the institution because otherwise you will never be able run it. The advantage of not being elected is that you don't have to please everybody. I closed all 13 departments, for example, in order to create five Schools. At the beginning, I would never have passed an election but today – maybe I'm wrong about his – I think I would have a chance.

One of your tough changes was to increase competition between Swiss universities. Most people probably agree that competition is necessary for progress. But was it maybe too much?

Aebischer: I don't think this is what I have done. I never said that I want to compete against Zurich. The ETHZ is a great institution. It's just that we were considered to be second-class and when my generation of scientists came back from the USA, we did not see ourselves as second-class. I am not doing this against anybody, I am just trying to attract the best scientists. Before, there was one world-class technical university. Now there are two. This is good for Switzerland and good for Europe as well. By the way, that's also how you attract the best students. Competition is inherent to science. It happens automatically between labs and even within labs – maybe too much so. You compete for funds and for publications.

There are worries that too much competition corrupts science. Many studies are not replicable, for example.

Aebischer: Yes, this is an issue that academia will have to deal with.

Is there something you can change as a president?

Aebischer: Yes, I think we need a broader perspective for our promotion criteria because that's what the faculty responds to. Before, it was one lab, one ego. Today, science is much more collaborative. The evaluation of publication has to change, the teaching and the technology transfer have to be taken into consideration as well. We have to get away from this publish or perish concept. In the ideal world, we should cease the salami tactics and publish less but more important papers. It's maybe a utopian idea but I suggest to our promotion committee not to look at the total number of publications but to take the five best and read them.

Are they doing it?

Aebischer: [laughs] I think they try. It's difficult to change the system all by yourself. The journal editors have vested interests as well. So yes, in many areas there is too much pressure to publish.

You are the godfather of the Human Brain Project, the best-known flagship of the EPFL. Is this sort of big science a model for science in the future?

Aebischer: Yes, it is. In the past, big science was the speciality of physics. CERN is a very good example in Geneva. Why did physicists do big science? Because they needed the big infrastructure. Life sciences started to become big with the Human Genome Project. The reason is that with all the omics technologies we are generating a gigantic amount of data. Of course, the hypothesis-driven investigation in your own lab will have to continue as well. The community has to learn to find the right balance between these two, the structured goal-oriented research and the hypothesis-driven approach. There has to be a little more top-down, more organisation in life science. Of course, the people doing the regular thing are afraid of big science.

The US National Institutes of Health (NIH) pumps about six billion dollars a year into neuroscience. The Human Brain Project costs 100 million euros a year. So this is certainly not putting the whole of neuroscience into disarray. You cannot continue to accumulate data lab by lab, some of it not being difficult to reproduce and not comparable. Some industrialisation of data gathering is necessary. That's exactly what the Human Genome Project did. Now, people would have a hard time to do human genetics without the results of the human genome database.

Isn't there a big difference between the Human Genome Project, which was organised by different leaders in the field, and the Human Brain Project, a one-man-show with Henry Markram?

Aebischer: The Human Brain Project is not a one-man-show. Henry Markram has a lot of visibility like Francis Collins or Craig Venter with the Human Genome Project. You always need people who think out of the box. Today, the Human Brain Project has 13 Principal Investigators with Henry Markram as the spokesperson. It's a highly coordinated and goal-oriented project, very similar to the Human Genome Project.

But the difference is...

Aebischer: There is a big difference. The goal set by the Human Genome Project was well-defined: read the three billion base pairs. The Human Brain Project is more open. Its ambition is to use the enormous amount of data generated by neuroscientists across the world to perform a simulation. There are more than 100,000 neuroscience papers published in a year. Somebody needs to integrate and normalise these data, to use them in order to understand the function of the human brain, so the generation of data that we can compare and use for the simulation is a first, well-defined goal. And very importantly, all this data will be shared with all interested scientists. It's not that we will fully understand the brain in ten years but I hope that in five years we will have better-defined goals to go to the next step.

