Friday, April 29, 2011

Interna

I am flying to Toronto on the weekend and will be visiting Perimeter Institute the coming two weeks. Since Superdaddy will be terribly busy and need all four hands for cleaning baby butts, and I'll be jetlagged and otherwise try to figure out what to write in the visit report, you're facing a slow time on this blog. When I'm back, Stefan, the babies, and I will pack our bags for we are spending 6 weeks in Stockholm so I can go back to work, at least temporarily.

Lara and Gloria are now exactly four months old. They are holding their heads well and grab everything that comes sufficiently close to their nose. They are also both trying to roll over, but haven't really managed yet. Lara is still the talker and Gloria is ceaseless in her daily workout. It seems that whenever I look at her, she's frantically waving her arms and kicking her legs. We've made some first attempts at spoonfeeding and were reminded that beta carotene is not water soluble.


The babies' paperwork is adding up. They're now registered also with the Swedish tax offices and have their own person-numbers. It turns out I'll have to apply for their residency in Sweden, so more forms waiting to be filled out. For the flight to Sweden, the girls got their own passport for which we had to get biometric photos. Even the babies manage to look like criminals on these photos.

Sunday, April 24, 2011

So you want to get a PhD?

... then here's some things to consider:

The recent issue of Nature has a News Report "The PhD factory - The world is producing more PhDs than ever before. Is it time to stop?" which summarizes the job prospects of PhDs in Japan, China, Singapore, USA, Poland, Germany, Egypt and India.
“In some countries, including the United States and Japan, people who have trained at great length and expense to be researchers confront a dwindling number of academic jobs, and an industrial sector unable to take up the slack. Supply has outstripped demand [...]”
The below graphic (from mentioned Nature News article), shows the distribution of academic post-doctorate jobs in science and engineering in the USA:
This shows the unfortunate trend towards more and more research done by scientists on temporary contracts that we talked about in my earlier post Short-term Thinking.
Germany by and large seems to be doing well as far as job prospects are concerned, though few PhDs remain in academia:
“[In Germany] just under 6% of PhD graduates in science eventually go into full-time academic positions, and most will find research jobs in industry [...] The relatively low income of german academic staff makes leaving the university after the PhD a good option.”
That agrees with my experience.

But back to the USA. One of the over-produced PhD-students from Illinois, Sergey Popov, has developed a model according to which top US universities have economic incentives to lower their standards, because the better their students' grades the better their students' job prospects and the better the university's reputation (and finances) in return. The Times Higher Education cheerfully titles Elite US students are securing top jobs 'despite being less gifted' and summarize Popov's model:
“Universities "choose [a] grading standard to maximise the total wages of [their] graduates". [Popov] said his theory suggested that grade inflation would be highest in top universities [...] the risk was that the process went so far that there were Harvard graduates in top jobs who would not have got an A at Illinois and who had fewer academic gifts and social skills than every Illinois A-student. This, he said, was not "socially optimal".”

Popov has his data online on a website called gradeinflation.com. His model is interesting but there doesn't seem to be sufficient data to tell how well it actually describes reality. Anyway, I'm sure though it will leave some people chuckling. Did I see you grin? Did I?

Actually, the Scolarly Kitchen reports that the whole higher education thing might just be the next bubble to burst! That's at least according to Peter Thiel, founder of PayPal:
“Thiel’s belie[ves] that higher education is the next economic bubble into which we’ve moved the air expelled from Web 1.0 and housing.”

The cited data shows that College tuition fees have increased 375% since 1982-84 (3 year average).

And Scientific American has an editorial, Dr. No Money about the unpleasant duties of those PhDs who dare to remain in academia:
“Most scientists finance their laboratories (and often even their own salaries) by applying to government agencies and private foundations for grants. The process has become a major time sink. In 2007 a U.S. government study found that university faculty members spend about 40 percent of their research time navigating the bureaucratic labyrinth, and the situation is no better in Europe. An experimental physicist at Columbia University says he once calculated that some grants he was seeking had a net negative value: they would not even pay for the time that applicants and peer reviewers spent on them.”

So you want to get a PhD...

