Thursday, November 29, 2007

Paul Davies: Taking science on faith

Update: a list of wrong assumptions about science was added at the end of this essay. The article written on 11/25 was moved to the top as the most discussed recent text.
As far as I can say, The New York Times remain by far the best source of science news and opinions among the English-speaking newspapers.
Paul Davies' op-ed
meditates about the controversial question concerning the difference between science and religion. I agree with most things he writes. He starts with the idealized picture that many people believe to be true - namely a picture in which science and religion are sharply separated. Skepticism belongs to science while blind belief belongs to religion.

He instantly adds his main thesis that this picture is an oversimplification because science has its belief system, too. The first thing that a typical scientist - and especially a theoretical physicist - believes is that the questions he is trying to answer have coherent, understandable, and universally valid explanation.

Needless to say, this belief is a fact in the case of approximate laws that describe classes of physical phenomena that have already been understood and whose theories have already been verified. But whether the "next step" is going to be comparably rational, mechanical, satisfactory, deterministic, unique, or otherwise similar to the previous theories is always a matter of extrapolation. It is really a matter of belief.


I primarily want to discuss this issue in the case of the anthropic principle but let me mention a few episodes from the past. How was science and religion separated during the debates about heliocentrism? Well, not too sharply. Of course, there were priests who were dismissing the champions of heliocentrism on religious grounds. But there were also many people paid as scientists who were frantically attacking heliocentrism as a bad science that contradicted insights of giants of science such as Aristotle or Ptolemaios and the basic pillars of the scientific method as understood at that time.

Some people like to forget about these things but geocentrism was not invented by the Catholic Church. It was constructed by Aristotle who lived 3 centuries before Christ and Ptolemy who lived 1 century after Christ. Aristotle was a philosopher and a scientist but Ptolemy was primarily a scientist. The belief in geocentrism was being defended by scientifically sounding propaganda, too. It couldn't prevent the theory from being incorrect.


The second example I want to mention is the interpretation of quantum mechanics. For centuries, people believed that the Universe was a deterministic system where every event has precise reasons why it occurs in one way and not another way. Many people believed that at least in physics, determinism was a universal requirement that a scientific theory had to satisfy. Well, they were proved wrong, too.

Today we have fully rational and scientific evidence that all the work on "hidden variables" and similar attempts to return to the era of determinism was incorrect. The best theories that explain virtually everything we observe inevitably imply that it is only the probabilities of different outcomes that can be scientifically calculated. The basic principles of quantum mechanics have been settled for decades and they show that the world is different than what the people have thought to be necessary for science as recently as a century ago.

In both cases, and many other cases in science (not so much in religion), these beliefs sold as "principles of science" emerge as a result of extrapolation of successes of a certain kind of scientific approach in the past. Such extrapolations can turn out to be wrong and they often do. Nature doesn't care how many times someone screams "it is essential for science!" when he or she tries to defend a wrong statement. If a statement is incorrect, it doesn't help to call it "scientific" as opposed to "religious". As we will discuss below, if someone screams "everyone must respect the holy principle that there is only one vacuum", it is equally religious and equally dumb a statement as the statement that "everyone in the world must worship the same God".

Of course that the "idealized scientists" are more likely to avoid irrational approaches and unjustified dogmas. But there is no universal and eternal definition of an "ideal scientist", partly because the beliefs of such an ideal scientist depend on scientific questions that have not been fully answered. A scientist must be honest, bright, hard-working but whether he should believe these philosophical principles or others must be left to himself: it is a dynamical question whose answer may be influenced and should be influenced by the scientific research itself.

In real science, opinions are being shifted by evaluating evidence rather than by frantically parroting philosophical dogmas. The same thing holds in many other situations in science, including the question about the multiverse.

The multiverse

For many centuries, people were working on theories that assumed that all parts of the world must be visible with the help of ordinary light and they have universal properties such as the spectrum of particles and their interactions. Virtually everything we have learned is compatible with this assumption. One is therefore often tempted to extrapolate. It must be true, otherwise it is not science, is it? If must be true, otherwise our belief system collapses, doesn't it?

Well, it may be true and it may be false. Some people who consider themselves "scientific" seem completely incapable to grasp the concept that their assumption may be wrong. Because the inability to understand that an assumption could be wrong is a defining feature of religious zeal, these people are living proofs of the fuzzy boundary between faith and reason. They apparently think that if they repeat 1,000 times in their books and their blogs that their assumptions are necessary for science, the assumption suddenly becomes true. But if it is wrong, if will remain wrong. By their frantic repeating of a hypothesis 1,000 times, they just reveal something about themselves, not about the Universe. They reveal that they are the same kind of zealots as those who have opposed heliocentrism.

Paul Davies correctly writes that the anthropic principle is becoming increasingly acceptable in the science circles. Of course, in this context, both of us are talking about real scientists - meaning people like Weinberg, Hawking, Susskind, Wilczek etc. and not Sarfatti, Woit, Smolin, his collaborators who have blogs, and similar shadow physicists who have almost nothing to do with the contemporary research in physics but who are extraordinarily capable to jam the Internet and the newspapers with garbage and to push the laymen's opinions exactly in the opposite direction than the direction indicated by the scientific research. Why is it becoming increasingly acceptable?