So you need to invest 500 million euros just to define the right goals?

Aebischer: The goals are clearly defined. But every so often you refine them as you move along.

One critique of the Human Brain Project is that, due to its size, it has to be a success. Is that compatible with the values of science, where you should honestly report negative results, too?

Aebischer: For the Human Genome Project, people said: we are going to cure all the diseases. Later, critics said: now we have the sequence and nothing has changed. But now it is impossible to think of cancer therapy without the sequencing capabilities. The Human Brain Project will become a platform, accessible to do neuroscience simulation. I see potential application to simulate deep brain stimulation [for Parkinson's disease], where we have no idea, how it works. We are used to thinking in two to three year terms. When you ask people to think in ten year terms, you get into a catch-22 situation between ambition and feasibility.

It's clear that the sheer number of published papers doesn't give more insight. But doesn't a healthy diversity of papers from different groups create better ideas than one single project?

Aebischer: Take ion channels as an example. We have to apply industrial scale approaches, like the Allen Institute has done with the connectome, so as to really understand how they interact. With comparable data you can simulate because it is reproducible at the macro-level.

One of your achievements was to involve industry much more at the university. This collaboration is no doubt necessary for innovation. Do you see a limit on how much industry should be involved?

Aebischer: Yes. That's an advantage in Europe that the state is involved in education and basic research. No company can afford that anymore. The Bell labs and similar institutions have all disappeared. So the public sector clearly has to support basic research. But then you need to pass the baton. We have the responsibility that our technology is going to be utilised. We need the patents and the people at the interface because at university we cannot develop drugs by ourselves. We also need to put things into perspective: industrial funding is less than ten percent of research funding at the EPFL.

So why can't the contracts be made publicly available?

Aebischer: They are; we were forced to hand them out. But I think this is problematic. In the case of research contracts, a company wants to sponsor some research in return for patents or licenses from it. If you tell everybody what the research is all about, the company will hesitate twice before collaborating with academia. It gives away their competitive advantage. Granting agencies are facing the same problem. If you don't collaborate with industry you'll end up doing academic engineering. We would never ever sign a contract, if the professor were to lose his or her academic freedom. And it would be naive to think that an industrial partner could dictate to Henry Markram what to do.

Why can't you publish the general conditions of a research contract without disclosing the topic?

Aebischer: The conditions are always the same. We give them maximum three months – usually one month – before publication of the results, so they can file a patent. Everything we do will end up in the public domain. Otherwise we do not collaborate with industry. It is important for our graduate students and postdocs to be able to publish. Sharing knowledge is at the root of academia.

What does a company get in return for sponsoring a chair?

Aebischer: Nothing. The name of the chair. Of course, they can talk to the professor and, indirectly, they get more people doing research in their area. If they want to get something directly they need a separate research contract.

I see that calling your library the Rolex Learning Center is not problematic for academic freedom. But don't you feel it's strange to call it that? Why don't you go ahead and sell the EPFL name as a whole?

Aebischer: It would be too expensive. [laughs] I wouldn't for the institution but for a mere building, I don't see the problem. We are Swiss and want to be proud of our watch industry. You know, who pays us? It's the taxes. We need a flourishing economy, otherwise there will be no taxes. Money doesn't grow on trees! Swiss-German journalists seem to be more sensitive to the issue. You are all thinking in the same box. I wish you would be more interested in talking about the 90 ERC grants we have received, since the launch of the programme.

I'll stay in my box: Is it necessary for you to be on the board of Nestlé Health Science and Lonza?

Aebischer: Would you put the same question to the president of Stanford?

Yes, absolutely!

Aebischer: We have the responsibility to help the companies to contribute to the wealth and development of Switzerland. Here, we have a very big company called Nestlé; why should they invest in a research centre in Beijing instead of Switzerland? If they ask me to join the board, I have no problem with that.

