I was stunned by the arrogance and dangerous proposals displayed in a talk by Sabine Hossenfelder
Peer Review and its Discontents [slide show].
That blog post contains a YouTube video with slides from a 1-hour-long talk she gave at a conference and her audio that was compressed to 23 minutes. She could have chosen a computer voice generator to make the audio sound more human and less like a computer-generated Orwellian EU bureaucratic speech but this is the smallest problem with the talk.
The bigger problem is the content. First of all, I can't understand where she finds so much arrogance to try to organize how the whole scientific community works given the fact that she has never contributed any valuable and correct insight to science and she was only allowed to stay in science because of the affirmative action. Physicists who have actually achieved something are far more modest in all these aspects.
Sometimes they are too modest but the modesty always has a rational core. It is important for everyone to realize that no one is infallible – and it's particularly difficult if not impossible to "predict" what sorts of discoveries will be made and considered essential in the future. For example, Edward Witten will always refuse to make prophesies about the distant future and he will always acknowledge that the important conceptual breakthroughs may be made – and have been made – by people who may be less famous than he is. Sabine Hossenfelder and her even more pseudoscientific readers are convinced that they can distinguish an important paradigm shift right away, and so on. This is absolutely preposterous. They don't have the expertise to decide about rather elementary questions in physics – let alone hard questions about the cutting edge and the far future of physics.
But I want to discuss the content of the talk more systematically because the number of dangerous delusions about the principles of science (and its sociological aspects) that were accumulated in the talk is almost astronomical.
Generally, she has observed that scientists are dissatisfied with the "system" – meaning various hiring and grant decisions and peer review in particular – and this dissatisfaction proves some "disease" and she offers some dramatic "cures" for that "disease".
Now, things are surely suboptimal but it is questionable whether in the real world, assuming the distribution, skills, and hard-to-change habits of the people who actually exist, the system could be much better than it is. I don't see any clear proof in either way. In particular, everything I see is compatible with the hypothesis that all the problems result from many people's not being good enough or the problems' being hard enough, and so on.
But let's start. At most slides, she discusses the peer review as if it were some holy authority. At the beginning, it looks like she is even criticizing scientists as soon as they dare to say something wrong about the peer review. At 2:53, we hear:
Peer review is necessary for good science.
Sorry but it is not. Such a claim is just a totally wrong nonsense spread by ultra left-wing collectivist ideologues. Good science is something that can generally be done by a good scientist or a group of scientists who decide to cooperate. Much of the good science – and perhaps most of the best science – was developed without any peer review.
Isaac Newton's 1687 Principia Mathematica was never peer-reviewed. Einstein's first 1905 paper on relativity was only "peer-reviewed" to the extent to which the editor Max Planck said "Jawohl". Even much more recently, James Watson's and Francis Crick's 1953 Nature paper about the structure of DNA was published without a peer review.
Newton didn't need a peer review. In fact, Newton couldn't really have afforded any peer review because he didn't have any genuine peers. To some extent, it's true about many of the best scientists of all epochs, too. And these peer-less scientists are obviously the most important ones. They're the class of scientists for which the claims about the importance of peer review fail miserably.
If the peer review helps in something, it's primarily in reducing the amount of published stuff near the opposite, low-quality end of the submitted articles. Peer review is a layer of filters that is meant to reduce the amount of stuff that someone has to read – and do so in such a way that it's more likely that the better and more important texts are not eliminated.
But peer review is in no way "necessary" for science. To do good science, one needs at least one sufficiently motivated, intelligent, honest, hard-working, and lucky scientist who just approaches some questions or phenomena from the right direction and uses the scientific method in one way or another. The idea that science has to be a collective enterprise or that it's an achievement of a whole community or a whole society is nothing else than political propaganda.
I am not saying that there are no scientific contributions that depend on many people and their interactions; I am only saying that there are tons of scientific advances that don't depend on such things and virtually any part of science may be done by individuals.
Moreover, Sabine Hossenfelder must have missed it but high-energy physics has basically "rejected" peer review in most contexts. People identified as eligible authors are posting the papers on the arXiv and the experts in hep-th etc. don't have a problem to go through the dozen or two dozens of titles and/or abstracts every day – even though they only end up reading at most several papers a day. They cite other work in their papers and care about the papers themselves – and perhaps the citations – when they're hiring others.
The publication in true peer-reviewed journals is just a formality that people sometimes need to co-exist with the institutionalized world outside the community of experts. But those rituals have become basically unnecessary for the progress in science itself. The number of papers isn't insanely high and the differing attention (and citation counts) that different papers attract is sufficient to play the role of picking the more important papers from the ensemble.
She repeats this totally wrong cliché that "if peer review is failing, the scientific process isn't working correctly" many times. But the repetition doesn't make it true. A dysfunctional peer review process may lead to a waste of money but a waste of money is something entirely different than the inability to make scientific advances. Science may be done – in many cases, even without any funding. And when good science is done by someone and when competent scientists exist, they will almost certainly notice. If a paper is right and important and not appreciated, it's because others just don't understand it or its importance. You can't fix this problem by a sociological rearrangement of the people.
Hossenfelder quotes some quotes from reviews that show that the process isn't always serious. And what? Of course, it isn't always serious. It isn't always critical, either. Some people take it more seriously – especially the careful people. Some people don't take it too seriously. People are generally not too motivated to write the reviews really carefully and read the relevant papers in detail. So the quality of the reviews isn't always spectacular.
But that doesn't automatically imply that people should be forced to spend much more time with reviews etc. It's time that they may use differently, especially by their own research. In most cases, the latter is arguably more important than writing essays about the work of someone else. There isn't clear evidence that the situation would improve by lowering or raising the time that people spent as reviewers.
