## Thursday, September 13, 2007 ... //

Musings from a French programmer convinced me to write another essay about the interactions of scientists and laymen.

The programmer's ideas are based on several basic myths about the very meaning of science in general and theoretical physics in particular:

1. The goal of science is to wait for a "new Einstein" or a savior
2. The job of scientists is to search for a "new Einstein" or a savior
3. It is possible to divide scientists to Einsteins and non-Einsteins
4. The main tool of science is to search for random sentences written by random people, waiting for one sentence that will revolutionize science
5. It is comparably likely for an outsider to find such a holy grail of physics as it is for an esteemed scientist; more generally, intelligence doesn't matter when it comes to cracking the secrets of the Universe
6. Peer review is always a bad thing
7. Societal pressures help scientists or science to make a scientific revolution; it is possible to social-engineer a new revolution in science
8. Interpretation of quantum mechanics is an unsolved problem that waits for a savior
9. More generally, what physicists are waiting for has something to do with Lee Smolin's proclamations in books and newspapers
10. Laymen have a good idea what scientists think, what principles they find reliable, and why they do so

The further a person is from the actual physics research, the more likely he or she is going to believe these myths. The more a person knows about science, the more ridiculous these myths sound to her. However, some bad physicists have a vested interest for these myths to spread.

Let's discuss these points one by one.

The goal of science is to wait for a "new Einstein" or a savior

The actual goal of science is to explain an ever greater set of natural phenomena ever more accurately, using increasingly coherent, robust, and meaningful theories. The particular insights that were solved by Albert Einstein can be viewed as random coincidences of the history of science. It is a complete accident that Albert Einstein has solved exactly the things that he has solved. There is no objective measure that would define the "current counterparts" of relativity and whose solution would make one a "new Einstein".

The job of scientists is to search for a "new Einstein" or a savior

More importantly, scientists are solving very different problems today and it is impossible to construct any isomorphism with the era of Einstein. And one more key thing has to be said: institutionalized science couldn't work at all if all of its members were trying to make the "revolution" only and no one would care about anything less than a revolution. In reality, of course, people work on more well-defined, more predictable tasks. Such an arrangement is completely necessary. It is analogous to the insurance industry: most of its resources and calculations are connected with relatively predictable events while the worst tragedies such as a collision with an asteroid play a secondary role.

Another important point is that scientists are supposed to make progress themselves rather than to wait for a mail from an outsider that will solve the problems for them. ;-)

It is possible to divide scientists to Einsteins and non-Einsteins

Some of these laymen - including Lee Smolin - have a completely religious attitude to Albert Einstein's name. Well, I find this attitude irrational. Einstein's story is captivating, he was unquestionably one of the most original physicists of all time, and his life differs from the lives of most of the other physicists in certain ways.

On the other hand, I don't think that Albert Einstein was the smartest physicist of all time and I don't think he was qualitatively different from other great physicists of the 20th century. What does it mean if we say that von Neumann, Max Planck, Richard Feynman, or Werner Heisenberg were no Einsteins? Equally importantly, there have been many aspects according to which Albert Einstein can be considered the ultimate mainstream physicist of his era.

He was very conservative, he considered the results and established principles of the contemporary physics to be extremely important and valuable, he knew them, he used them in his work more intensely than others, he considered his discoveries to be a minor update of the previous ideas, and he considered mathematical beauty to be the main feature that allows wise guys like himself to make progress. Also, 90+ percent of his papers were correct and most of them were even important.

In other words, Einstein was the exact mirror image of the people who like to use Einstein's name for populist purposes, e.g. Lee Smolin. Smolin dislikes established results of science, he doesn't understand them, he hates mathematics and the concept of mathematical beauty, and 90+ percent of his papers are pure junk. Still, he likes when people call him a "new Einstein". What an irony.

Is there a new Einstein alive? I don't know what this question is supposed to mean. With some definitions, Edward Witten would count as a new Einstein. With other definitions, another Einstein would be an oxymoron. With a third definition, you could pick a different new Einstein such as your humble correspondent. At any rate, "is there a new Einstein" is not a question that serious scientists would ask - except as a topic for poetry or jokes - because they know it is ill-defined and irrelevant. Only laymen who don't understand the difference between science and religion are obsessed with similar questions.