Nobody wants to stop you from advertising the EPFL at these companies. But being on the board and being paid by Nestlé comes with special responsibilities that also make it difficult for you to take a decision necessary for the EPFL but harmful to Nestlé. This results in a conflict of interest.

Aebischer: Is there anything in life without conflict of interest? If you do not have conflict, it means you are not doing anything that is interesting. I do not believe that I have ever been exposed to a conflict of interest since I have been on the boards. It may come, of course. What's important is how you manage them. I would not do this for a Chinese or an US-American company, even though there would be fewer perceived conflicts of interest. I want to be sure that the Swiss companies thrive. When I see Novartis closing down neuroscience in Basel and moving to Cambridge, I am worried as a Swiss citizen. When the majority of the managers of large companies are non-European, I start to be worried. When we have a popular vote on Europe, who speaks out publicly? Putting too much emphasis on the fear of conflicts of interests could be problematic for the country. You increase the risk of delocalisation by having top managers that are not rooted in the area you live in. Now, you and I will not be there but our children and grandchildren will live through this. I try to choose the fields where Switzerland can thrive and align the EPFL with the Swiss DNA. When I came to the EPFL 50 percent of the Swiss GDP was not represented in research. There was no life science and no finance. We have to align our interests. Maybe we've gotten too wealthy and we forgot how we got there.

If you are as worried as you say, why did you not speak out more against the popular initiative against mass immigration in February? There was only one small manifesto passing/slipping under the radar of most mass media.

Aebischer: Sure, we could have done more. Unfortunately, I was in Africa on a sabbatical, on massive open online courses at the time. The vote was not about research collaboration and it's always easier to judge afterwards. In addition, it is difficult for scientists to fight populism. For us, the ends don't justify the means. It's a delicate issue that worries me a lot. Being outside of the research programme Horizon 2020 and being submitted to contingents of scientists would be the end of Swiss science at the world level. This, for me, is much more an issue than links to industry.

What can you and your president colleagues do about it?

Aebischer: Let me first remind you that all the cantons from the French speaking part of Switzerland voted against the initiative. I did quite a bit since coming back from my sabbatical. We invite people to the campus and we try to show how openness is key. Switzerland was not only split between the French and the German parts but also between town and countryside. The influence of the EPFL on the German part is modest. It is the duty of our colleagues in Zurich to engage in trying to change the mentality about the fear of immigration.

Antonio Loprieno, the rector at the University of Basel, said that it is difficult for publicly-funded institutions to speak on political issues.

Aebischer: It's true and we even regularly receive directives by the government that we shouldn't get involved. I honestly just ignore it.

Are you admonished for this?

Aebischer: No. [laughs] I think we should have this autonomy, despite being part of the federal administration.

You are appointed by the government and you are on the ETH council that governs the Swiss Federal Institutes of Technology in Lausanne and Zurich. So, you are sort of a political figure yourself. How do you have to function on this level, in order to be successful?

Aebischer: All university presidents are scientists to start with. I kept my position as a professor. Yesterday, for example, I was in the lab and last night I corrected a paper. Without this I could not survive.

How do you behave with the politicians?

Aebischer: You have to be there, explain what we do. Several times a year I go to speak in parliamentary committees. They ask questions – often related to risks. It's perfectly normal as they provide us with the necessary funds. What is remarkable in the Swiss system is that we are highly autonomous. The interference from the executive branch is extremely limited. There have been very few cases in my 15 years as president.

When does it happen?

Aebischer: I typically interact with our federal councillors during trips abroad. It gives us the opportunity to talk in an informal manner. I have never had a federal counsellor calling me to do something or not to do anything, compared to France, for example, where they sometimes receive “directives”. We are highly independent. The autonomy is probably the biggest strength of the Swiss system. We are protected from direct intervention by the ETH board.

Last Changed: 03.07.2014

Information 4

Information 5