Monday, April 18, 2011

Robert Bosch Foundation: Seven points to improve research

In the aftermath of the plagiarism affair that led to the withdrawal of German defense minister zu Guttenberg's doctor title and, eventually, his resignation, the Robert Bosch Foundation invited a panel of experts to formulate ways to improve the conditions under which research is conducted. The outcome is a seven point paper "to assure integrity and quality in scientific research." You can download the paper (PDF) here. Since it's in German (and I realized Google translate doesn't cope well with academic-style German), here is a rough translation:

(All awkward grammar is entirely my fault.)

"1. Mitigation of publication flood

The number of publications around the world should be reduced (relative to the growing number of scientists) and thus - against the economic interests of publishers - also the number of journals. This is the only way to ensure that this important basis for assessing the quality of research will again consist of reflected and carefully evaluated results. And only then researchers will be able to again take sufficient note of relevant results and findings from their field.

2. Basic insights need permanent funding

Science needs durable and reliable funding, because the search for something new and for an increased understanding of nature follows radically different laws than a commercial enterprise. Of course academic institutions have to deal responsibly with their funds. We have to vehemently object however the expectation that academic institutions have to make direct financial profit or are evaluated by strongly economically oriented criteria. Rather, we should work together, even more than is already done today, to highlight the high intrinsic value of knowledge gain for the general public.

3. More emphasis on the content of scientific achievements

In the allocation of research funds it should be content that is assessed, not mindless promises of success of practical implementation. The qualitative assessment of the scientific work of a scientist or a researcher should at least equal in importance the quantitative bibliometric performance indicators. The sheer number of publications is not a valid criterion.

4. Proscription of strategic authorship

Authorship of a scientific publication requires substantial contribution to the content of to-be-published work. Authorship has become a currency of science, which is rewarded with money. The system for performance-based allocation of funds should therefore carefully investigate the actual contributions of the authors and proscribe a merely strategic authorship without substantial participation.

5. Researchers must write their own research proposals

External funding is an important competitive component of the academic system. Due to the trend to demand very high shares of external funding, the pressure has increased so much that a professional application system has formed, one in which scientists no longer write the research proposals themselves, but, in extreme cases, agencies formulate standardized applications. But scientific concepts need to be written by the researchers themselves. Ghostwriters must not be tolerated, not even in composite applications where parts written by different scientists are often "smoothed" by agencies.

6. Transparency in the presentation of data collection

Science needs transparency, despite the increasing complexity. Rapid technological progress, together with an excessive competition leads to more complex, and difficult to verify experiments. Without transparent and accurate representations of data collection and the scientific approach undertaken, more mistakes and improbities occur which jeopardizes the substance of science.

7. Good research takes time

Development and implementation of sound projects are not compatible with short-term contracts. The pressure generated by short-term contracts leads scientists and researchers to carry out small projects with no substantial knowledge gain and to publish fragments. Only contract terms that offer, through sensible conditions, the possibilities to plan long-term projects (esp. for young researchers) allow the quality of research indispensable for international competition."


This sounds very Germenglish, even to me ;-) Gee, all these many-syllable words and convulated grammatic constructs. I had to look up "improbity," and I'm not even sure I know what the German translation "Unredlichkeit" means (literally it means "something one doesn't speak of"). In any case, I hope it's roughly understandable. I think these are all very good points. However, I wasn't even aware that ghostwriting of proposals is an issue, I've never heard of this.

Do you have anything to add?

Tuesday, April 12, 2011

You are Ein Stein

Earlier this year, Gideon Rachman asked in the Financial Times "Where have all the thinkers gone?" Contemplating the Foreign Policy list of the Top 100 Global Thinkers 2010 and comparing it to who might have been on that list 150 years ago, he finds today's "crop of thinkers seem[s] unimpressive" and it gives him "the impression that we are living in a trivial age." Rachman proposes several explanations for this impression of his. We may only recognize great thinkers for what they are when enough time has passed. We may not appreciate them as long as they are living, breathing things who burp and dye their graying hair. Today's intellectual giants may live in China and the Financial Times hasn't heard of them. Or, the times of great thinkers are over due to specialized networks:
"In the modern world more people have access to knowledge and the ability to publish. The internet also makes collaboration much easier and modern universities promote specialisation. So it could be that the way that knowledge advances these days is through networks of specialists working together, across the globe – rather than through a single, towering intellect pulling together a great theory in the reading room of the British Museum."

