Wednesday, December 30, 2020

Puss' and Pup's cake: not a promising way to do physics

Josef Čapek (1887-1945) was a visual artist and the older brother of Karel Čapek, an essential playwright in the interwar Czechoslovakia. In 1920, Karel also wrote the play R.U.R. which introduced the word "robot" to the world. Well, that word was actually invented by Josef, the brother-painter. Aside from some paintings, Josef is also known for the 1929 series "Adventures of Puss and Pup" stories for kids.

After the 1938 annexation of Sudetenland by Germany, Karel was driven by some nasty pro-German trolls to pneumonia and death on December 25th (what a date to die). Josef's life continued for some time. Scholars may add places like Stanford, Yale, Princeton, and Harvard to their CV. Our German friends have exploited Josef's skills, too, so between 1939 and 1945, he added the following places to his CV: Dachau, Buchenwald, Sachsenhausen, and Bergen-Belsen. Do you have a similarly impressive CV? ;-) He was working as a sign painter (creating nice pedigrees for the Nazi apparatchiks), a painter of walls in these concentration camps, and died in the last one when an epidemic of typhus exploded there. Mortality of untreated typhus is between 10-60 percent but the Germans worried less than now they worry about Covid-19 which has below 0.2 percent.

But let me return to Josef's stories for children. They are about a cat and a dog who live together and do many things. The most famous story might be "How They Baked a Cake".

The 1951 cartoon (Czechoslovakia was already an industrial animation superpower) is almost 9-minute-long. But because many TRF readers are clever kids or adults above 80 years of age, I can give you a 30-second summary. To celebrate the puppy's name day, they decided to make a cake. They added everything they liked: vinegar, goose head, bones, sausages, chocolate, stinky cheese, cabbage.... What a nice cake. An evil black dog felt the unusual smell and ate the Pup's cake. He almost died – because Josef Čapek was a greater humanist than his German employees.

OK, this is just an innocent story but it carries a message:
Sometimes less is more. You won't get a good thing by mixing and combining too many desirable ingredients.
This wisdom is sometimes more true and important, sometimes less true and important. It may be understood as a recipe in mundane situations (such as cooking, indeed); but it may also be considered to be a metaphor for some more important human topics. It may be said to be a meme promoting modesty, simplicity, and purity.

Čapek's Puss-and-Pup story is also relevant in physics. Less is often more. Truly important and powerful theories are simple enough. They have a rather small number of independent assumptions and building blocks – but, on the other hand, the number isn't the minimum positive integer, either. The number of such ingredients or independent assumptions is really comparable to the number of ingredients in a good cake. There is some interval where the cake is more likely to be yummy. Too few is bad but too many is bad, too.

OK, most people don't get this point but a substantial percentage of the people interested in physics do understand this virtue of nice theories, some simplicity, agreement with Occam's razor, and so on. But there is another way to interpret the fairy-tale in physics:
The number of things that you should demand your theory to quickly predict, explain, and calculate shouldn't be too high, either.
Your ambitions should be nonzero but they can't be insanely high, either. As you learn at the end of Mafia when Tommy Angelo is eliminated, one needs a balance in the wishes.
The guy who wants too much, risks losing absolutely everything. The guy who wants too little from life, might not get anything at all.
As it turns out, this is a general lesson that almost no layman understands. This is a theme that I am colliding with all the time. It is somewhat similar to other aspects of "why the laymen completely misunderstand physics research" but I think that I have never written a full blog post about this precise angle. Well, the public has been brainwashed by some of the laymen who pretend to know something about physics but they know nothing. And they demand the infinite ambition:
Your theory must predict things (and sometimes everything) precisely right away, otherwise it's no science. Your ambitions and expectations must be maximized or infinite.
This is an attitude that echoes the expectations of the dog and the cat about their cake. In recent years, it's normal for the people to live in a heavily socialist society that is full of insane "entitlements" everywhere so virtually everyone behaves as an insufferable spoiled brat – and the "critics of physics" are some of the best examples of such spoiled brats. But the real point is that the resulting cake created out of a similar recipe is virtually guaranteed to be junk. The guy who wants too much risks losing absolutely everything. If you impose too many conditions on your hypothesis, the hypothesis will be internally inconsistent and/or inconsistent with all the data. It will be very easy to falsify it. It will be very likely to be shot dead – usually much faster than Tommy Angelo who has enjoyed many nice years because he was ultimately more balanced.

And the other aspect of these principles that the laymen don't get is that the death and falsification are irreversible. There is no water of life as collected by the Brothers Grimm that is capable of raising dead men to life; and analogously, in science, there is no water of life that may resuscitate theories that have really been falsified. They are dead, like the dead parrot. They are no more. It has ceased to be. It has passed on. You really need to watch the parrot or Mafia finale videos many times if you don't get the irreversibility of falsification.

