Friday, June 30, 2006

Dogmas vs. science

At a famous discussion forum of enemies of string theory, the previous letter to Wall Street Journal was exposed to severe criticism that I consider completely irrational and kind of dumb because the letter is right on the money. Let me say a few more words explaining what's wrong with the critics' opinions.

Technical and historical misunderstandings

First of all, the people over there have no idea about things like the AdS/CFT correspondence, black hole physics, or history of theoretical physics. They don't understand any of the calculational frameworks of string theory, which is why they can't understand Barton's obviously correct statement that string theory is essentially as rigorous as mathematics demands; there are no questions with ambiguous or even contradictory answers. But how could you know this fact before you actually know string theory? Of course that you can't. Why do the critics say so many wrong technical things about string theory? It should not be surprising because they are just laymen who have never learned these things.

So they also can't comprehend what it means to explain *why* black holes have entropy. Explaining why something has a non-zero entropy always means to identify the microscopic constituents of the physical system. What Hawking did in 1974 was to determine *that* black holes have a temperature, and consequently also an entropy, by a semiclassical and effectively a macroscopic and thermodynamic argument, but it did not explain where it comes from. It did not explain *why* the entropy is there and people have been waiting for a convincing calculation leading to the right result for more than 20 years.

Prejudices about the number of vacua

But the remaining points where the critics of physics give totally wrong answers to some questions look a bit different because I am convinced that every college science concentrator should simply be able to avoid this kind of flawed reasoning.

First of all, they view the existence of the large number of solutions to be an argument against a theory. They were apparently contacted by divine forces who informed them how many minima the potential energy relevant for this Universe has. Maybe they were told that the human beings are the only possible intelligent beings, too.

With this "knowledge" obtained directly from the skies, they enter the debate and try to attack everyone else who does not confirm their predetermined "knowledge". It's needless to say that their approach has nothing to do with scientific reasoning. What they promote is a very anti-scientific sort of philosophy or religion: it is bigotry.

In ancient Greece, it was believed that the whole world could have been constructed out of four elements. They just found it enough. In the modern era, it turned out that there are more than 100 elements (the word means something a bit different that the Greeks could not have quite understood) and hundreds of stable or metastable isotopes of their nuclei. Does it mean that we should abandon the theory that predicts these atoms and nuclei - many of which have not yet been seen or created?

No, it means that we should use the opportunity and *learn* something. It's exactly the whole point of every scientific theory that deserves the name that it can predict certain things that were not inserted as input. The position of the critics is completely inconsistent because they criticize string theory for not predicting anything new, but whenever it predicts something new, they hate it, too.

The same comment - that we should try to learn from our theories - applies to the number of anti de Sitter or Minkowski or de Sitter ground states of a theory of quantum gravity. It's just the case that whoever is thinking scientifically knows that a priori, we just don't know how many other solutions than those that are relevant for our life exist.

We just can't impose any dogmas about this question before we actually make some research. It is because the answer *is* unknown. Instead, we must try to apply scientific reasoning and collect arguments. If you think about it for an hour, it is rather obvious that there is nothing wrong about some laws of physics - dynamically or philosophically - if these laws also admit other solutions besides our Universe. We have seen analogous things many times. Our planet was not the only planet, and so on.

In the physics community, we are not sure whether there is a sufficient evidence that we live inside a "random" vacuum - that might be difficult to find - or whether there should exist deeper laws that will determine the right one more directly. There is some kind of controversy whether the answer to the previous question is obvious and what it is.

But there is not any controversy - and there can't be any controversy - about the fact that only detailed and careful mathematical reasoning based on known facts of the Universe may be used in science to determine how many vacua - and how many semi-viable vacua - there are and whether they are directly relevant for physics. We must take the most refined and accurate description of quantum gravity we have, apply its rules as carefully as we can, and derive these conclusions if we can.

In the previous sentence, the best description of quantum gravity is based on certain tools that fit together and whose union we call "string theory". Most of us are convinced that it is extremely unlikely that someone would ever find a semi-realistic description of the real world that would not be a part of string theory in any sense. This opinion is not a preconception: it is a consequence of thousands of technical papers about string theory and its conceivable alternatives or deformations that have had some impact to the question whether other possibilities exist or not. If we are brief, the answer is that the alternatives or deformations can't exist.

Even if you hypothetically imagine that the alternatives exist and will be found in the future, you can't use these hypothetical and currently unknown alternatives to make any scientific arguments. Science would become black magic or politics if we were approaching difficult questions with arguments based on hypothetical theories.

If we use the tools of string theory, we see that there are almost certainly googols of perfectly plausible anti de Sitter supersymmetric four-dimensional universes and everyone who follows (or works on) the latest developments knows that the evidence is overwhelming that these vacua exist. Quite possibly, there are also many metastable long-lived de Sitter universes but our certainty is not as high in this case. And we don't know how many important cosmological and other mechanisms that could be relevant for the vacuum selection problem we're still missing.

But the attempt of the critics of string theory to dictate Nature how many metastable states She should have is completely anti-scientific, and I am sure that no college physics concentrator who should be called "intelligent" would ever fall into this trap. Also, such a clever student would never think that the number of metastable vacua could be used as an argument against a theory or as an argument supporting the theory until we know what the number actually is.

