Wednesday, May 08, 2013

What do "most physicists" work on?

It always amazes me how skewed the image of physics research in the popular press is. To begin with, the amount of coverage is totally unrepresentative for the actual amount of research on a given topic. Controversial and outright fantastic topics are typically hotly discussed, so is everything that captures the public imagination. On the other hand, down-to-earth research like soft condensed matter or statistical mechanics rarely makes headlines.

The field I work in myself, quantum gravity, is among the over-represented fields. If you believe what you read, the quest for quantum gravity has become the "holy grail" of theoretical physicists all over the planet, and we're all working on it because the end of science is near and there's nothing else left to do.

Since coverage by the media is driven by popularity and not by relevance, one can expect such a skewed representation. It probably isn't much different in other areas of our lives. (Who actually wears those wacky clothes that fashion designers celebrate?) What bothers me much more than the skewed selection of topics is how their relevance is misrepresented even in these articles. I must have read hundreds of times that "many physicists" believe this or that, while in reality most physicists couldn't care less and probably have no opinion whatsoever.

Here are some examples:
"According to the current thinking of many physicists, we are living in one of a vast number of universes. We are living in an accidental universe. We are living in a universe uncalculable by science."
Alan Lightman, The Accidental Universe.
"The team’s verdict, published in July 2012, shocked the physics community."
Zeeya Merali, in a recent nature issue, Astrophysics: Fire in the hole!. We note in the passing that the article doesn't have much, if anything, to do with astrophysics.
"Most physicists believe that space is not smooth, but it is rather composed of incredibly small subunits, much like a painting made of dots. This micro-landscape is believed to host numerous black holes..."
Mihai Andrei, in an article titled Finding black holes at a quantum scale about a deeply flawed paper by Jacob Bekenstein. (Which, depressingly, got published in PRD.)

But why limit ourselves to physicists, let's be bold:
"Many scientists claim that mega-millions of other universes, each with its own laws of physics, lie out there, beyond our visual horizon. They are collectively known as the multiverse."
George F. R. Ellis, Scientific American, Does the Multiverse Really Exist? "They" presumably refers to the "other universes," and not to the "many scientists".

So then let's try to quantify "most physicists" by estimating an upper bound on the fraction of physicists who are working on these topics, a sub-area of quantum gravity. The topics under question here tend to appear on the arXiv under hep-th cross-linked to gr-qc or the other way round. That there is no subject category for "quantum gravity" should already tell you that there aren't all that "many" people working on it. First let us have a look at the arXiv submission rates

The left graph shows the total number of submissions, the right shows the percentage. Blue, which presently accounts for about 10%, is high energy physics and collects hep-th+hep-ph+hep-lat+hep-ex. Note that for historical reasons hep is likely to be over-represented in the arXiv statistics relative to the actual distribution of researchers. In hep, pretty much every paper goes on the arxiv, but the same is not true in other areas (at least not yet). Also, hep tends to be a very productive and communicative field, so looking at the number of arXiv submissions rather than researchers is probably an over-estimate. Be that as it may, the topics we are looking for almost certainly occupy less than 10% of researchers.

More data that tells you that the vast number of physicists aren't working on anything related to quantum gravity can be obtained from the number of members in sections of the German Physical Society. The section on Particle Physics (which includes beyond the standard model physics and quantum gravity) has about 2,500 members. The section on Quantum Optics and Photonics has more than 3,000 members, Physics of Semi-conductors 3,800Low Temperature Physics 1,450, Atomic Physics together with Hadronic and Nuclear Physics come to about 3,000, Material Physics together with Chemical and Polymer Physics and Thin Films another 3,500. Not all sections have membership numbers online, so this doesn't cover the full spectrum. But this already tells you that "most physicists" don't even do high energy physics, certainly not quantum gravity, and have no business with multiverses, firewalls, or "micro-landscapes of black holes".

But we can try to get a better estimate by seeing how many papers are cross-linked from hep-th to gr-qc, assuming that the opposite cross-linking is similarly frequent. For this, we look at the submission statistics of gr-qc for the first four months of the year 2013. It lists the submissions as well as the cross-lists. Click on any of the months, select "show all" and count the number of times "cross-list from hep-th" appears on the page. The numbers I get for January to April are: 70,71,52 and 67. If you look at the titles, you'll note that the papers you find this way fit well to the topics we're looking for.

Comparing these numbers with the total arxiv submissions per month (about 7500), we can estimate that it's about 1%. Multiply by two to account for gr-qc cross-linked to hep-th.

Now this is a rather crude estimate and I have mentioned several reasons why it's inaccurate: 1) Some fields of research are not as well represented on the arXiv as is hep-th. This means 2% is still an over-estimate. 2) Some fields might be more productive in paper output than others. If hep-th is on the more productive side, this means the 2% is even more of an over-estimate. 3) Not every paper in the area we're looking for might be hep-th cross-linked to gr-qc or vice-versa. This leads to an under-estimate. 4) On the other hand, not every paper cross-listed as such is about quantum gravity or related topics. 5) There are probably more people following the literature than actively working on it, which also leads to an under-estimate.

