Last week, Sabine Hossenfelder wrote a post entitled “Why not string theory?” In it, she argued that string theory has a much more dominant position in physics than it ought to: that it’s crowding out alternative theories like Loop Quantum Gravity and hogging much more funding than it actually merits.
If you follow the string wars at all, you’ve heard these sorts of arguments before. There’s not really anything new here.
That said, there were a few sentences in Hossenfelder’s post that got my attention, and inspired me to write this post.
So far, string theory has scored in two areas. First, it has proved interesting for mathematicians. But I’m not one to easily get floored by pretty theorems – I care about math only to the extent that it’s useful to explain the world. Second, string theory has shown to be useful to push ahead with the lesser understood aspects of quantum field theories. This seems a fruitful avenue and is certainly something to continue. However, this has nothing to do with string theory as a theory of quantum gravity and a unification of the fundamental interactions.
Here, Hossenfelder explicitly leaves out string theorists who work on “lesser understood aspects of quantum field theories” from her critique. They’re not the big, dominant program she’s worried about.
What Hossenfelder doesn’t seem to realize is that right now, it is precisely the “aspects of quantum field theories” crowd that is big and dominant. The communities of string theorists working on something else, and especially those making bold pronouncements about the nature of the real world, are much, much smaller.
Let’s define some terms:
Phenomenology (or pheno for short) is the part of theoretical physics that attempts to make predictions that can be tested in experiments. String pheno, then, covers attempts to use string theory to make predictions. In practice, though, it’s broader than that: while some people do attempt to predict the results of experiments, more work on figuring out how models constructed by other phenomenologists can make sense in string theory. This still attempts to test string theory in some sense: if a phenomenologist’s model turns out to be true but it can’t be replicated in string theory then string theory would be falsified. That said, it’s more indirect. In parallel to string phenomenology, there is also the related field of string cosmology, which has a similar relationship with cosmology.
If other string theorists aren’t trying to make predictions, what exactly are they doing? Well, a large number of them are studying quantum field theories. Quantum field theories are currently our most powerful theories of nature, but there are many aspects of them that we don’t yet understand. For a large proportion of string theorists, string theory is useful because it provides a new way to understand these theories in terms of different configurations of string theory, which often uncovers novel and unexpected properties. This is still physics, not mathematics: the goal, in the end, is to understand theories that govern the real world. But it doesn’t involve the same sort of direct statements about the world as string phenomenology or string cosmology: crucially, it doesn’t depend on whether string theory is true.
Last week, I said that before replying to Hossenfelder’s post I’d have to gather some numbers. I was hoping to find some statistics on how many people work on each of these fields, or on their funding. Unfortunately, nobody seems to collect statistics broken down by sub-field like this.
As a proxy, though, we can look at conferences. Strings is the premier conference in string theory. If something has high status in the string community, it will probably get a talk at Strings. So to investigate, I took a look at the talks given last year, at Strings 2015, and broke them down by sub-field.
Here I’ve left out the historical overview talks, since they don’t say much about current research.
“QFT” is for talks about lesser understood aspects of quantum field theories. Amplitudes, my own sub-field, should be part of this: I’ve separated it out to show what a typical sub-field of the QFT block might look like.
“Formal Strings” refers to research into the fundamentals of how to do calculations in string theory: in principle, both the QFT folks and the string pheno folks find it useful.
“Holography” is a sub-topic of string theory in which string theory in some space is equivalent to a quantum field theory on the boundary of that space. Some people study this because they want to learn about quantum field theory from string theory, others because they want to learn about quantum gravity from quantum field theory. Since the field can’t be cleanly divided into quantum gravity and quantum field theory research, I’ve given it its own category.
While all string theory research is in principle about quantum gravity, the “Quantum Gravity” section refers to people focused on the sorts of topics that interest non-string quantum gravity theorists, like black hole entropy.
Finally, we have String Cosmology and String Phenomenology, which I’ve already defined.
Don’t take the exact numbers here too seriously: not every talk fit cleanly into a category, so there were some judgement calls on my part. Nonetheless, this should give you a decent idea of the makeup of the string theory community.
The biggest wedge in the diagram by far, taking up a majority of the talks, is QFT. Throwing in Amplitudes (part of QFT) and Formal Strings (useful to both), and you’ve got two thirds of the conference. Even if you believe Hossenfelder’s tale of the failures of string theory, then, that only matters to a third of this diagram. And once you take into account that many of the Holography and Quantum Gravity people are interested in aspects of QFT as well, you’re looking at an even smaller group. Really, Hossenfelder’s criticism is aimed at two small slices on the chart: String Pheno, and String Cosmo.
Of course, string phenomenologists also have their own conference. It’s called String Pheno, and last year it had 130 participants. In contrast, LOOPS’ 2015, the conference for string theory’s most famous “rival”, had…190 participants. The fields are really pretty comparable.
Now, I have a lot more sympathy for the string phenomenologists and string cosmologists than I do for loop quantum gravity. If other string theorists felt the same way, then maybe that would cause the sort of sociological effect that Hossenfelder is worried about.
But in practice, I don’t think this happens. I’ve met string theorists who didn’t even know that people still did string phenomenology. The two communities are almost entirely disjoint: string phenomenologists and string cosmologists interact much more with other phenomenologists and cosmologists than they do with other string theorists.
