Thursday, February 26, 2015 ... Français/Deutsch/Español/Česky/Japanese/Related posts from blogosphere

Nature is subtle

Caltech has created their new Walter Burke Institute for Theoretical Physics. It's named after Walter Burke – but it is neither the actor nor the purser nor the hurler, it's Walter Burke the trustee so no one seems to give a damn about him.

Walter Burke, the actor

That's why John Preskill's speech [URL fixed, tx] focused on a different topic, namely his three principles of creating the environment for good physics.

His principles are, using my words,

  1. the best way to learn is to teach
  2. two-trick ponies (people working at the collision point of two disciplines) are great
  3. Nature is subtle
Let me say a few words about these principles.

Teaching as a way of learning

First, for many of us, teaching is indeed a great way to learn. If you are passionate about teaching, you are passionate about making things as clear to the "student" that he or she just can't object. But to achieve this clarity, you must clarify all the potentially murky points that you may be willing to overlook if the goal were just for you to learn the truth.

You "know" what the truth is, perhaps because you have a good intuition or you have solved similar or very closely related things in the past, and it's therefore tempting – and often useful, if you want to save your time – not to get distracted by every doubt. But a curious, critical student will get distracted and he or she will interrupt you and ask the inconvenient questions.

If you are a competent teacher, you must be able to answer pretty much all questions related to what you are saying, and by getting ready to this deep questioning, you learn the topic really properly.

I guess that John Preskill would agree that I am interpreting his logic in different words and I am probably thinking about these matters similarly to himself. Many famous physicists have agreed. For example, Richard Feynman has said that it was important for him to be hired as a teacher because if the research isn't moving forward, and it often isn't, he still knows that he is doing something useful.

But I still think it's fair to say that many great researchers don't think in this way – and many great researchers aren't even good teachers. Bell Labs have employed numerous great non-teacher researchers. And on the contrary, many good teachers are not able to become great researchers. For those reasons, I think that Preskill's implicit point about the link between teaching and finding new results isn't true in general.

Two-trick ponies

Preskill praises the concept of two-trick ponies – people who learn (at least) two disciplines and benefit from the interplay between them. He is an example of a two-trick pony. And it's great if it works.

On the other hand, I still think that a clear majority of the important results occurs within one discipline. And most combinations of disciplines end up being low-quality science. People often market themselves as interdisciplinary researchers because they're not too good in either discipline – and whenever their deficit in one discipline is unmasked, they may suggest that they're better in another one. Except that it often fails to be the case in all disciplines.

So the interdisciplinary research is often just a euphemism for bad research hiding its low quality. Moreover, even if one doesn't talk about imperfect people at all, I think that random pairs of disciplines (or subdisciplines) of science (or physics) are unlikely to lead to fantastic off-spring, at least not after a limited effort.

Combinations of two disciplines have led and will probably lead to several important breakthroughs – but they are very rare.

There is another point related to the two-trick ponies. Many breakthroughs in physics resulted from the solution to a paradox. The apparent paradox arose from two different perspectives on a problem. These perspectives may usually be associated with two subdisciplines of physics.

Einstein's special relativity is the resolution of disagreements between classical mechanics and classical field theory (electromagnetism) concerning the question how objects behave when you approach the speed of light. String theory is the reconciliation of the laws of the small (quantum field theory) and the laws of the large (general relativity), and there are other examples.

Even though the two perspectives that are being reconciled correspond to different parts of the physics research and the physics community, they are often rather close sociologically. So theoretical physicists tend to know both. The very question whether two classes of questions in physics should be classified as "one pony" or "two monies" (or "more than two ponies") is a matter of conventions. After all, there is just one science and the precise separation of science into disciplines is a human invention.

This ambiguous status of the term "two-trick pony" seriously weakens John Preskill's second principle. When we say that someone is a "two-trick pony", we may only define this proposition relatively to others. A "two-trick pony" is more versatile than others – he knows stuff from subdisciplines that are further from each other than the subdisciplines mastered by other typical ponies.

But versatility isn't really the general key to progress, either. Focus and concentration may often be more important. So I don't really believe that John Preskill's second principle may be reformulated as a general rule with the universal validity.

Nature is subtle

However, I fully agree with Preskill's third principle that says that Nature is subtle. Subtle is Nature but malicious She is not. ;-) Preskill quotes the holographic principle in quantum gravity as our best example of Nature's subtle character. That's a great (but not the greatest) choice of an example, I think. Preskill adds a few more words explaining what he means by the adjective "subtle":
Yes, mathematics is unreasonably effective. Yes, we can succeed at formulating laws of Nature with amazing explanatory power. But it’s a struggle. Nature does not give up her secrets so readily. Things are often different than they seem on the surface, and we’re easily fooled. Nature is subtle.
Nature isn't a prostitute. She is hiding many of Her secrets. That's why the self-confidence of a man who declares himself to be the naturally born expert in Nature's intimate organs may often be unjustified and foolish. The appearances are often misleading. The men often confuse the pubic hair with the swimming suit, the \(\bra{\rm bras}\) with the \(\ket{\rm cats}\) beneath the \(\bra{\rm bras}\), and so on. We aren't born with the accurate knowledge of the most important principles of Nature.