The main tool of science is to search for random sentences written by random people, waiting for one sentence that will revolutionize science

The French programmer tells us that he is sending his physics theories not only to Lee Smolin - who doesn't respond - but also to lots of other physicists including many Nobel prize winners - who usually don't respond either. Everyone privately uses the word "crackpot" for these people but when a journalist puts such people against you as peers, you're in trouble and it becomes a taboo to say that they are crackpots.

Are scientists supposed to be searching through similar mail for valuable ideas? I don't think so.

What the crackpots usually misunderstand is that good science is not supposed to generate a short slogan. It is supposed to produce a coherent theory of something. A coherent theory is composed of at least 2 pages - but more typically 10-200 pages - of text where sentences and formulae fit together. It is next to impossible to write a meaningful scientific paper "by chance". On the contrary, it takes a few seconds to see that a mail from a crackpot is not a meaningful scientific paper and it can therefore be ignored.

The French guy tells us that the "savior" doesn't have to have EDU in his e-mail address and the famous physicists should read all his mail.

Well, of course that a future science star doesn't have to have EDU in her e-mail address (even though it is more likely that "she" or more likely "he" will). But I am afraid that the EDU is not the real reason why the French guy's mail was ignored.

What about if I try to give you a new idea, Mr French guy? Why don't you try to consider the possibility that your mail is actually junk because it reveals some basic ignorance of yours that the recipients of your mail are able to spot within a few seconds? What if the actual message of these stories is that you should learn a great deal before you dare to talk to people who are just five levels above you? It is an inconvenient theory for your pride, isn't it? There are thousands of pompous pools and self-confident cranks like you and it has become nearly dangerous to suggest that they are not real scientists.

Most people in science ignore this correspondence. I have almost certainly been the only person who has replied to virtually all of them in my whole life but I assure you, it is almost always a waste of time. These people often have huge basic problems with scientific thinking. They are constrained by dozens of wrong dogmas that simply prevent them from understanding extremely trivial things - things that an intelligent person must be able to get within a few minutes.

These people are never able to focus on the real essence of their ideas and to rationally decide whether it is correct or not. They don't want to listen to things that could disagree with their dogmas. They prefer to live in a fog and to invent verbal pseudoarguments that "confirm" their wrong dogmas. It takes roughly 10 times more time to explain them a simple thing than what it takes to teach the same thing to average students. They never feel that they are wasting your time.

The French guy complains that not even Lee Smolin wants to talk to him about the French theories - despite Lee's proclamations that he is searching for a new Einstein. Lee Smolin must realize that the relationship between him and the French guy is completely analogous to the relationship between leading physicists and Smolin. But he thinks that leading physicists are obliged to listen to him even though he is not obliged to listen to his French colleague. The French guy asks:

• How would I know that I am a crackpot, if nobody tells me?
He also writes: "At the same time, I am painfully aware that, from a technical standpoint, I am sorely lacking, and have no hope to understand more than one word out of five in a physics discussion with Ed Witten. So I don't need to be reminded of that. Does that necessarily invalidate my idea? I don't think so."

I think that the answer is effectively Yes. I say "effectively" because as a matter of principle, it is surely possible that someone finds an answer to a difficult question without knowing any maths that is assumed to be needed. It is unlikely but conceivable. But because of some infinitesimal probabilities of such miracles, people like the French guy simply can't expect to be treated as peers of Edward Witten. They can't expect leading theorists to dedicate hours to their insane theories. And I am not exaggerating: this guy really thinks that he is a peer of Edward Witten as his following paragraph reveals:

"If Ed Witten came to me saying 'I have this great idea for a new piece of software', it is quite likely that the idea might be good even if he probably doesn't know crap about C++ template metaprogramming, compiler implementation for coroutines, or Itanium speculative loads (which are the kind of technical topics I do need to master in my work)."

I have no idea whether we should expect Edward Witten to revolutionize programming, but if you let me guess, I think that Witten's revolution would be exactly based on his detailed new knowledge of C++ template metaprogramming, compiler implementation of coroutines, Itanium speculative loads, and similar issues. The revolution would follow after he would learn these things and put them together in an entirely new way. Don't expect Edward Witten to generate vague slogans. In Witten's case it is obvious but even more generally, I think that this is how similar revolutions could occur.