Around 1:37, she shows the results of her 2008 survey among 1800 researchers. One-third said that the selection process works well, one-third said it worked badly, one-third was neutral, and a small fraction didn't know. She makes a big deal out of it. Why? What's so shocking about the fact that people get divided to thirds in this way? When people are free, they almost always do. Ask how any nation is satisfied with its government. Offer them three possible answers. About one-third will answer "very good", a third will say "neutral", a third will say "bad". The only exceptions will be countries such as North Korea where everyone has to love their Kim. OK, Russia gets pretty close to the inner harmony. But this enthusiasm is in no way a necessary condition for a country to work well, quite on the contrary.
Even more importantly, if one finds those people who are dissatisfied, they are dissatisfied for tons of reasons. They have been harmed by someone – maybe someone unrelated to the question. They are in trouble. They see legitimate problems with this minister or another. The problems that people see often contradict each other. People have different opinions. The implicit idea that the existence of one-third of the participants who are dissatisfied is evidence in favor of her particular "cures" only shows that she is a wannabe dictator.
I simplified her poll. There were two questions – both results were close to 30:30:30:10, however, and whether some of the numbers were 25% or 40% is clearly absolutely irrelevant. One was asking about the ability of the system to recognize "necessary research topics"; the other was about "transformative research and paradigm shifts". Holy cow. I hate these petty small people who love to repeat these big words. She has been very far from making any research that was good enough to be distinguishable from šit. So why does she even talk about paradigm shifts? She has nothing to do with those. She isn't able to make them and she isn't able to appreciate them when she sees them, either. How is one supposed to define the boundary between great research and "transformative research", anyway? These labels are subjective ones. People give different ratings. And in the case of the media, almost all the usage of the big words like "transformative research" is pure hype. So please calm down with this overheated rhetoric.
More or less by definition, paradigm shifts come rather rarely. We might reasonably demand that the number of paradigm shifts in physics is 1-3 per decade or 5-30 per century or whatever. When we calibrate things in this way, it is tautologically obvious that the paradigm shifts are rare. You can't really "improve" the process so that there would be 10 paradigm shifts per decade because they wouldn't really be paradigm shifts if they were so frequent.
Instead of big words about paradigm shifts, a scientist wants to learn the truth about scientific questions he finds interesting. Ambitious scientists want big enough questions to be answered. But good scientists are not looking for breakthroughs according to the breakthroughs' ability to produce big words. They look for solutions to problems that seem important or interesting yet doable to the scientists. Some of those will lead to many implications and will be retroactively considered revolutions sometime in the future. But no one can know this whole future fate of a discovery right now. And that's why it's complete nonsense to "plan them". You just cannot plan breakthroughs. Pretty much by definition, breakthroughs and paradigm shifts are events that destroy the business-as-usual which means events that weren't planned. If a social engineer like Sabine Hossenfelder can plan something, it is clearly not a paradigm shift.
The mindless collectivist ideology penetrates pretty much every slide of hers. Around 3:10, we read:
The scientific system is not able to self-optimize "good science" because individual incentives are not aligned with the collective goal.
Holy crap. This is like one of the boring clichés we would read in the summaries of the conventions of the communist party. There is absolutely no content in this sentence beyond the wordy intimidation of the reader who is pushed to be terrified by the powerful collectivist ideology. How could this sentence contain anything wise or useful?
The sentence implicitly assumes – it really wants you to not even doubt it – that the scientific knowledge is a "collective goal". But if someone is an actual curious person or a scientist, the scientific knowledge or understanding is primary his private goal. After all, this is true even for a scientist who is not so curious but who was hired to make a certain work.
A group of people may agree that they're interested in some scientific questions so the scientific understanding is also their collective goal, indeed. But even if that's the case, why would one talk about the scientific understanding as about a collective goal? It just doesn't make any sense whatsoever. It doesn't matter whose goal it is – also, most ordinary citizens almost certainly don't have the goal to understand some difficult questions in fundamental physics.
So if we strip this totally off-topic ideology that clearly can't have any value aside from the propagation of the collectivist ideology by repetition, her sentence should have replaced the misleading and irrelevant description "whose goal it is" with the actual name of the goal:
The scientific system is not able to self-optimize "good science" because individual incentives are not aligned with the scientific progress.
Is this sentence true? Well, it depends on whether you think that the system works well because the purpose of the system of institutionalized science is clearly to push individuals in such a way that the scientific progress emerges. In other words, the sentence above is exactly equivalent to the sentence "the system doesn't work because it doesn't work". If you assume that it doesn't work, the sentence is a vacuous tautology. She gets so totally and hopelessly lost in the Orwellian jargon that she doesn't realize it.
So the situation is completely analogous to a paper by a sociologist who dared to ask Feynman to read his paper at a weird interdisciplinary conference. The first sentence sounded similar to Hossenfelder's sentence:
The individual member of the social community often receives his information via visual, symbolic channels.
For a while, Feynman thought that he must have been inadequate before he decided to decode the text. The first sentence above said "People read", he realized. And once he got the self-confidence, he could have decoded the whole essay. It continued: "Sometimes people read, sometimes they listen to radio." And so on. When decoded, it was clear that the text had no nontrivial content. The same is true for Hossenfelder's slides except that her verbosity looks even worse and her claims are much more disputable. The long sentence above only says "the system doesn't work" which is less true than "people read".
OK, so are scientists motivated to do scientific progress? Sometimes they are, sometimes they are not. When a sponsor pays his own money and really wants some progress to come out of it, he almost certainly tries to give the right incentives to the scientists. But it's not easy or straightforward especially because the nontrivial results – unpredictable advances in science that go beyond "mechanical work" – just can't be planned. Only statistically, people see that certain strategies have apparently led to better results than others so they try to pursue these strategies and perhaps modify them. And sometimes, some sponsor may try completely new strategies to motivate the scientists etc. Those things may be as ambiguous as science itself. And paradigm shifts may occur in methods to organize and fund science as well, not only in science itself.
