Monday, April 07, 2008 ... Deutsch/Español/Related posts from blogosphere

Is theoretical physics possible?

It surely sounds as a silly question but dozens of discussions of mine during the recent years have shown that a vast majority of people wouldn't give the right answer. In fact, I am convinced that even a majority of college-educated people would answer incorrectly: they would answer "No".

I am not exclusively talking about the aggressive and mostly anonymous scum that gathers at the world's crackpot discussion forums such as "Not Even Wrong". I am literally talking about the majority of the world's population. Even during many maximally friendly meetings, too many people showed me that certain simple ideas are just way too difficult.

The misunderstanding - something that proves the failure of science popularizers to convey the very basic and the most critical ideas, rules, and principles of the scientific method (more important things than particular insights such as the existence of quarks) - can be summarized into several widespread and related myths.

Myth: Reliable answers to questions about Nature can only be obtained by a direct experiment or observation.

In other words, what you don't see is what you can't get.

Needless to say, the opinion above expresses a universal disbelief in any kind of science except for some of the most primitive portions of zoology and botany in which one only takes photographs of plants and animals, if you allow me to oversimplify a little bit. Nevertheless, way too many people believe the myth and some of them even think that this belief - an extreme empiricism promoted to a quasi-religion - makes them "scientific".

As science keeps on improving its understanding of the real world, increasingly vast groups of phenomena and objects that used to be considered as independent are identified as related or as exhibiting the same underlying principles or rules.

If two phenomena are related and if they can be proven to follow the same laws with the same accuracy, it really means that if we know the structure of one of them, we also know the structure of the other one. On the other hand, it doesn't follow that both of these phenomena are equally difficult to be observed or measured experimentally. The similarity of the underlying structure and the similar "ease" of verification are two different things.

In fact, it is not only possible but really usual for one of those two (or more) related phenomena to be much more accessible to tests, for non-fundamental but scientific, technological, financial, or ethical reasons. Technologically speaking, one kind of experiment testing e.g. "one channel" of a Feynman diagram may be much easier to be performed by an existing collider than another one. Or it can be much cheaper to build a new gadget to realize a task. Concerning the third adjective, "ethical", it is well-known that certain hypotheses can be tested with animals even though one can be pretty sure that - and why - many of the results apply to human beings, too.

High-energy physics is an extreme discipline as far as its ability to unify different phenomena goes. In fact, it justifiably claims to unify pretty much all physical phenomena that have ever been observed into an equation - I mean the Standard Model Lagrangian - with 30 free real parameters only (counting neutrino mass matrices). Zoology and botany are close to the opposite extreme - you need terabytes of disk space to store the movies about the life stories of known organisms and you still feel that you are missing a lot.

Quite obviously, zoology and botany is easier to understand for most people. It is the procedure of relating phenomena that used to look as independent where the real source of the myth lurks.

I am convinced that the people who have ever been successful - at least a little bit - in solving real puzzles of science or, at least, in solving problems in a physics olympiad know and "feel" very well why the myth is a complete nonsense. Architects and engineers can calculate the properties of many buildings that haven't been built or cannot be built, at least not with realistic budgets of the present world. It is not hard to add hundreds of similar engineering examples.

And theoretical physicists can not only reliably extrapolate the known laws of physics by many orders of magnitude - much further than an overly skeptical layperson could think - but they can also say a huge amount of details about completely new phenomena such as the behavior of tiny enough black holes even though none of them has ever been directly observed.

Simply speaking, it is possible if you do it right. And the brighter, hard-working, and (to a lesser extent) luckier a scientist is, the higher chance he or she has to figure out the correct answers to questions that are separated from the world of direct experience. Such a success is not merely a matter of chance. The "right" people can produce the right predictions and answers again and again. The popular books about physics often try to explain that our intuition based on everyday experience is not ready to understand the world at the fundamental level.

But virtually no one has tried to explain that even if you include all expensive professional experiments, scientists must still routinely use the known laws of Nature outside the precise contexts where they have been directly verified. And whenever it happens, we are not necessarily dealing with a "shaky extrapolation". In fact, one can employ thousands of extrapolations that can be shown, by scientific arguments, to be more or less guaranteed to be right. Yes, it is possible. Many other extrapolations can be seen to be unjustified or plain wrong but it doesn't mean that there cannot exist extrapolations that are demonstrably correct, at least with an overwhelming probability.

But who are those people who can say or find the right answers to questions that are separated from the world of experience?

Myth: If I cannot settle or understand a question, no one else can do it either.

First of all, if you are an average human being or even an average TRF reader and we talk about a particular but randomly chosen specialized discipline, such as the mutations of insect or subleading corrections to thermodynamic properties black holes, it is extremely unlikely that the right person is you.

Nevertheless, the idea that it must be you who must be able to decide - otherwise a theory is inevitably meaningless and all arguments are inevitably unsubstantiated - is a favorite populist talking point among many ideologues such as the critics of evolution, starting from William Dembski, or critics of theoretical physics, such as Peter Woit or Lee Smolin.

Their messages are directed to the least reasonable people in the population - namely to those people who are, in fact, so incredibly intellectually challenged that they can't even realize or admit their own limitations. These people, especially once they become visible, are a real existential threat to science in particular and the modern civilization in general.

