Wednesday, February 15, 2017

A story about Roger Penrose

Philip Ball visited Roger Penrose (85) in Oxford, talked to him, and wrote the profile
Roger Penrose and the vision thing
in the Prospect Magazine. It's fun reading but I surely have mixed feelings. Penrose is a very creative guy who has done some cool things and I agree with many of his views about the organization of the research and "style" that is being suppressed. On the other hand, his views on many important technical questions – and not just difficult ones – are childishly wrong and the self-congratulatory tone of the article is undoubtedly excessive.

Much of the article is about the funding and researchers' freedom to think. I sort of agree although my agreement has its limits, as I will discuss momentarily. People are being overwhelmed by bureaucracy and the expectation to publish regularly which is why they spend lots of time by writing papers, often papers that almost nobody reads, instead of working on potentially bigger things with an X Factor that could wow everybody – and they could do these things in a more relaxing atmosphere.

Penrose or Ball also complains that things are too polished, you need pizzazz, and state-of-the-art facilities. Well, I don't think so. I – and others I know – didn't have a problem to largely denounce polish and pizzazz. And state-of-the-art facilities aren't that bad. They just naturally come with the growing wealth of the society. I assure you that I would be doing just fine as a homeless guy – and this is not meant to be an exaggeration or a joke. On the other hand, I don't see how state-of-the-art facilities could hurt.




People publish papers because there's some competition and they need to show that they have something to be hired or rehired or given grants. You can't eliminate this feedback entirely. To some extent, one must always get research-related jobs or grants because he has already shown some ability to produce interesting results. The system just couldn't work well if it were employing random people as researchers, could it? So if I take the article literally, I think that Penrose is throwing the baby out with the bath water.

Some relationship between a typical researcher's material well-being or job sustainability and his output has to exist which will always place some pressure on his or her production of some results. The society or special sponsors could fund many more people "without too many questions". I think that it could work better for some folks, usually top folks. But I don't think it would work well for a majority of the researchers. For the median researcher, research is just another job and it has to be verified that he or she is actually doing something.




I think that the Harvard Society of Fellows was a nicely balanced job environment. You may be nominated, a shortlist is picked, some candidates undergo interviews – with a bunch of senior Harvard scholars in diverse fields – and about 10% of them get hired as Junior Fellows. Their salary is higher than the regular postdoc salaries. For three years, their only duty is to attend a monstrous dinner on Monday with the senior fellows and guests (often Nobel prize winners and other VIPs) and somewhat more regular Tuesday, Friday lunches with junior fellows only.

The dinners have about 7 courses and to guarantee that the special cooks of the Society have the resources, some junior fellows volunteer and serve as wine stewards, cheese stewards, and chocolate stewards – buying wine, cheese, and chocolate sometimes for $100 a bottle, pound, or bar. ;-)

And they have to interact with everyone else – because the founder of the Society, a former president of Harvard, believed that the interaction between smart folks from different fields is very important. It was surely fun and intellectually stimulating. At the end, I am not quite certain whether these interactions are so important for the progress of the disciplines themselves. A junior fellow isn't obliged to write any bureaucracy over 3 years and doesn't have to teach. It's really "forbidden" at some level for anyone to use him or her in the teaching process but in practice, exceptions are often granted.

I think that this setup is the most relaxed or generic one that makes sense. It was a good enough environment for me to do research and other things. I do believe that it could spread – and similar jobs could be established even at less famous universities, perhaps with lower salaries. And I think that a similar setup could be established even for achieved "mature" scholars. But I don't believe that most of the young people trying to do some scholarship could be funded in this way. The room for abuse is just huge. If you have tens of thousands of such people in the world, it's just impossible that each of them discovers some results of global importance.

Ball's article admits that funding issues didn't get "bad" just in recent decades. Albert Einstein didn't turn out to be a complete loser as a physicist but he still had to work as a bureaucrat in a Swiss patent office during the most productive year of his career. Well, it's unavoidable. Einstein was really zero in science before 1905. It changed abruptly and the revolution couldn't have been quite predicted before it took place. And after all, I think that his job in the patent office was convenient enough and it could provide him with a better environment to think about important ideas than most other jobs or environments, including jobs at universities.