You want to talk about sociology? Sociologically, people choose careers and fund research because they expect something to happen soon. People don’t want to be left high and dry by a dearth of experiments, don’t feel comfortable working on something that may only be vindicated long after they’re dead. Most people choose the safe option, the one that, even if it’s still aimed at a distant goal, is also producing interesting results now (aspects of quantum field theories, for example).
The people that don’t? Tend to form small, tight-knit, passionate communities. They carve out a few havens of like-minded people, and they think big thoughts while the world around them seems to only care about their careers.
If you’re a loop quantum gravity theorist, or a quantum gravity phenomenologist like Hossenfelder, and you see some of your struggles in that paragraph, please realize that string phenomenology is like that too.
I feel like Hossenfelder imagines a world in which string theory is struck from its high place, and alternative theories of quantum gravity are of comparable size and power. But from where I’m sitting, it doesn’t look like it would work out that way. Instead, you’d have alternatives grow to the same size as similarly risky parts of string theory, like string phenomenology. And surprise, surprise: they’re already that size.
In certain corners of the internet, people like to argue about “punching up” and “punching down”. Hossenfelder seems to think she’s “punching up”, giving the big dominant group a taste of its own medicine. But by leaving out string theorists who study QFTs, she’s really “punching down”, or at least sideways, and calling out a sub-group that doesn’t have much more power than her own.
I find your, erm, “analysis” of my supposed opinion utterly bizarre.
To begin with, I used to go to the string pheno conference and know pretty well what’s the size and status of that community within string theory.
Second, I never said anything to the extent that string theory is “crowding out alternative theories like Loop Quantum Gravity”. You should try to to find out why you think this is something I said (or wrote). Is it maybe possibly that you have heard someone else proclaim that this is an opinion I hold? If you look at what I actually wrote (see eg most recent post) I actually expressed there very clearly that the LQG community suffers from much the same problems as string theory. I would really appreciate if you stopped assigning opinions to me I do not hold.
Third, it’s hilarious that you think I’m leaving out the QFT/Holography arch because I supposedly believe it’s a small area. I sincerely wonder what would make you think that I might have missed its dominance, given that I recently posted a postdoc position in the field? Do you think I’d get multiple research grants in an area I know so little about that I don’t even know its position within hep-th? Do you think I’m unable to see what’s posted on the arXiv or what the applicants to our positions have in their CV? I wrote explicitly that I’m leaving it out because at least for the purpose of that post I was interested only in the toe-claim. (It’s the claim that Richard Dawid focuses on in his book.)
Your assertion that I am “punching” anyone I can only interpret as a projection. Clearly you believe you have to “fight” me in some way. What are you afraid of?
You’re welcome to contact me by email for further clarification.
LikeLiked by 1 person
I don’t mean to imply that you, personally, are ignorant of the size of the communities involved. (In my experience, most of the lay audience is, and very few discussions of this subject clarify things for them. Similarly, I mentioned LQG not because I think you advocate for it, but because it’s the only “alternate route” most layfolk have heard of.)
That said, you have straightforwardly said that string theory is overpopulated. The question becomes, overpopulated with respect to what? If you are indeed just focusing on the TOE-claim, the community you are describing as overpopulated is (as you’ve seen by going to their conferences) extremely small, and not especially powerful.
(I wouldn’t count people like Gross and Schwarz here, since their recent work isn’t really along these veins and their public statements of confidence in string unification don’t seem to translate into increased funding/positions for fields like string pheno. Dawid seems even more irrelevant. But maybe this is where we disagree?)
In a fair world where people weren’t influenced by sociological pressure, roughly how big do you think string pheno, string cosmology, and the like would be? If you can clarify that, maybe I can get a better idea of what you were arguing.
LikeLiked by 2 people
Unfortunately I get the impression you either haven’t read or not understood the point I was trying to make. I am saying that string theory is almost certainly overpopulated because academia is currently organized so that the rich get richer and the not-so-rich never get anywhere. The consequence of this is that the population of research topics does not accurately reflect the promise of this topics.
Why you want me to tell you how large research areas “should” be? I don’t know. And if I would tell you what I expect, of what use would my opinion be? What I am asking for is simply that scientists wake up and recognize that social dynamics and cognitive biases affect all humans – and scientists are not immune to them. It is ridiculous to deny that these play a role in how physicists select the topics they work on. If we would put into place measures to prevent these effects from distorting research then the population of topics should relax to a more accurate distribution. What that would be, and whether it would differ much from the present one, I can’t tell you.
I’ve said it previously but just to be clear, I am in that not concerned with string theory in particular, it’s a much more general problem that exists in pretty much all areas in science where theory develop is pursued independently of experimental assessment.
LikeLiked by 1 person
Mmm, yea the argument is unconvincing. You could say there is a cognitive bias towards the opposite as well. Afterall, almost all crackpots have the same ‘we are up against the establishment’ belief system, and there appears to be a certain psychological draw to being the primary expert in your small subfield. I certainly don’t know how you would ‘correct’ such things without reintroducing bias in another way. The simplest system is still the one we have! Eg let the students decide what they want to work on. We can certainly advise them and give them our opinions on what the most plausible paths are, but it’s their career choice and the onus lies with them.