What is easier to see is that some system doesn't work at all if it doesn't work at all. If most people get lazy and corrupt, something is probably seriously wrong. But even though Hossenfelder will deny it, this is really not a good description of the best places in physics. These places do decisions according to conventional, common-sense recipes. A group of people who know the field uses all the available data and reports from outside to decide who is the best hire or who wrote the best applications for a grant, or whatever.
The basic principles of such decisions are no rocket science – and for two reasons. First of all, they are not rocket science because the precision of such decisions simply cannot be as high as the precision in the location of rockets and their parts. There's lots of uncertainty and lots of "incomparable aspects" of people's and papers' quality. Second of all, they are not rocket science because the only truly general rule is a matter of common sense: choose the good or best ones, leave the worse and bad ones. You don't even need Howard Wolowitz's rocket science degree from MIT to get this point, do you? On the other hand, to apply this recipe "good is good, bad is bad" in practice, you do need the expertise about the particular field about which you are deciding!
Around 3:23, she criticizes scientists who dare to criticize any officials in funding agencies or whatever for their own mistakes. Unfortunately, she doesn't say what those scientists' mistakes are. In other words, I think it's obvious that she just says that "we have to stop" criticizing the authorities because they are the only ones who can manage science well.
But this is largely nonsense – and a reflection of her fanatical pro-bureaucratic world view. Science is ultimately not done by bureaucrats. Science is done by scientists. In particular, a bureaucrat who loses the ability to understand science to the extent that he can't know what is good and what isn't so good will simply be unable to move science in the right direction. And in fact, he will be almost guaranteed to harm science when he intervenes into the details he doesn't understand.
Just to be sure, when a decision is done by a working scientist, it may also be a bad decision, especially when the scientist is not so good. But if there is a working scientist, there is a much higher chance than in the case of the bureaucrat that the expertise will be available and correctly applied. A manager may do a much better job of monitoring which scientists are actually cheating or doing things like that – you don't need a degree to be a cop. But to place bureaucrats above scientists who haven't been guilty of anything is just sick. I think that at this moment, and we're just in 1/7 of the talk, you will already wholeheartedly agree that she is a despicable communist activist who talks about "collective goals" and what she really means by that is that science is reduced to a huge bureaucratic monster governed by bureaucrats.
She says that funding agencies, journals etc. rely on peer review and that's right because scientists use it, too. But as I said at the top, it's not necessarily the case. The high-energy physics community doesn't really depend on the old-fashioned peer review process this much. But the task for a funding agency is to distribute the money wisely; and the task of the journal is to print the best articles. It's their job to find out how to find the good applicants or authors and papers!
They may be forced to rely on peer review and the peer review may be found to be of a low quality. But this is ultimately not the fault of the scientific community as she tries to pretend. Unless it's a part of their contract, scientists aren't obliged to spend lots of time with the review of other people's papers. So if an agency or a journal needs some quality reviews, it's their task to find the right people who are sufficient to do the job; and if necessary, these reviewers have to be compensated in some way. Alternatively, they may find other criteria or algorithms to do their job well. If the funding agency or the journal fails in this work and relies on some low-quality work of reviewers, it is indisputably the fault of the funding agency or the journal.
The idea of Ms Hossenfelder's that the scientists should feel some guilt (collective guilt) because some funding agencies or journals apparently rely on a lousy reviewing process is absolutely illogical and unfair. Some scientists would perform – and do perform – the job very carefully. And others don't, sometimes for sensible reasons. The funding agencies and journals simply have to learn how to distinguish well done reviews from the others. It's their job. Scientists' job is to do science and it's absolutely unacceptable when an unfriendly woman says that "scientists are the system" and criticizes scientists because some of those hired by journals or agencies do a bad work. The job of a scientist simply isn't collective in this sense and one cannot apply the collective guilt.
When she says that we aren't allowed to criticize the system because we are the system, I couldn't avoid thinking about the 1984 ad by Apple. She's really a system, someone who hides either behind the masses or behind some powerful ones all the time, but genuine scientists don't do that.
At 4:22, she impresses us with the communist tale that in a burning opera house, people are egotist so they try to escape and therefore jam the door, so they need a bureaucratic leadership. Please, give me a break with this stinky Bolshevik junk. It's totally natural for people to try to save their lives and trying to escape from the fire is the best strategy among the obvious ones. Sometimes, people will decide to cooperate. But it is simply not true that it's generally a good idea to cooperate or that thinking collectively increases the benefits for the average member. The greatest advances in the human society have occurred within the framework of capitalism which is based exactly on the type of self-serving optimization that Hossenfelder demonizes.
Good scientists should follow their own personal quality and ethical standards and try to do work regardless of anything that is done by others – in this sense, they should be analogous to the visitors of the opera house who just look for the best route out of the building (instead of waiting for some collective miracle that will save everyone), something that many Parisian music fans recently found to be a good recipe, too. This is perhaps the most important point that she tries to deny – and the denial of this crucial rule seems to be the point of pretty much every slide of her talk.
She says that without the bureaucratic control, people behave rationally – which either means that they are individually escaping the opera house; or they cite themselves. One may perhaps use the word "rational" for both activities but the goals of both activities are completely different. The visitor of the opera house tries to save his life while the self-citing researcher only wants to increase his number of citations regardless of their "type".
Whether the latter is "rational" depends on the broader goal of this behavior. Why does he want to increase the number of "any" citations? The ideal scientist – who should also be picked – really cares primarily about the scientific progress so it's "rational" for him to do things that do lead to progress, and such scientists are ideally preferred (and they're often preferred in the real world, at least at good places, too). But OK, let's assume that the scientist doesn't really care about science so much but he cares about his career. Is it useful for him to cite himself? It's only useful if some people or computers evaluate the citation number "blindly". But that's virtually never the case. When any good enough place is hiring someone, they will be able to identify people whose "big fraction of citations", whatever "big" means, is composed of self-citations.
