Wednesday, October 31, 2018

Dark matter. Or what? [Video]

Fixed some problems, created some new ones. Sorry for the glassy stare at some places; I had to redo the recording but had a gigantic headache. I promise not to do that again.


Click on “CC” to get English subtitles.

I encountered the same problem with YouTube as with the previous video, that the automatic transcription falsely identifies the language as German. (This time I left it there for your amusement.) I have no idea why this is. I have all my language settings in English and really I don't think my pronunciation is so bad you’d mistake my English for German. The only thing I learned by Googling the issue is that others have it too. In case you know a fix, please let me know.

Oh, and Happy Halloween :)

Update Nov 2nd: Now with Italian and German subtitles. Click “CC” to turn on captions, click on settings/gear icon to change language.

Sunday, October 28, 2018

Physics may be up against the wall, anno 1986

I thought you’d enjoy this gloriously prescient NYT article from 1986 which my husband has dug out. In it, Malcom Browne reports that Alan Chodos (then at Yale) worries about the future of particle physics:
“Unable to mount experiments that would require energies comparable to that of the Big Bang genesis event, Dr. Chodos believes, growing numbers of physicists will be tempted to embrace grandiose but untestable theories, a practice that has more than once led science into blind alleys, dogma and mysticism.

In particular, Dr. Chodos worries that “faddish” particle physicists have begun to flock all too uncritically to a notion called “superstring theory.” […] Deprived of the lifeblood of tangible experiment, physicists will wander off into uncharted regions of philosophy and pure mathematics,'' says Dr. Chodos, leaving true physics to wither.””
Chodos later became the Associate Excecutive Officer of the American Physical Society, a position from which he retired in 2014.

Friday, October 26, 2018

Will it end? [In which I have breakfast with John Horgan]

Taking a selfie with a book on
your face is more difficult than
you may think.
I had breakfast with John The-End-Of-Science-Horgan two weeks ago, and I’m beginning to think it was a mistake.

I had backed out from an after-lecture-dinner two days earlier for which I felt guilty already, so I may have forgotten to mention I actually don’t eat breakfast. To make matters worse, I arrived late that morning because once I stepped into the shower, I noticed there were no towels in the hotel room. And when I had finally managed to dry my hair and find the place, I had to prevent an excited New Jersey taxi driver from having John pay my bill. Then we watched the taxi-man write down my credit card information in sloth-motion.

To celebrate this shitty start into the day, I ordered a coffee, just to learn that John doesn’t drink coffee. Which I should have known because he wrote about his coffee-fast on his blog. Evidently, I didn’t read this. Or maybe I did but immediately forgot about it. Either way, I’m a bad person. Even more so because John promptly also ordered a coffee. Caffeine-free, but still, now I had become somebody’s bad influence. And caffeine-free coffee, I hope y’all know, isn’t actually caffeine-free.

Luckily, the morning improved thereafter. John turned out to be a really nice guy who will cheerfully explain why science is over, which reminds me of the time I accidentally sprinkled herbal salt on a strawberry-jam sandwich. Indeed, he turned out to be so nice that now I was feeling guilty for spending Saturday morning with a nice guy somewhere in New Jersey while my husband watched the kids 4000 miles to the East.

If that makes you think my brain is a pretty fucked-up place, it gets worse from here on. That’s because to work off all that guilt, I did what you do to make authors happy: you go buy their books. So, once back in Germany, I went and bought “The End of Science,” 2015 edition. It was not a good idea.

Horgan’s book “The End of Science” was originally published 1996. I never read it because after attempting to read Stuart Kauffman’s 1995 book “At Home In the Universe” I didn’t touch a popular science book for a decade. This had very little to do with Kauffman (who I’d meet many years later) and very much to do with a basic malfunction of my central processing unit. Asked to cope with large amounts of complex, new information, part of my brain will wave bye-bye and go fishing. The result is a memory blackout.

I started having this in my early 20s, as I was working on my bachelor’s degree. At the time I was living in Frankfurt where I shared an apartment with another student. As most students, I spent my days reading. Then one day I found myself in a street somewhere in the city center without any clue how I had gotten there. This happened again a few weeks later. Interestingly enough, in both cases I was looking at my own reflection in a window when my memory came back.

