Luboš Motl, January 26, 2015
S.H. of Tokyo University sent me a link to another text about the «problems with physics». The write-up is one month old and for quite some time, I refused to read it in its entirety. Now I did so and the text I will respond to is really, really terrible. The author is Sabine Hossenfelder and the title reads
Does the prevalence of untestable theories in cosmology and quantum gravity require us to change what we mean by a scientific theory?
To answer this, No. Only people who have always misunderstood how science works – at least science since the discoveries by Albert Einstein – need to change their opinions what a scientific theory is and how it is being looked for. Let me immediately get to the propositions in the body of the write-up and respond.
Here we go:
Theoretical physics has problems.
Theoretical physics solves problems and organizes ideas about how Nature works. Anything may be substituted for «it» to the sentence «it has problems» but the only reason why someone would substitute «theoretical physics» into this sentence is that he or she hates science and especially the most remarkable insights that physics discovered in recent decades.
The third sentence says:
But especially in high energy physics and quantum gravity, progress has basically stalled since the development of the standard model in the mid 70s.
This is an absolutely preposterous claim. First, since the mid 1970s, there have been several important experimental discoveries – like the discoveries of the W-bosons, Z-bosons, Higgs boson, top quark, neutrino oscillations; non-uniformities of the cosmic microwave background, the cosmological constant, and so on, and so on.
But much more shockingly, there have been long sequences of profound and amazing theoretical discoveries, including supersymmetry, supergravity, superstring theory, its explanation for the black hole thermodynamics, D-branes, dualities, holography, AdS/CFT correspondence, AdS/mundane_physics correspondences, and so on, and so on. Many of these results deservedly boast O(10,000) citations – like AdS/CFT – which actually sometimes beats the figures of the Standard Model. Which of those discoveries are more important is debatable and the citation counts can’t be treated dogmatically but some of the recent discoveries are unquestionably in the «same league» as the top papers that have led to the Standard Model.
It is silly not to consider these amazing advances «fully important» just because they’re primarily theoretical in character. The W-bosons, Z-bosons, Higgs bosons etc. have been believed to exist since the 1960s even though they were also discovered in 1983 or 2012, respectively, and they were «just a theory» for several previous decades. The beta-decay was known by competent particle physicists to be mediated by the W-boson even though no W-boson had been seen by 1983. Exactly analogously, we know that the gravitational force (and other forces) is mediated by closed strings even though we haven’t seen a fundamental string yet. The situations are absolutely analogous and people claiming that it is something «totally different» are hopelessly deluded.
One can become virtually certain about certain things long before the thing is directly observed – and that is true not only for particular species of bosons but also for the theoretical discoveries since the mid 1970s that I have mentioned.
Yes, we’ve discovered a new particle every now and then. Yes, we’ve collected loads of data.
In the framework of quantum field theory, almost all discoveries can be reduced to the «discovery of a new particle». So if someone finds such a discovery unimpressive, he or she simply shows his or her disrespect for the whole discipline. But the discoveries were not just discoveries of new particles.
But the fundamental constituents of our theories, quantum field theory and Riemannian geometry, haven’t changed since that time.
That’s completely untrue. Exactly since the 1970s, state-of-the-art physics has switched from quantum field theory and Riemannian geometry to string theory as its foundational layer. People have learned that this more correct new framework is different from the previous approximate ones; but from other viewpoints, it is exactly equivalent thanks to previously overlooked relationships and dualities.
Laymen and physicists who are not up to their job may have failed to notice that a fundamental paradigm shift has taken place in physics since the mid 1970s but that can’t change the fact that this paradigm shift has occurred.
Everybody has their own favorite explanation for why this is so and what can be done about it. One major factor is certainly that the low hanging fruits have been picked, [experiments become hard, relevant problems are harder…].
Still, it is a frustrating situation and this makes you wonder if not there are other reasons for lack of progress, reasons that we can do something about.
If Ms Hossenfelder finds physics this frustrating, she should leave it – and after all, her bosses should do this service for her, too. Institutionalized scientific research has also become a part of the Big Government and it is torturing lots of people who would love to be liberated but they still think that to pretend to be scientists means to be on a great welfare program. Niels Bohr didn’t establish Nordita as another welfare program, however, so he is turning in his grave.
Ms Hossenfelder hasn’t written one valuable paper in her life but her research has already cost the taxpayers something that isn’t far from one million dollars. It is not shocking that she tries to pretend that there are no results in physics – in this way, she may argue that she is like «everyone else». But she is not. Some people have made amazing or at least pretty interesting and justifiable discoveries, she is just not one of those people. She prefers to play the game that no one has found anything and the taxpayers are apparently satisfied with this utterly dishonest demagogy.
