In physics, is experiment the only criterion for testing truth?

An argument about the nature of physics: Is experiment the only criterion for testing science?

String theory established to unify general relativity and quantum mechanics is closer to "a general mathematical framework" than a theory that can be proved or falsified.

A few decades ago, Karl Popper, one of the most influential philosophers of science, stated a conclusion: general relativity is a reliable science, while Freud's psychoanalysis is pseudoscience. Popper is interested in how to divide the boundaries between science and non-science, especially between science and pseudoscience. He called this issue "demarcation problem". After a long period of painful thinking, Popper put forward a simple criterion to distinguish between "science" and "non-science": falsifiability. In other words, if an idea can be proved wrong-at least in principle-then we can say it is scientific, even if it is finally found to be really wrong.

19 19 total solar eclipse proved the correctness of Einstein's theory under the witness of the whole world, and also deeply influenced Popper. He thinks Einstein's theory is a model of "good science". The following is taken from his book Conjecture and Refutation (1963). He took Einstein's general theory of relativity and Freud's and Adler's theories as examples to illustrate the difference between science and pseudoscience:

We can clearly see that Einstein's theory conforms to falsifiability's standards. Even if our existing measurement means can't let us determine whether its conclusion is right or wrong, at least its conclusion may be negative.

Freud's and Adler's psychoanalytic theories are different: they cannot be tested or falsified-no conceivable human behavior can refute them and prove them wrong. Personally, I don't deny that some of their theories may be important, nor do I deny that these theories may play some roles in the experimental science of psychology one day. But it doesn't stop there. I don't think the theories drawn by these analysts on the basis of "clinical observation" can be better than the conclusions drawn by astrologers every day.

However, it was later discovered that Popper's total solar eclipse experiment in 19 19 was probably not as impeccable as he imagined. When some scholars review the long history and explore the details at that time, they found that there was actually a calculation error in the early conception of Einstein's theory-this calculation error made the angle predicted by the theory that light was bent by a massive object (such as the sun) twice the true value, and the physical quantity measured by scientists in the total solar eclipse and cited as the theoretical proof was the curvature of light. Therefore, if the total solar eclipse experiment was conducted in 19 14 (originally planned for this year, but postponed due to World War I), Einstein's theory at that time was proved to be wrong. Moreover, after Einstein corrected the calculation error and put forward the final version of general relativity, the total solar eclipse experiment in 19 19 actually contained considerable errors. In order to get the best experimental data, Sir Arthur Arthur Eddington, one of the chief astronomers who presided over the experiment at that time, probably screened the collected experimental results manually. This story tells us that science is as complicated as life, and perfect "clean results" are hard to get.

Let's stop here about Popper's theory, but readers can't help asking, no matter how famous Popper is, he put forward this theory at the beginning of last century, so why can his theory attract so much attention until today? This has to mention another concept: string theory. This term may be familiar to you. In recent decades, basic physicists have been circling around the "string theory" and regarded it as the ultimate theory to explain all the substances in nature and their interactions-that is, the candidate of Steven Weinberg's "theory of everything". Obviously, the word "theory of everything" can't be understood literally. Strictly speaking, "string theory" can't even be regarded as "theory"-if we define the word "theory" as a mature concept system, such as evolution theory and continental drift theory. In fact, at present, string theory is closer to "a general mathematical framework" than a theory, which is the most complicated framework we have, known and mathematically, and its purpose is to solve the fundamental problem of modern physics: how to unify general relativity and quantum mechanics. General relativity and quantum mechanics are both very successful scientific theories, but when they solve some physical problems under certain conditions, such as black holes or singularities (the origin of the universe before BIGBANG), there will be sharp conflicts between them.

Physicists have come to the conclusion that since there are irreconcilable contradictions between general relativity and quantum mechanics, at least one of them must be wrong or imperfect. One of the solutions they found is string theory: it puts the two in a larger theoretical framework (see the picture above).

This is a beautiful idea, but there is only one problem: some basic physicists swear that string theory is not only a promising scientific theory, but even the only truth; Other scientists scoffed at this. They think that "string theory" is not a science at all. Their arguments are also very strong: string theory has nothing in common with any experimental evidence-whether it is vibrating superstring, folding multidimensional space-time or other parts of the theory-all of which cannot be verified by experimental methods. String theory is like metaphysics in the field of mathematics. Scientists never give anything labeled as "metaphysics" a good look. Unexpectedly, more and more public comments and scathing slanders are directed at only one philosopher-karl popper. What's going on here?