You really can't get too far without getting ludicrous. I think that all the people for whom self-citations significantly helped (according to the percentage of citations that are self-citations) either end up with a small overall number of citations; or they must have lots of ludicrous papers that contain 50 references citing themselves so that they will become legendary with this activity and in every hiring or similar committee, someone will know about them.
Just realize that with \(N\) papers citing each other causally, you will only get \(N^2/2\) citations or so. If you want to reach 1,000 citations (and you're still far from Witten's 100,000), you need some 45 papers and the 45th paper will already cite the previous 44 papers or so. By that moment, you will already lose the credibility of everyone who matters. People will know about you. And even if they don't know about you, they will laugh and throw the paper to the trash bin once they see the last pages with those 44 references. There have been such people but all of them are "jokes" to the extent that they're not considered members of the research community.
So it's clear that all these problems are pretty much non-existent. She only talks about them because they are used as justifications of her plan of bureaucratization of all of science.
At 5:20, she shows a slide that is absolutely shocking:
I work in theoretical high energy physics.
For various reasons I believe this field is particularly bad: high competition, traditionally fad-driven, very tightly connected.
It's a good case study.
You can't be serious, can you?
This incompetent would-be boss has gotten a job at an institute founded by Niels Bohr himself – a job that should have obviously gone to a great young string theorist – and she can't compare with the average member of the high energy community, not even remotely. But she has the arrogance to give talks all over the place and claim that the field of theoretical high energy physics is "particularly bad". She should be grateful enough to lick the rectum of every genuine high energy theoretical physicist who would be perverse enough to agree.
It's a great example what kind of pathologies you get when you allow people to be hired because of their non-convex reproductive organs. And if she finds the field so horrible, why the hell doesn't she leave it?
The criticism is absolutely nonsensical. High competition always tends to increase the quality in which the competitors compete. It may lead – and, as we discussed with Bill, it has led – to the admission of many people who aren't good according to other criteria but certain very specialized skills are really needed in that field which is why the competition has to exist. At the end, I think that the number of string-theory jobs in the world is way too small given the importance of the problem (as measured e.g. by the interest of the smart young people) and this is obviously the main reason of the excessively intense competition.
It's also nonsense to criticize the "fads". A "fad" is just a negatively sounding description of the excited atmosphere that follows an interesting enough advance that opens the possibility that many more low-hanging fruits may be picked. When stupid or dishonest people launch a "fad", it's probably pathetic. But when careful and smart people follow a "fad", it's mostly because it's something really valuable, exciting, and important. This is particularly shocking in combination with her calls for "transformative discoveries and paradigm shifts" because a "really big fad" and a "paradigm shift" are completely synonymous. One term sounds positive, one term sounds negative, so they are used by people with different opinions about the content of the advance. But you know, it's because Edward Witten or Andy Strominger are vastly smarter and more honest than Sabine Hossenfelder, you should better trust their and not her description of some major dynamics in the field. If lots of people comparable to Strominger join a "fad", it's probably because the right word for the "fad" is a "cool minirevolution". I think it's not always like that ;-) but as a statistical rule, one must understand that such a fad should be probably expected to be a good thing.
I wrote about "fads" a month ago.
Also, the field is "tightly connected" because it indeed unifies all phenomena in the Universe. There are clearly some insights and methods etc. that are absolutely central and everyone who competently works in the field knows them. Because of the data and principles that unify them that all these people have to be familiar with, the people end up being intellectually "tightly connected". They must be interested in similar ideas and new proposals. It just cannot be otherwise. Who is not connected just can't be near the cutting edge.
Around 7:30, she complains that there are too many papers and no one can read them. But as I mentioned, the number of papers is low enough so that people may be familiar with the existence of all new papers in their field. Ten or twenty papers on hep-th appear every day, for example. They read a small fraction because they are specialized. And to some degree, specialization is obviously unavoidable and the extent of specialization is guaranteed to increase with time. It's been like that and the trend will continue simply because the total body of scientific knowledge keeps on increasing while the capacity of one scientist doesn't grow this much.
I do believe that the specialization is too high and people should feel like experts in "at least 1/10 of papers in hep-th", if I have to say a number. But the numbers are variable and no canonical numbers exist. What's more universally important is that the system has worked and may work even with the specialization. Every scientist is a full-fledged expert in "some patch" of the manifold of his field; he knows a lot about the adjacent subfields; and he knows something about everything.
This hierarchical knowledge focusing on one's immediate expertise but reaching very far is enough to make even those decisions that need to compare projects from different subfields or even different disciplines. People will often disagree but that's normal. After all, there can't be any objective way to universally compare the value of two projects in different disciplines.
Around 8:00, she says that the lack of time to review papers leads to superficial criteria – like judging papers according to the fame of authors – which leads to the "rich-get-richer trend". And it would be even worse if reviewing were voluntary. Well, this is another purely socialist ideological slide. In many cases, some famous people get even more famous because they keep on producing. Sometimes, it becomes easier for them to produce new things when they're already famous.
This fact isn't a bug of the system in any sense. It's totally right to make the conditions better for the people who have already shown that they have a higher chance to find interesting things – because those people seem to be the best place to invest! The only mistake is when the right people are incorrectly identified; or when the credit is unfairly assigned. They can be (and often are) incorrectly identified if the "reviewers" in the system are not honest or smart or independent enough. But if that's so, there's no easy solution. And when some people manage to take credit for other people's work, it's bad. But such questions are universal topics about the human nature that have no specific relationship with research in physics.
People in some fields completely lack the independence of thinking and integrity to do things right, so the whole field ends up being corrupt. You surely know which fields may be quoted as examples and high energy theoretical physics certainly isn't one of them.