So even if you imagine that the analogy between physics and programming is a good analogy, the correct answer confirms that the French guy's ideas about making breakthroughs is a naive one. However, there is of course another important subtlety here: the analogy is simply not a good one because the situation is not symmetric.

The French guy is not a mirror image of Edward Witten even though he tries to create this impression - and hopes that everyone will be polite not to tell him that this analogy is absurd. Sorry but I find it very important to say that the analogy is absurd. Even if pure intelligence were the only thing that mattered here, he is roughly 25 points below Witten. The high-energy theoretical physics community is more than 1000 times more selective than the programming community. If someone uses completely crazy assumptions in his analyses, it is not shocking that he ends up with ludicrous conclusions.

Is it comparably likely for an outsider to find such a holy grail of physics as it is for an esteemed scientist?

Finally, I want to return to the question whether one can make a breakthrough without knowing the necessary maths. Once again, the effective answer is No. The outsider doesn't have to know the words and terminology but he needs to know whatever content that is necessary. Even the knowledge of an outsider must be effectively isomorphic to the key portion of the knowledge of the experts. Otherwise his contributions are impossible.

There is no universal rule but on the other hand, there is a lot of demonstrable patterns that make certain people much more likely to make a certain kind of progress. Even among the outsiders, there are people who are much more likely to contribute. What will actually happen will often differ from these statistical expectations but if there is way too much stuff to listen, a scientist must obviously make some selection and the selection should be a qualified one.

I think that all sane people realize that physicists aren't waiting and can't be waiting for some opinions of a random person among six billion people. The problem with some people is that they're unable to figure out that they're essentially average people. They have irrational reasons to think that their opinions are important. Of course that there can be an outsider who can make progress. But such an outsider will typically be able to find a common language with an insider and convince the rest.

It is much more likely for an insider to make progress. But even if you're an outsider who finds something that is very important, it is still easier for you to find an insider who will understand your breakthrough. If the people who support you are other complete outsiders who have never made anything that requires similar skills that are expected from someone who solves your problem, you can be pretty much certain that you are just another crackpot.

The idea that masses of laymen are routinely making revolutions in science is completely absurd and as far as I know, there is not a single example of this kind in the history of science. The authors of the revolutions, either insiders or outsiders, always had to make an intellectual step that was difficult for normal scientists and even more difficult for average people. The previous statement is not a controversial one; it is rather a mathematical assertion that can be trivially proven. If such a difficult step were not needed, someone else could have done it much earlier and the result would not be viewed as a revolution.

Is peer review always a bad thing?

The French guy offers us a typical example of an irrational argument that he uses to fool himself into thinking that he is just like Einstein:
• Einstein apparently did not lack peer review either. That makes two of us, then.

Well, Einstein has had problems with particular reviewers. Because we talk about Einstein, it was much more likely that the reviewers were worse physicists than he was. ;-) But every physicist can have a good experience with some reviewers and bad experience with others. In average, one of the two sides is correct if there is any disagreement and it is not always the same side. ;-)

But what I find important in this context is that peer review is just a sociological setup that tries to increase the average quality of articles or keep it above a certain threshold. It is not perfect but it was necessary for journals that need to protect their good name and that need to save paper to survive. Peer review has played a similar role as the natural selection. All working arrangements that don't have any peer review have something effectively equivalent instead.

But even if the French guy had similar opinions about peer review as Einstein had, this fact is absolutely irrelevant for judging the French guy's physics. I claim that the French guy thinks that he is closer to Einstein because of some of these similarities - even if he doesn't state this assumption explicitly. No, he is not. Good science is not about opinions about peer review.

The French guy thinks that it is impossible for an author to defend a good idea because of peer review. This opinion just shows that the French guy doesn't have a slightest idea how the reviewing process actually looks like.

