I have to admit that I am not entirely happy about the way I was dragged into the discussion by Lubos, who concluded, based on a comment I made on his blog, that what I suggested would make any country 'scientifically inferior' and 'much like the Nazi Germany'.
Though the question of democracy in science is a topic I have annoyed my friends and colleagues with since at least my MS, I can't give you a working proposal. What I try to advocate is simply that we need a reasonable discussion how science in the 21st century works best. It is probably because of the books by Peter Woit (Subtitle: The failure of String Theory...) and Lee Smolin (Subtitle: The Rise of String Theory, The Fall of a Science,...) that the present discussion returns again and again (and again) towards the string theory community, and the group theory of string theorists.
Therefore, I want to start by saying that the reason for me being concerned about democracy in science was and is not the string theory community.
Instead, the reason for me becoming concerned was that I (and many in my generation) felt that research funding in Germany was severely dominated by nuclear physics and (considerably older) nuclear physicists. Generally, it was easy to continue ongoing projects, almost irrespective of outcome and prospect, whereas new research projects were very hard to establish.
- Side remark:
The reason for me still being concerned also is not string theory, but the general question of whether our community has appropriately adjusted to the demands of the modern world.
So, this post is not about string theory.
In a comment over at Cosmic Variance, Eugene Stefanovich expressed this more clearly than I could ever have done:
Eugene Stefanovich on Jun 20th, 2006 at 1:13 pm
"Dear string theorists,
Please understand that these are not deeds of some evil antistringy types like Woit or Smolin. In the absence of deliverables, sooner or later the field would come to the same point even if Peter and Lee didn't write their books and blogs.[...] they are just messengers of the inevitable change. Please appreciate the fact that the message was delivered early and you have some time to make a graceful exit out of this situation. Don't shoot the messengers."
2. Why now?
As I pointed out in my earlier post, I think the present perception of the so-called-crisis in theoretical physics is not surprising. It is due to the changes that research has undergone during the last decades:
- The increasing amount of people working together in larger groups
- The large output of publications (papers as well as books, including popular science books)
- The long time the education takes until you can contribute to the front of research
- Or maybe just the fact that it takes only seconds to send an email - or to write a silly comment on a blog.
Then there are also the changes in the society that we are a part of, it's demands on scientists, and it's notion of progress. Money, power and being famous have replaced ancient values like wisdom and tranquility. These changes have not been accompanied by appropriate changes in the administration of our communities research.
But whether we like it or not, these changes have taken place.
3. My concerns
A phrase that has been used by Sean as well as by Lubos is the 'free market' of ideas which compete among each other until it is clear which one is the fittest and which survives.
But where do the ideas come from, and how is the 'freedom' of their competition guaranteed?
Always being the optimist, I actually agree that some self-regulating mechanism will eventually set in, and the so-called-crisis will be replaced by a so-called-revolution. It might however take quite some detours before this self-regulation sets in. Waiting for a market to collapse costs time, is a waste of effort, and money. Not to mention, that it is frustrating. It could very well be that the present discussion about string theory is such a case of delayed collapse, making it a prime example to analyze the failure of the present regulating process (or its absence).
- Side remark:
To my eyes, the focus on string theory is a dominating topic mostly in the US-part of the community. I can't avoid having the impression that the quarrel strings/loops could well be translated into US/Canada. When I made my MS in Germany (in 2000), string theory was not at all a dominating topic in any regard, and it still isn't. I recall it was considered to be 'breadless' (brotlos) and detached from reality. A lecture on string theory held at my university was mostly attended by mathematicians. Around the same time, the maths department had seminars on Quantum Gravity about the Ashtekar formalism. This also was very suspicious for the physicists, simply because the mathematicians liked it (physicists and mathematicians hardly spoke to each other, as far as I know, they still don't).
Meanwhile, I was sitting among nuclear physicists with the reality constraint to do 'butter-and-bread' physics, and sneaked out to the maths seminars every now and then, before we set up our group on LXDs. At latest by the time the RS model came up (98) it seemed pretty clear that the string-community was going to be deflated. It is surprising for me that it took from 98 until now to happen.