Well, simply because the alternative explanations keep on failing. As a person who finds the anthropic principle inconvenient, I would like to tell you something else but I won't do it because it wouldn't be true and I happen to prefer the truth over wishful thinking. The details of possible non-anthropic explanations simply don't seem to work so far.

It is pretty clear that if a truly viable - and correct ;-) - non-anthropic explanation of the tiny cosmological constant and other constants emerged, it would be abruptly accepted. The anthropic explanation seems far less attractive, far less accurate, and far less convincing which is why it cannot be accepted too quickly but that doesn't mean that its validity would cause the end of science. It doesn't mean that science would cease being scientific. It doesn't mean that this answer can't be the correct one.

As we are excluding non-anthropic scenarios by seeing that their details don't work, the likelihood that the unattractive anthropic explanation is correct is inevitably increasing whether someone likes it or not. It is increasing in the same way and for the same reason as the likelihood that the probabilistic approach to quantum mechanics was correct. That likelihood was also increasing when new problems with various "hidden variable theories" were being discovered.

The analogy between the probabilitistic interpretation of quantum mechanics and the multiverse could turn out to be very accurate - but it could also be wrong. We just don't know for sure. It is a matter of belief. If the analogy turns out to be correct, people in the future will emphasize that we should have known it from the beginning: in both cases, one can talk about "parallel universes" - something that shouldn't exist according to the widespread prejudices but something that is a possible interpretation of the newer and correct scientific answer.

Strategies to find the answers

Physicists should sensibly divide their capacity to different approaches that can turn out to be right. They should also rationally divide their energy between all the approaches to the vacuum selection problem on one side and other problems where progress is more likely on the other side. But the worst thing that could happen to science would be if some intellectually inferior zealots such as those who gather at Peter Woit's blog were intimidating scientists and if they were preventing them from reaching a certain kind of "inconvenient" results.

I am not sure whether the ultimate explanation of the Universe completely avoids the anthropic principle but I am certain about many other things, especially the fact that science has much greater chance if it is pursued by intelligent and hard-working people who don't have to be afraid of the reactions of the laymen. One should expect much worse results if we fill science with less gifted people and one may expect comparably bad results if the scientists are afraid to communicate the conclusions of their work because they could be attacked by Peter Woit or similar aggressive zealots.

And that's the memo.

Bonus: a list of incorrect "essentials of science"

Finally, I want to end up with a slightly more extensive list of seventeen wrong beliefs that were once considered to be essential components of any scientific explanation of the world. I am confident that this list instantly falsifies the naive viewpoint of many shallow thinkers, such as those at, who deny that science itself requires a sort of belief system that is often unprovable and that can turn out to be incorrect.