It’s known as dissociative fugue, and not entirely uncommon. According to estimates, it affects about one or two in a thousand people at least once in their life. The actual number may be higher because it can be hard to tell if you even had a fugue. If you stay in one place, the only thing you may notice is that the day seems rather short.

These incidents piled up for a while. Aside from sudden wake-ups in places I had no recollection of visiting, I was generally confused about what I had or had not done. Sometimes I’d go to take a shower only to find my towel wet and conclude I probably had already taken one earlier. Sometimes I’d stand in the stair case with my running shoes, not knowing whether I was just about to go running or had just come back. I made sure to eat at fixed times to not entirely screw up my calorie intake.

Every once in a while I would meet someone I know or answer the phone while my stupid brain wasn’t taking records. For what I’ve been told, I’m not any weirder off-the-record than on-the-record. So not like I have multiple personalities. I just sometimes don’t recall what I do.

The biggest problem with dissociative fugue isn’t the amnesia. The biggest problem is that you begin to doubt your own ability to reconstruct reality. I suspect the major reason I’m not a realist and have the occasional lapse into solipsism is that I know reality is fragile. A few wacky neurons are all it takes to screw it up.

What has any of this to do with Horgan? Nothing, really, but it’s why I didn’t read his book when it came out. And then, when I met John, he unwittingly reminded me of times I’d rather have entirely forgotten.

Back then I took records of my episodes. It looked like it was primarily popular science books that would bring them on, so I stopped reading those. This indeed mostly solved the problem. That and some pills and a few years of psychotherapy. I can only guess why I never had issues with textbooks, maybe because those tend to be rather narrowly focused.

In any case, for ten years the only thing I read besides textbooks was cheap novels, notably Dean Koontz, whose writing is so repetitive and shallow that I have blissfully forgotten what those books were about. Then, in late 2005 Lee Smolin handed me a draft of his book “The Trouble With Physics” which would appear the following year. And what was I supposed to tell him? So I read Lee’s book. My memory lapses came back with a few months delay, but they were nowhere near as disruptive as earlier. And so, thanks to Lee, I slowly returned to reading popular science books.

With the self-insight that age brings, I’ve noticed my mental health issues are strongly stress-related. I’ve also learned to tell first signs of trouble. The past months I’ve worked too much, traveled too much, and said “yes” too often. It took me two weeks to make my way through the first 50 pages of Horgan’s book. It’s not going well. And so I think for now I better go back to reading Chad Orzel’s new book “Breakfast With Einstein.” Because that’s an easy read about things I know already. I’m sorry, John. Don’t take it personally.

Having said this, I thought it would be good to write down some thoughts about the supposed end of science before reviewing Horgan’s book (should I ever manage to finish it). But first let me show you an advertisement:


I don’t particular like American comedy (neither the intended nor the unintended kind) because I tend to find it unfunny. But this guy with his blender makes me laugh every time. Not sure why, maybe it’s his glassy stare.

In case you’ve never encountered these videos before, it seems to be an advert series featuring an old white guy shredding electronics with his awesome blender. “Will it blend?” he asks and infallibly ends up with a pile of grey dust.

I now picture Horgan stuffing science in his blender, pushing the button asking “Will it end?” This thought-experiment teaches us that science will end as infallibly as the Amazon Echo will blend. Because everything will end. You, and I, and John Horgan, and, yes, even Donald Trump’s complaints about the evil media. Entropy increase will get us all, eventually.

So, yeah, science will end.

But that’s not the interesting question. The interesting question is whether it’s ending right now. On the death bed, flatlining as we speak.

As most scientists, I am willing to argue the opposite, though not because I see all that much progress. On the opposite, it’s because I see so little progress. Scientific research today works extremely inefficiently because scientists waste time and money chasing after well-cited publications in high-impact journals. This inefficiency is problematic, frustrating, infuriating even. But it implies that science has untapped potential.

Whether making science more efficient is possible and whether it would actually make a difference I don’t know. I’ll see what John has to say about that. Which I should have done before I wrote my book.

I’m a bad person. And I promise I’ll read his book, eventually.

Monday, October 22, 2018

String theory pros and cons [video - no singing!]