At this 8:20 slide, she is against voluntary peer review because some papers wouldn't get read at all. I am sorry but if a paper isn't read at all, it's probably because it isn't interesting for anybody. Or it's so revolutionary that the existing mankind isn't able to get it at this moment, anyway. Or because the author hasn't "sold it" sufficiently creatively. All these reasons mean that the mandatory reviews of every paper just won't be useful.
Scientists ultimately must be free to spend most of their time with things that they consider valuable. If a bureaucratic system forced top scientists to spend their time "equally" by reading the preprints by Lee Smolin or Sabine Hossenfelder, I assure you that the scientific progress would be over. It's very important that scientists discriminate at every moment. It's good for most of them to avoid wasting time with garbage.
At 8:30, we hear that the process is not optimal because
good ideas may go unnoticed
fads are amplified
errors remain uncorrected and repeat
What a shock. People aren't perfect. Incidentally, she also failed to mention that paramount ideas may be criticized by the totally wrong people (like when she talks about supersymmetry later). What the negative people call "fads" or want to go "unnoticed" may be, and often are (and, in the history of science, often have been), the true paradigm shifts.
People aren't perfect so it is unavoidable that the system won't immediately and exactly separate the important science from the good one and from the bad science. But what's important is that the system works at least slowly and people are gradually getting closer to the truth. The mills of god grind, yet they grind exceeding small.
Ideologues like hers want to transform totally innocent and vacuous observations about the real world – like the observation that the people and the systems aren't perfect – to some kind of a revolution that results with a system that doesn't work at all.
In a refinement at 8:52, she says that "there's nothing wrong about fashions but she points out that they are amplified above their value". Sometimes they are, sometimes they are not – sometimes they are underrated, right? Individual scientists have some guesses about what is more important and what is less important. They differ from other scientists. Some of them turn out to be much more often right than others – sometimes because they're simply smarter.
But if Sabine Hossenfelder finds it so easy to quantify what's the right amount of attention that every idea should be getting, why doesn't she write a valuable paper herself? It must be comparably difficult, right? I am simply flabbergasted when someone who is obviously a below-the-average member of the field criticizes the collective decisions to follow fashions etc. When someone is worse than the average, then her opinions about the "right weight" to associate with a fashion are probably less accurate than the opinions of the average, too, right?
In other words, why don't you close your super-arrogant mouth? Even for the best people in the field, it would be way too arrogant to demand the right to decide about the "right amount of attention" that is given to every idea. But you?
After 9:00, we hear that science is an "adaptive system" just like the free market economy and the natural selection. Yes, no. There are some similarities and there are some important differences, too. In particular, her 4-step circular process of science is just a naive stupidity. We almost never know in which stage of this circle we should be. What I really want to say is that science isn't running according to some mechanical algorithm with steps. At every moment, evidence may arrive – from expected or unexpected sources. At every moment, someone may modify an experiment or a theory in a way that looks promising. At every step, some ideas from the past may be suddenly revived or imprinted. It's just silly to say that four types of steps regularly alternative and it's even more silly to think that the path towards the proposal of new viable hypotheses is as straightforward as the simple arrows suggest.
To propose a viable theory is an extremely creative enterprise that depends on many regimes of thinking that are not easily describable and that most people cannot reproduce. That's why we often say that some true breakthroughs depended on geniuses.
The "circularly alternating 4 steps" model of science may be helpful to convey some simple points but in the real-world conditions, it is so spectacularly unrealistic that every recommendation built upon such a model is bound to be stupid. At 12:20, she proposes some new bureaucratic regulations that scientists need to follow. Give me a break. Someone who is hired as a postdoc by a university knows what science is, and if he doesn't know it, he will be left a few years later – unless he is a woman, of course. It's insulting to suggest that people like Hossenfelder or her beloved bureaucrats should impose lots of new regulations on the scientists.
At 12:30, she complains that a grant application was rejected because it wanted to build "big things" based on a preprint that had 0 citations and wasn't even published. She thinks it was not a professional review. Please, give me a break. If applications like that weren't immediately rejected, then nothing could be rejected. She has actually previously said that she understood that grant agencies were relying on the pre-existing judgement of scientists about each other. A paper's having zero citations is a clear judgement. You can't expect an "okay" for a big project if the proposal is such that it isn't even capable of earning one citation.
As I said before, it's totally legitimate for the funding agencies to use any kind of criteria they find OK to distribute the money wisely. The argument about zero citations is a sociological argument but it is totally OK for grant agencies to rely on sociological arguments. It's only the real scientific work that has to rely on true meritocratic criteria. But funding decisions are not pure scientific work. After all, everyone can pay his money to any cause he wants, according to any rules, and the high citation counts surely belong among the better criteria to distribute money.
At 13:00, we learn that science has an invisible hand like the free market but it needs:
objective judgement: no bias, no marketing, no distortion of incentives by artificial pressure.
The hypocrisy is absolutely breathtaking. Just a minute ago, she screamed – in fact, at a conference where it was completely off-topic – that a reviewer who just politely pointed out the obvious fact that her papers were worthless junk that shouldn't waste anyone's time and money has certainly done a "bad thing". She effectively blackmails the reviewers because whenever she gets rejected, she can whine in similar talks that she has been hurt. And now, a minute later, she is trying to teach the people about the ethical rules that allow no bias and no marketing?
Are you serious, Ms Hossenfelder?
Around 14:00, there was a slide about "governments" whose role in the talk remained totally incomprehensible. She must simply love this obnoxious bureaucratic jargon. At 14:20, she says that it was "shockingly high" to find out that 20% or 25% of the people would like to change their field but they cannot.