Anyway, you are still making a judgement based upon what you think the correct value distribution of research funding ought to be, for all I know the future might prove the opposite to be the case and hence that we were seriously underfunding string theory/pheno. Quite honestly I find that it’s pretty silly to be in the prediction market in this business. As an example, during the 60s most people in HEP were either in the SMatrix program, Current algebras or in Field theory. At first glance, you could say that it was a major mistake to have been in anything but Field theory. But then, well, the Smatrix program did father string theory and quite a lot of material from this blog, but then other parts of the program did leave a lot of wastebin papers. So what would have been the correct and optimal ‘distribution’ in your opinion with the benefit of hindsight? Would any ‘institutional corrections’ to the system have made things better or worse? Does it even make sense to ask the question?
I completely agree that there are relevant sociological pressures, and I said as much in my post. I just disagree as to their end result.
In the end, neither of us are sociologists, and as you point out in your post the actual sociologists don’t seem to be making much progress on this either. The best we can do as amateurs is try to identify natural experiments, situations where one pressure or another is absent.
That’s why I latched on to the LQG example. I thought you were arguing for equality between different approaches, and thus LQG vs string pheno seemed like the natural experiment.
Now that I understand that that wasn’t your point (and I apologize for the condescending tone that resulted), I’m trying to tease out what you would consider a plausible natural experiment to be, hence the question about sizes of fields.
Is there an academic field where you think convention exerts less pressure? Or an institute? (Perimeter springs to mind.) That might be a useful start.
Since about 10 years, it can rather be observed that in particular by well planed negative campaigns and crusades of well-known protagonists played out in rather popular physics blogs, popular media, etc students are getting more and more intimidated and discouraged from taking up anything related to string theory as their field of study even if they would be interested in investigating such topics. So claims that string related research fields are overpopulated are obviously nonsense and part of this negative campaigns aimed at eradicating parts of legitimate theoretical/fundamental physics for not scientifically backed up reasons.
From looking at some of the latest posts on Backreaction, I am shocked and disappointed that you obviously have joined these non-scientifically motivated campaignists and crusaders, as I once liked to read stuff on Backreaction from time to time.
Well, my summary of your comment is that you only like what I write if it agrees with what you think. Tough luck. Hope you find someone else to support your convictions then.
String theory can never work in its entirety, because it considers at the same time the Lorentz symmetry postulate and the existence of extradimensions, which would manifest itself just with Lorentz symmetry violation experimentally. In this way, whole the phenomenology of string theory must be missing by the very definition of string theory subject. Not to say, that this intrinsic inconsistency of theory leads into mathematical uncertainty and fuzziness of interpretation, which is known and diplomatically called as a “landscape of solution”. All these things make string theory untestable in its rigorous sense.
You shouldn’t consider it as an attack of your pet theory – but my kind and solely disinterested explanation of this famous weird string theory behavior. BTW Mrs. Hossenfelder can stay at rest, her loop quantum gravity suffers with similar inconsistency problem too… 😉
LikeLiked by 1 person
Hossenfelder doesn’t do LQG, I was just using it as an example of an “alternative” approach.
More or less you are right but the boundaries are not always that well defined; Witten for example has such a deep knowledge of the theory that he can easily switch between areas and subfields whenever he feels he can make a contribution or solve a specific problem.
Another example is Kachru; he moved from flux compactifications and inflationary models (pheno) to Moonshine.
Polchinski is doing all sort of things in String theory and similarly Susskind and even Douglas or Silverstein.
String pheno at present is underpopulated, no doubt about it, but it was not always the case. Back in the day getting the standard model or a GUT from String theory was the dominant activity. Similarly with the flux compactifications/moduli stabilization/KKLT era. Admittedly after the Multiverse epoch String pheno lost its appeal but there are always ups and downs for all the various subfields. For example after the Strominger-Vafa paper black hole research was the thing to do.
So it depends really, what I see is that whenever there is a breakthrough in some subfield the focus is shifted relatively quickly and the relevant subfield becomes dominant.
Of course there are always people with strong preferences and inclinations.
Here is how Nekrasov explains it:
“SIENA ORISTAGLIO: Does it ever bother you that you cannot test your research experimentally?
NIKITA NEKRASOV: I have been studying a similar area of physics since my PhD. I understand things much better now than 20 years ago, but I cannot say that I drastically changed my field of research. It’s still quantum field theory, gauge theory, string theory, this whole area of physics. It’s complicated because in a sense you are asking fundamental questions about nature but there are no experimental tools to confirm or disprove your theories. Part of the challenge is to find some predictions or indirect consequences of the theories that would be testable and experimentally verifiable or falsifiable, but, no, I am not particularly interested in that, actually. I find my main driving force is actually the mathematical beauty of the equations I am writing. There are other people who are more interested in connecting theoretical research to experimental physics, but this is my way.”
But anyway what needs to be emphasised is that there is only one theory really that keeps on giving highly non trivial results, expanding so fast that in the future there will be no distinction between hep-th/QG and String theory.
Not to mention that it is the only theory you can use to do any sensible calculations beyond QFT.