As I mentioned, I used the prize money I got from the recent FQXi essay contest to buy a new video camera. The main reason for doing this was not better music videos (though there’s that), but that my old camcorder doesn’t have a mic port and the audio quality of the built-in mic is miserable to say the least. Turned out, however, that none of my microphones actually worked with the new camera, so you haven’t heard any more of me in the past months than in the prior 10 years.

After too much time in various forums, I finally found a microphone that works with the camera. Miserable failure that I am, I evidently managed to live to the ripe age of 42 without knowing the difference between a TRS and a TRRS plug. But, hey, now I am enlightened! Next problem was that I used the default mic settings of the camera and the result was so incredibly terrible I threw the recordings away entirely and with it went a whole weekend’s work.

Now I believe I’ve found a mic setting that works okay-ish, which is to say that listening to myself does no longer make my toenails roll up. There is still too much echo but that’s an issue which is unfixable for the time being as I have no other place to do the recording. I could have put some filter and compressor on the audio, which I will try next time.

Then of course I noticed once again too late that my white-balance is jumpy. I suspect this isn’t actually something I do, but something I don’t do, probably also some automatic setting that I’d better turn off. Also, I have screwed up my make-up, but you’re probably used to that. All of this is to say it’s more complicated than it seems and really I’d rather stick to writing.


And, erm, “Yadis and Nau” should have been “Nadis and Yau.” Sorry about that. Click on CC in the tool bar to get English captions.

Thursday, October 18, 2018

First stars spell trouble for dark matter

The HERA telescope array
in South Africa 
[img src].
In the beginning, reheating created a hot plasma of elementary particles. The plasma expanded, cooled, and emitted the cosmic background radiation. Then gravity made the plasma clump, and darkness was upon the face of the deep, whatever that means.

Cosmologists call it the “dark ages,” the period in the early universe where matter is already too cool to emit radiation, but not yet clumpy enough to ignite nuclear fusion. At this time the universe was filled almost exclusively with rather dilute hydrogen gas. It’s not until a billion years after the Big Bang that the first stars light up, an epoch poetically called “cosmic dawn.”

We cannot directly measure light emitted from those first stars, but we can indirectly infer the stars’ presence by studying the cosmic microwave background. That’s because the early stars emit UV radiation which couples to the hydrogen gas, and for some while this coupling enables the gas to absorb light of a specific wavelength – at about 21cm. This leaves a mark in the cosmic microwave background.

The wavelengths of light stretch with the expansion of the universe, so what was 21cm back then is now deep in the radio regime. That makes it difficult to find cosmological signals because other sources – both on earth and in our galaxy – can contaminate the data.

In February, the Experiment to Detect the Global Epoch of Reionization Signature (EDGES) announced they had measured the absorption that stems from the first stars. They found it at the expected wavelength – a few meters – but stronger than the predictions said it should be.

Astrophysicists can make predictions for this absorption by using the concordance model for cosmology. This model has 8 free parameters – one of which is the amount of dark matter –  and the physics of the first stars follows from this straight-forwardly. Besides the cosmological dynamics, it’s only well-known thermodynamics and atomic physics. Compared to the large variety of today’s stars, the first stars were fairly simple. Or at least that’s what astrophysicists thought so far.

It took me a while to get around and read the EDGES paper, and I’ve since tried to find out what, if anything, astrophysicists think about the mismatch with the predictions. The answer is: not much. Most of them think it’ll go away. Maybe a measurement error. I have even been told to not pay attention to the EDGES result because the paper has not been cited all that often. Seriously.

Well, as you can tell, I looked at it anyway. I’m not an astrophysicist and I can’t judge the experimental design of the EDGES collaboration. I can only say that I don’t see obvious flaws with their data analysis. The paper seems fine to me.

Besides the possibility of a measurement error, the theoretical explanations for the signal have so far focused on what type of dark matter could possibly make it work, as the commonly considered ones don’t do the trick.

To explain the EDGES result, dark matter would need a stronger interaction with normal matter than presently assumed. If so, that could shift the temperature between gas and radiation and give rise to more absorption. However, a more strongly interacting type of dark matter is difficult to make compatible with other observations that rule out such interactions.