What's so shockingly high about it? It's surprisingly low. To say the least, it's obvious that this percentage can never be tiny. People have lots of dreams that simply cannot be true. Some high percentage of women would love to change their occupation to the job of a celebrity in Hollywood but due to some petty hiring decisions, they are not offered the job of Angelina Jolie. Well, many of these women surely think that they're as good as Jolie if not better. And a tiny percentage of them might even have a point. But for a vast majority of them, it's unrealistic. They won't be hired for very good reasons. The people in Hollywood typically don't think that the new women would be as good as Jolie. Believe me, I am no uncritical fan of Hollywood but I surely think – or, I would say, see – that many things over there work very well and have very sensible reasons.
The logic behind Hossenfelder's "shocking" poll is absolutely identically trivial. People who have gotten a PhD (or achieved any other impressive threshold) in semiconductors can't be hired as string theory professors tomorrow (or choose any pair of occupations instead of these two) simply because it seems obvious or at least very likely that they're not equipped for the new job.
It wasn't really trivial to get the PhD. People had to write a thesis that made some sense. To do so, they had to learn new things for 4 years or so. About a half of this learning was directly linked to their narrow specialization. And most of this learning – perhaps not the specific content but at least the suffering of going through some "similar" things – was surely necessary for them to be ready for the research in the field of their degree.
I think that most PhDs wouldn't even have the energy or desire to spend the same amount of concentrated energy once again and get another PhD. Only a very small fraction of people has several PhDs. These people are smart in some way – this versatility is usually one of the values that drives them from the beginning. But a majority of PhDs just won't double the amount of work they needed for a PhD. At some moment of their life, most people have hit their apparent limits. Again, I am not saying that everyone is limited by one PhD. But what I do say is that all the statistical groups Hossenfelder talks about are dominated by the people who do have such limits – ordinary enough people.
And in the case of some people, it's not true. They can get new degrees and learn totally new disciplines and switch and so on. But it's obvious that this is not the case for a large majority of the 25% in Hossenfelder's poll. Most of these 25% are analogous to the women who would like to be Angelina Jolie – but it's obvious to every other sane person why they are not Angelina Jolie.
So this result of the poll means absolutely nothing. At 14:40, she adds
A quarter of physicists work on topics they don't consider most promising.
This is a distortion. The question she asked was different – whether they would like to switch. A topic may be promising but it may still be too much work or boring work etc. to work on it and people may want to switch for lots of reasons. But even if we ignore this distortion, it's just absolutely unsurprising that some visible fraction of people would prefer a different job they can't get and some of them have become disillusioned with their work. Such people obviously exist in every single occupation in the world.
It's likely that these 25% don't belong among the main drivers of research in their field. But that's not a crime. You may find it cruel but exactly 60% of the people in a discipline fail to belong to the top 40% of the drivers in their field. ;-) Obviously, many of the people near the bottom may feel relatively dissatisfied with their job or specialization. And some of the dissatisfied ones may be closer to the top, too. Hossenfelder keeps on attaching absolutely unjustified and nonsensical emotional and ideological implications to all these self-evident facts and tautologies. She's a hardcore demagogue.
The percentage of people employed as scientists who are not contributing much to science is much higher than 25%, of course. But it is not a tragedy. The system simply works in such a way that one needs to hire a sufficiently high number of people and some of them will do important things. It's like buying a certain number of lottery tickets. People who are unhappy with their job should be allowed to leave and perhaps encouraged to leave. But they simply can't be "helped" to the extent that they could get a job that they are not ready for, a job for which they haven't passed the exams and tests and interviews that those jobs demand. That would mean a complete collapse of the system and evaporation of all the meritocracy. Moreover, if people were guaranteed to be able to switch in this easy way, most people would obviously switch to the most well-paid discipline or occupation.
A "diagnose" at 15:12 says:
Scientists fail to protect their objectivity. The result is the misalignment of collective and individual incentives. I call it the irritable scientist syndrome. Scientists want good science but good science isn't good for their career.
How can she write – and pronounce – these infinitely hostile and mostly untrue, oversimplified slogans about science? Lots of scientists, especially the good ones – which is highly correlated with the "famous ones" (at least if we mean "famous among other scientists") – do protect their objectivity very carefully. Some don't. But the former group is more important than the latter. The first group are the "true" scientists. It is absolutely unacceptable to insult the first group because the second group exists.
I won't discuss the collectivist ideology again. It's all over her talk. She would definitely be an apparatchik in the top committees of a communist party if she had the opportunity.
Now, good science isn't good for scientists' career. What a pile of cr@p. In almost all fields of science in the real world, good science is extremely good for the scientist's career. Perhaps unfortunately, there are lots of other things that help to decide about careers as well. But if all these things are equal, good science surely makes things easier, and not harder, for one's career! This is particularly the case in high energy theoretical physics but as I say, it's true in most disciplines. The correlation between the career and good science is very strong.
At 15:50, she divides the scientists according to their attitude to the would-be "problem" she claims to exist. All those who disagree are evil heretics, we learn. They either do well in the system which must automatically be a crime, she argues, or they are conservatives who point out that things should be and are basically the same as they worked in the past, or they say that the underlying reasons can't be really dramatically improved, or they at least think it's not a wise investment of time for them to try to "solve it". I would subscribe to all these assertions to one extent or another. They are not really mutually exclusive – it's just another piece of her demagogy that this clumping to boxes is meaningful.
She says insulting lies about those who believe every single thing like that. I mean, is this possible? It really leaves me speechless. Does she really believe that most of the scientists who are active and celebrated by their colleagues for some results are so evil? Does she really believe that instead of focusing on their research, scientists should be like her and pretend to be saviors of the world who must introduce some new bureaucratic regulations that will surely make science better? Does she believe that something about scientists' attitudes to science has changed so qualitatively? Does she really believe that bureaucratic committees and regulations constructed by visceral haters of science such as hers will help science? I am speechless, shocked, disgusted, it's just incredible. I don't understand how she can be in the same room with any good and successful scientist if she hates them in this way.