Societal pressures help scientists or science to make a scientific revolution; it is possible to social-engineer a new revolution in science

This myth is popular with some postmodern philosophers of science but there are many people who buy this nonsense. I am always amazed when I learn that many people who should be rightwingers believe it but they apparently do. To my frustration, this list includes many climate skeptics. My theory is that they buy this nonsense because they place themselves into the position of an unjustly discriminated person who is waiting for a help from above.

OK, let me describe the idealized science. In idealized science, you find ideally honest, intelligent, and productive scientists who ideally listen to others. They choose the best ideas in the most efficient way they can, according to the most rational arguments one can imagine. They are not constrained by any taboos or social limitations and develop the ideas in the most spectacular and speedy way. :-)

Fine. Idealized science doesn't exist. ;-) We can't have idealized science but we can say what arrangements make science better and what arrangements make science worse. First, if you flood science with people who are just not as talented or productive or honest, you will make science worse. Second, if you define any social criteria or taboos that constrain scientists in such a way that certain rational conclusions are discouraged or prohibited, you will always make science less honest, less scientific, and thus less efficient, to say the least.

Any social-engineering framework for institutionalized science is always designed to help a particular group of scientists. This statement is true not only in science. Is such a framework helpful? Well, if a new policy helps a better group of scientists instead of a worse group of scientists, the society benefits - and vice versa. It is very likely that the scientific community itself knows what people are good and what people are bad and their opinions are more likely to coincide with reality than the opinions of random laymen manipulated by the media or other means or the opinions of bureaucrats.

But of course, the attitudes of the experts don't have to be true. One can have fields or subfields that are temporarily filled with intellectually weak or dishonest people. But even if this is the case, I think it is important to let the intelligent and honest people influence the situation more than the dumb and dishonest people.

One more fact is important here. If you design policies that systematically restrict what scientists are allowed to think, you are devastating science. If a biologist is not allowed to think (to know) that creationism is junk or if a theoretical physicist is not allowed to think (to know) that loop quantum gravity is junk, it is simply a huge problem. And be sure that those who know what's going on have these opinions.

The social-engineering projects have different "philosophies" behind them - philosophies that are often contradictory. Sometimes they try to increase the influence of a "majority" - e.g. in "consensus science" that is crippling climate science these days - and sometimes they try to increase the influence of a "minority" - e.g. to promote crackpots in theoretical physics. These explanations are completely contradictory. Nevertheless, they are often promoted by the very same people.

Needless to say, the basic goal is identical in both cases - to help a particular group of scientists by using political arguments that are irrational from a scientific viewpoint. If the "favorite" group is a majority, they use the former argument, if it is a minority, they use the latter argument. Such an influence is a bad thing in both cases and all cases. Such an influence always moves us away from the idealized science. Such an influence always increases the power of dirty politics relatively to rational thoughts. The people who promote any of these policies are jerks.

Free scientists who are allowed to think whatever they can about science - and who are allowed to ignore whomever they consider stupid or irrelevant - are not always 100% efficient but in the long run, the arrangement based on free scientists who decide about the equilibrium of the market of ideas are surely the best arrangement to find the correct answers that one can get. Every science that has been influenced by political arguments involving "consensus" or, on the contrary, "diversity" is more likely to have deteriorated. Idealized science never uses these arguments, only political movements do. These pseudoarguments always decrease the signal-to-noise ratio.

Interpretation of quantum mechanics is an unsolved problem that waits for a savior

This is a somewhat technical myth but it shows that the people who don't follow the current picture of the world as painted by science not only offer wrong answers but they are often asking wrong questions, too. For virtually all practical purposes, the correct interpretation of quantum mechanics has been known since the late 1920s. By mid 1980s, all remaining details - such as decoherence - that are relevant for a unified understanding of all physical systems were clarified.

The only "known" open interpretational questions about quantum mechanics appear in the context of quantum gravity.

What the people who dream about "solving" this "problem" actually want is to make quantum mechanics go away and return to a pre-quantum picture of the world. Every single one of them wants it. But that will never happen. The leading quantum physicists in the late 1920s already knew it would never happen but during the last 80 years, a lot of theorems and much more specific insights have been collected that make it clear that it won't ever happen.

The fact that Einstein didn't understand the inevitability of quantum mechanics in the last 30 years of his life showed that he was no God but also had human limitations. The fact that some people are not able to understand these facts about quantum mechanics even 50+ years after Einstein's death shows something very different - namely that they're stupid bigots.