Jonah Lehrer from The Frontal Cortex speculates that the apparent dearth of intellectuals is due to the "lessened importance of the individual" because "the era of the lone genius is coming to an end," for which he cites a study showing that teamwork and collaboration is on the rise in modern research.

There's the obvious thing to say about Lehrer's argument. There has never been something like a lone genius. You don't contribute to a society's knowledge and well-being without being part of that society. Scientific research has never been done in intellectual vacuum. All the great thinkers had their friends, their correspondences, their mentors and colleagues. But maybe more important, that teamwork is on the rise doesn't lessen the importance of the individual. It just integrates it better and, truth be said, makes it less apparent. But either way, it is questionable that the number of peer reviewed articles from large collaborations has anything to do with intellectuals to begin with.

I think the reasons for Rachman's impression are more mundane. He is probably right with the suggestion that it is difficult to recognize a great thinker while they're still thinking. There's 7 billions people on the planet and all of them have something to say. A lot of them say smart things occasionally, some say smart things most of the time, but all that smartness may turn out to be bullshit anyway. Take that guy Kurzweil with his Singularity prediction for 2045. Chances are, in the year 2045 he'll be little more than a curiosity. And some people that today might appear completely nuts will turn out to be right on the spot. Time will tell, so give it time.

The only real possibility there won't be no intellectuals in the future (aside from stupidity spreading and progress stagnating) is that thinking indeed becomes truly collective. However, as I argued in my earlier post on Collective Intelligence, we are far from that. In today's collaborations knowledge is not emergent. It is not something that really happens on the collective level. It is simply an assembly of many small parts. Yes, the parts profit from the other parts' contributions and if you put a group of smart people together they can work with each others contribution faster, but it's still a piece-by-piece work.

The prototypical example for a system that is more than the sum of its pieces is a frog. If assembled correctly, it croaks and jumps and that's emergent features. The prototypical example for a system that is the sum of its pieces is lot of bricks. It gets you a wall, alright, and maybe even a house. But it doesn't actually acquire new abilities. Today's specialist networks are brick walls, not frogs. The thinking still has to be done by the individual. We're all just bricks in the wall.

"Ein Stein" is German for "a stone."

So what do you think? Do you share Rachman's impression that today's intellectuals are disappointing? And if so, what do you think the reason is?

Wednesday, April 06, 2011

Multiversing around

    multiverse [noun]

    From L. multus “much, many” and L. versare “to busy oneself,” lit. “to turn to.” Tech: A hypothetical collection of different variants of -> universes. Colloq: A large collection of no apparent purpose. Expl: “She has a whole multiverse of shoes,” “His essay received a multiverse of comments.”

I am considering to consider to read Brian Greene’s new book The Hidden Reality on the multiverse. On the pro side there’s a likely readable synopsis of an interesting topic. On the con side there’s two of Greene’s books in my shelf that I never finished reading. A writing therapy exercise I thought might be useful.

First, let’s get over with the terminology. Yes, multiverse is a disingenuous nomenclature. If the universe is by definition all that exists, then anything to the multiverse more than the universe does by the same definition not exist. But it’s moot to complain about terminology that has already become common use. What exactly the multiverse is depends on the context, but in either case it’s something that exists in addition to what the presently most widely accepted theories let physicists expect to observe. Some theories seem to imply the existence of “more,” of a multiverse of “more,” and that in other ways than “more of the same.”

Multi versus uni

The central question is what does it mean “to exist?” As a particle physicist I’d say something that can’t be observed doesn’t exist. (Observation doesn’t necessarily mean a direct interaction.) Talking about the “existence” of something that can’t be observed opens the door to fairy tales. Though my invisible friend disapproves, from a scientific point of view I am interested in the multiverse only if it’s observable. And even then my interest is very limited since I find the presently discussed possibilities of observation remote and implausible. But yes, there are versions of the multiverse that may have observable consequences. Eg. we recently discussed a paper on signatures of bubble collisions in eternal inflation, one possible multiverse scenario, and there’s Laura Mersisni’s superhorizon entanglement giving rise to the giant void, and related stories.