Wise scientists - and I would say that wise people in general - always understand most of their own limitations or at least their possible existence. Especially the brightest scientists on Earth are often modest and many of them eagerly wait for someone who is brighter or luckier to crack a puzzle that they couldn't have cracked themselves. Sometimes for decades.

Myth: If my preferred scientist couldn't crack the question, no one else can do it either.

This is a variation of the previous myth. It is perhaps less selfish and more realistic than the previous myth - because it deals with another person and usually a brighter one - but it is still incorrect as a general principle.

While it is true that a single "top" scientist is likely to give the right answer to an otherwise mysterious question than other average people, there never exists and there cannot exist any monopoly here.

If we talk about someone's preferred scientist, one of the problems is that the preference itself is often based on wrong information. For example, if we talk about cosmology, most laymen would almost certainly choose one of the figures who are well-known from the media - and who have written their own popular books - as the ultimate experts. It is not hard to check that the people who have been driving the progress in cosmology during the recent decades would make a different choice.

If one wants to make a sensible choice of this kind, he or she should not only know which scientist is actually "in" but also which scientist has some of the specialized know-how or talent to answer a particular question. Even if your only task is to choose the most qualified people to settle a certain question, some knowledge of the field and of the structure of the relationships between its main ideas (and ideas of the adjacent fields) is extremely useful if not essential.

Outsiders can solve many problems and they often do so. On the other hand, it is usually not about generic people. The people with a chance to do something extraordinary have certain positive characteristics that make them more likely to solve the task whatever the task is. But they don't have to have all the characteristics you are thinking about. Even if the probability for someone to succeed is significantly smaller than 100%, the existence of many people in a similar ensemble can (over)compensate for the lower probability.

The result is that major breakthroughs are often made by people who are not too famous at the relevant moment. But once they do something, it still doesn't follow that they are or they must be the brightest people in the discipline. The expectation value of their "quality" exceeds the "quality" of the people without achievements but there is still a lot of chance and other factors that have led to their success. Science simply doesn't work by searching for the single one "best person" who has to be right.

Myth: If a majority of people or holders of a scientific degree has a certain opinion, it must be the right one.

This very silly myth is being actively promoted in certain contexts, especially in the context of the climate hysteria.

Superficially, it looks like the evil sister of the previous myth because it relies on huge majorities instead of selected individuals. But if you think about the social dynamics in detail, these two myths are almost identical. Why? Well, if someone wants to assign special people (or one person) with some extraordinary influence, such a choice of the special people inevitably depends on some social phenomena that have "made" the person special. At the end, everything depends on the majorities anyway.

Outside the world of science, we can study similar examples. Most notoriously, Adolf Hitler was the ultimate dictator. On the other hand, the fact that he could have become one depended on the irrational behavior of large groups of people - that ultimately included a majority of the German nation.

So the extreme approaches - in which either a selected "holy soul" or the whole population should have their say - are both irrational and they are pretty much isomorphic at the fundamental level. Both of them are scientifically unjustifiable. Neither votes nor religious worshipping of individuals is a part of science.

In politics, a 55%:45% victory in the elections is viewed as a pretty safe one. But this is only due to a rule that was deliberately and "artificially" introduced by humans to the election system, a rule that makes the "55%" majorities very strong.

If the numbers 55% and 45% represent probabilities, you certainly cannot be sure that the "event 45%" won't materialize. And if the numbers 55% and 45% encode the people who believe in certain things, the result means even less than in the case of the 55% vs 45% probabilities simply because the counting of the people includes a lot of noise and nonsense and the numbers of people who believe something differ from the actual underlying "probabilities" or anything else that matters in science. It is completely wrong to "amplify" the counting, to say that "55%:45%" is the same thing as "99%:1%" or even as "100%:0%", and to import the "winner takes it all" principle from politics to science.

Ideally, one must always rely on solid and technical scientific arguments. Realistically, this is usually not possible. But if someone needs to rely on some social heuristics before he decides about a scientific question, the heuristics must be put together calmly and rationally, too. Once a group of people starts to amplify the influence of uniform masses or the holy Führers, it is becoming very clear that their behavior reduces the signal-to-noise ratio and introduces irrational bias to a debate. The groups where these methods play a key role should be disregarded altogether.

For sensible people, it makes sense to completely rely on some brighter people only if one is sufficiently certain that there don't exist other bright people who are more likely to be right than your favorite thinker and who have a different opinion. And it only makes sense to assign the importance to large groups of people if you can demonstrate that most of them have reached the opinion honestly and independently i.e. if there exists a body of independent scientific work and insights that is comparably large to the size of the large community whom you intend to trust.

If a group's opinions are being amplified because "the group is so large because it is so large" or if the opinions of an individual are being promoted because "he is so holy because he is so holy", you are a witness of sectarian rituals, not procedures that at least remotely belong to the scientific method.

And that's the memo.

Add to Digg this Add to reddit

snail feedback (0) :

(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o), m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m) })(window,document,'script','//','ga'); ga('create', 'UA-1828728-1', 'auto'); ga('send', 'pageview');