Sunday, August 20, 2006 ... Deutsch/Español/Related posts from blogosphere

Mountain climbing

Many people including David Gross, Brian Greene, Lisa Randall, and Lee Smolin have compared theoretical physics to mountain climbing. The idea is that the Physics Mountains around us represent the possible ideas about the physical world and the theoretical physicists' goal is to get to the top of the highest peak called Mount Nature.

What is the best strategy to get there? I think that the answer is obvious. If you already have experimental data that tell you everything you need, these are your maps. In that case, you think for a few hours and find the most straightforward path towards the beloved peak: you simply construct the right theory that matches the experimental data.

If you don't have any maps but you want to get to the peak, it is also clear what you need to do: you try to climb as much as you can. If there is a cliff on your way that you can't make, you try to find a local bypass and circumvent it. But you must still realize that your direction is upwards. If you're forced to move downwards for a while, you must have evidence that such a motion is temporary. This strategy will bring you to a local peak. If you're lucky, this peak is nothing else than Mount Nature. If you're less lucky, you will still benefit: it's because you will be able to see the scenery of the Physics Mountains from the top of the local peak.

Elementary facts about the extrema of functions guarantee that you will be able to see at least one peak that is higher. If it is so, you can sketch your map and prepare another trip for tomorrow. Chances are that you may even be able to see which of the other peaks is the highest one. You will also see that the tomorrow's peak is similar to today's peak, and you will be able to apply many skills that you have learned today. If you do things right, you will indeed reach a higher peak tomorrow than the peak that you have conquered today. If you repeat this strategy several times, we can eventually reach our destination.

Lee Smolin has a very different suggestion how to get to Mount Nature. Start to take the risk and chaotically probe the bridges and valleys. The more I think about it, the more silly it looks to me. It is an effective advise to jump down from the hill. It is a suggestion not to try to climb up anymore. Once you give up climbing up, it is the end of the story. If you don't break your skull by your jump from the cliffs, you will get lost in between the small rivers and bridges. Your tourism will have nothing to do with the search for Mount Nature.

There is only one Mount Nature which is one point on the map. The closer you are to the top of Mount Nature, the smaller region on the map describes your possible current location. The closer you are to the ultimate theory, the more determined your next steps are. Quantum gravity is extremely constraining and those who think that it is not constraining only think so because they are miles below the peak. When you are miles below the peak, there are many rivers and many roads. However, the vicinity of the peak is different.

Also, if you don't have a complete map (the experimental data you need), it is even more important to climb up and be very systematic about it. If you're not, it is hard to guarantee that you are making any progress. You might switch to a chaotic Brownian motion four miles below the peak. Smolin's suggestion is completely wrong, and the less we know which direction we should take assuming that the current direction does not lead to the Mount Nature peak, the more wrong the suggestion is.

Lee also likes to say that the revolutionaries in physics have always been taking risk in the sense that they were ready to publish wrong papers - to go up and down in the mountains - and only a certain (small) fraction of them turned out to hold the right answer. I think that this theory contradicts everything we know about the history of science. Among the great physicists, almost everything they have written in their most famous periods is true and a large fraction of it is composed of a rather systematic path towards their famous discoveries.

Take, for example,

You will see that all 27 papers listed by SPIRES are either about special relativity or the photoelectric effect, or steps towards general relativity or the first calculations using general relativity. Or consider

You will see 12 papers where he discovers the path integral (his thesis) or constructs the QED Feynman rules. Other papers with John Wheeler were games that were however crucial for him to obtain the right Feynman propagator.

You can try other famous physicists, too. None of them will confirm the hypothesis about risk-taking. There will be virtually no nonsensical papers in the list. There will be no copulating universes with mutated children and no octopi pretending to be the neutrino. All of the heroes have been writing true papers most of the time - and in fact, most of the papers were relevant for the important insights associated with these physicists. This fact that the papers have almost always been correct was one of the primary sources of the physicists' authority. The great physicists have always been able to find logical arguments that eliminate uncertainties and avoid ambiguities whenever it was possible.

No one would take the physicists too seriously if they were publishing rubbish most of the time.

In fact, you cannot really find the truth by random scribbling in which you take the risk of being wrong most of the time. You would become a random monkey typist. If a revolutionary paper we need is 2 kilobytes long, you will need to write 256^{2048} random papers before you get the right one. That's a rather large number but even if you work for 10^{5000} years to achieve this goal, no one will recognize that the last paper you wrote is actually the correct one because after all these events, everyone will unlearn how to distinguish a solid idea from nonsense.

Clearly, random guessing is not how progress in physics works. Physics, after all, is a rather systematic activity in which you want to make progress in virtually every decade or every year or every time period that happens to be relevant for your particular goals. You may try to guess but you must immediately look which of the guesses are reasonable (climbing up) and which of them are not (falling down). Without this filter, you can't really do any science. If you don't have experimental maps that guarantee such progress, you must rely on theoretical reasoning. You must simply focus on the cliffs around you - the mathematical consistency and details of the theories that you study right now - and try to climb up. There is no other way.

Those who are lost in the valleys and forests can make no progress - or make negative progress - for centuries. Most of them are making loops tens of miles away from the peak. In the context of gravity, the search for the theory of quantum gravity by doing loops tens of miles from the peak is called loop quantum gravity.

In the second half of the 19th century, some physicists were taking the risk to design microscopic theories of the aether. None of these ideas was supported by experimental data; none of these things was mathematically elegant either. They were just trying to reconcile well-established Maxwell's equations with their silly philosophical dogma that space should be made of atoms if it supports the electromagnetic waves.

They should have been able to see that this assumption was unjustifiable philosophical crap because Maxwell's equations could clearly work - and do work - even in the vacuum. At least, they could eventually construct models of the aether that could in principle underlie Maxwell's equations even though we know that they don't. FitzGerald has created a working prototype of such an aether.

Their modern counterparts are doing very similar things today. They try to create the gravitational aether. Again, it is not a way to reconcile two well-established insights of physics; it is a way to reconcile general relativity with a philosophical preconception that the space should be made out of discrete atoms: the very same preconception that Einstein humiliated by his revolutionary discovery of special relativity.

In this particular case, Einstein's equations instead of Maxwell's equations are supposed to emerge from a discrete starting point. They call it spin network, spin foam, loop quantum gravity, causal sets, dynamical triangulations, trinions, and so on. You can have dozens of names but it is still just a single unjustifiable, un-original, obsolete, and discredited idea whose measure in the space of interesting theories is close to zero. Another difference is that in this case, no FitzGerald will be able to create a spin foam with the desired behavior. Other physicists know very well why it can't work.

How is it possible that many physicists end up looking at models that are nothing else than increasingly dysfunctional versions of bad ideas that did not work even 120 years ago? It is exactly because they have lost the sense of altitude. They no longer want to climb up. They don't want to follow the basic rule of science that you must work on promising ideas (up) as carefully as you can and abandon the ideas (down) that have already been falsified and discredited. Instead, they want to dedicate the same amount of time to patently wrong discrete theories of gravity as they dedicate to the right theory that works, if not much more. They want all of us to follow this strategy; they want to abolish the very principle of science according to which different ideas are not equally good once you look at them. They call it risk-taking; I call it a postmodern anti-scientific confusion.

It is not a route to Mount Nature; it is a route to hell. 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');