If you have the feeling that the money paid to the research is not spent optimally, you may be right but you may want to realize that it’s thanks to the likes of Hossenfelder, Smolin, and others who do nothing useful or intellectually valuable and who are not finding any new truths (and not even viable hypotheses) about Nature.
Especially in a time when we really need a game changer, some breakthrough technology, clean energy, that warp drive, a transporter! Anything to get us off the road to Facebook, sorry, I meant self-destruction.
We don’t «need» a game changer now more than we needed it at pretty much any moment in the past (or we will need it in the future). People often dream about game changers and game changers sometimes arrive.
We don’t really «need» any breakthrough technology and we certainly don’t need «clean energy» because we have lots of clean energy, especially with the rise of fracking etc.
We may «need» warp drive but people have been expressing similar desires for decades and competent physicists know that warp drive is prohibited by the laws of relativity.
And we don’t «need» transporters – perhaps the parties in the Ukrainian civil war need such things.
Finally, we are more resilient and further from self-destruction than we were at pretty much any point in the past. Also, we don’t need to bash Facebook which is just another very useful pro-entertainment website. It is enough to ignore Facebook if you think it’s a waste of time – I am largely doing so 😉 but I still take the credit for having brought lots of (more socially oriented) people who like it to the server.
So every single item that Hossenfelder enumerates in her list «what we need» is crap.
It is our lacking understanding of space, time, matter, and their quantum behavior that prevents us from better using what nature has given us.
This statement is almost certainly untrue, too. A better understanding of space, time, and matter – something that real physicists are actually working on, and not just bashing – will almost certainly confirm that warp drives and similar things don’t exist. Better theories will give us clearer explanations why these things don’t exist. There may be some «positive applications» of quantum gravity but right now, we don’t know what they could be and they are surely not the primary reason why top quantum gravity people do the research they do.
The idea that the future research in quantum gravity will lead to practical applications similar to warp drive is a belief, a form of religion, and circumstantial evidence (and sometimes almost rigorous proofs) makes this belief extremely unlikely.
And it is this frustration that lead people inside and outside the community to argue we’re doing something wrong, …
No, this is a lie, too. As I have already said, physics bashers are bashing physics not because of frustration that physics isn’t making huge progress – it obviously *is* making huge progress. Physics bashers bash physics in order to find excuses for their own non-existent or almost non-existent results in science – something I know very well from some of the unproductive physicists in Czechia whom the institutions inherited from the socialist era. They try to hide that they are nowhere near the top physicists – and most of them are just useless parasites. And many listeners buy these excuses because the number of incredibly gullible people who love to listen to similar conspiracy theories (not so much to science) is huge. And if you combine this fact with many ordinary people’s disdain for mathematics etc., it is not surprising that some of these physics bashers may literally make living out of their physics bashing and nothing else.
The arxiv categories hep-th and gr-qc are full every day with supposedly new ideas. But so far, not a single one of the existing approaches towards quantum gravity has any evidence speaking for it.
This is complete rubbish. The tens of thousands of papers are full of various kinds of evidence supporting this claim or another claim about the inner workings of Nature. In particular, the scientific case for string theory as the right framework underlying the Universe is completely comparable to the case for the Higgs boson in the 1960s. The Higgs boson was discovered in 2012, 50 years after the 1960s, but that doesn’t mean that adequate physicists in the 1960s were saying that «there wasn’t any evidence supporting that theory».
People who were not embarrassed haven’t said such a thing and people who are not embarrassing themselves are not saying a similar thing about string theory – and other things – today.
To me the reason this has happened is obvious: We haven’t paid enough attention to experimentally testing quantum gravity. One cannot develop a scientific theory without experimental input. It’s never happened before and it will never happen. Without data, a theory isn’t science. Without experimental test, quantum gravity isn’t physics.
None of these statements is right. We have paid more than enough attention to «experimental quantum gravity». It is a vastly overstudied and overfunded discipline. All sensible physicists realize that it is extremely unlikely that we will directly observe some characteristic effects of quantum gravity in the near future. The required temperatures are around 1032 kelvins, the required distances are probably 10−35 meters, and so on. Max Planck has known the values of these natural units since the late 19th century.
So we have paid more than enough attention to this strategy.
It is also untrue that the progress in theoretical physics since the mid 1970s has been done «without experimental input». The amount of data we know about many things is huge. To a large extent, the knowledge of one or two basic experiments showing quantum mechanics and one or two experiments testing gravity is enough to deduce a lot. General relativity, quantum mechanics, and string theory largely follow from (subsets of) these several elementary experiments.