Stacy McGaugh – the modified gravity dude – had the brilliant idea to see what the absorption signal from the first stars would look like if there was just no dark matter. Turns out this would fit remarkably well with the EDGES data. I say “remarkably well” because the parameters that enter his calculation are known from other measurements already, so no freedom to adjust them.

Fig 1 from McGaugh, PRL 121 (2018) 081305.


The reason why the absorption is stronger without dark matter isn’t hard to understand. The more matter there is in the universe, the faster the expansion decelerates. This means without dark matter, the period in which the gas can interact with the radiation is longer, allowing more absorption.

Now, I have recently developed a soft spot for modified gravity, but I am not terribly convinced of Stacy’s argument. It’s one thing to say that galaxies probe different physics than cosmology and thus a new type of force may kick in on galactic scales. It’s another thing to just throw out dark matter from the concordance model because that screws up the whole fit from which the other parameters stem to begin with. You have to self-consistently extract the whole set of parameters from the data – you need a different model entirely.

Indeed, to recover the benefits of dark matter, Stacy employs rather heavy neutrinos. The masses are on the upper end of what is still compatible with constraints. (That’s not counting the cosmological constraints, which are tighter, because these constraints assume the concordance model, and hence don’t apply for modified gravity.) The neutrinos don’t make a difference for the EDGES signal. Still, the dark-matter-less model does not account for the third acoustic peak of the cosmic microwave background. So you have to choose, get the EDGES absorption right or get the third acoustic peak right. Frankly I’d rather have both.

As we speak, other experimental groups are trying to reproduce the EDGES result. One of them is the Hydrogen Epoch of Reionization Array (HERA) in South Africa that is supposed to be completed next year, another one the Square Kilometer Array (SKA) in Australia, planned to be in full operation by 2024. If they confirm the EDGES result, this may be decisive to distinguish dark matter from modified gravity.

Monday, October 15, 2018

Dear Dr B: What do you actually live from?

Some weeks ago a friend emailed me to say he was shocked – shocked! – to hear I had lost my job. This sudden unemployment was news to me, but not as big a surprise as you may think. I was indeed unemployed for two months last year, not because I said rude things about other people’s theories, but simply because someone forgot to renew my contract. Or maybe I forgot to ask that it be renewed. Or both.

In any case, this happened a few times before, and while my younger self wouldn’t normally let such a brilliant opportunity for outrage go to waste, I now like to pretend that I am old and wise and breathe out bullshit.

After some breathing, I learned that this time my sudden unemployment originated not in a forgotten signature, but on Wikipedia. I missed the ensuing kerfuffle about my occupation, but later someone sent me a glorious photoshopped screenshot (see above) which shows me with a painted-on mustache and informs us that Sabine Hossenfelder is known for “a horrible blog on which she makes fun of other people’s theories.”

The truly horrible thing about this blog, however, is that I’m not making fun. String theorists are happily studying universes that don’t exist, particle physicists are busy inventing particles that no one ever measures, and theorists mass-produce “solutions” to the black hole information loss problem that no one will ever be able to test. All these people get paid well for their remarkable contributions to human knowledge. If that makes you laugh, it’s the absurdity of the situation, not my blog, that’s funny.

Be that as it may, I have given a lot of interviews in the past months and noticed people are somewhat confused about what I actually work on. I didn’t write about my current research in my book because inevitably the physicists I criticize would have complained I wrote the book merely to advertise my own work. So now they just complain that I wrote the book, period. Or they complain I’m a horrible person. Which is probably correct because, you see, all that bullshit I’ve been breathing out now sticks to them.

Horrible person that I am, I don’t even work in the foundations of physics any more. I now work on quantum simulations or, more specifically, on using weakly coupled condensed-matter-systems to obtain information about a different, strongly coupled condensed matter system.

The relation between the two systems stems from a combination of analogue gravity with the gauge-gravity duality. The neat thing about this is that – in contrast to either the gauge-gravity duality or analogue gravity alone – we are dealing with two systems that can (at least in principle) be prepared in the laboratory. It’s about the real world!

This opens the possibility to experimentally test the validity of the gauge-gravity duality, or its applicability to certain systems, respectively. Current experiments (like Jeff Steinhauer’s) aren’t precise enough to actually do this, but the technology in this area is rapidly improving, so I’m hopeful that maybe in a decade or so it’ll be doable.