LikeLiked by 2 people
Yeah, that’s a good way of putting it. For the most part, it’s not a matter of some people being “string phenomenologists” and some people being “formal string theorists”, it’s a matter of what people choose to work on at any given time. Right now, the phenomenological side of things isn’t making the kind of progress that draws interest, and so it’s smaller. But as long as people have broad enough training, the field will be flexible enough to pivot when the situation changes.
LikeLiked by 1 person
Exactly! I agree completely ☺
Sorry, here is the correct link to Nekrasov’s interview:
Deepest problem in quantum gravity phenomenology (not just string theory or LQG) is the misunderstanding of the dimensional scale, at which these theories are supposed to operate. This scale logically exists just at the middle of distance scale of general relativity (which applies at the 10E10 meters) and quantum mechanics, (which applies at the 10E-10 meters) – i.e. just the human observer scale. If we’re developing theory, which is supposed to reconcile the quantum theory and relativity, why we are looking for its confirmation at esoteric Planck scale or cosmological scales? These questions are simple, but quite principal one. https://www.reddit.com/r/Physics_AWT/comments/3y52kd/is_string_theory_not_a_science/cyall6m
zephirawt, this is pretty close to a spam comment. I’m letting it through, but in general you should strongly consider whether posting links to your own discussions are actually on-topic.
LikeLiked by 1 person
I can’t agree on the sociology. Most of what goes under the name of ‘formal string theory’ (including the majority of what goes under the name of QFT) is far closer in spirit and motivation to what goes on in mathematics departments than in physics departments. While people working here like to call themselves ‘physicists’, in reality what is done has very little in common with what goes on with the rest of the physics department.
From my perspective in string pheno/cosmo/astro, people go into formal topics because they are afraid of real physics – they want to be in areas that are permanently safe, and where their ideas can never be killed by a rude injection of data.
For those of a certain generation – who did PhDs before or within the 10-15 years following the construction of the Standard Model – this is less true, and they generally have a good knowledge of particle physics. But I would say probably >90% of formal people under the age of 40 have basically zero ability to contribute anything in the pheno/cosmo areas; I have talked to enough to know that most have little real knowledge of how the Standard Model (of either particle physics or cosmology) works, how experiments work, or how ideas to go beyond the SM work.
I have heard the ‘when something exciting happens we will move in and sort it out’ attitude of formal theorists the entire time I have been in the subject – and it’s deluded BS.
I think you meant this in reply to the discussion with Giotis, since I highlight essentially the same point in my post: people work on what’s safe.
I disagree that “safe” means “no connection to experiment”, though. Remember, pheno/cosmo/etc. aren’t just bigger than their string counterparts, they’re bigger than formal theory too. Non-string pheno isn’t vastly bigger than any of these fields because it’s further from experiment, it’s vastly bigger because there’s a lot more in the near-term that’s available to do.
I mean look, you’ve worked with Edward Hughes before. Do you really think he’s working on amplitudes just because he’s afraid of data?
I don’t think it’s going to ever be as simple as formal theorists waltzing in and taking over, though. I agree that that’s condescending bullshit.
What I do think is that people are more flexible than you realize. Physicists are reasonably good at reinventing themselves in general. You won’t see everyone making the jump, formal theory won’t just vanish. But if string pheno is ascendant again, people will pivot: older folks more easily than younger folks, fresh grad students more easily than either…but if you guys actually manage to connect with experiment, the field isn’t going to remain tiny. All the sociology in the world couldn’t stand up to that.
I get the impression this sort of thing is happening with LIGO-related work now. It wasn’t too long ago that LIGO was the red-headed stepchild of experiments, whittled down to a few die-hards. Now, there’s a huge influx of people trying to jump back in.
LikeLiked by 1 person
I think we completely disagree on the ‘what can be done in the near-term’. I don’t see that there has been anything of large interest or importance (for physics) coming from formal theory for quite a number of years. Sure, people make narrow technical progress in areas that are mainly of interest to other experts working in similar areas. I think the significant decline in (physics department) faculty hires in formal theory across North America reflects a sense that over the last decade formal theory has not made great progress.
What I like about string pheno is that it gives you so many options: you can work on stuff related to particle colliders, or cosmology, or astrophysics, or the more formal mathsy stuff associated to compactifications. Trust me, there is no difficulty in having plenty of interesting topics to work on with short-term feedback.
The other general comment I have is that from formal theorists I get the sense that they feel interaction with experiment is like BICEP or LIGO: experimentalists (them over there) announce some big huge clean result, and THEN theorists can come in and start working on what it means.
But interpretation of data generally isn’t clean, and there’s always a lot of messy work to get and understand data.
I think people who go into formal theory and take up residence there tend to be people who like clean problems. Nothing wrong with this – but certain area require mucky hands (in the metaphorical sense; i.e. understanding how a calorimeter works vs understanding N=4 SYM), and those who prefer cleanliness will never switch topics under any circumstances.
Regarding the near-term: the pie chart in the post shows string pheno and string cosmology as together about 1/12 of the field. This is probably higher than reality, conferences are going to over-represent minority fields to make sure there’s a variety of topics discussed. But let’s pretend it’s accurate.
In order for string pheno and cosmo to grow to the size of formal strings, then, you’d be growing by about a factor of twelve. So let’s say that one day you and everyone else in your sub-field suddenly got eleven eager, fully-funded young grad students. Do you think you could find projects for each of them? Projects that are distinct, and original, and interesting? And that most of the rest of your sub-field could too?