On the other hand, it is not true that scientific progress cannot be made without (new) experimental input. Einstein found special relativity even though he wasn’t actively aware of the Michelson-Morley experiment. He could have deduced the whole theory independently of any experiments. Experiments had previously been used to construct e.g. Maxwell’s equations but Einstein didn’t deal with them directly. Einstein only needed the equations themselves. More or less the same thing occurred 10 years later when he discovered general relativity. But the same approach based on «nearly pure thought» has also be victorious in the case of Bekenstein’s and Hawking’s black hole thermodynamics, string theory, and in some other important examples.
So the idea that one can’t find important things without some new experiments – excluding experiments whose results are old and generally known – is obviously untrue. Science haters can say that this or that important part of science «is not science» or «is not physics» but that doesn’t change anything about the fact that certain insights about Nature may be found and have been found and supported by highly convincing bodies of evidence in similar ways. Only simpletons may pay attention to demagogue’s proclamation that «something is not science». This emotional scream is not a technical argument for or against any scientifically meaningful proposition.
I will omit another repetitive paragraph where Hossenfelder advocates «experimental quantum gravity». She thinks that tons of effects are easily observable because she’s incompetent.
Yes, experimental tests of quantum gravity are farfetched. But if you think that you can’t test it, you shouldn’t put money into the theory either.
This is totally wrong. It is perfectly sensible to pay almost all of the quantum gravity research money to the theorists because whether someone likes it or not, quantum gravity is predominantly a theoretical discipline. It is about people’s careful arguments, logical thoughts, and calculations that make our existing knowledge fit together more seamlessly than before.
In particular, the goal of quantum gravity is to learn how space and time actually work in our Universe, a world governed by the postulates of quantum mechanics. Quantum gravity is not – and any discipline of legitimate science is not – a religious cult that trains its followers to believe in far-fetched theories. The idea that you may observe completely new effects of quantum gravity (unknown to the theorists) in your kitchen is far-fetched and that really means that it is extremely unlikely. And its being extremely unlikely is the rational reason why almost no money is going into this possibility. This justification can’t be «beaten» by the ideological cliché that everything connected with experiments in the kitchen should have a priority because it’s «more scientific».
It’s not more scientific. A priori, it is equally scientific. A posteriori, it is less scientific because arguments rooted in science almost reliably show that such new quantum gravity effects in the kitchen are very unlikely – some of them are rather close to supernatural phenomena such as telekinesis. So everything that Ms Hossenfelder says is upside down once again.
And yes, that’s a community problem because funding agencies rely on experts’ opinion. And so the circle closes.
Quantum gravity theorists and string theorists are getting money because they do absolutely amazing research, sometimes make a medium-importance discovery, and sometimes a full-fledged breakthrough. And if or when they don’t do such a thing for a few years, they are still exceptional people who are preserving and nurturing the mankind’s cutting-edge portrait of the Universe. The folks in the funding agencies are usually less than full-fledged quantum gravity or string theorists. But as long as the system at least barely works, they still know enough – much more than an average human or Ms Hossenfelder knows – so they may see that something fantastic is going on here or there even though they can’t quite join the research. That’s true for various people making decisions in government agencies but that’s true e.g. for Yuri Milner, too.
As Ms Hossenfelder indicated, the only way how this logic may change – and yes, I think it is unfortunately changing to some extent – is that the funding decisions don’t depend on expert opinion (and on any people connected with knowledge and progress in physics) at all. The decisions may be done by people who hate physics and who have no idea about contemporary physics. The decisions may depend on people who are would-be authority and pick winners and losers by arbitrarily stating that «this is science» and «this is not science». I don’t have to say how such decisions (would?) influence the research.
To make matters worse, philosopher Richard Dawid has recently argued that it is possible to assess the promise of a theory without experimental test whatsover, and that physicists should thus revise the scientific method by taking into account what he calls “non-empirical facts”.
Dawid just wrote something that isn’t usual among the prevailing self-appointed «critics and philosophers of physics» but he didn’t really write anything that would be conceptually new. At least intuitively, physicists like Dirac or Einstein have known all these things for a century. Of course that «non-empirical facts» have played a role in the search for the deeper laws of physics and this role became dramatic about 100 years ago.
Dawid may be confused on this matter because physicists do, in practice, use empirical facts that we do not explicitly collect data on. For example, we discard theories that have an unstable vacuum, singularities, or complex-valued observables. Not because this is an internal inconsistency — it is not. You can deal with this mathematically just fine. We discard these because we have never observed any of that. We discard them because we don’t think they’ll describe what we see. This is not a non-empirical assessment.