There are surely lots of things that both Penrose and I have always found annoying. Like writing one's CV. Holy crap. But again, I think that at the end, Penrose tends to go too far and miss the importance of some mechanisms that can't be quite eliminated. Ball worships Penrose almost as a unique and singular living genius – well, the words are probably too much – but we also see criticisms of some inventions such as the h-index:
One widely-used metric is the so-called h-index, which measures consistency of citation: if you have published 20 papers that each earn at least 20 citations, your h-index is 20. Even though this number is, by construction, bound to rise over the course of a career, Penrose’s h-index is still nothing special today—which shows it is not a measure of everything that matters in a creative and influential scientist.
The h-index is a rather arbitrary metric. There are many situations in which I find the overall number of citations a better metric – that's surely the case of a researcher who makes a revolution and writes 1 or several papers only. But when I look at the INSPIRE's value of Penrose's h-index, I see he has 36 or 38 (published in journals or all). Sorry but I find this number to be an utterly fair and reasonable quantification of Penrose's contributions to research in related fields of science. So you know, I would dare to say that there is an alternative to the "conclusion"
Penrose’s h-index is still nothing special today—which shows it is not a measure of everything that matters in a creative and influential scientist.
The alternative is that the index at 36 or 38 shows that Penrose is overhyped by journalists, book publishers, and some filmmakers. Quite generally, Penrose's image and not just his cold hard results is cool. I am attracted to that image of a genius that doesn't need to rely on anyone else, too. But I am still sane enough to realize that much of it is "arts", a creation of the writers of popular articles and reviews such as Philip Ball himself.

Penrose is very broad but there are other broad people, too. A problem with some very broad people is that they're shallow, that there is no specialization in which their thinking and knowledge really reaches the greatest depths. Think about some skillful people who have many hobbies. They can do lots of things but they're very far from the top professional or Olympic athletes in all of them. Now, you know, I think that Penrose is ultimately also rather shallow in the fields that I know. I will discuss it at the end.

Ball describes some circumstances of big intellectual events in Penrose's life. He comes from a highly intellectual family. The Escher pictures, impossible staircase and Penrose triangle etc., were invented by Roger Penrose along with his father. Ball also discusses how Penrose got into general relativity when it wasn't really hot – which is nice evidence that Penrose really evaluates the substance independently and wasn't "in it for the money". And he brought Hawking to general relativity etc. The fresh film about Hawking is mentioned, Penrose isn't too satisfied with the actor who was supposed to be Penrose (and not quite with the film's "Hawking", either).

Finally, I want to mention Penrose's criticism directed against some pillars of contemporary theoretical physics. Recall that his words fashion, faith, and fantasy that appear in the title of his latest book refer to string theory, quantum mechanics, and inflationary cosmology, respectively.

These three F-words are just three ways to insult important and ingenious ideas that Penrose decided to dislike, for wrong, irrational, as well as personal reasons. He picked "fashion" for string theory because it still seems like a doable task to convince others that string theory should go away soon. But you know, it hasn't. String theory will soon be 50 years old and there still doesn't seem to be any alternative how to go beyond quantum field theory. String theory is still the only game in town. Ball and Penrose may work hard to deny this fact but that's the only thing they can do against it. They know that if they actually had a counterexample, another comparably promising game in town, they could name it and boast it. Even 47 years after the birth of string theory, there's nothing of the sort. When some theory is studied for 47 years – by some three generations of physicists – does it really make any sense to call it "fashion"?

Quantum mechanics was insulted using the word "faith" because Penrose knows that it is unlikely that he will convince others that it's a fad that will soon die away. Instead, it's already pretty old – over 90 years – which is why Penrose chose to compare it to some religions that are also old. If you believe in quantum mechanics, you must be analogous to religious bigots, Penrose implicitly and sometimes explicitly says. He proposes that general relativity should survive in its exact form while quantum mechanics has to be "bent". This contradicts pretty much everything that physics has learned about these matters, especially in recent decades. The postulates of quantum mechanics are as rigid as the basic rules of logic or probability calculus – they're really just an "update" of the classical ways to think. On the other hand, general relativity is just another effective field theory, bound to be corrected by numerous corrections of many types when studied more accurately. It's clear which of them has to "bend" to become more compatible with the other. Penrose has no justification for his opposite assertion which is why he hasn't published any – after all, there isn't any. An average researcher with the h-index at 38 surely knows better than Penrose. I think it's embarrassing for Penrose to say such things.

Inflation is insulted as "fantasy" simply because Penrose relies on the irrational reaction by the laymen who find some numbers in cosmology and especially inflation too large, impossible to imagine, which is why they tend to buy such demagogic labels. Again, there is nothing scientifically justifiable about this criticism.