If you don't want to be a soldier who just runs under the enemy's tank, you simply need to have this balance – and some modesty. It is amazing if one can make a nonzero valid contribution to physics, especially fundamental physics where the room for new rules is constrained so much that claustrophobia is almost unavoidable. Needless to say, the people who demand the incredibly specific and precise results from others are among those who have not contributed a damn thing in their life – and, rather certainly, they never will. They are just whining parasites, useless spoiled brats that are the most characteristic products of the entitlement society.

Such spoiled brats often say that they're dissatisfied with the achievements of science – ignoring the fact that their own achievements are strictly zero. They criticize science for not predicting something or for remaining ambiguous. But when they propose their own ideas, they are immediately dead as the parrot and the spoiled brats constantly demand the water of life to be poured on the dead bodies. But there is no water of life and the people whose results are falsified and people who demand the water of life all the time clearly end up being worse than those who find theories with some uncertainty and wiggle room.

So some people may have their toy theory. Imagine someone has a very similar theory to Stefan Tungsten: everything in the real world is some generalized, local conditions-dependent Mandelbrot set. The Mandelbrot set vaguely looks like things in the real world, it is complex enough, so it must be behind the real world. That's an ambitious theory, you know, a far-reaching claim. Now, tell this person, let me call him Mr Tungsten Oxide, that he should better reproduce Yang-Mills fields and interactions, not things that look like raspberries – because raspberries are clearly not fundamental, they are made out of Dirac and Yang-Mills fields' quanta.

The key Pupp and Pus moment arrives now.

Mr Tungsten Oxide will tell you: Our world is still a breathing Mandelbrot set but it is one that is improved by having the \(SU(3)\times SU(2)\times U(1)\) gauge symmetry. So now it is really great, Mr Tungsten Oxide argues. We have the beautiful Mandelbrot set (which is far more clever and nicer than the pictures by Mr Stefan Tungsten, by the way!) and we are also compatible with the boring insights of particle physics such as the gauge symmetry.

It's just like combining the sausages, chocolate, vinegar, and goose heads. The only problem with that is that it is not possible to have a theory like that. Why? Because you just can't build a theory by saying what the elementary building blocks are and do and by adding the list of virtues that the theory should have. Why? Because once you sufficiently non-vacuously specify what the theory is, and sometimes very little qualitative information is more than enough, it already decides whether the theory has some characteristics and others – and when it comes to the modern particle physics virtues such as the gauge symmetry, you may be pretty sure that a theory of the Tungsten, Raspberry, Mandelbrot, or Cake types won't have any of these virtues! Does a picture of a raspberry imply the \(SU(3)\times SU(2)\times U(1)\) gauge symmetry? It seems insanely unlikely to me but even if I agreed that it is possible in principle, your raspberry may only be admired once you actually present the evidence that the gauge symmetry (shockingly) follows from the fruit. Otherwise it's just a fruit, not a contribution to physics! Your wish or declaration that the raspberry has something to do with Yang-Mills theory is worth absolutely nothing, it just shows that you have no idea what you are talking about.

If you decide to determine your theory "constructively", by explicitly saying what the elementary degrees of freedom or ingredients are, then the symmetries (including possible or explicit gauge symmetries) of your theory are determined by the ingredients or rules that you decided to have. They are derived consequences, not additional options you can choose. You just can't change the characteristics. You can't demand "virtues" in addition to the properties that you have heavily restricted because your heavy restriction implies that your theory just doesn't have the virtues. If you describe a species that seems to be... an elephant, it just won't fly. You can't say that it also flies because elephants don't fly. You would need something totally different, e.g. something that quacks like a duck. Again, the Tungsten, Raspberry, Mandelbrot, or Cake theories (just like loop quantum gravity, causal dynamical triangulation, the surfer dude board, the octonions having coitus with quaternions, and tons of similar "ambitious" ideas from the pop science books) demonstrably don't have the right gauge symmetries, the equivalence principle, and most of them don't even follow the uncertainty principle (because the authors' thinking about quantum mechanics is stuck in the 17th century). They are non-starters. To say that these theories also have the Standard Model's most powerful principles isn't "additional information" to construct a theory. Instead, it is just a totally invalid statement about the properties of a theory that you have already described pretty accurately (enough to be sure that it is hopeless).

So successful theories often (and probably usually) make many predictions ambiguous, dependent on the environmental details or random historical events, and they make it extremely hard to observe many of their features. The point that the laymen just don't seem to get is that this is a good thing. This is what is needed for those theories to escape the immediate falsification (the death). This is what has allowed Tommy Angelo to live for another decade or two after he joined the Mafia. Take Darwin's evolution.