Around 17:30, she says that we need "incentives for researchers to switch field". As I said, it is a very expensive thing for a trained scientist to switch fields. A worker who operates a machine may learn how to work with a new machine after a day or at most a month. It may cost hundreds of dollars. But a trained geneticist has studied for a decade or more. In the West, his training costs hundreds of thousands of dollars. You can't just give "incentives" for scientists to switch fields because it's just terribly expensive. She is aware of this very good reason why scientists don't frequently switch their fields but she pretends that she doesn't understand it.
If he doesn't perform his job well, deserves to be fired, and may be fired, he should be fired instead of "switching field". Just because someone doesn't do a good work doesn't make the adjacent disciplines "obliged" to hire him. If a whole discipline or subdiscipline gradually deteriorates, students just stop going into it. Why should there be some artificial extra "incentives" to create this chaos?
At 17:50, she criticizes scientists from nurturing the culture of no criticism. That's really cute for one reason. Whenever I have explained some of her work to be completely wrong (and yes, everything that is at least slightly original that she has ever written is wrong), I had trouble with it. She would whine and try to character assassinate me by posting all kinds of stuff. And now she claims that scientists are guilty of the culture of no criticism?
Also, she complains that people don't respect her blog to be on par with scientific work. Neither do I. Obviously, I don't consider my blog to be "equivalent" to scientific papers, either, even though dozens of blog posts have been rated by some scientists in this way. Some colleagues told her that the public criticism isn't a part of the proper scientific discourse and they mentioned that she is aggressive or attention-seeking.
Whatever her behavior or motives are, it is obvious that in general, the criticism backed by a readership of a blog like her BackReaction certainly can't be considered on par with responses in scientific papers for a simple reason: the papers are written by at least somewhat qualified people while all the people actively following her blog are complete crackpots. You find Uncle Al, Leo Vyuk (sorry, Gentlemen), Arun, and tons of these people who don't have the slightest clue about any of the relevant questions but who want to pretend that they are very smart and important.
Science critically depends on some quality control and Hossenfelder's writings – especially in their pure form without some moderating sensible co-authors etc., i.e. her writings on her blog – is a canonical example of something that is trash according to science and that science eliminates for this reason. She can barely survive when she writes "moderated" papers with no valuable content – although, I assure you, all achieved physicists know very well that Hossenfelder isn't one of them. But the pseudoscience on her blog is just too much.
Whenever she's not the target, she worships the cult of criticism. But the problem is that for it to play a role, criticism must be allowed to go in both directions. Chronic grumpy critics like herself believe that they are uncriticizable themselves because they're the holy celebrated critics. The left-wing journalists and similar stuff that has poisoned portions of the Western society is thinking in the same way.
But science isn't just a sequence of infinite or repetitive meaningless arguments. Science is actually a process that leads to some process. It compares the evidence and arguments and may decide what claims can be elaborated upon and which ones are gone. The situation with Ms Hossenfelder is simple: you are already dead in this sense. You have lost all the arguments. You may find crackpots who praise you but no real scientist does because they have already figured out that your opinions are not backed by any expertise and they are just wrong. They are nonsense. Trash. You may keep on criticizing but that won't change that you are already on the dumping ground of science, Ms Hossenfelder.
I agree with the comment that the attitude "live and let live" is excessive. But Ms Hossenfelder is one of the best examples showing how harmful this attitude may be. If that attitude hasn't run amok, should couldn't have completed her PhD let alone something else.
She repeats slogans about tons of biases and things like that but she never realizes that she's one of the worst examples of all these diseases, one of the best examples of what needs to be eliminated for science to get healthier. She simply wants to place herself among all scientists, including folks like Witten, and decide about them. She is plain insane.
At 19:30, she makes you laugh by saying that supersymmetry is an example of all the evils and the research of supersymmetry should be banned. In reality, her inability to understand the absolutely critical role that SUSY plays in contemporary theoretical physics is one of the reasons why it's absolutely outrageous that she has received her PhD. But even those who can tolerate the PhD won't be willing to go much further with their tolerance. They know that her knowledge is totally inadequate for a theoretical physics career.
SUSY may or may not be seen at the LHC but it will still be essential in theoretical physics because the models with SUSY are, in some profound sense, simpler than those without SUSY. They're the "default" options according to the properly understood Occam's razor. They reduce the number of independent fields by relating bosons and fermions. They eliminate lots of corrections by the cancellations and non-renormalization theorems. This just cannot go away.
She basically believes a conspiracy theory in which all these people, including the young students and postdocs and professors who start a phase of their career by studying SUSY, believe that SUSY is worthless but they do so anyway because of some bizarre pressures. Why would they do it? How do you make some 1,000 people study some very demanding models with lots of indices and fermionic variables and components if they believe that it's rubbish? It would be much simpler for them to write that the spacetime is discrete or the Hawking radiation doesn't exist or... whatever is in Hossenfelder's papers.
But they want to do actual physics motivated by rational arguments, evidence, and hints and SUSY is both an unavoidable aspect of all realistic vacua in string theory; as well as the single most well-motivated extension of the Standard Model. I am terrified of the idea that sometime in the future, a politician with a bunch of bureaucrats will listen to this female crackpot and they will ban the most critical directions in science by the same method by which the Nazi Germany banned the "Jewish physics" and the Soviet Union banned genetics.
The fundamental thing that Ms Hossenfelder misunderstand is that she is the personification of all the corruption, hypocrisy, egotism, and dishonesty and it is people like her who have to be eliminated for science to get healthier and stronger. The institutionalized science mostly does it but a few counterexamples such as hers are striking, indeed.
At 20:10, scientists are said to be under four pressures:
financial pressure
time pressure
peer pressure
public pressure
Scientists are "not free" for that reason. Or are they? It depend which scientists. If a scientist doesn't find the money too important or has already earned enough, he's free from the pressure #1. If he has lots of energy and is willing to spend lots of time with his work, he may avoid the pressure #2. If he's independent, he avoids pressure #3. If he doesn't care about the public, he avoids the pressure #4.
