Thursday, February 01, 2018

Bogus arguments in favor of "dwindling" physics

Like Stalin and Mao, affirmative action screws the world and kills. People who have tried to make the world a better place by placing many more women, people of color, and other privileged groups into physics and other prestigious fields have hurt both these fields and the members of these privileged groups, too.

Mrs Sabine Hossenfelder is a stunning example.

She was basically forced to pretend that she is a physicist literally for decades even though she has no innate aptitude for physics and she hates the discipline viscerally (especially its more theoretical subdisciplines where she is supposed to belong). The latest shocking manifestation of this hatred was shown in her tirade Physics Facts and Figures.

She has compared some apples and oranges in various disciplines – including the number of papers, its annual growth rate, power law relationships between the number of papers and number of authors, typical numbers of authors per paper – and irrationally interpreted all these things as being "bad" for physics. All of this pathetic theater was only presented in order to justify her predetermined conclusion. The last paragraph says:
So this is what physics is, in 2018. An ageing field that doesn’t want to accept its dwindling relevance.

Let me first mention her main talking points that are meant to be bad for physics. To show the omnipresence of her anti-physics bias, I will start from the end. We learn that physics is an "ageing" field with "dwindling" relevance that "denies" this trend. Whether physics is "dwindling" and whether it's right to "deny" such an adjective will be discussed later.

But the most emotional adjective that is supposed to make a difference is "ageing". What does it mean for physics (or anything or anybody) to be ageing? If you are interested in an emotionally neutral definition of this adjective, it means for its age to increase. Let me tell you something incredible. Everything and everyone in the world that is still alive is "ageing". The only reason why a speaker picks XY (a person or a discipline) as someone or something that is "ageing" is that the speaker actually dislikes or hates XY (either XY's appearance or something deeper) and/or he or she dislikes XY's current appearance relatively to the past one. There is simply no other justification.

We may say that some generalized physics has been around for thousands of years – ancient civilizations have studied "it" along with astronomy – but even if we adopt a narrower "Western civilization's" definition of physics, it started with Galileo Galilei or Isaac Newton and those lived some 3-5 centuries ago, too. It's been a long time in either case.

You know, because physics is this old, relatively to some newer disciplines, it's actually ageing less quickly, relatively speaking. If we decide that the age of physics is 2,000 years, physics only ages by 0.05% every year these days! Younger disciplines are "ageing" much more quickly. So is there some meaningful beef in the usage of the word "ageing"? Well, there is a subjective one. Mrs Hossenfelder would love her readers to link physics with the stereotypes about some useless, unproductive pensioners who are waiting to die.

Is there some real justification for this way of understanding physics? Not at all. Physics is a vibrant field and it will always remain the most fundamental natural science because that's how physics is basically defined and separated from other disciplines. So the precise topics that physicists and scientists focus on may change with time and will change with time but there will always be something that is fundamental-like about the world, it will be studied by some scientists, and they will naturally consider themselves physicists and heirs to the giants who have already been identified as physicists.

So the adjective "ageing" is just a reflection of her negative sentiments.

Is physics "dwindling"? Hossenfelder offers some numbers. The doubling time for the annual number of papers is 18 years in physics but lower (faster) in other fields, e.g. 8-10 years in electrical engineering and technology. Is that a justification to describe physics as "dwindling, ageing"?

It's obviously not. The annual growth rate of the number of papers is 3.8% which is somewhat higher than the world's real GDP growth rate. So even relatively to the average economy, the production of physics papers goes up more quickly. It's not dwindling but growing.

However, more importantly, counting papers is just a terribly stupid criterion to evaluate the value of anything. According to INSPIRE, Lee Smolin wrote 208 papers but his contribution to physics remains basically at zero. Papers aren't created equal. And the value or importance of the average paper doesn't necessarily stay constant in time, either.

So she discusses the brute number of papers as if it could be directly used as a measure of "value" or "impact" and their evolution over time. Well, those just can't be used in this way. An average paper's importance may double or get halved every 10 years, too. She completely ignores these "intensive" factors in the game which is why her conclusions are guaranteed to be nothing else than worthless and demagogic expressions of her prejudices (mainly her hatred towards physics).

She also criticizes the large number of authors in some physics papers. Well, the number of authors has certainly grown once the ATLAS and CMS Collaborations at the LHC in CERN began to write hundreds of papers where the names of the whole collaboration (3,000 members in either case) are included as parts of the paper. Needless to say, she says that this trend is bad, too.

People want to write papers with many co-authors because that's how they may have a higher number of publications for less effort, she claims. Holy cow. This is such a stupidity.

If you write many papers that have 10 authors each, you may have a large number of papers but you won't be considered an important physicist waiting to be hired by the best places and given the best grants. Those who make the actual hiring and funding decisions are not idiots – unlike Mrs Hossenfelder. They know that a physicist who manages to write an important paper by himself is probably smarter or more important or valuable than one of 10 physicists who belong to a large group of authors that shares an equally important paper. And you know, even in larger lists of co-authors (starting from 2: the best work is teamwork but 2 is often already too many), people who decide about hiring and funding almost always care "which co-author has contributed how much".

So at most, she is unmasking how lame and idiotic her own thinking about the quantity – the number of authors of a paper, in this case – is. But her being a complete idiot does not imply that many actual physicists or most physicists are idiots. On top of that, the large number of authors is simply a great idea in many cases.