Last year, I happened to be lucky enough to witness such a seminar on the spot. At that time, I was invited to Munich to attend a seminar on basic physics. The specific discussion content of the meeting was "String Theory Debate". The organizer of the conference, Richard Dawid of Stockholm University, is a philosopher of science with a profound background in theoretical physics and a supporter of theoretical epistemology. He affirmed the efforts of string theory physicists and tried to defend "string theory": string theory is not just a gorgeous mathematical model divorced from real science. My task in this meeting is to ensure that every participant (half a philosopher, the other half a scientist, and one of them is a Nobel Prize winner) can fully understand the content of the introduction course of philosophy of science: what did Popper say and why-since many physicists in this group use Popper's falsification theory to deny the scientific nature of string theory, while others disdain it, it is necessary to let them know whether Popper's theory is scientific or not.

As a matter of fact, as early as one month before the meeting, the leading scientists from both sides have tried their best to launch an overwhelming public opinion offensive. Some people published an article declaration in the serious magazine Nature, and some people expressed their views on grounded Twitter, all in order to win the offensive and defensive war at the core of modern physics. In order to let everyone feel the atmosphere at that time, I excerpted several confrontations. "What scares me most is that if a theory that can't pass the experimental test can become science, then science is no different from those nonsense or science fiction." This comes from cosmologist George Ellis, who is obviously criticizing string theory; Followed by Swedish physicist Sabine Hossenfelder, he said: "The word' science without experimental proof' is self-contradictory." Peter Galison, a professor at Harvard University, hit the nail on the head and pointed out the core of the debate between the two sides: "This is a debate about the nature of physics." On another occasion, people who support string theory are not to be outdone. Cosmologist Sean Carroll wrote on his Twitter: "We can't predict in advance which theory can describe the world correctly." Then, he wrote: "Only scientists who lack philosophical literacy will regard falsifiability as the Bible." Limited by space, I can only extract a wonderful response from a scientist-Leonard Susskind, a theoretical physicist at Stanford University, invented a new word "Popoola" to satirize the scientist who is tearing the tiger skin of Pop ("Popoola" can be translated as "follower of Pop", and "Azi" or "Raz" means "stalker").

This frank and heated public debate is not surprising. It completely reflects the attitude of well-known scientists in today's academic circles towards philosophy: some people try to use some philosophical concepts, while others simply give up all philosophical thinking, and neither the former nor the latter really understand these philosophical concepts. In the war of string theory, they have a little knowledge of Popper's philosophy of science and its application in "boundary problems". At present, quite a few scientists (especially physicists) don't think highly of philosophy, which is very ironic for me, a "former scientist" who became a philosopher after the middle-aged crisis. As early as a few years ago, Stephen Hawking declared that philosophy was dead. Physicist Lawrence Kraus ridiculed that philosophy would remind him of the black humor of American comedian Woody Allen in the last century: "People who can't do great things will run to teach; If you can't teach books, teach physical education. " Popular science workers Neil de Grasse Tai Sen and Bill Nye also questioned philosophy. They think that young people are wasting their precious youth by choosing philosophy as their major.

Fortunately, this view did not appear until recently, and only a few physicists have this idea. In sharp contrast to the arrogant attitude of modern physicists to philosophy mentioned above, Einstein's attitude towards the same topic in his letter to his friend Robert Thornton in 1944: "I agree with you very much, and scientific methodology, history of science and philosophical thinking of science are of great significance and educational value. Nowadays, too many people-even many professional scientists-only see trees, not forests, only parts, not the whole. Even a little knowledge of historical or philosophical background can give them a broader vision and enable them to transcend the limitations of their contemporaries. In my opinion, it is this philosophical perspective that separates the' truth seeker' from the' professional and technical personnel who master a certain technology'. " In other words, if measured by Einstein's standards, most people in the field of modern physics may be "poor technicians", but there are very few real truth seekers.

But objectively speaking, Einstein's view on philosophy was not representative even at that time, so it may not have any practical significance today, not to mention that string physicists only account for a small part of the physics community, and they will not leave messages on Twitter-they may just be a more talkative part of this small group of string physicists. It can be seen that their attack on philosophy does not actually represent the overall attitude of the physics community. Unfortunately, because the top talents in all academic fields participate in this discussion, they can easily mislead the public-because ordinary bystanders who don't know this field (even physicists who study other topics) can only understand the grand topic of string theory through heated debates among scientists in public media and popular science magazines.

In other words, it seems that there are only two kinds of people in the field of physics that the public comes into contact with today: those who openly despise philosophy, and those who think they know the true meaning of relevant philosophy and think highly of themselves. From this point of view, not only the small circle of researchers in this field has fallen into crisis, but the whole society is gradually losing its praise and respect for humanity and science-not to mention the millions of scientific research funds allocated to physicists every year (of course, scientific research funds will only be distributed to physicists, not philosophers). Therefore, in today's situation, it is necessary for us to reopen the historical archives, carefully scrutinize the deep meaning of Popper's philosophy, and answer a question: What is its relationship with science?