I found another comment from someone who apparently shared this impression:
EU on Jun 19th, 2006 at 3:35 pm
"here in Europe the topic "failure of string theory?" is informally discussed since 8 years at least. Discussing this topic turned out to be much less dangerous than what I was alerted. Although a few string theorists prefer avoiding discussing this issue, most string theorists agree with the main points. The ones that successfully moved towards less stringy physics didn't kill their academic career."
In the comments to my previous post, I have been taught that Capitalism is an example how the free market works. However, the economical system needs a political counterbalance to guarantee the freedom, and the fairness of competition.
"What drives real progress of the society - and what was necessary for you to write your communist utopias today - has always been the free market of ideas and products. Exactly the things that you want to attack, deny, and abandon.
The reason why most political utopias fail is that they require an idealized, utopian type of humans, or maybe just inhuman humans.
- E.g. Utopia - Thomas Morus, Christianopolis - Johann Valentin Andrea, The City of The Sun - Thomasso Campanella, The New Atlantis -Francis Bacon, The Law of Freedom - Gerrard Winstanley, Code of Nature - Morelly, The Theory of the Four Movements - Charles Fouriers, Travels in Icaria - Eitenne Cabet, Pictures of the Socialistic Future - Eugen Richter, The Dispossessed - Ursula K. Le Guin, and yes, I admit that I read them all.
Most of those who have worked in theoretical physics know that in practice researchers are not entirely logical in their believes and convictions. At least I am sometimes quite irrational and stubborn. Sooner or later however, evidence or mathematical proof should sort out the crap. This process has worked for centuries, and it did so fairly well. In a certain sense this is natural selection -- no matter if achieved with or without method.
But the availability and quality of positions does without doubt influence people who work, or want to work, in theoretical physics.
a) The Hierarchy Problem:
For example let us have a look at an average non-ideal postdoctoral researcher in the 21st century. Being in an early stage of his education (meaning, less than 10 years after his MS), and due to the complexity of modern research, he is unlikely to have an overview on the whole field of theoretical physics. He might have his own ideas for research projects, but he needs a job. So, he looks for a field that seems interesting to him, and hopefully also has available positions. Let's assume he is lucky and gets a position.
If his supervisor tells him he wants him to investigate the stability of higher dimensional bumpy thingees in 7 dimension, will he say: "Uhm, well, we don't even know there are extra dimensions, how they are stabilized, or whether there are black holes in these extra dimensions. And, actually, I am not so into thingees. I'd rather think about why we live in 3+1 dimension?"
If his supervisor tells him to run code alpha-beta-pi with the k-factor set to 2.8, to leave out the results that don't fit the data and just keep those which do, because publications are needed for the grant a whole group is paid of. Will he say "I'd rather spend the next 10 years trying to find an analytical approach to non-perturbative QCD. I will probably fail, but thanks for paying me meanwhile."
He probably wouldn't.
He will of course have in mind to work for his employer for some time, but eventually to come back to his own ideas. After some years. Or after the next position. Or after that. Or after his first evaluation. Or after that.
I know many physicists, very many, who think so.
I know many physicists, who tell me they would rather work on something else than they do. Some have their own ideas, unfinished drafts in a drawer, some would prefer other fields which lack funding. Most feel the constant pressure to produce output in a fashionable field.
- For fairness I should add that almost all of them are under the age of 40. Those I know older than that seem to be constantly busy with lectures, administration, giving interviews, or writing books. (Forgot to mention, some of them have a life as well).
Okay, I crudely exaggerate: please don't send me your list of exceptions, I am aware they exist. I hope you get the point.
This is what concerns me most: the large detour. The economical pressure on young researchers and the resulting conformity. The waste of time. The waste of ideas. Even though I don't like to hear it, it's a fact that the human brain has it's best time in the early to mid twenties. Why do we waste these best years?
Lee Smolin on Jun 22nd, 2006 at 1:59 pm
Few think about these questions long and hard enough to get anywhere before the next “hot topic” takes everyone’s attention away.
This was a very productive style of research when high energy theory was driven by many new experimental results but it has clearly failed over the last 30 years to go beyond the standard model. [..]"