The LHC Collaborations could be smaller but they couldn't be tiny. There's simply a lot of work at many levels that are needed for an LHC paper to be completed. To paint the growth of the collaborations in fundamental experimental physics as a "bad trend" only tells us about the speaker's prejudices – there's nothing rational about this judgement. High-energy physics needs high energy, therefore big gadgets, those need many "hardware" and "software" layers, and lots of people must work on them. It's inevitable and it's just a symptom of progress in some specific enough situation.

More reasonably, we could look at the number of people who work there and the funding they're getting. Physics is doing fine. Of course you find disciplines that are growing faster. But many of them may turn out to be temporary flukes that may end soon – they may be bubbles or fads. A discipline may double each 5 years but you may be pretty sure that it hasn't done so for 2,000 years. Readers who are uncertain about my statement should try to calculate how much is 2 to the 400th power. ;-)

So the implicit assumption that one should abandon disciplines that need 18 years to double, and only work on some fads that double more quickly, is a sign of someone's extremely superficial character, her desire to join every bubble etc. Science and the mankind just couldn't work if this attitude were very widespread.

There's another problem with all of this. The separation to the disciplines is arbitrary to a large extent. The first example of the "faster, more dynamic" discipline she mentioned was electric engineering. It's fine we call it "engineering" so it's not "physics" but it's equally sensible to think that it's just a subfield of applied physics. Why should a particle physicist feel the duty to consider an atmospheric physicist to be his "close colleague" while the electric engineer has to be treated as a "distant alien"? It makes no sense. A particular physicist may very well find some fields classified as non-physics to be closer to his definition of good physics-like science than some disciplines classified as parts of physics.

The general point of the previous paragraph is that this whole pissing contest in which we compare clumped collections of disciplines is an example of stupid identity politics.

Instead of counting the number of papers, we may compare the total number of citations or the h-indices which still make much more sense for a meritocratic evaluation of disciplines. You know, I have served as the Harvard Junior Fellow which means that I know a lot about the top people from very diverse disciplines and their interactions. My fellow Fellows – our main duty was to talk to each other – included top historians of chemistry, philosophers of morality, experts in Slavic or Chinese literature or medieval Icelandic poetry, condensed matter physicists, Yau-like mathematicians, a chemist of early life on Earth, and so on. Some comparisons may be made, others can't because the disciplines culturally differ and have different expectations, different units of "output" and "excellence", and other things. Lots of "calibration" has to be made for any comparison of the disciplines to be at least slightly professional.

But assume that the number of citations or the h-index (the maximum integer "h" such that the author has at least "h" papers with at least "h" citations) may be applied across the scholarly disciplines. Where do we get? Pick this amusing list. On top of the list, you find Sigmund Freud ;-) with the h-index of 272 and almost half a million citations.

Sigmund Freud was a neurologist but he was mainly pseudoscientist working in psychology, too. He has accumulated half a million of citations but I am confident that most physicists – and even scientists in other disciplines that appreciate hard science – will agree with me that most of these citations are unavoidably rather soft. Followups of Freud's papers are something in between science, pseudoscience, and popular articles. It is not necessarily wise to take the number too seriously.

Beneath Freud, you find lots of people in medicine, some people in genetics, some sociologists, economists such as Stiglitz, and postmodern philosophical crackpot Derrida, among others. You need to go to the place #47 to find Edward Witten with h-index of 188 and 180,000 citations or so. So the theoretical physicists' citation guru Witten only made it to the "top 50" which doesn't look like a stellar result.

But again, every sensible person knows that the citation from a Freud fan isn't quite the same "deal" as a followup to Witten's paper. It takes a lot of intelligence and hard work to do the calculation for a typical followup to Witten's papers. On the other hand, it's enough to write another paper saying that "your sexuality reflects the pressures from your parents during your childhood" for Sigmund Freud to win another citation. That's really not quite as precious as a new paper on M-theory or anything else that Witten has worked on (or discovered).

Jacques Derrida – who is 2 spots above Witten – has written only garbage, I think, and all the people "working" on similar stuff are more or less ideologically driven morons.

So again, of course you will find other disciplines above physics (especially theoretical physics) in similar tables. But some of these disciplines – especially medicine – are considered "crucial applied sciences" by most humans on Earth (and I largely share this respect towards medicine – just to be sure, I wouldn't write an overly general negative article about almost any discipline of natural sciences) which guarantees that their importance is and will remain high. Others are postmodern or otherwise pseudoscientific rubbish. Others are about technological breakthroughs – like basic concepts of some new photovoltaic cells – that are considered important in the industry but may be considered branches of applied physics, too.

There's a lot of diversity, some of the fields are closer to each other than others, some of the fields are considered worthless by researchers in other fields, some of the fields make it much easier to collect the citations. So it's still a problem to draw any conclusions from such lists – but such lists based on h-indices and citations are still vastly better than anything that Mrs Hossenfelder has discussed (based on the demonization of the number of papers and authors). Mrs Hossenfelder's criteria are all about comparing things that are obviously neither commensurable – comparing apples to oranges – nor constant benchmarks in time and assigning labels "good" and "bad" to these criteria in a way that is done to confirm her prejudices, anyway (especially "physics is dwindling").

It's terrible for someone to be forced to be a physicist when she actually hates it so much (especially theoretical physics). Affirmative action is evil. Please, someone who matters, end this farce and liberate her from this curse.

No comments:

Post a Comment