More generally, what physicists are waiting for has something to do with Lee Smolin's proclamations in books and newspapers

Again, the more you know about physics, the more you know that the statement above is ridiculous. Good graduate students may be influenced by the media and the caricatures of physics offered by them. They may believe these things in the first year of their graduate school. But if they're good, they will realize that something is wrong with these "ideas" during the second year.

If they can't show themselves that Lee Smolin's "popular physics" is based on wrong science and irrational arguments by the third year, they are just bad physicists themselves. It is not a disaster if one or two such physicists get a PhD in a decade. But flooding the scientific community with these severely limited people is dangerous.

Laymen have a good idea what scientists think, what principles they find reliable, and why they do so

Unfortunately, it is not the case. Laymen typically not only misunderstand what are the correct answers to technical questions. They also fail to get a correct sociological picture of science - what questions are important, how difficult it is to work on one problem or another, who is believed or respected and why, what is "normal" to say in different contexts, and so on.

This failure of laymen to get a realistic picture of science is not just a fault of the media and science communication: it is largely a fault of the laymen themselves who often prefer colorful conspiracy theories about the interactions of people over the actual technical content of science. The truth found by science is OK but it is arguably too boring for too many people. Myths are sometimes better even if they are entirely untrue.

And that's the memo.

#### snail feedback (4) :

I generally agree with your points with the exception of point 8:

"Interpretation of quantum mechanics is an unsolved problem that waits for a savior"

I'm not sure about the "savior" thing but I otherwise agree with the above statement.

On short, I'll explain why.

Usually Bell's theorem is presented as a proof that classical local determinism (or realism if you want) is incompatible with QM so those who hope to find a local hidden variable theory behind QM's formalism are necessarily wrong.

However, Bell's theorem is based on the assumption that the detectors' orientations are free parameters (the “freedom assumption” in Bell's words). It is easy to show that such an assumption is most probable wrong in the context of a strictly deterministic theory.

As an example we can think about gravity. Can we assume that we can freely move Pluto's orbit without changing the rest of the solar system? Would have Shoemaker-Levy collided with Jupiter if Pluto were in other place? I think not.

Therefore a deterministic hidden variable theory that denies the freedom assumption is not affected by Bell's theorem.

I would further propose a toy mechanism that could bypass Bell’s theorem. Three assumptions are needed:

1. a pair of entangled particles is only emitted if two suitable absorbers (detectors) exist. That is, the source will only emit after a signal from the two detectors (containg information about the position/momenta of each quark/electron in them) arrives at the speed of light.
2. the source can "extrapolate" the state of the two detectors at the time of detection given the available information (this is not something unheard of in physics, we see such extrapolations in electromagnetism – for uniform motion – and general relativity – for both uniform and uniform accelerated motion)
3. the source emits a pair of entangled particles, according to Malus’s law, towards the extrapolated position of the two absorbers.

Whether these assumptions can be used to build a credible interpretation of QM I do not know, it is possible that the math does not allow such an idea to go too far. But as an example of a mechanism that it is both local and realistic and gives the required results it should be good enough.

Regards,
Andrei Bocan

Dear Andrei, what you propose is like curing flu by the HIV/AIDS virus because the latter is bigger and overshines the flu.

Your idea that entangled pairs are only created if they can check that there are detectors positioned somewhere in the Universe with certain properties and that the humans who control them are ready to do something is an idea that contradicts not only locality but even the basic features of causality.

I wouldn't call it an assumption but rather a mad conspiracy theory. I have no idea why you want to pursue these mad theories - probably because the 20th century progress in physics is something that you simply can't stand - but I am ready to bet 10000:1 that this theory won't ever explain anything.

Dear Lubos,

“Your idea that entangled pairs are only created if they can check that there are detectors positioned somewhere in the Universe with certain properties and that the humans who control them are ready to do something is an idea that contradicts not only locality but even the basic features of causality.”

Can you please explain why is that? At what point did I assumed a faster than light information transfer? The detector sends information about its state at the speed of light. The source “calculates” the future state of the detector based on the past state available + the determinist evolution law. The particles are sent towards the detectors with a well defined spin/polarisation that was chosen based on the above calculation.