We have to learn them by carefully studying Nature and we should always understand that any partial insight we make may be an illusion. To say the least, every generalization or extrapolation of an insight may turn out to be wrong.

And it may be wrong not just in the way we can easily imagine – a type of wrongness of our theories that we're ready to expect from the beginning. Our provisional theories may be wrong for much more profound reasons.

Of course, I consider the postulates of quantum mechanics to be the most important example of Nature's subtle character. A century ago, physicists were ready to generalize the state-of-the-art laws of classical physics in many "understandable" ways: to add new particles, new classical fields, new terms in the equations that govern them, higher derivatives, and so on. And Lord Kelvin thought that even those "relatively modest" steps had already been completed, so all that remained was to measure the parameters of Nature more accurately than before.

But quantum mechanics forced us to change the whole paradigm. Even though the class of classical (and usually deterministic) theories seemed rather large and tolerant, quantum mechanics showed that it's an extremely special, \(\hbar=0\) limit of more general theories of Nature (quantum mechanical theories) that we must use instead of the classical ones. The objective reality doesn't exist at the fundamental level.

(The \(\hbar=0\) classical theories may look like a "measure zero" subset of the quantum+classical ones with a general value of \(\hbar\). But because \(\hbar\) is dimensionful in usual units and its numerical value may therefore be changed by any positive factor by switching to different units, we may only qualitatively distinguish \(\hbar=0\) and \(\hbar\neq 0\). That means that the classical and quantum theories are pretty much "two comparably large classes" of theories. The classical theories are a "contraction" or a "limit" of the quantum ones; some of the quantum ones are "deformations" of the classical ones. Because of these relationships, it was possible for the physicists to think that the world obeys classical laws although for 90 years, we have known very clearly that it only obeys the quantum laws.)

Quantum mechanics demonstrated that people were way too restrictive when it came to the freedoms they were "generously" willing to grant to Nature. Nature just found the straitjacket to be unacceptably suffocating. It simply doesn't work like that. Quantum mechanics is the most important example of a previously unexpected difficulty but there are many other examples.

At the end, the exact theory of Nature – and our best approximations of the exact theory we may explain these days – are consistent. But the very consistency may sometimes look surprising to a person who doesn't have a sufficient background in mathematics, who hasn't studied the topic enough, or who is simply not sufficiently open-minded or honest.

The lay people – and some of the self-styled (or highly paid!) physicists as well – often incorrectly assume that the right theory must belong to a class of theories (classical theories, those with the objective reality of some kind, were my most important example) they believe is sufficiently broad and surely containing all viable contenders. They believe that all candidates not belonging to this class are crazy or inconsistent. They violate common sense, don't they?

But this instinctive expectation is often wrong. In reality, they have some evidence that their constraint on the theory is a sufficient condition for a theory to be consistent. But they often incorrectly claim that their restriction is actually a necessary condition for the consistency, even though it is not. In most cases, when this error takes place, the condition they were willing to assume is not only unnecessary; it is actually demonstrably wrong when some other, more important evidence or arguments are taken into account.

A physicist simply cannot ignore the possibility that assumptions are wrong, even if he used to consider these assumptions as "obvious facts" or "common sense". Nature is subtle and not obliged to pay lip service to sensibilities that are common. The more dramatic differences between the theories obeying the assumption and those violating the assumption are, the more attention a physicist must pay to the question whether his assumption is actually correct.

Physicists are supposed to find some important or fundamental answers – to construct the big picture. That's why they unavoidably structure their knowledge hierarchically to the "key things" and the "details", and they prefer to care about the former (leaving the latter to the engineers and others). However, separating ideas into "key things" and "details" mindlessly is very risky because the things you consider "details" may very well show that your "key things" are actually wrong, the right "key things" are completely different, and many of the things you consider "details" are not only different than you assumed, but they may actually be some of the "key things" (or the most important "key things"), too!

Of course, I was thinking about very particular examples when I was writing the previous, contrived paragraph. I was thinking about bad or excessively stubborn physicists (if you want me to ignore full-fledged crackpots) and their misconceptions. Those who believe that Nature must have a "realist" description – effectively one from the class of classical theories – may consider all the problems (of the "many worlds interpretation" or any other "realist interpretation" of quantum mechanics) pointed out by others to be just "details". If something doesn't work about these "details", these people believe, those defects will be fixed by some "engineers" in the future.

But most of these objections aren't details at all and it may be seen that no "fix" will ever be possible. They are valid and almost rock-solid proofs that the "key assumptions" of the realists are actually wrong. And if someone or something may overthrow a "key player", then he or it must be a "key player", too. He or it can't be just a "detail"! So if there seems to be some evidence – even if it looks like technical evidence composed of "details" – that actually challenges or disproves your "key assumptions", you simply have to care about it because all your opinions about the truth, along with your separation of questions to the "big ones" and "details", may be completely wrong.

If you don't care about these things, it's too bad and you're very likely to end in the cesspool of religious fanaticism and pseudoscience together with assorted religious bigots and Sean Carrolls.

Add to Digg this Add to reddit

snail feedback (19) :

reader kashyap vasavada said...