This was actually the only paragraph I fully read when I replied to S.H. in Tokyo for the first time – and this paragraph looked «marginally acceptable» to me from a certain point of view.
Well, the paragraph is only solving a terminological issue. Should the violation of unitarity or instability of the Universe that would manifest itself a Planck time after the Big Bang, or something like that be counted as «empirical» or «non-empirical» input? I don’t really care much. It’s surely something that most experts consider consistency conditions, like Dawid.
We may also say that we «observe» that the Universe isn’t unstable and doesn’t violate unitarity. But this is a really tricky assertion. Our interpretation of all the observations really assumes that probabilities are non-negative and add to 100%. Whatever our interpretation of any experiment is, it must be adjusted to this assumption. So it’s a pre-empirical input. It follows from pure logic. Also, some instabilities and other violations of what we call «consistency conditions» (e.g. unitarity) may be claimed to be very small and therefore hard to observe. But some of these violations will be rejected by theorists, anyway, even if they are very tiny because they are violations of consistency conditions.
I don’t really care about the terminology. What’s important in practice is that these «consistency conditions» cannot be used as justifications for some new fancy yet meaningful experiments.
A huge problem with the lack of empirical fact is that theories remain axiomatically underconstrained.
The statement is surely not true in general. String theory is 100% constrained. It cannot be deformed at all. It has many solutions but its underlying laws are totally robust.
This already tells you that the idea of a theory for everything will inevitably lead to what has now been called the “multiverse”. It is just a consequence of stripping away axioms until the theory becomes ambiguous.
If the multiverse exists, and it is rather likely that it does, it doesn’t mean that the laws of physics are ambiguous. It just means that the world is «larger» and perhaps has more «diverse subregions» than previously thought. But all these regions follow the same unambiguous laws of physics – laws of physics we want to understand as accurately as possible.
The comment about «stripping away axioms» is tendentious, too, because it suggests that there is some «a priori known» number of axioms that is right. But it’s not the case. If someone randomly invents a set of axioms, it may be too large (overconstrained) or too small (underconstrained). In the first case, some axioms should be stripped away, in the latter case, some axioms should be added. But the very fact that a theory predicts or doesn’t predict the multiverse doesn’t imply that its set of axioms is underconstrained or overconstrained.
For example, some theories of inflation predict that inflation is not eternal and no multiverse is predicted; other, very analogous theories (that may sometimes differ by values of parameters only!) predict that inflation is eternal and the Universe emerges. So Hossenfelder’s claim that the multiverse is linked with «underconstrained axioms» is demonstrably incorrect, too.
Somewhere along the line many physicists have come to believe that it must be possible to formulate a theory without observational input, based on pure logic and some sense of aesthetics. They must believe their brains have a mystical connection to the universe and pure power of thought will tell them the laws of nature.
There is nothing mystical about this important mode of thinking in theoretical physics. It’s how special relativity was found, much like general relativity, the idea that atoms exist, the idea that the motion of atoms is linked to heat, not to mention the Dirac equation, gauge theories, and many other things. A large fraction of theoretical physicists have made their discovery by optimizing the «beauty» of the candidate laws of physics. People like Dirac have emphasized the importance of the mathematical beauty in the search for the laws of physics all the time, and for a good reason.
That’s the most important thing Dirac needed to write on a Moscow blackboard.
And the more recent breakthroughs in physics we consider, the greater role such considerations have played (and will play). And the reason why this «mathematical beauty» works isn’t supernatural – even though many of us love to be amazed by this power of beautiful mathematics and this meme is often sold to the laymen, too. One may give Bayesian explanations why «more beautiful» laws are more likely to hold than generic, comparable, but «less beautiful» competitors. Bayesian inference dictates to assign comparable prior probabilities to competing hypotheses and because the mathematically beautiful theories have a smaller number of truly independent assumptions and building blocks, and therefore a smaller number of ways how to invent variations, their prior probability won’t be split to so many «sub-hypotheses». Moreover, as we describe deeper levels of reality, the risk that an inconsistency emerges is high and ever higher, and the «not beautiful theories» are increasingly likely to lead to one kind of an inconsistency or another.
Sabine Hossenfelder’s denial of this principle only shows her lack of familiarity with physics, its logic, and its history.
You can thus never arrive at a theory that describes our universe without taking into account observations, period.
Whether someone has ever found important things without «any observations» is questionable. But it is still true and important that a good theorist may need 1,000 times less empirical data than a worse theorist to find and write down a correct theory, and a bad theorist will not find the right theory with arbitrarily large amounts of data! And that’s the real «period», that’s why the mathematical beauty is important for good theoretical physicists – and the others have almost no chance to make progress these days.