I would also like to point that the ideas I presented in my previous post are by no way original. They are based on Bell’s opinion, as expressed in his book “Speakable and Unspeakable in Quantum Mechanics”. I don’t have that book now but I would like to reproduce Bell’s words, from a BBC interview (http://en.wikipedia.org/wiki/Superdeterminism):

“There is a way to escape the inference of superluminal speeds and spooky action at a distance. But it involves absolute determinism in the universe, the complete absence of free will. Suppose the world is super-deterministic, with not just inanimate nature running on behind-the-scenes clockwork, but with our behavior, including our belief that we are free to choose to do one experiment rather than another, absolutely predetermined, including the "decision" by the experimenter to carry out one set of measurements rather than another, the difficulty disappears. There is no need for a faster than light signal to tell particle A what measurement has been carried out on particle B, because the universe, including particle A, already "knows" what that measurement, and its outcome, will be.

The only alternative to quantum probabilities, superpositions of states, collapse of the wave function, and spooky action at a distance, is that everything is superdetermined. For me it is a dilemma. I think it is a deep dilemma, and the resolution of it will not be trivial; it will require a substantial change in the way we look at things.”

The mechanism I proposed in my previous post is naïve and completely ad-hoc. Nevertheless, it was intended just as a counterexample to the common claim that such a mechanism cannot exist. It speculates just the “superdeterminism loophole” described by Bell above.

Now, logic tells us that if a logically valid counterexample to a claim exist, no matter how naïve or silly it looks, the claim must be false. One cannot conclude that because this mechanism is silly, every possible mechanism based on superdeterminism must be silly as well. So, unless there is something more that EPR and delayed choice experiments (which pose no problem for the above mechanism) the claim that QM is proven to be fundamentally probabilistic has been proven false.
“I have no idea why you want to pursue these mad theories - probably because the 20th century progress in physics is something that you simply can't stand - but I am ready to bet 10000:1 that this theory won't ever explain anything.”

I don’t understand what the “20th century progress in physics” has to do with the fundamental deterministic or probabilistic nature of QM. Even if a valid local hidden variable interpretation of QM is found, the Heisenberg uncertainty still applies for all practical setups. I don’t think that the present day technology stands or falls on one particular assumption of the Copenhagen interpretation. One has only to gain and nothing to loose from replacing a probabilistic picture (which means basically the absence of an explanation for the particular experimental results obtained) with a classical deterministic picture.

Regards,
Andrei

reader Wanderer in the country of blindfolded said...

The French story is funny and sad at the same time. Funny for those who see the joke, sad, once you realize that this is probably better than most of the humanity thinks. At least the guy thinks about science and not about dance competitions on TV. (BTW: in Polish high schools, history of dance is an equivalent topic on the final matura exam to physics or biology).

The general view of science is severely distorted by the lack of understanding of both the results (important) and the method (much more important). In the course of writing a book which is a sort of personal record of someone who wanders around science's domain, I have been surprised how typical is the view represented by the French programmer, even among the intelligentsia'. And how little bona fide scientists do to combat it.

Though I left physics long time ago, and by necessity my understanding of the details of current research is limited, I find very few people who do apply even basic rules of research to understand phenomena around us. This is at the time when the environment we live in (at least in industrialized countries) is now more and more a product of scientific thinking!)

For example, while I do not have the resources nor time to rework every theoretical model of climate change, I - like you - oppose blind following of marketing claims of (nonskeptical) environmentalists. The activists that issue (in rare moments of clarity) open statements of we must scare the public to death because otherwise they would not react'. Even if they blatantly lie.

Last, but not least, the issue of QM, taken by a few other commentators: recent experiments by Zeilinger and co-workers show that Nature is stranger than we imagined it to be. The `easy' loophole out of broken Bell inequalities, via admitting that reality is there but simply nonlocal seems to be insufficient. Legget's arguments that limited forms of nonlocality are not enough, and that the problem touches the core of realism are very interesting. But... how many people who dare to talk about quantum mystery have read the papers of Leggett and Groblacher?

See me on http://countryofblindfolded.blogspot.com/