I know a lot of people in a lot of different fields who couldn’t do that. There’s a fundamental limit to how big a field can get, and that’s how many projects are available. There’s some flexibility in that, sure…but when I look around at the QFT folks, the fields that are the biggest always seem coincidentally to be the ones with lots and lots to do.
Again, you could be recruiting from regular phenomenologists as well. They certainly understand experiment design, and they absolutely understand the Standard Model. They won’t know string theory, but you could teach them. But I get the impression they’re not biting.
I’m not trying to say that string pheno is useless, or a waste of time, or anything like that. I’m trying to say that it’s not safe, that it’s risky, and not just because it interfaces with experiment. Rather, it’s because the low-hanging fruit has been picked, because the easy, flashy, obvious projects have been done. I’m glad that there are people who stick around after that. But academia is fickle enough that you shouldn’t be surprised when many don’t.
LikeLiked by 1 person
Definitely I could not handle eleven graduate students, I think the maximum amount of graduate students I could manage at any one time would be about four. But that’s not really relevant, as it is independent of how hot a topic is.
BTW, On numbers, one reason the String Pheno conference was founded was because people working on pheno topics weren’t getting a look-in at the Strings conference and so set up their own conference. I think the Strings conference is most accurately regarded as the Princeton view of the world (broadly, every year it reflects subjects popular at the IAS and a couple of other similar places).
Generally, the ratios vary with place. A small fraction do string pheno in the US, a significant number in Europe, almost none in India, quite a few in Korea…
Ok, so replace the grad students with experienced researchers looking to switch sub-fields. Same question.
And if you think the 1/12 number is off (which it well might be, I don’t have as much experience with the European community), what do you think is a more accurate percentage?
LikeLiked by 1 person
And it gets the same reply. In normal science, subjects grow over periods of years. There hasn’t been a single drop-everything-switch-topic-oodles-of-low-lying-fruit period in the time I have been in physics (since 2003). From what I hear, AdS/CFT and the 1995 D-brane/duality period were the last time this happened.
(I am excluding the ambulance-chasing pheno phenomenon, which I class as a separate category)
“There hasn’t been a single drop-everything-switch-topic-oodles-of-low-lying-fruit period in the time I have been in physics…”
There will be.
First shot fired in the string civil war!
I do not understand the disparaging comments on formal theorists. Formal string theory is not mathematics. While there may be some who would be comfortable in a mathematics department, I know I wouldn’t be. I am much happier in conversation with physicists than mathematicians.
When grad students study \phi^3 theory in 6 dimensions in their first QFT course, are they doing physics or mathematics? The answer should be obvious to everyone. \phi^3 in 6D is a completely unphysical model, but it has almost zero interest for a mathematician.
Formal string theory is quite analogous to \phi^3 in 6D.
It is true that I prefer clean problems, but so what? Physics is a big tent that covers a lot of people with different talents and preferences.
With all due respect, the reason why string theory is dominated by formal theory is because people know how to make scientific progress on formal theory questions. There is a general, and I think accurate, feeling that it is difficult to make scientific progress in connecting string theory to experiment. Where progress is difficult, there will be fewer scientists. That being said, nobody ever got a Nobel Prize for avoiding difficult problems. So it is great that you and others are devoted to these kind of questions which everyone knows are of supreme importance.
While there is still progress to be made in formal string theory, I think probably the subject will die from lack of oxygen unless it starts making more convincing contact with “real physics.” The death throes are probably already well under way.
LikeLiked by 1 person
Let me try and illustrate a bit more what I mean by physics vs mathematics.
Suppose one studies scattering amplitudes in N=4 SYM at large N. Here are two reasons to study it,
One, with the aim of being better able to compute scattering amplitudes in QCD, so as to describe experimental data better.
Two, because N=4 SYM is a beautiful, integrable theory and it is really lovely to understand it deeply.
Both reasons are valid. The first reason I would characterise as basically a physics reason, and the second as basically a maths reason.
Based on my observations and interactions with those in formal theory, i am happy to assert that it is the latter type of reason that dominates. If one is never willing to take the step towards the theories that actually apply to this world, then I don’t see ‘physics’ as the right word.
NB: this isn’t saying reason two isn’t a good reason to work on N=4 SYM! I just don’t think this should be called physics.
My opinion here is that what “should” be called mathematics or physics depends pretty heavily on what one expects to do with the classification.
Should formal theory be in math departments, or physics departments? In the UK, it’s increasingly the former, but I don’t think this is a model that will work universally. Formal theorists and phenomenologists mostly need grad students from the same background, with the same introductory training, training that’s much harder to find in a math department. Math-trained physicists certainly have made important contributions, but I don’t think the field would work if it was just them.
What about journals, or funding? arXiv already distinguishes, I doubt that dividing JHEP into mathematical and physical sections would accomplish very much. As for funding, in the US at least the DOE currently has people with expertise in evaluating both formal theory and pheno grant proposals, and I suspect the same is true to one extent or another elsewhere. I mention the DOE in particular because it’s pretty clear that the original intent of the agency doesn’t matter very much: formal theorists won’t do much for the energy future, but neither will you.