However, I also want to point out that guidance by a supervisor is one of the best ways to learn science. Some months ago, Petr Haijeck wrote to me in an email (and I hope he excuses that I cite it here):
"Of course, one could object that young people wishing to learn a craft are to serve for the first time their masters similarly as Tizian had to ground colours for Bellini. There is also something true in that. I think that there should be some equilibrium, but I do not know, how this could be achieved."
Neither do I. But I do think that the situation as it is now is not optimal. And that it can be improved.
b) The Backlash Problem
There is no point in switching from one extreme to the other, and repeating the same mistakes again and again. This is exactly what is likely to happen if we don't figure out what goes wrong, and how to avoid it in the future. I can already notice that there is a sometimes quite violent demand for instant falsifiability of theories, for closeness to experiment, etc.
I don't think pushing this too extreme is a good solution either. Phenomenological models might be more applied, but in some regards it's questionable what is there to learn from them. Just tuning some extra parameters to describe a data set is not what I think 'understanding' means.
Peter Woit on Jun 21st, 2006 at 1:47 pm
I strongly disagree with your idea that the answer to overhyping string theory is to overhype other ideas. The respect that the public has for science is based on the fact that scientists have been able to sort out what is true about the world and what isn’t. If we decide that upholding standard scientific norms about this is less important than generating enthusiasm for what we do, we will ultimately destroy our credibility and turn science into science fiction, a subject which lots of people are enthusiastic about, but is very different indeed.
c) The Selection Problem:
Besides the above hierarchy and backlash problem, there are the apparent weaknesses in the selection process. Hiring decisions are currently in a non-negligible manner based on criteria like
- Top-ranking university were the PhD was obtained
- Renowned supervisors and letters from them
- Numbers of publications in high-impact journals
- Famous co-authors on these publications (which basically guarantees a high cite-index)
- Or in generally: relations to influential people.
Taken together, this means also that the land of origin is an important factor. It also implies that it is complicated, if not impossible, to change the field after the PhD, because you will know nobody in the new field, have no connections, no letters, no publications anybody will know of. You are nobody, you have to start all over. This makes interdisciplinary work almost impossible, amplifies the specialization and incest in sub-fields.
A part of this problem is the quality of peer reviewed papers, i.e. the question in how far the publication list, or cite-index, actually is a measure for scientific excellence. However, though this is an important issue, taken alone, it won't solve the problem.
4. What now?
It is pretty obvious, isn't it? After all, we are scientists and know how to analyze problems. Here are my propositions:
- Formulate goals of research in theoretical physics.
These are not static goals, but are necessarily influenced by sociological questions like Where do we come from?, What are we made of? etc. Questions that in one way or the other were the reasons why we studied physics in the first place, and we should not loose them out of sight.
- Formulate ways to best reach above goals by supporting researchers.
- a) How:
This means not only positions, but also the quality of available positions (e.g. contract length), technical support and equipment of institutions, libraries etc. One major point that is often neglected are grants for international cooperation (for the US-citizens, with 'international' I mean 'worldwide'). It has become hard, if not impossible, to do good research without inviting seminar speakers, travelling to conferences/workshops, working visits in maybe far away places. The importance of which is often underestimated.
- b) Who:
Investigate whether the currently applied criteria to select researchers are successful. Lee Smolin has more to say about this, so I will just give you a link to his Physics Today Article and ask you to think about it, if you haven't yet done so. If you have problems thinking on your own, read Lubos or Peter's comments, to jump start your brain.
- a) How:
- On a regularly base, evaluate current research fields as to their progress regarding the goals from point 1.
I do not think evaluation merely by experts on a sub-field is a good way to objectively judge on progress in a wider sense as given by point 1., under consideration of points 2a) and 2b). As I mentioned earlier, I would find it a good idea to have an advisory committee to provide an annual report, and to make recommendations.
In my experience, most scientists are reasonable people, and act to their best believes to ensure progress in research. However, currently it is poorly understood how science works best in the 21st century, and I think it is necessary to formulate some guidelines. Not as laws, but as recommendations.
Thanks to everyone who took the time to contribute to the discussion, on this or other blogs.