This depends on individual's ability. Some are good teachers, some are good researchers and some are both. But let me take this opportunity to compliment you Lubos. You are both. By running this blog and answering questions extensively, you are proving that you are excellent teacher. And your research credentials are well known too.

reader Luboš Motl said...

A woman simply cannot be raped remotely.

reader Rehbock said...

No. he could not and did not have a sexual relationship by exchanging photos or words. Also he is not her professor because he has no power or authority over her. Also he is not harrassing when the exchange is continued as it was without objection.
I am reminded of a joke popular long ago, the professor of biology is lecturing. He states that the sensory responses of a person having an orgasm are no different than those of one taking a shit. A young lady interrupts " professor either i don't know how to shit or you don't know how to screw.
Seems this is same problem. I. Think I know what sexual relationships are and are not. Unlike Bill Clinton it doesn't really depend on the meaning of is but like Dilaton said it can't happen spatially separated. So whether it offends some that not all behavior they disapprove of is not forbidden sexual misconduct it is not.

reader Rehbock said...

When a woman is raped from thousands of miles away over an email link it will be interesting to find out whether she liked it and even more interesting how it was managed. call me names but the correct response to unwanted email is not to cry rape.

reader Rehbock said...

You are right that birds have rational minds.too bad that you are not as gifted.

reader Andy Everett said...

The link for the speach might not work. Is this the link,

reader Rehbock said...

No tuition and no credit and nothing he did affected whether such a certificate is provided.

reader lukelea said...

At first I thought you wrote that we are apt to confuse "the pubic hairs with the swimming pool," which would have been funnier.

reader Steve Presse said...

I agree with Lubos on points 1 and 3. Let me add my thoughts on point 2. J.W. Gibbs ultimately got his Ph.D. in engineering. One can imagine that his engineering background (he also had strengths elsewhere) certainly motivated his later ambitions in statistical physics. Contrary to Boltzmann (who seemed much more narrowly focused both in his writings and interests), Gibbs seems to have had just the right mix of practical knowledge and theoretical background to make a profound contribution to the theory of statistical physics. Of course, this is an extreme example of success. But I do think it should be possible to use the machinery from one area to make a meaningful contribution in another area and have cross-disciplinary remain of high quality.

reader Luboš Motl said...

Yes, Steve, I agree it's possible! I just think that doing it without that is possible, too.

reader Dilaton said...

Cool, the birth of a new theoretical physics institute is always good news :-)

Before reading whay Lumo says, I was thinking that point two is not bad because under the two tricks of the pony I imagine things like particle physics+cosmology, condensed matter+HEP theory, and other combinations that produced interesting and sometimes surprising results by applying the methods and mathematics of one field to another one.
I hope point two is not meant in the other-low level site.

Anyway I wish to the new institute all the best for the future


reader John Harley said...

"New theory" of the Universe says no Big Bang or something. Universe existed forever. They use hidden variables QM. They claim relativity and QM can't be reconciled. Maybe it's something you want to evicerate.

reader kashyap vasavada said...

Hi Lubos: Mentioning of two tricks ponies(!) reminded me about statements on some blogs that ideas which came up in ST such as CFT, ADS/CFT, Virasoro algebra etc have been verified experimentally in condensed matter physics.
When you have time can you write a blog on this? I would like to understand what was actually done and what does it mean for ST. If you have already written on this and I missed it, please give a reference. Thanks.

reader Luboš Motl said...

Hi, I have never been a real expert on that, and see e.g. Andreas Karch's guest blog on that,

reader Don said...

Its nice to see hear your thinking about how you have come to the conclusion that realism is no more. First it was God, now its realism who is dead. I come at it from a different angle, but hey, all roads lead to Rome.

Still, once realism is tossed out, then the low wattage light bulbs who accept that realism is dead fall into a post-modern delirium of "anything goes". It's a fine balancing act to deal with the lack of realism in a sober way. It's too easy to fall into "the bewildering metaphysics born of Ignorance". In a way, this seems to paint the current intellectual situation pretty well. Those goofs talking about "post-modern science" have indeed fallen into the pit of bewildering metaphysics. Others are reactionaries trying to return to the "golden age" of classical realism. Finding the way forward without realism, but also without nonsense is no joke.

Once again, Professor, thank you for the provocative post. -Don

reader Don said...

Hi...couldn't resist to say that every protein that has folded correctly has no idea what you mean by "protein-folding paradox"! :)

reader Like a G-6 said...

Ohp! Here comes the feminazi anti-sex league to ruin another prestigious MAN!

This crap makes me think Saudi Arabia has slightly saner intersex policies!

reader Rich said...

Thank you for having the courage to make this statement. Americans have lost touch with common sense. There is no possible email communication between an 80 year old man and a woman that would provoke anything more than a chuckle or a roll of the eyes. The net result of fascist policy like that imposed by MIT against Lewin will be this: brilliant minds will despise and avoid the United States, like Bertrand Russell did. Or perhaps they will even be denied entry, as in the case of Erdös. I would never recommend that anyone attend MIT at this point. Let the small minded fascists have the place.

reader RAF III said...