Yes, in the UK most of this type of work certainly occurs in mathematics departments.
I think it’s a good point that what you call it depends on what you want to do with the classification. My emphasis is to argue against the idea that most formal theorists could be working on pheno; they choose not to, but could switch at any time if they wanted.
I don’t claim to understand the motivation of everyone who studies string theory. But if you just want to understand beautiful structures, you should be a mathematician. If you are interested in solid physical results confirmed by experiment, you should study one of the many fields of physics where there are actual experiments.
Where does that leave string theory? String theory is for people who want to understand nature at the most remote and fundamental level, and are willing to take a chance on a unproven idea in the hopes that it is “close enough” to the right answer to give a taste of what the universe is really about before they die. To me, this is basically the curiosity of a physicist. Perhaps a frustrated, morally degenerate physicist. But physicist all the same.
I don’t know how this fits in with your example of N=4 super Yang-Mills. People who study this kind of thing are probably not wholly motivated by computing amplitudes in QCD. But if these people could not be allowed to imagine that AdS/CFT is an important test case for the fundamental laws of nature, they might not be so enthusiastic.
On the other hand, I know a few formal string theorists who are almost completely motivated by understanding structures. They annoy me too.
Well , The distinction between physics and mathematics is not very sharp. For me , the separation between them is largely sociological. Now concerning the issue of experimental validation of mathematical investigations. As far as I know , there is no fundamental principle of nature that say that the theories of physics must be immediately falsified by doing dirty work experimentation in a short time period after the theory in question is proposed.
It’s worth emphasizing to laymen why there is a dirth of string pheno’s (whereas in the 80s they were everywhere). Right now the state of the field involves a cookbook of recipes for producing classes of compactifications that have been previously known to produce ‘quasi’ sensible looking physics. Unfortunately there are a lot of these. Worse, there is a sort of cursory way of describing them that leaves a lot of details either uncomputed, or more commonly only approximate (and often poorly controlled approximations).
So now lets say you are dealing with a particular class of compactifications that have 10’s of thousands of solutions, and you want to study just one of these in more detail to try to get really specific answers. A good graduate student can spend most of his 5 year career working out the details, and even then he/she will be unsure of some of the resulting details, likely due to computational restrictions. The immense majority of the time, these solutions won’t look like the real world, or worse you won’t even know if it does or doesn’t.
So instead people rarely get this specific, and instead jump back a level and keep things more general (like these classes of compactifications are interesting b/c they tend to produce 3 generations of particles, like in the standard model).
Anyway long story short, a pretty common belief in the string theory circles that I know, is that string pheno is an extremely difficult field, that has a lot of things going for it, but is simply too difficult to approach at this time. The better solution is to look for a pure theory constraint that might simplify the lives of the pheno community and point them in the right direction, and then reattack the problem, rather than looking for the proverbial needle in the haystack and wasting a lot of time.
LikeLiked by 3 people
This is not ‘the state of the field’. Most people in this area do not work on reproductions of the Standard Model through particular classes of compactification, and those that do are not searching for needles in haystacks but for understanding general properties of compactifications.
Could you give us a general picture of what’s popular in string pheno right now?
LikeLiked by 1 person
String models of inflation. Particularly the question of whether there is a maximal bound to the inflationary field range at O(M_P), and what are the possible forms of inflationary potentials that generate large-field inflation. This is interesting because it relates to the observability of tensor modes (cf BICEP) and it can only be addressed within a quantum theory of gravity (as you need to control M_P suppressed operators).
There is a lot of debate as to whether a bound exists and if so where it would lies. If large-field models exist, there is also a lot of interest in the form of corrections they would lead to in the CMB power spectrum.
Also related is whether the structure of string inflation models (particularly with lots of moduli) is likely to lead to non-gaussianity in the CMB, and if so what one can say about the resulting form of non-gaussianity.
Particle physics and cosmology of moduli, axions and other light stringy relics
Compactifications often give rise to light, very weakly coupled particles such as moduli or axions. These can significantly affect cosmology (e.g. the cosmological moduli problem and non-thermal dark matter production) and also lead to additional dark radiation.
Light axion-like particles are a generic prediction of string compactifications. How do you search for them, how do you bound them, what observations/experiments can you use to find them?
Classic TeV BSM physics – interpretation of the 750 GeV boson as a stringy excitation, more generally stringy models for new physics at the TeV scale. There has traditionally been a lot of work on gravity-mediated supersymmetry breaking and the resulting structure of soft terms, but this is reducing as the absence of evidence for SUSY grows.
There is also the more traditional type of work on understanding compactifications and constructing consistent theories with interesting properties going from 10 to 4 dimensions.
In terms of quantity of effort over the last few years, I would say that (1) and (2) are topics which have received the most attention.
Indeed, if someone with more formal background wants a pheno problem to think about, the question of inflationary field ranges in de Sitter space is a good topic, because it brings together more classic clean formal questions (properties of quantum gravity in de Sitter space) with observational relevance (is there a maximal amount of primordial B modes that an experiment can observe in the cosmic microwave background?).
LikeLiked by 2 people
My numbering seems to have disappeared on posting – ONE was inflation, TWO was moduli/axions, THREE particle physics stuff.