Their idea that they can deduce far-reaching conclusions about string theory by comparing its properties with their religious preconceptions is a sign of profound ignorance and flawed reasoning. If we have two theories among which one explains the parameters of the Standard Model in detail, we will prefer the more predictive theory. But if we have theories that don't yet predict the exact parameters of the Standard Model, we must be choosing according to different criteria. According to the optimal application of scientific arguments that we can realize right now, it seems obvious that quantum gravity has a lot of supersymmetric anti de Sitter vacua, whether or not someone likes it.

Distribution of attention

There exist many more aspects of their criticism that I consider to be incompatible with the thinking habits a bright college science concentrator. One of them is their attempt to assign the weights of different fields and subfields *before* one actually studies the subject and its inner structure.

In the context of theoretical physics, they would like to argue that a higher fraction should study "alternatives" to string theory - because they don't want to waste their precious time by figuring out whether any consistent alternatives actually exist. In the context of string theory, they believe that everyone should study AdS/QCD. This acronym only represents the gravitational dual of the actual SU(3) QCD we know from the strong itneractions, not the general AdS/CFT enterprise.

Everyone who actually knows something about the state-of-the-art theoretical physics knows very well that AdS/QCD is just one among dozens of research directions. Much like others, it seems promising because of many reasons. Much like in the case of others, there are reasons to think that the amount of truly interesting applications and insights might be approaching the point at which the "excitement potential" has been largely exhausted.

People in science are actually choosing their topics according to their qualified guess which topic is the most likely one to lead to some valuable outcomes. It is completely obvious that if AdS/QCD were by far the most interesting direction in string theory and if most people were familiar with the arguments why it is the most interesting direction, the proportion of the people working on this topic would jump almost instantly because of the very laws of supply and demand.

Science has seen many examples of this kind. When it became obvious in 1984 that string theory was clearly the most interesting framework for physics beyond the conventional local quantum field theory, hundreds if not thousands of people switched to string theory. When many of them got the feeling that the progress got stuck in the early 1990s, they were switching to different topics. In the middle 1990s, people jumped on D-branes, M-theory, dualities because it was clearly the new "gold mine" filled with exciting ideas, and so on, and so on.

The market equilibrium is reached quite rapidly after a credible piece of information that reveals the importance of a particular direction becomes available to most scientists.

Those "Not Even Wrong" people believe in many conspiracy theories - not too different from the theories that NASA never landed on the Moon. They believe that throughout the last two or three decades, some evil invisible hands have been systematically leading theoretical physics in a completely incorrect direction. I find these conspiracy theories as likely as in the case of the theory about NASA, and those who are immediately ready to believe this kind of theories are not too bright as far as I can say.

Even in the most politicized parts of science where the standards are poor - for example the climate science - it only takes a few years for everyone to see that a scientific paradigm was wrong. When it was proposed in 1998 and 1999 that 1998 was the hottest year in the millenium and that the human influence on the climate exceeds the natural background essentially by an order of magnitude, it was endorsed by thousands of dishonest journalists, politicized scientists, and politicians for political reasons. It took less than 5 years until serious bugs were described in detail - in peer-reviewed journal articles - and it took 8 years before it became a generally known fact to the mainstream climate science community that there exists no evidence that 1998 was the hottest year in the millenium and that the natural variations are at least comparable to the human influence, if not much bigger.

If there are no serious political pressures, wrong papers or directions are abandoned much earlier. It typically takes less time to debunk a wrong paper than the time you need to write it. Think about it: it can't be otherwise.

Why do the critics of string theory exactly single out AdS/QCD? It's simply because QCD is the last topic in theoretical physics whose meaning they can at least vaguely understand. Imagine an ordinary guy who likes beer and who noticed how bubbles are created in the bottle when you shake it. He does not want to understand anything else besides beer. Such a guy might think that all physicists should be working on hydrodynamics of beer (because he does not realize that hydrodynamics is not really enough to say everything about the bubbles).

The situation of those who propose AdS/QCD to be the only "recommended" topic in string theory is completely analogous to the beer lover. Rather severe limitations of their education and, indeed, of their intelligence just make it impossible for them to imagine that there could be other interesting topics in string theory. They just don't want to hear about other topics, they don't want to learn them, and they don't want to listen to any scientific arguments about them. They are already decided, much like the beer lover.

Meanwhile, those who actually know the subject and who know what's going on have no doubt that they could redirect all of their efforts to AdS/QCD if they wanted but most of them don't do it because they realize that there are many other interesting topics around.

If the science critics focused their attention on chemistry, they would surely argue that too many people study compounds with many carbon atoms. It is unfair, the critics would say. However, the people who understand chemistry know very well why a significant fraction of chemistry is organic chemistry where the carbon atoms play a rather important role. My most important point is that without knowing how the science works internally, one is ignorant not only about the answers to detailed technical questions but even about the rough hierarchy of importance of different topics.