They deny that scientism is not the same thing as science and they deny that a person who uses the magic word "science" (or "falsifiability") in every other sentence may still be wrong (and easily "falsified"). Here is my list:
  1. Isaac Newton believed that his impressive physical explanation of the world required everything to be composed of particles. In his opinion, they were necessary for a scientific description of the real world. This principle was applied to light - that had to be explained as a flow of particles - and even more seriously, to heat. Newton thought that the heat had to be carried by a special kind of material substance.
  2. It was believed by Maxwell et al. in the 19th century that every wave requires, much like sound, a material substance to propagate; they believed that the existence of the aether was a crucial and inevitable component of any scientific or "materialistic" explanation of electromagnetic waves. Lorentz's and Einstein's liquidation of the luminiferous aether was one of the essential steps in the development of relativity.
  3. Marxists were selling themselves as the champions of a "scientific" world view. Everyone knows that virtually all of their beliefs about the society were incorrect but that doesn't mean that their beliefs about natural sciences were always right. For example, Lenin believed that the electron was as inexhaustible as the atom: he thought that it had to be composed of many other particles that were also composites and this hierarchy continued indefinitely. We know that the substructure certainly stops at the Planck scale and probably earlier and Lenin's belief was in no way important for the essence of science.
  4. It was believed by Lord Kelvin that a scientific description of the Sun implied that the Sun couldn't be older than 30 million years or so, because of an upper limit on the amount of energy that the Sun could have emitted. The severe lord viewed the alternative assertion of Charles Darwin, involving a much longer and much more accurate life expectancy, to be associated with an inferior scientific approach.
  5. Lord Kelvin offers us a lot of examples like that. But instead, let us look at Auguste Comte who argued in 1835 that the chemical composition of the stars would remain outside the realm of natural sciences forever because it was impossible to travel there and take a sample. The hypothesis about the hydrogen over there was untestable and not even wrong. It only took seven years for spectroscopy to be discovered and Comte's prophecy to be shown profoundly incorrect.
  6. It was once believed that a scientific description of reality requires us to think about the world as events that occur on the background of a Euclidean geometry. It was first realized by mathematicians that other geometries can be defined and analyzed. Later, it was understood by general relativity that a non-Euclidean geometry is actually essential to describe the gravitational force.
  7. In mathematics, it was believed by many people that every assertion can either be proved or disproved. Kurt Gödel has demonstrated that every sufficiently powerful system of axioms allows one to construct a statement that can be neither proved nor disproved. Moreover, the consistency of such a system of axioms cannot be proven within the system. There have been many shocks in mathematics - for example when Bertrand Russell proved that the old set theory was inconsistent.
  8. A century ago, physicists believed that the Universe had to exist from minus infinity to plus infinity. It couldn't have had a beginning because such a "Big Bang" would be resembling the creation according to many religious systems. It wasn't just a wrong belief but it had also led to some wrong research. For example, Einstein designed his model of a static Universe and introduced the cosmological constant for a wrong reason. The Big Bang remains deeply controversial among the Marxists. Contradictions between the Big Bang and Lenin's guesses about infinities play against the theory; on the other hand, Eric Lerner, an anti-Big-Bang theorist, argues that the Big Bang is bad according to Karl Popper. Because Popper is viewed as an anti-Marxist, this argument supports the commies' belief in the Big Bang. Amazing way of thinking, isn't it?
  9. As I have already mentioned, it was believed that determinism was a key component of a physical description of the real world. Many well-known scientists were thus led to reject the postulates of quantum mechanics and/or construct theories of hidden variables that were later falsified.
  10. Thirty years ago, it was believed that causality can never be violated, not even in the presence of black holes and not even infinitesimally. Such an assumption implies that the information is being lost during the evaporation of black holes. As we know today, the information is preserved and the reason why the old argument was incorrect is an exponentially small violation of causality and locality inside black holes, a kind of "tunelling of information".
  11. It was believed by some people that only particles that can be observed in isolation exist in the scientific sense. Today, quarks are known to exist. They are also known to be confined and they can never be observed in isolation.
  12. It was believed that the equations that describe the properties of elementary particles and forces had to have a unique solution. Now we know that there exist discretely many vacua of quantum gravity. Whatever are the rules that determine the correct vacuum, it is clear that the assumption that solutions must be unique has been turned into an irrational prejudice. This fact has also killed the original bootstrap program that assumed that consistency was sufficient to find the unique laws of quantum field theory.
  13. It was believed that continents were so huge that they simply couldn't spontaneously move. The motion of such a huge object sounded ludicrous or perhaps religious: God could perhaps move them but the laws of mechanics couldn't. It took several decades before continental drift became acceptable. Anti-religious sentiments may partially be blamed for this slow process.
  14. It was believed that non-white races had to be phased out for the human species to survive in a decent form: such a belief was argued to be an inevitable consequence of Darwin's scientific theory about the origin of species. Most of us no longer believe these things. But these beliefs were not confined to the Nazi Germany: they had many advocates in the U.S. and elsewhere, too.
  15. On the other hand, people believed - and some people still believe in 2007 - that there can't be any innate differences between groups of people because such distinctions are artifacts of religious hierarchies and science respects symmetry and creates all people equal. Or you should look what their exact wording is: they certainly defend a patently false scientific assertion by the jargon of scientism.
  16. The Soviet officials thought that conventional genetics was a burgeois pseudoscience. It was banned in the Soviet Union for quite some time. Instead, they were supporting "real progressive" scientific approaches such as Lysenkoism that led to food shortages in the Soviet Union and famines in China instead of substantial improvements in agriculture.
  17. In the first decade of the 21st century, it was believed by many people that a radical reorganization of the functioning of the society was needed to save the Earth from a global warming apocalypse. This profoundly religious viewpoint was advocated as the most important insight of science and those who disagreed were labeled as unscientific.

Once again, if there were an ideal scientist who would know the right answers not only to these old questions but also all questions that will be answered in the future and who could therefore choose the correct assumptions and avoid wrong assumptions 100% of the time (these decisions must be constantly made when someone actually applies the "scientific method" in practice), he could define a sharp boundary between science and religion.

But because the only being capable to do these things is God, assuming that He exists, the boundary between faith and reason remains inevitably blurred. There are no ideal scientists in the real world. And someone's quasi-religious focus on some dogmas of scientism simply cannot prevent him from believing deeply incorrect things about the reality. The more often the anti-religious jargon of scientism is being used to defend wrong beliefs, the more similar "real religion" and "real science" become.

Mechanical work may be defined easily but creative work and real cutting-edge research always requires some intuition and an appropriate choice of beliefs about the extrapolation of the lessons from the past.


  1. Lubos,
    I don't seem able to reach you except by this post. As it turns out, "Taking Science On Faith" is quite appropriate.

    I recently made a Post to John Brignell's Forum at The Title:-

    Exploding The Myth of AGW(Hard Science Thanks To Dalton) & (Izaac Asimov To The Rescue)

    I feel that you as physicist will be intrigued with the radical issues raised. If you care to comment, could you do so on our Forum as well as here.

    P.S. Your fellow countryman Vaclav Klaus will see the political ramifications, I am sure,( many here in UK consider him to be a very brave man, as well as Right)


  2. "As far as I can say, The New York Times remain by far the best source of science news and opinions among the English-speaking newspapers."

    Beg to differ with you there. You'll get news and opinion, all right, but it'll be painfully one-sided. Balance it with the Wall Street Journal, for instance.