Wonder if your views have changed, now that the formal theorist Cumrun Vafa has proposed a radical solution to the pheno problems you listed above?
I think that there isn’t as much of a clear division for the older people in the community. Most of these are people who date back to when string pheno was a more hopeful enterprise, when people thought the right compactification was just around the corner. Vafa is no exception, if you look at some of his older papers.
For the most part, that community didn’t give up on the idea that string theory could describe the real world. They just ran out of ways to productively investigate that, and worked on other things where they could make more progress. When they do find something that (they think) gives them a handle on “string pheno”, though, they’re still perfectly happy to work on it. Hence Vafa’s conjectures about the string swampland, the weak gravity conjecture, and his more recent discussion of quintessence.
The swampland appears to be a trendy topic at the moment, but even so I don’t think there’s a huge number of people working on it compared to the community of people working on “formal stuff”. There were two swampland-related talks at Strings this year. I don’t expect more than that next year.
I think that the last paragraph in the comment by Haelfix describes quite well the unfortunate dismissive feeling of some formal string/field theorists towards string phenomenology. There is a view, developed in the 80s, that string phenomenology is expected to somehow find a specific vacuum that leads to the Standard Model and further explain why that particular vacuum is our universe. At our current level of understanding of string theory this is a completely unreasonable expectation – this is what people mean in saying string phenomenology is too difficult or that it is searching for a needle in a haystack. Dismissing the whole field of thinking about what string theory has to say about our universe because of this particular unachievable target, for me, is throwing the baby out with the bath water.
String theory still has a lot to teach us about our universe, and I think top string phenomenologists think about general ideas, properties, and mechanisms coming from string theory, with specific string models playing the important role of a proof of concept. In my opinion, good string phenomenology is like Beyond-the-Standard-Model (BSM) physics but with inspiration from string theory and with the bonus of concrete reasons why some particular idea is well motivated from the ultraviolet. And the field is still producing valuable insights in this direction.
Let me give an example. If you extend the Standard Model symmetry groups with a global U(1) symmetry to forbid some bad operators, then often right-handed neutrino masses are also forbidden by this symmetry. However in string theory there are string instantons associated to the U(1) which can regenerate these masses re-enabling the neutrino see-saw mechanism. In Beyond-the-Standard-Model physics one usually only considers instantons for non-Abelian gauge theories, like QCD, and this idea was not explored. String instantons are gravitational in nature and are an example of non-perturbative quantum gravitational effects that can play a crucial role in particle physics and can only be calculated precisely using string theory. Now we probably will not know if this is how right-handed neutrino masses are generated in nature for a long time, and one can not take any specific string model with this effect as a serious candidate for our universe, but learning that this is a possibility at all as a general mechanism, as well as having the tools to quantitatively analyse it, I think is worthwhile as a physical insight.
There are many such examples coming from progress in string phenomenology over the past years – string theory is often more imaginative in its physics than BSM physicists. I think the feeling that some future formal qualitative breakthrough will completely revolutionise our understanding of physics from string theory, and that we should wait for that before thinking about such physics, is misguided. Firstly, we would miss out on insights such as the example above. Secondly, such a breakthrough might revolutionize our thinking about which string vacuum is the relevant one for our universe (a vacuum selection principle), but the key general physical mechanisms that lead to observable physics in that specific vacuum would probably be not very far from ones being explored already in current research: string theory is not fully understood but neither is it completely mysterious. Thirdly, are we really more likely to reach such a breakthrough by thinking about condensed matter applications or mathematical questions than string phenomenology? While I agree that progress in physics sometimes works in mysterious ways, it seems to me the answer is no. It is not unreasonable to expect that if no work is done towards connecting string theory with our universe the chances of establishing such a connection decrease dramatically. Maybe taking small steps forward in a difficult and long journey is more worthwhile than standing waiting for a magical bird to arrive and fly you to the destination.
LikeLiked by 1 person
Consider the folly of regular physicists who use a completely unproven model of a particle as a supposed moving point to do physics. And they have the nerve to expect funding for this nonsense? Anyone who has studied physics knows that Quantum mechanics says that an electron is everywhere not a moving billard ball. It’s actually described by a wave(function).The point? No model of particles from strings to billiard ball models can actually explain quantum mechanical phenomena fundamentally without the addition of probabilistic receipes which is more like a religion (that works) than a science. The gap between “strings” and classical models of the electron may be far less than most imagine if a more general theory were found. My gripe with string theory is that it does not address what a fundamental quantum mechanism is, not because I think strings are weird or unphysical, and I cannot hold string theory to a higher standard on this point than any other theory or model of space-time.
You seem to be misunderstanding how both particle physics and string theory work. These both include wavefunctions just as ordinary QM does, though the theory is usually not formulated in those terms. The “particles” of particle physics are not little billiard balls, and haven’t been for nearly a century.
If you’re looking for a “fundamental quantum mechanism” you’re probably out of luck in the first place. The constraints are pretty severe and despite a nontrivial number of people wanting one there hasn’t really been much progress. Decoherence is a good answer for every experimental question you could dream of asking here, and is certainly not a random “probabilistic recipe” handed down from on high. Expecting quantum mechanics to go away in the fundamental theory seems like expecting relativity to be replaced by a Newtonian explanation of a “fundamental relativistic mechanism”. It doesn’t look like there’s any value in moving backwards this way.