Let's go back to Popper's theory mentioned at the beginning of the article: the criterion for judging "science" and "non-science" is falsifiability. This standard seems simple, but the bad news is that it is just an illusion. Popper's theory was summarized into a few sentences and sent to Twitter, which successfully deceived many smart commentators and made them seriously underestimate the complexity of the philosophy behind the theory. This is like turning Popper's philosophy into a small advertisement posted on a telephone pole: "What can't be falsified is not science, so don't waste the public's time and money!"

However, excellent philosophical thoughts can't be summarized by simple slogans like small advertisements, so people who intend to really understand this kind of thoughts can't be satisfied with such simple sentences. Even Popper himself, under the baptism of peer criticism and counterexample, has revised his theory of falsification and demarcation many times-this is what every prudent thinker will do. For example, he ruled out the role of verification in the process of establishing a theory from the beginning, because if people actively look for evidence to confirm it in the process of establishing a theory, it would be too simple to pass the test. In the lives of ordinary people and in the research process of scientists, similar tendencies are very common. Modern psychologists have given this phenomenon a proper term: confirmation bias.

But later, Popper admitted that testing is part of a reasonable scientific method, especially for theories that can produce bold and novel predictions. After all, Einstein was among the top scientists in 19 19 because astronomers confirmed his expected data in the total solar eclipse. But for Popper, this doesn't mean that Einstein's theory (general relativity) is "true", but it just means that people don't have to argue about it in the short term. Indeed, considering the conflict between general relativity and quantum mechanics mentioned above under certain conditions, today we no longer think Einstein's theory is true. However, it cannot be ignored that the general theory of relativity has successfully withstood the storm of a century, and just a few months ago, scientists first detected gravitational waves-the theory of relativity was confirmed again.

In addition, other theories of Popper have also changed. He once regarded Marxist historical materialism as a representative of pseudoscience, but later at least admitted the feasibility of Ma Zhe's historical view. Another similar example is Darwin's theory of evolution. At first, Popper mistakenly thought that evolution was based on meaningless tautology, so he was always skeptical about evolution, but later he gradually accepted this view. At the same time, he also admitted that even the best scientific theory may have falsification loopholes-considering that a complete theory always contains some auxiliary secondary theories and conjectures as the background. For example, if someone wants to verify Einstein's theory by "observing the sun with astronomical telescope and camera", then in fact, he must also ensure the lens focal length theory of camera, the basic optical principle of designing astronomical telescope, the mathematical model and hypothesis used to process experimental data, and the common sense and background that a large number of other scientists take for granted-in fact, it is equivalent to verifying so many theories at the same time, and he must also ensure that there are no problems. Therefore, when the actually observed experimental results can't meet the relevant expectations of the theory, we can't rush to deny the whole theory, because it is likely that only one secondary theory will go wrong. This is why we must go through repeated verification under different conditions before we can be sure of a theory to a certain extent.

Popper's job is not so much to answer the "demarcation question" as to put it on the map of philosophy of science, calling on philosophers to try their best to define what is science and what is not for the sound development of philosophy. This pattern lasted until 1983, when Larry Laudan, a famous philosopher of science, published a paper entitled "The Death of Delimitation". In this influential paper, louden claimed that it was a complete waste of time for philosophers to solve the demarcation problem. He listed a series of reasons in the article to prove his point of view. The most important thing is that it is absolutely impossible for anyone to find a set of necessary and sufficient conditions to define "science", "pseudoscience" or other similar things-and if these necessary and sufficient conditions are not found, it is impossible to completely distinguish science from pseudoscience.

"Sufficient necessity" (abbreviated as "sufficient necessity") is actually not a scientific philosophical term, but a logical philosophical term, but it is very important for us to understand Louden's theory. Louden believes that Popper and others' attempts to accurately define "science" and "non-science" are actually similar to the definitions in elementary geometry. For example, people define a geometric figure with an internal angle of 180 degrees as a triangle. This definition is sufficient and necessary: as long as the sum of the internal angles of a geometric figure is 180 degrees, we can say that it is a triangle (sufficiency); And if the sum of the internal angles of this geometric figure is not 180 degrees, it is not a triangle (inevitable). In louden's view, there is no such answer to the boundary problem, because concepts such as "science" or "pseudoscience" are inherently complex, multidimensional and vague, and there is no clear dividing line. In a sense, physicists who accuse "Popper followers" (that is, supporters of string theory) actually coincide with Lauden's idea: Popper's falsification standard is a "one-size-fits-all" standard far from the objective situation, which is not suitable for distinguishing "science" from "pseudoscience", but thinks that it can distinguish sound and imperfect scientific theories in frontier disciplines such as theoretical physics.