The attempt to reduce axioms too much just leads to a whole “multiverse” of predictions, most of which don’t describe anything we will ever see.
I have already said that there is no relationship between the multiverse and the underdeterminedness of the sets of axioms.
(The only other option is to just use all of mathematics, as Tegmark argues. You might like or not like that; at least it’s logically coherent. But that’s a different story and shall be told another time.)
But these Tegmark’s comments are purely verbal philosophical remarks without any scientific content. They don’t imply anything for observations, not even in principle. For this reason, they have nothing to do with physical models of eternal inflation or the multiverse or even specific compactifications of string/M-theory which are completely specific theories about Nature and the observations of it.
Now if you have a theory that contains more than one universe, you can still try to find out how likely it is that we find ourselves in a universe just like ours. The multiverse-defenders therefore also argue for a modification of the scientific method, one that takes into account probabilistic predictions.
Most people writing papers about the multiverse – more precisely, papers evoking the anthropic principle – use the probability calculus incorrectly. But the general statement that invoking probabilities in deductions of properties of Nature is a «modification of the scientific method» is a total idiocy. The usage of probabilities was not only «allowed» in the scientific method for quite some time. In fact, science could have never been done without probabilities at all! All of science is about looking at the body of our observations and saying which explanation is more likely and which explanation is less likely.
And of course that a theory with a «larger Universe than previously thought» and perhaps with some extra rules to pinpoint «our location» in this larger world is an OK competitor to describe the Universe a priori.
Every experimenter needs to do some calculations involving probabilities – probabilities that a slightly unexpected result is obtained by chance, and so on – all the time. Ms Hossenfelder just doesn’t have a clue what science is.
In a Nature comment out today, George Ellis and Joe Silk argue that the trend of physicists to pursue untestable theories is worrisome.
I agree with this, though I would have said the worrisome part is that physicists do not care enough about the testability — and apparently don’t need to care because they are getting published and paid regardless.
I don’t get paid a penny but I am still able to see that the people whose first obsession is «testability» are either crackpots or third-class physicists such as Ms Hossenfelder who don’t have an idea what they are talking about.
The purpose of science is to find the truth about Nature. Easy testability (in practice) means that there exists a procedure, an experimental procedure, that may accelerate the process by which we decide whether the hypothesis is true or not. But the testability doesn’t actually make the hypothesis true (or more true) and scientists are looking for correct theories, not falsifiable theories, and it’s an entirely different thing.
One could say that the less falsifiable a theory is, the better. We are looking for theories that withstand tests. So they won’t be falsified anytime soon! A theory that has already resisted some attempts to be falsified is in a better shape than a theory that has already been falsified. The only «philosophical» feature of this kind that is important is that the propositions made by the theory are scientifically meaningful – i.e. having some non-tautological observable consequences in principle. If this is satisfied, the hypothesis is perfectly scientific and its higher likelihood to be falsified soon may only hurt. If one «knows» that a hypothesis is likely to die after a soon-to-be-performed experiment, it’s probably because he «knows» that the hypothesis is actually unlikely.
See, in practice the origin of the problem is senior researchers not teaching their students that physics is all about describing nature. Instead, the students are taught by example that you can publish and live from outright bizarre speculations as long as you wrap them into enough math.
Maybe this is what Ms Hossenfelder has learned from her superiors such as Mr Smolin but no one is teaching these things at good places – like those I have been affiliated with.
I cringe every time a string theorist starts talking about beauty and elegance.
Because you are a stupid cringing crackpot.
Whatever made them think that the human sense for beauty has any relevance for the fundamental laws of nature?
The history of physics, especially 20th century physics, plus the Bayesian arguments showing that more beautiful theories are more likely. The sense of beauty used by these physicists – one that works so often – is very different from the sense of beauty used by average humans or average women in some respects. But it also has some similar features so it is similar in other respects.
Even more important is to point out that this extended discussion about «strings and beauty» is a straw man because almost no arguments referring to «beauty» can be found in papers on string theory. Many string theorists would actually disagree that «beauty» is a reason why they think that the theory is on the right track. Ms Hossenfelder is basically proposing illogical connections between her numerous claims, all of which happen to be incorrect.
I will omit one paragraph repeating content-free clichés that science describes Nature. Great, I agree that science describes Nature.
Call them mathematics, art, or philosophy, but if they don’t describe nature don’t call them science.
The only problem is that all theories that Ms Hossenfelder has targeted for her criticism do describe Nature and are excellent and sometimes paramount additions to science (sometimes nearly established ones, sometimes very promising ones), unlike everything that Ms Hossenfelder and similar «critics of physics» have ever written in their whole lives.