If that was too much terminology, I’m developing new methods to describe how large numbers of atoms interact at very low temperature.

Today, Tobias Zingg and I have a new paper on the arXiv that sums up our recent results. And that’s what I’ll be working on until my contract runs out for real, in November next year. And then what? I don’t know, but stay tuned and we’ll find out.

Thursday, October 11, 2018

Yes, women in science still have a disadvantage.

Women today still face obstacles men don’t encounter and often don’t notice. I see this every day at my front door, in physics, where women are still underrepresented. Among the sciences, it’s physics where the gender-balance is most skewed.

While women are catching up on PhDs, with the ratio now at roughly 20% (US data), women are more likely to leave higher education for good. Among faculty, the percentage of women is down to about 10%. The better the job, the more likely it’s occupied by a man.

Source: APS.

The reasons for this “leaky pipeline” are manifold and no one presently knows for sure which factors are most relevant. Therefore, the question what, if anything, to do about it is hotly debated.

Source: AIP statistics.
On one end of the spectrum are those who think that the current gender-balance correctly reflects qualification and nothing needs to be done. People in this camp explain differences between genders by women’s lack of performance and not discrimination. On the other end are those who think the world will only be a good place if half of physicists are women. In this camp, any under-performance by women is caused by discrimination.

I think both of these extreme positions are unreasonable.

The current gender balance almost certainly does not correctly reflect qualification because a) women have to push back harder on gender stereotypes and b) are more likely to have difficulties combining academia with family life.

Consider this: Last time I got my hair cut, I was informed that women can’t do physics because they can’t think logically. This insight was delivered to me matter-of-factly by a middle-aged, female hair-dresser after I answered her question what I’ve been up to lately. (Wrote a book – About what? – Physics).

I didn’t pursue the matter because I don’t like to argue with people who use sharp instruments near my eyes. But I was glad I didn’t have my daughters with me. They’re still at an age where by default they believe what adults say.

Gender stereotypes like this are everywhere and they almost certainly influence career choices. We all want to belong, and if women feel that science is for men, they’re less likely to pursue this avenue. They lose motivation more easily. They call it quits sooner. This means we are missing out on qualified women and instead fill up the pool with lesser qualified men.

Men tend to underestimate how pervasive these stereotypes are, and how tiresome it is to always be the weird one. It wears you down; doubly and triply so for women who are also members of minorities.

Example: I am constantly accused of being rude, aggressive, and snarky, and have been advised multiple times (by men) to not express myself with certainty. Because that’s offensive, you see. If men behave this way, they’re brilliant geniuses, and one cannot blame them for what is merely an expression of their enthusiasm. I’m left to constantly apologize for being who I am. Or figure this: Last time I won an award, the speaker who was supposed to summarize my achievements felt the need to point out that I do great work despite being short.

Yes, I have a lot of anecdotes, and they can be summarized as a pronounced lack of respect. I’m as much a professional physicist as the men around me, yet I’m not treated the same way. I’m LOOK-A-WOMAN!

However, I try not to draw conclusions from my own experiences because the very fact that I’m still working in physics is evidence I’m not suffering all that greatly. I’d go so far to say most of my colleagues are nice guys, and even the assholes don’t normally mean to be assholes. But I hear what my female colleagues have to say, and most of them are really frustrated about the nonsense they have to deal with. And, yeah, I too have been mistaken for the secretary.

The other major disadvantage that women face is that they are hit harder by the family-unfriendliness of academia. Turn it how you wish, women are still responsible for procreation, and female fertility rapidly declines past the age of 40.

Unfortunately, the years between 35 and 40 are also critical to establishing yourself as researcher. Most physicists presently don’t secure permanent positions until their early 40s. Up to then, job-hopping and frequent international moves are the norm. Taking time off to raise kids is difficult, and the inevitable decrease in academic output and flexibility is a competitive disadvantage. A priori this disadvantage exists both for men and women, but women are on the average the younger part of the couple, and, needless to say, pregnancy and nursing is an extra burden.