“You seem to be misunderstanding how both particle physics and string theory work”.
-no-all require some inherent probabilitic mechanism and string theory relies on the sum of paths formulation which is still a “probabilistic recipe handed down from on high” as is non-relativitic QM, field theory etc.
“If you’re looking for a “fundamental quantum mechanism” you’re probably out of luck in the first place”.
-that misses the point if true than a meaningful difference between a “string theory” and a typical “billiard ball type model” of zero dimension say for an electron in standard QM has not been elucidated and quantum mechanics does make a physical representation of the electron despite your insistence that it doesn’t otherwise the theory would merely describe thin air.
“The constraints are pretty severe and despite a nontrivial number of people wanting one there hasn’t really been much progress. Decoherence is a good answer for every experimental question you could dream of asking here”
-What creates the probabilistic aspect of measurement? See the problem, until you understand what causes the probabilistic mechanism-and decoherence does not tell you how the measurements are random but merely provides an alleged guided tour of the process you cannot truly evaluate one starting physical model of an electron from another.
Consider this example. At one time it was assumed that planets moved in circular orbits but no one new why. When it was proposed that a planet moves as an ellipse this was thought to be absurd, clearly whatever celestial mysterious mechanism supported circular movement would not support a flattened out orbit to infinity? But without knowing what the mechanism holding the planets in place was the appropriateness of a circular vs elliptical model of movement could not be evaluated-although this was held to be irrelevant to the powers at that time-circualr orbits were held to by symmetrical and perfect in keeping with “heavenly” motion. However once that mechanism was understood-gravitational attraction it was seen that the vast differences attributed to elliptical orbits from circular orbits really were not at all-just as without really knowing what causes random behavior in measurements-one cannot truly evaluate a particle model of an electron vs a string theory type model.
Haelfix comment is spot on and reinforces my basic argumentation i.e. that the main reason string pheno is currently underpopulated is that at this point in time it is hard to make significant progress. I have little doubt in my mind that whenever there is a breakthrough the focus will be shifted rather quickly. String theory is a marathon not a sprint.
Saying that the field is underrepresented because “formal string theorists” are afraid of facing the data is totally unjustified and unfair.
Saying that their work is just mathematics that have nothing or little to do with physics is shocking and basically it is what the average hater of the theory advocates on the net.
It is wrong to fractionise people so brutally and totally against the spirit of the theory itself which intertwines so many apparently different areas.
Take Gobakumar-Vafa invariants for example, they were discovered by “formal” String theorists and they are intensively studied by Mathematicians but they are important for so many areas of Physics like Topological String theory, 4d Supersymmetric theories, 3-fold Calabi-Yau compactifications, Knots and Chern-Simons, Black Hole entropy, Gauge/Geometry correspondence, Geometric transitions, M-theory M2/M5 branes and so on.
Implying that Witten who recently published a 120 pages paper on Gobakumar-Vafa invariants is not doing Physics is absurd.
You find it shocking huh? Let me repeat again:
Most of formal string theory has more in common with mathematics than physics.
And again, just in case my views are unclear:
Most formal string theorists have little interest in data.
Look, mathematics is great. One of the many fun things about string pheno is that on the way you learn lots of nice mathematics about Calabi-Yau manifolds. But mathematics is not physics, and when you start exalting a topic like Gopakumar-Vafa invariants as important for so many areas of physics, I think you are losing touch on what physics is.
I don’t know who you are or what your background is, but if you are in a physics department then think about whether your colleagues could keep a straight face if you said that Gopakumar-Vafa invariants are important for many areas of physics.
And yes, I think it is completely bleedingly obvious that some of Witten’s papers are better described as mathematics than physics.
LikeLiked by 1 person
First you said:
“people go into formal topics because they are afraid of real physics – they want to be in areas that are permanently safe, and where their ideas can never be killed by a rude injection of data.”
Now you are saying:
“Most formal string theorists have little interest in data.”
Quite different statements don’t you think?
In any case I don’t see a real argument in your comment.
Topics in “formal” String theory like Gopakumar-Vafa invariants, AGT correspondence or even Langland’s duality and Moonshine is perfectly legitimate Physics because the true goal of this “formal” work from String theorists’ point of view is to understand the ultimate foundations of Physics.
The fact that such topics attract Mathematicians’ attention too doesn’t change this basic truth.
LikeLiked by 1 person
If you want to regard monstrous moonshine – which is beautiful work – as physics, go ahead, but this has no relation to what the rest of the world means by physics.
I don’t have anything against physical mathematics – to use Greg Moore’s expression – but physics has to be rooted in experiment. Data matters.
And of course I misspelled it; it is Gopakumar-Vafa invariants 😊
LikeLiked by 1 person
I agree. Gromov-Witten Invariance e.g. could be successful to help to find a pure geometric theory. Maybe it’s the key for the 3rd revolution. I hope this too.
Pingback: String Sociology | Not Even Wrong
Yes, I agree. It could be that Gromov-Witten invariants in P_M Branes leads us to a pure geometric theory.
It’s my hope too.
Please do not engage with Lubos Motl:
Not sure who you’re responding to here. Perhaps you meant to post it on one of the comment threads where people do in fact engage with Lubos Motl?