In the case of the criticisms of string theory, Penrose boasts to have "different reasons" to criticize string theory than your median Šmoits in your city's sewerage system. He knows that the criticisms concerning testability are bogus – string theory obviously is testable in principle and basically none of the ideas he has ever worked on in his life are testable in practice, either, which is why such a criticism would be silly and suicidal even from his perspective. Unfortunately, the bulk of the stupid people prefer to hear the criticisms by Šmoits and not Penrose. So not only Penrose has made zero contributions to string theory: he hasn't tangibly influenced the demagogic anti-string movement, either.

Penrose says that the extra dimensions are "unstable" and the total number of dimensions above four is bad because "twistors are cool and only possible in four dimensions".

This simple sentence summarizes basically all the reasons why Penrose thinks it is legitimate for a theoretical physicist in 2017 to reject string theory. But a good theoretical physics graduate student knows why these comments are just plain rubbish. String theory's vacua are stable. This is most obvious if you start with the vacua that have a high degree of supersymmetry. Supersymmetry generally guarantees stability – either of the spacetime or objects. It's basically because in supersymmetric theories, the energy may be written as the square of a Hermitian supercharge operator,\[

H = Q^2

\] or a sum of such squares etc. This formula guarantees that the energy is non-negative and it can't drop any deeper from the ground state – and a decay or an instability would need the energy to drop deeper. This basic idea may be applied to states in the whole spacetime or excitations of an object (such as a brane) etc. Penrose's comments that the stringy extra-dimensional spacetimes are unstable is just stupid. If he revealed these "views" while being hired as a formal theoretical particle physicist today, he would indeed be rejected, not hired, and it is absolutely sensible that he would be rejected because his "opinions" are just plain stupid and the worshiping by some journalists who don't have a clue about the stability of spacetimes (and much simpler things, too) can't change anything about these matters.

His comment that "twistors only work in four dimensions" is also misleading. First, spinors obviously exist in every dimension, the group \(SO(m,n)\) has a locally identical cousin \(Spin(m,n)\) which has spinor representations. Twistors are more than spinors, they allow you to replace the spacetime coordinates by some spinor-like coordinates instead. We know how it's done in \(D=4\) but qualitatively similar constructions exist in other dimensions, too. In some sense, this "coincidence in \(D=4\)" arises because a spinor has (about) the same number of components as a spacetime vector. But similar coincidences also exist in \(D=3,4,6,10\), the dimensions in which superstring theory exists classically. Quantum mechanically, one of those cases is favored but it is \(D=10\), not \(D=4\). The spacetime dimensionalities that actually offer us the maximum "synergy" between vectors and spinors are the maximally dimensional supersymmetric spacetimes in \(D=11\) (M-theory) and, with some disclaimers, \(D=12\) (F-theory). And contemporary theoretical physics knows many "hybrid" descriptions that use some spinorial degrees of freedom for some dimensions and vector degrees of freedom for others, perhaps in curved spacetimes, and other related things. The opinion that Penrose's minimal, non-supersymmetric, \(D=4\) twistor is the only thing of the kind that is worth studying is an indefensible symptom of his egotism.

It's questionable whether the "existence of a twistor description" is a law of physics one should expect. It's not quite clear whether twistors are totally sufficient to describe quantum gravitational phenomena in \(D=4\). All such things have been intensely studied in recent 15 years. It's too bad that Penrose wasn't able to register any of this progress. I have often dreamed about the resuscitation of Isaac Newton – we could see his happy eyes if he were taught relativity, quantum mechanics, string theory that he would surely understand. I think that Newton was really smart and let me admit, yup, I would love to resuscitate him to finally have a real peer. ;-)

But you know, Penrose calms down my desire to resuscitate Newton. Maybe Newton's eyes wouldn't be happy at all. Maybe he would react just like any average man. That's surely how Penrose reacts to much of the progress in string theory – but even in the twistor minirevolution. He clearly doesn't have a clue what was going on. It's sad because what has been happening was and still is extremely exciting. The amount of ingenious ideas found by some ingenious people – and those who are "in" surely know that there are other people who are about as ingenious as Penrose if not more so – has greatly surpassed the original ideas realized by Penrose. The amount of related stuff that just works is already large.

While I have still lots of respect for Penrose, I see how much of his fame has been fabricated. On one hand, Penrose is worshiped as a natural genius who doesn't need to rely on any "infrastructure" let alone grants, CVs, and h-indices. On the other hand, his being a top physicist increasingly depends on journalists, book publishers, and perhaps filmmakers. Arts are exciting but science is something else and at the end, one's importance to science is more tightly correlated with boring and imperfect indices than with the enthusiasm of the folks in the humanities and the entertainment industry.

No comments:

Post a Comment