Arguments that women just perform worse than men are fundamentally flawed because they’re based on shaky interpretation of data that don’t quantify what they’re supposed to show. Data say, for example, that women in science publish less and their papers are cited less often (references here). Does this mean women are less capable of doing science? No, it means that they publish less and their papers are cited less often. To put it differently, it means that women’s papers are less popular with their – predominantly male – colleagues. How about not judging women by how much their work appeals to men?

The recent case of Alessandro Strumia is an example for such shaky interpretations of data. There’s no simple way to measure whether women are having a harder time with their research. There could be all sorts of reasons, from the lack of role models to difficulties getting funding to being more frequently asked to sit in committees because, well, we need at least one woman, you see? None of these difficulties will reflect in publication records in any obvious way.

Besides, there’s no agreed-upon measure for academic success. Indeed, it is well known that rewarding a high number of publications and citations creates perverse incentives, and therefore many scientists now compete to excel on meaningless performance indicators. So why are we even talking about this?

This is not to say that the differences between publications of men and women are not interesting or should not be studied. Just that one shouldn’t jump to conclusions from them.

But I am also not an advocate of a gender-balance of one-to-one. Biological factors, such as muscle strength, arguably play a role for some professions. I find it highly questionable that a profession like physics, which mostly deals with abstract ideas, is much influenced by genetic factors. But it’s a controversial subject and, as they say, more research is needed. Regardless of whether the reason is nature or nurture, however, women demonstrate preferences different from men, and it doesn’t make sense to push them into disciplines that they might not currently feel well in.

For this reason I cannot support policies that aim to increase the ratio of women based on the premise that 50:50 is the “correct” target. It’s not only that this risks we’ll hire women who are less motivated or qualified than some men who don’t get jobs, it also pisses off the men who feel like they are now the ones at a disadvantage. And this creates yet another bias, namely the belief that women now have an easier time rather than merely having less of a disadvantage. Again the recent Strumia-case is a good example.

Having said this, some people tell me it’s justified to risk hiring lesser qualified women, at least temporarily, to restore what they consider fairness in the long run. But at this point we are down to a value-decisions. What is more important to you: That science works most efficiently, or that women catch up with men as quickly as possible? I think this is the key question which no one wants to discuss.

Be that as it may, the easiest way to disqualify yourself from any discussion on the matter is to simply disregard the existing hurdles that women face. The problems are real, and they’re far from vanishing.


If you want to make a difference, check out the Practical Guide to Improving Gender Equality in Research Organisations by Science Europe, and raise awareness for the therein proposed changes among your colleagues.

Friday, October 05, 2018

Gender-bias in Academia: The Case Strumia

Dr Alessandro Strumia is a physicist working at CERN, where he is Principal Investigator of an ERC Advanced research grant on the topic “Understanding the mass scales in nature.” Ie, he spends your tax money on the kind of research that I criticize in my book. He also recently published an analysis of publication and citation-rates in his field. At a workshop on gender diversity at CERN last weekend, he used this analysis to argue that women are underrepresented in physics because they are underperforming.

I did not attend the workshop and have not seen a recording of the talk, but I have seen the slides (a PDF version of which is here). The slides contain statements that are both inaccurate and exceedingly unprofessional.

For example, he begins his talk by stating that “smarter people are less affected by implicit bias,” but this is wrong. Studies have shown repeatedly that intelligence does not protect from thinking biases. Yes, intelligence is useful to overcome certain types of biases (mostly those that can be exposed with mathematical reasoning), but only once people are aware they are biased to begin with.

Strumia’s mistaken belief that intelligent people are less affected by cognitive biases does not remotely surprise me. I have encountered this very same attitude (“We are too smart to be biased!”) among almost all high-energy theorists and phenomenologists I have spoken with about the issue. That in itself is a bias, known as the “bias blind spot.”

But that Strumia is ill-informed about the very topic he speaks about at a scientific workshop is not the biggest problem with his presentation. Far worse is that he names and attacks two women, apparently because he is annoyed he did not get a job that he was shortlisted for. Nonsense like this just does not belong in a research presentation.

After complaints ballooned on social media on Monday, CERN pulled the slides from the net quickly and has since suspended Strumia. What will happen to his ERC grant is unclear. A large number of members of the particle physics community have meanwhile signed a statement declaring that they distance themselves from the content of Strumia’s talk.