For the more entertaining part I’ll now take off my physicist’s hat (okay, it’s an Einstein wig really) and put on my hobby philosopher hat (if you really want to know, it’s actually a tea pot lid).

    multiversal [adj]

    From noun -> multiverse. Colloq: Of confusing variety. Expl: “By the year 2010, social networking had become multiversal,” “The promises during the election campaign were multiversal.”

A lot of effort has been spent on the search for a “Theory of Everything.” Commonly meant to be a theory unifying General Relativity with the Standard Model of particle physics, it is another misnomer in common use: It is unlikely that a reductionist approach will ever be able to actually explain everything, not in practice and maybe not even in theory. I will however refer here to a TOE in the more general sense as a theory that leaves us with no “Why” questions and reduces all of science to a question of “How” and, knocking on the teapot lid, I’ll refrain from pointing out that we can never know if we’ve found it.

What may such a TOE look like? None of the currently pursued approaches to grand unification or quantum gravity comes even close. Even if string theory or something similar would allow us to compute all the parameters in the Standard Model and in the ΛCDM model, and so on, Nobelprizes would be handed out for certain, but it would just move the Why’s elsewhere, for all these theories have other unexplained assumptions: Why are strings/ loops/ E8/ networks fundamental? Why causality? Why these initial conditions? Why quantization? Why a semi-classical or classical limit? Why matter? Why in fact anything instead of nothing?

For this TOE we cannot use an assumption that constrains the theory to reproduce observation. The only guidance eventually left is mathematical consistency. Most occurrences of the multiverse actually still have additional assumptions, but already the problem is the same: too many possibilities. If you don’t want to settle for a “just because,” a question without answer, an unexplained final cause, you have to swallow that all that can exist, according to current theories, does exist. That doesn’t answer the question, but it removes the need for an answer. That is, in a nutshell, the reason for the recurrence of the multiverse in various branches of theoretical physics: Mathematical consistency just isn’t enough.

(A probabilistic approach for the multiverse with the assumption that our universe is one of the common ones, trying to derive some features of our universe at least as probable, is a reentry into the question-room through the backdoor. It just rephrases the question why our universe is special and what theory allows us to derive the details, to the question why our universe is especially unspecial and what measure allows us to derive the details, and it makes additional assumptions about how to compute probabilities rspt. about the logic used etc. I’m not dismissing the attempts to define a probability measure on the multiverse as useless since sometimes looking at an old problem from a new direction is fruitful. But the attempt in itself isn’t actually progress.)

Mathematical consistency is not a strong requirement. The complex plane and holomorphic functions on it for example are mathematically consistent (unless you insist on some wrong theorem), so is linear algebra on n-dimensional vector spaces. What sort of a universe is that, you might ask. But if mathematical consistency is all that you’re left with, that’s what you get: Everything that’s mathematically consistent “exists” in the same sense as the world around us, a notion of “existence” not in agreement with that put forward by the strange person with the funny wig. This thought then brings us straight to Tegmark’s Mathematical Universe: All of mathematics is real, and all that’s real is mathematics. There is no distinction because there’s no other meaning to “existence.”

My problem with the Mathematical Universe is not that I dislike the idea of being made of math (whatever that might mean). In fact, I quite like the idea (up to a face factor). My problem is that for all I can tell it’s not of use for anything (oops, lid slipped off) and it is based on an assumption I don’t find particularly plausible: That humans in the 21st century have already found the language to describe the fundamental nature of reality.

Formal mathematics is a quite recent achievement in mankind’s evolution. Sure, its precision and usefulness in the description of nature is vastly superior to that of, say, the English language. But 50,000 years ago our ancestors have thought of their precise spoken language as the ultimate tool to describe nature, vastly superior to grunting and waving with paws. So how sure really can we be mathematics is so intimately connected to nature that nature is mathematics?

Versus multi

Now let us turn the argument around. Searching for a TOE we were ultimately left with mathematical consistency as only guidance and it’s not enough of a constraint. It offers too many possibilities and eventually doesn’t explain anything. Unless, that is, mathematical consistency is not the only requirement. (There is of course the requirement to reproduce observation, but that’s too pragmatic for my tea pot.) The only way to avoid a multiverse then seems to be that mathematics is not sufficient to describe the fundamental nature of reality.