However, even if you'd add up all these errors, you would still be left to conclude that the above quoted uses of "most physicists" or "physics community" are extremely inaccurate and misleading.


Navneeth said...

Most physicists would agree. :-)

What do the 'widths' or each "band" represent in the graphs?

Sabine Hossenfelder said...

In the left graph it's the total number of submissions, in the right graph it's the fraction. The color code is on the arXiv page that I linked to, together with some more information (and further stats).

Nemo said...

Amen ... :-)

George Musser said...

It's a point well taken. Few writers conduct opinion polls to justify these remarks anyway. Another issue is: what does "many" mean? More than three? But these are not really meant literally, but as convenient shorthands to assign an idea to some group of people while avoiding the use of passive voice. In fact, your post does something similar! Although it's careful not to say "most writers", the post definitely carries the implication that the use of "most physicists" is extremely common, even though the post produces only anecdotal evidence for this.

Sabine Hossenfelder said...

Hi George,

Yes, point well taken. But then I didn't say that "most writers" do this :) I also understand the need to be brief and to get to a point quickly. But sometimes it seems to me the accurate phrase would have been something like "I talked to three string theorists and they thought it was interesting" and not "most physicists think".

I guess what I am missing is some qualifier for the sample, something like "among researchers studying this or that question". Or why not "in the quantum gravity community". I know it's clumsy, but this irks me for two reasons. First, readers get the impression that "most physicists" are working on things that are totally irrelevant for the lives of most people on the planet, and are dangerously close to philosophy. Second, they come to believe that the topic of quantum gravity is much better populated, and thus supported, than it actually is. As a matter of fact, as I guess "most" of my colleagues can confirm, quantum gravity is not a topic funding agencies get very excited about because it's so detached from application. That's a problem, I think, because this really basic research is essential for progress in the long run, and I wish that this problem was better known. Best,


Giotis said...

Hi Sabine,

But I guess you must have a picture of the reader in mind for such conclusions. The average reader (I hate the average characterization BTW but anyway) interested in science and physics would I understand that when we say 'many physicists', it is implied that we are talking about the physicists working in the specific field.

But if the reader is the average Joe well then all bets are off; but yet again such reader would read science columns? I don't know...

Uncle Al said...

Physics disallows non-interaction (mesonic heavy nuclei), shielding (emvelope, idstrong), and cloaking (meta-materials) for gravitation. Even Wesley Crusher dark matter obeys gravitation. String theory rescales gravitation's magnitude by allowing leakage Elsewhere. Given the multiverse, everywhere is Elsewhere. (If Elsewhere is really Elsewhen, we are in for a pranging. No, wait! Big Bang. More studies are needed.)

Religions and economics, their literature and their subscribers, are legion. Each is locally inerrantly rigorous and globally empirically wrong. When riding a horse...

Plato Hagel said...

Hi Bee,

It sounds like your gearing up for,"Upcoming Science Writers Workshop at Nordita."

Bee:"[A]n environment with a very high interaction rate thermalizes quickly, and can be very destructive in the early stage of an idea's development. A highly connected community means we’ll have to watch out very carefully for sociological phenomena that might affect objectivity, and work towards premature consensus. We will have to watch out for fads that grow out of proportion, and we will have to find a way to protect the young ideas that “you have to ram down people's throats,” in Atkin's words, until people are ready to swallow them. There is no reason to assume scientists are immune to sociological effects."

With the wisdom I gathered out of Surowiecki's book, the point I was trying to make is that sharing too much information and being too tightly connected will actually lead to a dumb rather than a smart community.
See: And how open would you want your science?

The good thing about science bloggers is they think beyond their profession?:) You have been appreciated beyond words.

What may be thought as trendy may in fact be an attractor in society(as it goes top down from the social background of the 3% scientists holds in society[you have further divided the science population up]) bottoms out to a society of those who are truly mystified, and feel a mystery needs to be explained.

The tax basis conferred to research and development needs an explanation? Give it your best shot. What is the pearl for such presentation, in the mean time, you have to make a living.


Plato Hagel said...

In effect are algorithm constraints which are ruling our search features?


George Musser said...

Sabine: Plato Hagel makes a good point. Let's bring up this whole issue at the workshop. Another troublesome word is "recently", which basically translates as "my editor insists that I make the subject sound timely".

I think it’s overly self-deprecating to say these issues are "dangerously close" to philosophy or "totally irrelevant" to the lives of most people. (There's that word "most" again!) What does it mean to be "relevant"? Is art relevant? Music? Anything other than oxygen, food, water, and sanitation? The reason science journalists disproportionately report on quantum gravity is that they, and readers, love it. A lot of people clearly do find it relevant, presumably in the same way they find art and music relevant as ways to give life meaning beyond the purely utilitarian.

Rastus Odinga Odinga said...