Now, as you know, I have also recently taken up bibliometric analysis. And I admit I found some of the data Strumia showed interesting. We did, in our paper, also look at gender differences, but not for citation counts. We looked at an entirely different quantity, that of research broadness, and for this we did not find any gender differences.

The gender difference that Alessandro Strumia and his co-author Ricardo Torre find is huge. It’s a more than 100% difference in the total number of citations that researchers accumulate throughout their career.

I don’t think that the number of citations is a good measure for scientific performance, but if the difference between the genders was so large, it might mean that women and men chose their research projects in distinctly different ways. That would be interesting. I thus decided to look into this for a bit.

The key figure that Strumia presents on his slides is the total number of citations that researchers accumulate since the publication of their first paper:

Figure from slide 16 of Alessandro Strumia’s talk.

That the horizontal axis is labeled “scientific age” is unfortunate because this term has been coined to emphasize that the scientific age might differ from the chronological time passed since PhD or first paper. If a researcher takes a career-break, for example because of health reasons or for parental leave, their scientific age goes on hold. However, there is presently no standardized way to determine the scientific age, and in any case, you couldn’t do it from publication data alone, you’d also need biographic information.

Since women are more likely to take leave for child-raising, their citations should on the average increase somewhat slower, simply because they have more breaks in which they don’t publish. However, it seems unlikely that this would make such a huge difference. So, while the label on the axis is inaccurate, I don’t think it’s all that relevant.

When I saw this graph, however, another worry came to my mind immediately. When we did our previous analysis, we found that the vast majority of people who use the arXiv publish only one or two papers and are never heard of again. This is in agreement with the well-known fact that the majority of physicists drop out of academic careers.

I am not sure why this surprised me when it showed up in the data. Maybe because, if you work in the field, the drop-outs are pretty much invisible. They leave and you forget about them. But they are there, in the stats, big and fat.

Now, the total number of citations for such drop-outs will accumulate very slowly because they don’t publish new papers any more. And we know that women are more likely to drop out – that’s the “leaky pipeline” and reason why I find myself increasingly often, if not the only woman in the room, then at least the oldest woman in the room. And I’m only 42.

If you leave the drop-outs in the citation analysis, the leaky pipe will pull down the average of female authors more than of male authors.

I hence asked one of our PhD students, Tobias Mistele, to plot the same quantity as Strumia did for our data sample, but to only keep authors who have more than 5 papers in total, and who have published a paper in the last 3 years. This is sloppy way to shrink down the pool to “active researchers only.” It’s maybe not the most sophisticated way to do it, but it should give us an idea how large the contribution from the drop-outs is.

If we plot the number of citations for active researchers only, we see no noticeable difference between men and women:
arXiv data, active researchers only

When normalized to the number of authors per paper (as Strumia did), there is also no noticeable difference between men and women.

I must add a warning here. We do not use the same data set as Strumia and Torre. They use data from inspire, we use data from the arXiv. This means our data set does not reach back in time as far, and it includes disciplines besides high energy physics. So the absolute numbers are not directly comparable.

Another caveat I must add is that we are using a different method to identify male and female authors. We use the author-id algorithm that is explained in our earlier paper, and then try to match first names with a database for common anglo-saxon first names. Naturally, this means that the authors who remain in our sample are most likely to be of Western origin. By this method we assign a gender to 19% of authors. This is in contrast to Strumia and Torre who use a more elaborate gender-id procedure that allows them to match 60%. The remaining authors in our sample break down to 70,295 male und 53,165 female researchers. After applying the above mentioned cuts, we are left with 12,654 male and 8,177 female. That’s not a huge number, but decent.

Let me also mention that probably a similar effect is behind another finding in Strumia’s talk. He points out that women, on the average, are hired into faculty positions earlier. A paper that appeared on the arXiv yesterday argued that this is not a signal that women have an unfair advantage, but simply a consequence of women leaving at a higher rate. If they aren’t hired early, they’ll not be hired at all, which means the average age of hiring is smaller.

Finally, to state the obvious, this is a blogpost, not a paper. The above is a quick and dirty way to check whether removing dropouts significantly affects the large difference between men and women, and the answer seems to be yes. However, we will have to do a more careful analysis to arrive at definite conclusions. I haven’t checked my biases.