OK, humans have had common ancestors not only with apes and monkeys but with all other organisms on Earth. When you study what the theory can predict and what it can't, how it can be made compatible with other things that we know about life and geophysics, you will find out that the traits of the species in the "tree of life" are "almost" unrestricted (although the general Darwinian theory tends to imply lots of correlations in the traits); and the shared ancestors may be found millions, tens of millions, or hundreds of millions years in the past (OK, the oldest ones lived billions of years ago). It is very hard to test what life forms have lived such a long time ago. Does this fact should make you abandon (or spit on) the evolutionary theory?

Not at all. The age of the Earth and the age of life and separate species is whatever it is. Whether you would prefer (for some quasi-ethical reasons, something that you predefined to be "good science") the ancestors to be more recent and/or the Earth or species to be younger is irrelevant. If you don't have the actual evidence that these ancestors (or competing processes) had to exist just recently, then you have no valid complaint against the evolution that has lasted for millions of years. You should shut your mouth. And indeed, if you had a would-be "more ambitious" theory that is more testable, it could place the origin of species to a more recent epoch, e.g. 6,000 years into the past. A problem with such an ambitious, "more testable" theory is that it is more testable, indeed. And a sane biologist knows that if such a test of species-changing processes just 6,000 years ago is performed, the conclusion will be that it contradicts the empirical data. Or to say the least, it makes the observed patterns in the dinosaur and other bones vastly less likely (by dozens of orders of magnitude or more) than in the evolutionary explanation that postulates a very long time for the evolution.

The number of vacua in string theory follows the exact same logic as the duration of the evolution and the diversity of the biological species. A layman could prefer all these quantities to be smaller (because smaller numbers are easier to imagine and it is easier to look into the recent past): species "should be" born just recently, species should be almost unique, and the string vacua should be unique. But Nature just isn't your slave. It doesn't fudging care about your wishes. In science, Nature is in charge and She is deciding about the validity of theories. If your expectations contradict the evidence, the problem is with your expectations. You may fool yourself into saying that the higher predictability or the highest ambitions are great things but if these expectations lead you to predictions that contradict the evidence, then it is your theories (and you) that are in trouble! You are the loser! If you don't understand this basic point – that Nature and not your would-be ethical stances is what ultimately matters – then you have absolutely no idea what science is.

So the number of the string vacua is whatever it is. If you don't have real evidence that it is higher or lower than what some other people have derived, then you should shut up because your wishes and would-be ethical expectations are worth less than one millišit. There is surely a room for a further drift in physicists' opinion about the size of the landscape (and the relevance of its bulk for the realistic cosmology) but as long as we keep on doing science, the opinions will shift because of the evidence, not because of the prejudices and screaming by spoiled brats. As we understand it today, string theory, while extremely constraining, still leaves a high amount of ambiguity about the precise low-energy particle spectrum and interactions – many of these properties of particle physics are largely reinterpreted as coincidences. The landscape of possibilities in string theory is indeed analogous to the landscape of possible species which is why Leonard Susskind imported the word "landscape" from biology.

Even without knowing the precise vacuum (and analogously the precise DNA of a species and its implications for the anatomy and physiology), both theories or paradigms make highly nontrivial predictions – old predictions (that were implied by older theories but it is still equally important to be verified in any newer theories that replace the older ones!) and new predictions, indeed. Clearly, the paradigm shifts to Darwin's framework and to string theory are two of the deepest advances in all of science. That's true despite the fact (and perhaps even because of the fact) that these theoretical frameworks declare many pieces of information that we observe to be accidents, to one extent or another. Accidents clearly exist in the Universe (outcomes of measurements are intrinsically random according to quantum mechanics) and it's important to know what patterns are accidental and what patterns are really uniquely implied by some laws (and especially the fundamental laws) of Nature. You can't predetermine the answers to these questions "without science". You can't "make the definition of science more detailed" by imposing your wishes on "what science should say" about all these matters because your wishes are bound to be wrong. The correct answers to all these questions are subject to science as well and a changing answer to such questions is indeed a typical form of a deep scientific revolution.

So if you think that the right way of doing physics is to guess a random list of ingredients and interactions for a cake, then you will end up with a catastrophe, an inedible cake that almost kills another dog, with a dead parrot, with a Tommy Angelo who is shot into his head. You are really closer to the overhyped stupid brother-soldier in Mars Attacks who basically commits suicide when the Martians start their attack. You are ludicrously childish to propose theories like that. Random ingredients for a dog-and-cat cake end up being no good, with almost no certainty. To be a real scientist, or a scientifically thinking layman, you really need to tame your unjustified self-confidence about "how Nature should be" and "how many things and what things a theory should unambiguously or precisely predict". You need to calmly and impartially compare the evidence for and against particular propositions and collections of propositions (such as theories). Otherwise you will be stuck at being a ludicrous, clueless layperson troll, like the critics of Darwin's theory or string theory who complain that they're not given everything they are entitled to.

And that's the memo.

No comments:

Post a Comment