So the options are: a) Accept a final cause. b) Accept the multiverse. c) Accept that there’s a way to describe nature better than with mathematics.

If you don’t like a) and b) and therefore have to sympathize with c) you are however left wondering what may describe nature even better than mathematics? Well, you can. Tegmark’s Mathematical Universe irks people because they believe there is a distinction between reality and mathematics, between platonic ideas and the world out there. The common point of view is that the math used in theoretical physics is a description of nature, but humans provide the map between reality and the math. It is possible that this mapping is itself a purely mathematical process. I can imagine there to be an algorithm that searches for mathematical definitions whose properties fit to observed data. Yet presently there is no answer to the question whether there is in fact such an algorithm able to do science like a human. Gödel’s incompleteness theorem is happily waving its tail, waiting for a chance to pee on your leg.

In summary this means if there is neither a final cause nor a multiverse, there likely won’t be any Singularity in 2045 either since no computer algorithm will be able to go beyond math. And vice versa, if a computer algorithm, coded in the language of math, is able to map every aspect of reality to a mathematical structure, then you’re likely stuck with the multiverse, Tegmarkian version, subsuming all other versions. It might then just be that the next revolution in physics comes from neuroscience.
    multiverse [verb]
    From noun -> multiverse. From L. multus “much, many” and versus, pp. of vertere “to turn.” To make many turns. Colloq: To act or talk incoherently. “She spent the afternoon multiversing around,” “His job interview was a disaster; he totally multiversed it.”

I’ll finish with a quotation from a wise physicist, who wants to remain unnamed but reportedly reads this blog: “The multiverse, the simulation hypothesis, modal realism, or the Singularity –it’s all the same nonsense, really.”

Friday, April 01, 2011

Citation Ponzi Sheme discovered

Berlin, April 1st 2011: The Federal Intelligence Service discovered a Ponzi scheme of academic citations lead by an unemployed particle physicist. A house search conducted in Berlin last week revealed material documenting the planning and administration of a profitable business of trading citations for travel reimbursement.

According to the Federal Intelligence Service, the hint came from researchers at Michigan University, Ann Arbor, who were analyzing the structure of citation networks in the academic community. In late 2010, their analysis pointed towards an exponentially growing cluster originating from a previously unconnected researcher based in Germany's capital. A member of the Ann Arbor group, who wants to remain unnamed, inquired about the biography of the young genius, named Al Bert, sparking such amount of activity. The researcher was easily able to find Dr. Bert scheduled for an unusual amount of seminars in locations all over the world, sometimes more than 4 per week. However, upon contacting the respective institutions, nobody could remember the seminars, which according to Prof. Dr. Dr. Hubert at The Advanced Institute is "Not at all unusual." The network researcher from Ann Arbor suspected Dr. Bert to be a fictitious person and notified the university whose email address Dr. Bert was still using.

It turned out Dr. Bert is not a fictitious person. Dr. Bert's graduated in 2006, but his contract at the university run out in 2008. After this, colleagues lost sight of Dr. Bert. He applied for unemployment benefits in October 2008. As the Federal Intelligence Service reported this Wednesday, he later founded an agency called 'High Impact' (the website has since been taken down) that offered to boost a paper's citation count. A user registered with an almost finished, but not yet published, paper and agreed to pay EUR 10 to Dr. Bert's agency for each citation his paper received above the author's average citation count at the time of registration. The user also agreed to cite 5 papers the agency would name. A registered user would earn EUR 10 for each recruitment of a new paper, possibly their own.

This rapidly created a growing network of researchers citing each others papers, and encouraged the authors to produce new papers, certain they would become well cited. Within only a few months, the network had spread from physics to other research fields. With each citation, Dr. Bert made an income. The algorithm he used to assign citations also ensured his own works became top cites. Yet, with many researchers suddenly having papers with several hundred citations above their previously average citation count, their fee went into some thousand dollars. On several instances Dr. Bert would suggest they invite him for a seminar at their institution and locate it in a non-existent room. He would then receive reimbursement for a fraudulent self-printed boarding pass, illegible due to an alleged malfunctioning printer.

Names of researchers subscribed to Dr. Bert's agency were not accessible at the time of writing.