I want to thank Tobias Mistele for doing the graphs so quickly and Alessandro Strumia and Ricardo Torre for helpful communication.

Thursday, October 04, 2018

You say theoretical physicists are doing their job all wrong. Don’t you doubt yourself?

This is me with John Horgan, yesterday.
This photo is only here so
the share widgets work properly.
One of the most frequent critical remarks I have gotten on my book is that I seem confident. I was supposed, it seems, to begin each paragraph with “I’m sorry, but.”

But I am not sorry. I mean what I say. Yes, in the foundations of physics we are financing some 15,000 or so theorists who keep producing useless scientific articles because they believe the laws of nature must be beautiful. That’s exactly what I am saying.

Let us leave aside for a moment that you have to skip half the book to not notice I question myself on every other page. Heck, if you ask me to sign the book, I’m afraid I’ll misspell my own name. I’m a walking-talking bag of self-doubt. Indeed that was the reason I ended up writing this book.

See, I don’t understand what’s going on with this community. Everyone knows there’s no reason that a scientific explanation must appeal to the human sense of beauty. Right? Doesn’t everyone know this? Science is about explaining observations, regardless of whether we like these explanations.

But if it’s clear that putting forward new hypotheses just because they are beautiful doesn’t mean they’re likely to be right, then why do theorists in these fields focus so much on beauty? Worse, why do they continue to focus on the same type of beauty, even though that method has demonstrably not worked for 40 years?

At first I considered there might be a mathematical basis to their arguments which I was missing. That there is a solid reason why a theory must be natural, or that the fundamental forces must be unified, or that the mathematics of a theory must be “fruitful” and “have deep connections” and be “rigid” – to quote some expressions people in the foundations of physics commonly use. But there is no mathematical basis. Arguments from beauty are additional assumptions, and they are unnecessary to make a theory work.

Indeed, some philosophers have suggested I speak of “metaphysical assumptions” rather than “aesthetic arguments”, but I think the latter captures the historical origin better. These arguments trace back to tales about God’s beautiful creations. Also, if I’d call it metaphysics no one would know what I am talking about.

I then considered that using criteria from beauty is justified because it has historically been successful. This would leave open the question why that would be so – I cannot think of a reason such a connection should exist. But in any case, history speaks against it. Relying on beauty has sometimes worked, and sometimes not. It’s just that many theoretical physicists prefer to recall only the cases where arguments from beauty did work. And in hindsight they then reason that the wrong ideas were not all that beautiful. Needless to say, that’s not a good way to evaluate evidence.

Finally, the use of criteria from beauty in the foundations of physics is, as a matter of fact, not working. Beautiful theories have been ruled out in the hundreds, theories about unified forces and new particles and additional symmetries and other universes. All these theories were wrong, wrong, wrong. Relying on beauty is clearly not a successful strategy.

So I have historical evidence, math, and data. In my book I lay out these points and tell the reader what conclusion I have drawn: Beauty is not a good guide to theory-development.

I then explain that this widespread use of scientifically questionable but productive methodology is symptomatic to the current organization of academic research, and a problem that’s not confined to physics.

Now, look, just because I cannot find a reason that beautiful theories are more promising than ugly ones doesn’t mean that relying on beauty cannot work. It may work, if we get lucky. Neither, for that matter, do I think that if we find a new law of nature it must be ugly. Chances are we will come to find a successful new idea beautiful simply because it works. But our sense of beauty changes and adapts, and therefore I do not think that using criteria of beauty from the past is a promising route to future progress.

Needless to say, making a case against a community of some thousands of the biggest brains on the planet has not been conducive to my self-confidence. But I have tried to find a scientific reason for the methods which my colleagues use in theory-development and could not. I wrote the book because I think it’s my responsibility as scientist to say clearly that I have come to the conclusion what goes on the foundations of physics is a waste of money, and that the public is being misinformed about the promise of this work.

I do not think that this will change the mind of people in the field. They have nothing to worry about because the way that academia is currently organized there is safety in numbers.

So, yes, I doubt myself. But I have written a whole book in which I explain why I have arrived at my conclusion. Rather than asking me, you should ask the people who work in these fields what makes them so certain that beautiful ideas are promising descriptions of nature.