For different scientists, those considerations are relevant to a very different extent. Good scientists are generally driven by strong internal forces that beat most of the pressures.
But a more important point that Ms Hossenfelder doesn't want to see – even consider – is that all these pressures have good reasons and do improve the system in many and probably most cases. They're not something that should be eliminated.
People are under financial pressure. What is happening is that their desire to earn some money or be wealthier is forcing them to do things that other people want them to do. Is it good or is it bad? It depends on whether other people want good things. If other people – the potential payers – want good things to be done, the financial pressure will generally encourage the scientist to do useful things.
If a scientist may either do a calculation or have some fun and a sponsor pays some money so that the scientists picks the "calculation", this will surely increase the number of calculations that may be done. So most such pressure ultimately help the system to achieve more. Hossenfelder always assumes that the system doesn't work at all – basically that there shouldn't exist money and salaries and things like that. But within the existing system, it is spectacularly obvious that her recommendations are completely wrong – if the wages weren't correlated with what the sponsors think about the value of the work, it is obvious that the value of the work would go down.
So her claims may only be right in a completely and fundamentally different system, perhaps in a world where money doesn't exist at all, some kind of a communist utopia. She is never saying what this alternative is supposed to be – thankfully because it would be guaranteed to be totally idiotic.
Similar comments apply to the other three pressures. People, not only scientists, may feel some time pressure because one day has 24 or at most 25 (or 23) hours – when the summer time is switched. So people may be forced to think how to use the 24 hours a day – and several decades of their life – to be happier or more productive or whatever adjective they find desirable. What's wrong about it? It's obvious. Why should it be included in a talk about the management of science? When people have too much work – they have embraced too many duties etc. – they will feel too busy. What a shock.
Peer pressure. There is peer pressure. It's obvious because other people have their ideas, too. If two sides of a pressure disagree with each other, both of them may be right. The pressured person may be right but the pressuring person may be right, too. Two days ago, Stephen Wolfram sent me a new text about the cellular automaton models of the spacetime. I pressured him to stop with this stuff and learn proper string theory because I feel that a very smart brain's CPU time is wasted with something that could be used far more efficiently. You may count it as a case of peer pressure. But just because it's a case of peer pressure doesn't mean that it's bad.
It's a totally one-sided demagogic thing to identify any peer pressure with group think. Peer pressure may equally well be directed to "remove" someone from some other kind of group think – that the pressured person would otherwise share with another group of "peers" or any people. There is clearly no universal recipe to figure out who is right. The idea that the peer pressure "shouldn't exist" is absolutely dumb.
And the public exerts some pressure, too. They want certain practical solutions to be available so they may care about some kinds of applied science more than another kind. And in pure science, they usually prefer the science to confirm their pre-existing metaphysical beliefs. Most of these pressures are indeed counterproductive for a scientist and a good scientist escapes them. But sometimes the public may be right as well. And more importantly, not the whole life is science. The public is composed of free people as well – they have the right to live as they want, believe the values they like, and exert any pressure that is legal and they find helpful.
All those things and interactions exist and people deal with them in one way or another. Good scientists can generally ignore the money but they often earn a lot of them, anyway; they can do lots of things in a given period of time because they are hard-working; they are independent enough not to surrender to the pressure of their environment just because it's loud and strong, but they are wise enough to listen to others and change their mind when they are persuaded by some arguments or evidence. Good scientists also try to improve the world around them by exerting pressure on their peers and other people. There's nothing wrong about it. Sometimes, the pressure is going in the right direction, sometimes it isn't. Hossenfelder's propaganda claiming that all such pressures are "evil" is a terribly one-sided misunderstanding of how the whole human society works.
At 20:50, she says that people who judge whether science is good sometimes use proxy criteria which may be helpful because of the correlations. But when someone tries to optimize the proxies only, bad science results. I mostly agree with that. Well, in some cases, bad science results. Sometimes or often, the proxy criteria still remain good enough. A paper with Witten is a proxy criterion for a good paper. Instead of doing good science, you may try to manage to write a paper with Witten. Well, I think that if you achieve that so that Witten actually signs something you did together, it's very likely to be a good achievement simply because Witten won't sign a generic piece of mediocre work. So the proxies often continue to be applicable much further than Hossenfelder pretends.
But it is true that sometimes, they break down and the correlation evaporates, and it's more likely to evaporate once the researchers start to fully exploit the differences between "good science" and the otherwise imperfectly correlated "proxies for good science".
At 21:10, there's another repetition of the obnoxious communist burning opera house metaphor. Now, two people are running from a bear and one points out that he doesn't have to escape the bear, it's enough to outrun the other man. She presents it as an example of a would-be failure of a free society but it's rubbish again. This behavior is a good enough template for good work in science and elsewhere. You don't need to make a paradigm shift according to some metaphysically overhyped judges. It's enough to be better than everyone else. The bear story actually suggests a very important and true point: One must ultimately always evaluate the quality of some scientific work relatively to the other work in the real world, not according to some unrealistic perfectionist criteria. Hossenfelder clearly misunderstands this simple point, too.
Another list of problems and diseases appears at 21:40. These aspects amplify problems:
community grows
it's better connected
fewer long-term positions
more public exposure
The idea that all those things are "evil" is totally wrong, too. In the latest issue of Laissez Faire, a Libertarian magazine I am getting, they published a wise essay by Radim Smetka titled "I need seven billion people for my life". It nicely explains how the increased population has allowed a better specialization and optimization of lots of things so the quality of our lives is higher than it would be in a world with a much smaller population. Both the increasing population and the increasing total number of relationships (cooperation, competition, division of labor) are important for the system to get better.
Hossenfelder's points #1 and #2 directly contradict this impor