One just can't judge any of these things before he or she learns them at a sufficient level. When all these critics are asking whether one can be a legitimate critic of string theory without actually knowing any details about it, my answer is No. It is really not possible. All of us who are critically evaluating loop quantum gravity are doing so after we studied the arguments and technical details in many papers. It is not a "different field" that we criticize. It is a wrong set of ideas in our field. I can't imagine how science could work if string theorists and loop quantum gravitists were viewed as different, complementary fields within quantum gravity. Whoever is asking profound questions about quantum gravity must have faced the question whether the concepts of string theory or loop quantum gravity are correct. And this person must again use rational arguments to decide.

How do the critics envision their forced, Inquisition-style redistribution of the attention within science? I think that it is obvious that they must be eventually think about attracting people who don't like string theory and who think it is important to do other things. Sorry to say, but this really means people who are not capable to learn string theory and the rest of the state-of-the-art image of the world according to theoretical physics. In other words, they want to replace physicists by complete or partial ignorants.

It is important for the internal structure of science and physics not to be directly manipulated by the beer lover in our story and other "critics" who have no idea what's going on because such a direct influence contradicts all moral standards of science and it can only lead to codification of bigotry rather than scientific progress. In science, it is critically important that scientists are not being forced to believe ideas that they find demonstrably and patently false. And that's the memo.


  1. Dear Lubos:

    What you are dealing with here is a recurrent problem in the sociology of knowledge: people whose native intelligence is 5 sigmas above the norm cannot understand what those who are 6 or 7 sigma above the norm are talking about, and are pissed off. The same thing happens all up and down the academic pecking order. Luckily, in the hard sciences at least, these "differences" are not settled by majority rule or in the media. So keep up the good fight.

  2. Lubos,

    I think you should avoid making sweeping generalisations like "all of those people over there have no idea about...[such and such]". You do know the individuals you may be slandering, and you cannot be sure that your claims are even true. Just because someone disagress with you on a point does not mean that they are clueless. Unless you think that you are God, of course. Well, do you?

    If a physical model were to correctly predict the equation for black hole entropy for all types of black hole, then that might be considered a consistency check for that model with respect to Hawking's semiclassical results in the context of quantum gravity. But that (a) would not prove that the physical model is correct (b) would not even provide a consistency check for the model unless Hawking's calculations were proven to be correct from experimental observations. Unfortunately we have not even gotten to the stage where we know that black holes even exist for sure, let alone 'measure' their entropy! (Any ideas on how one might do that?)

    The claim made was that there are so many possible vacua that string theory effectively becomes useless for prediction. Your opinion seems to be that there is nothing wrong with this. Well, I would agree with you if there seemed to be any hope that of finding the underlying principle or boundary condition that selects our universe as the only possible one, but as the number of potential vacua grows both in number and complexity, do you not see that his becomes a more and more distant hope? As Zwiebach says, we have a rigorous 'framework', but unless there is found a way to isolate our universe from the multitude of others, do you not see that the arguments for lack of predictivity are correct? The burden is on string theorists to concentrate on figuring out what this underlying selection principle might be.
    Given all the complex manual selection of parameters that needs to be carried out to get something which even vaguely resembles our universe, does it not strike you as extremely unlikely that our universe will turn out to be special some way in the context of string theory? Even the anthropic arguments are on dubious ground as we cannot possibly know what all possible universes might be like and can only make non-rigorous hand-waving arguments of dubious validity. (Max Tegmark's work is a classic example of such dubious logic in action).

    Its pretty obvious that you were pretty upset (were these 'bitter emotions' perhaps?) about my response to your review of Woit's book. But maybe you should consider the possibility that you do not have the right to insult people just because (you think) you are more intelligent than they are. Any maybe just maybe you could try being a little humble and consider the possibility that you are not necessarily right?

    After all, while your mathematical ability and knowledge seem to be very strong, your logical deductive ability seems to be relatively weak, and in particular, is weaker than Peter Woit's, which may explain why he has such a large following despite his lack of technical knowledge. You may not like to hear that Woit's faculties of reasoning are more acute than your's but I am afraid that it is a fact nevertheless. Peter can see the wood for the trees without examining each tree in great detail, whereas you analyse as many trees in as great detail as you can, but fumble badly when you try to describe the wood, leaving gaping logical inconsistencies which others can pick up on.

    I do agree with you however, that the situation is unlikely to change unless people actually do come up with alternative theories to challenge string theory (and now LQG), but it is a vicious circle, in that the more people that go into string theory (and usually these are the brightest and most talented individuals), the fewer that will ever have the opportunity to come up with radically new ideas. String theory has become so vast and complex, that there is a strong argument for us to go back to basics and start thinking once again about the foundations of relativity and quantum mechanics - and look for new possibilities or for ideas which have been overlooked or forgotten. There are plenty of them around, but unfortunately the intellectual investment required to learn string theory makes it rather difficult to turn back and look at completely different ideas. How much more mathematically complex and distant from physical reality does string have to get before people finally realise that there must surely exist simpler and more elegant alternatives?

    And I have to say, Lubos, your childish attempts at character assassination of any who disagrees with you are quite pathetic. Grow up, and maybe just maybe, you will be taken a little more seriously.