Among people who love music, there is more interest in Bach than in Korngold. Does that surprise you? Why then are you surprised that more people are interested in deep questions than in superficial ones which, in many cases, are of "interest" only because money is to be made from studying them?

Anonymous Snowboarder said...

Bee - 4 out of 5 dentists surveyed recommend sugarless gum for their patients who chew gum!

Couldn't help but remember that while reading your rant about "many".


Sabine Hossenfelder said...

Hi George,

If you'd do a global survey with representative samples from each nation and you'd ask them what is relevant to their well-being, I guess you would find art and music listed quite frequently. I seriously doubt that a significant amount of people would list quantum gravity.

I don't think I'm "overly self-deprecating" when I say quantum gravity is not relevant to the lives of most people. Yes, the topic has its readership, I'm not denying that. But it is abundantly clear that the work done in this area isn't going to improve anybody's life anytime soon. The problem is to communicate why it's relevant in the long run to pursue this research. There isn't enough awareness for this necessity and there isn't remotely enough support for long-term projects in basic science. That's what I'm saying. And that's why I find it unfortunate if articles covering work in this area create a different impression.

Yes, we can keep this in mind as a topic for the discussion sessions (though this wasn't my intention when I wrote this post). Best,


Hamish said...

Bee, I'm struggling to understand what you are saying in the context of topics that you have covered in your blog. On one hand you seem to be berating journalists for suggesting that, for example, a polymer physicist could have a meaningful opinion about quantum gravity. Yet in the previous blog entry you (a quantum gravity specialist) are sharing your insights into polymer physics!

George Musser said...

@Sabine Let's talk about this more in two weeks. I have traveled a lot in developing countries, including poor and remote villages, and, although people are unlikely to list quantum gravity per se, I have encountered fascination about the Big Questions of science everywhere I go.

Plato Hagel said...

IN recognition of the algorithmic foundation of the search function in google, how will you appear as "number 1?"

Relevancy is not constructed as to appear in such a location but is more to the question of what you offer as an explanation of the understanding about gravity.

So, "beyond the affect of geometrical insight" as to confer the realization that this exists as an abstraction in society. It is not just some "mathematical construct" without some relevance to our understanding. So that, is a starting point in my view.

It is accepting a foundational change not just in the way we had viewed reality in our historical past, but of how such paradigmatic change had brought society forward to think in such a casual and assuming way. The realization then, that experiments have progressed that casual conversation and experience. GPS. Relativity and Muons. New views of the earth and moon.

For some it might mean more....and to think of probabilistic bath of opportunities in context of society, what are the chances that an answer will come forth from such counting? It's about opportunity? Not me:)

An anomaly being witness then may defy everything that is being taught, but what is taught cannot be ignored. It just leaves, the ever unending mystery displaying a route to possibly solving that experience, or not.:)


Plato Hagel said...

While mirror neurons are important that is not explicitly what I wanted to show. What I wanted to show is this.

Sabine Hossenfelder said...

Hi Hamish,

I think you're misunderstanding me. I'm not saying that a polymer physicist can't have an opinion on recent research in quantum gravity. I'm saying it's unlikely he does, and in addition it's exceedingly unlikely he'll be asked for it by some journalist. Re my previous blogpost, microfibre cloths might be many things, but they're not recent research. I have no clue, really, what people are presently working on in this area. Besides this, how many quantum gravity researchers do you know who blog on occasion about soft condensed matter? Best,


slipper-mystery said...

see arxiv.culturomics searches for "quantum gravity":

Zephir said...

Because string theory was never accepted by mainstream in its entirety due the [1] general lack of falsifiability, the string theorists had started to use arXiv portal as their alternative publishing platform like squatters, thus by-passing standard process of peer-review of mainstream physics. The gradual decline of (experimental successes of) high-energy physics and era of stringy/SUSY theories manifests itself with gradual decline of preprints dedicated to this topic at ArXiv.

eddie mccloy said...

If a string theorist falls into a black hole. Is any information actually lost?

Zephir said...

This is a good comparison. The society inside of informational black hole can still exchange the information in longitudinal waves. In terms of informational space it means, the other people will essentially f*ck their quantitative predictions (bacause they don't work anyway) - but the qualitative insights could be recycled into another theories and models.

The problem is, the quantitatively thinking physicists tend to throw out their babies with the bath water at the moment, when it turns out, that these formal theories don't fit the LHC or WIMPs data at the quantitative level. This leads to somewhat paradoxical situation: despite I'm not proponent of blind formal approach to physics in any way, I can see the predictions of formal theories even at the places, which the strictly formally thinking physicists didn't considered so far. I can still see some relevance of these theories at the qualitative level and/or inside of systems of low dimensionality.

For example, you can study the superstring theories with nested density fluctuations inside of condensing supercritical fluids, which are simple enough for to keep this schematic theory working. And I don't think, that the physicists disproved the extradimensions with lack of formation of black holes at LHC, because they prepared the artificial atom nuclei instead during this.