However, please allow me to refute the point I mentioned above again-I know I changed my mind many times in this article (laughs). Popper is by no means as miserable as Louden, Sean Carroll and Leonard Susskind mentioned above. Similarly, the issue of demarcation is by no means as useless as they say. Many researchers (including my long-term partner Martin Boudrie and I) have recently insisted that Louden's denial of demarcation is a bit too fast. Moreover, we all think that Twitter is not a good place to discuss the philosophy of science (here, of course, it refers to the criticism of Popper's theory by Sean Carroll and others on Twitter), and many detailed and subtle discussions can't be held there.

The good news is that there is no way to study the demarcation problem-as long as you are willing to abandon the limitation of necessary and sufficient conditions, the road will appear in front of you immediately, and in fact, even Popper himself has not asked that the demarcation problem must be studied under necessary and sufficient conditions. Then we can't help asking, if the necessary and sufficient conditions are abandoned, what will be replaced? The answer to this question is Wittgenstein's' family similarity' theory. We believe that "science" and "pseudoscience" are concepts with the characteristics of "family similarity". Ludwig Wittgenstein is another great philosopher in the 20th century. On this issue, Wittgenstein is like another Popper, although the temperament and socio-economic background of the two philosophers are completely different (translator: I won't tell you that Xiao Weijia is super rich), and even their research interests are completely different. [If readers want to know more about the differences between these two philosophers, they can refer to an interesting reading: Wittgenstein's Poker (Changchun Press, 2003), written by journalists David edmonds and John Edno)]

Interestingly, Wittgenstein left no words about philosophy of science, let alone basic physics or historical philosophy. Wittgenstein is really interested in "language", including language itself, language logic and language application. Wittgenstein pointed out that many practical concepts in life actually do not obey the clear definition mentioned by Louden. His favorite example is the seemingly simple word "game". If someone wants to define the word "game" by defining a triangle, there is only endless frustration waiting for him-if readers don't believe it, you can try it yourself, and you will find that defining "game" itself is an impossible "game". Wittgenstein wrote: "How do we explain to others what' game' is? We can only describe a few examples of games to each other, and then add,' something like this, or something similar, is a game.' This is not because we are ignorant, but because no one has ever deliberately drawn such a line. We can draw artificial boundaries, but can this make this theory more practical? I don't think so! "

The crux of the problem is that in many cases, boundaries do exist, but we can't find them, which makes "defining games" and "defining normative scientific methods" platonic ideals and always exist in the metaphysical dimension. We often create boundaries for a certain purpose, and then test whether they are effective according to the purpose of creation-just as when defining science and pseudoscience, we subjectively think that there are significant differences between them, so we must try to draw a boundary between them and emphasize their differences. I believe that no matter for scientists or philosophers, there should be no objection to the statement that astronomy and astrology are essentially different, but the question is, where is this difference? Or at least, where?

Similar to the above question, many scientists who attended the Munich seminar (or, more broadly, participated in the debate on string theory) did feel that there were some important differences between the traditional concepts of basic physics and what string theorists claimed. Richard David, the organizer of the seminar, is against using post-empirical science to describe string theory, which is easily ridiculed. He advocated the use of "non-empirical theoretical evaluation" instead. However, no matter how others choose these names, David and his colleagues found that physical theories such as string theory have deviated from the traditional scientific channel established in Galileo's time. Even if the truth is that Galileo himself pays more attention to theoretical research (he participated in a lot of theoretical debates and thinking experiments, and probably never finished the experiment of free falling on the leaning tower of Pisa), his ideas are absolutely falsifiable, and scientists like to design experiments to verify them (perhaps the most famous one is David Scott's experiment when he landed on the moon in Apollo 15: he landed a feather and a hammer on the surface of the moon without air at the same time.

Then, a bigger question is, does this mark the beginning of a brand-new scientific field, or is it just a short pause in the long history of science in the eyes of future generations? Is it possible that the end of basic physics is not because we already know everything we want to know, but because our intelligence and technical level have reached the limit? These are serious questions that not only physicists and philosophers should think about, but also the general public should think clearly. After all, it is the general public who provide funds for basic physics and other research.

To my surprise, scientists and philosophers are actively attracting more supporters from society in this string theory debate about the application and misuse of the principles of philosophy of science. I can't help but regret that it would be great if they could stop arguing needlessly with each other and cooperate to give full play to their collective wisdom. Instead of attacking each other with all kinds of unprofessional terms, they should have cooperated sincerely-of course, it is not enough to pursue better science, and they should also expel real pseudoscience: for example, homeopathy (this theory means that if a substance can cause a patient's disease in a healthy person, it can be treated after being diluted and shaken) or telepathy, and so on, such obvious pseudoscience still lingers. These are indeed worthy of serious criticism and analysis by academic circles and made public. Moreover, intellectuals and researchers (whether in physics or philosophy) also have the responsibility to try their best to improve society, because it is such a society that supports them and enables them to concentrate on such a difficult subject as epistemology or theoretical physics.

By Massimo Pigliucci.

Translated leaf xuanjia

Revised ding