Wednesday, October 04, 2006

The Inverse Problem

There have been so many reviews about Lee Smolin's (The Trouble with Physics) and Peter Woit's book (Not Even Wrong) in the last weeks, that I have kind of lost track who said what about whom and why. Besides the reviews, there are comments on the reviews, comments on the comments, psychoanalytic examinations of the author's intentions, or otherwise people who take the opportunity to comment on whatever they think the problem is, or isn't, or what other peoples problems are, or aren't. And if there is no problem, how boring, let's go make one, preferably for somebody else.

Today for instance, I read George Ellis review in Nature (Nature 443, 482, 5 Oct. 2006) about Lee Smolin's book "Unburdened by proof - String theorists are setting a worrying trend by downplaying the need for experimental evidence."

The review itself is very reasonable in my opinion, the biggest part being a summary of the books content, with only spare judgement ("an excellent presentation of the foundations of fundamental physics", "Smolin crystallizes what many in the physics community feel about these extravagances of string theory"). The last paragraph then adds the opinion of George Ellis (so I presume)

What is crucially needed in developing string theory is a serious attempt to engage with the philosophy of science, developing an approach to theory validation that is adequate where insubstantial evidential support has to be supplemented by other principles of inference. So far, this has not been done.

Though I seriously hope that evidential support for or against string theory will arise soon and it won't be necessary, I do agree that in the absence of any contact to reality, we should think about where to draw the line between theoretical physics and mathematical physics. However, this does not only account for string theory. I should also add that I certainly don't mind mathematics, having started up as a student of maths, but one shouldn't sell the one for the other.

The same issue of Nature also features an editorial titled "Power and particles, String theories dominate for good reason." This is probably meant as kind of an antidote, but is so weak that it does completely fail in its intention, ending with the sentence "Critical-mindedness is integral to all scientific endeavour, but the pursuit of string power deserves undaunted encouragement.", which can be easily read as a support for the accusation about the exclusiveness of the string community.

Also in this same issue is a commentary about Lee's and Peter's book by Geoff Brumfiel, titled "Theorists snap over string pieces- Books spark war of words in physics." which quotes Joe Polchinski:

In recent years the theory has contributed significantly to heavy-ion physics, according to Joe Polchinski, a string theorist at the Kavli Institute for Theoretical Physics in Santa Barbara, California. When the Relativistic Heavy Ion Collider at Brookhaven National Laboratory in Upton, New York, first produced a hot quark gas, it was string theory that correctly predicted, retrospectively, some of the gas's properties. "In many ways, I feel the boundaries with other areas of physics are coming down," Polchinski says.

Though I am not sure whether I would call the application of the AdS/CFT correspondence to heavy ion physics a direct prediction of string theory (?!??), I do agree that I think the boundaries to other areas of physics are coming down, and I see this as a very good development. More importantly, it is one that is already taking place.

And this is the purpose of my writing today. In my perception, it is not even a very recent shift in priorities that many of those working on string theory have realized that a connection to what Lee calls 'the real world out there' should be a prime goal in their research. That is, unlike the exploration of all mathematical features of the allegedly beautiful theory, string phenomenology has grown to be an important field, maybe sparked by the idea of large extra dimensions in '98.

I myself know numerous people working on string phenomenology, and even though I personally don't believe that string theory is the-one-and-only theory of everything, this is not a vacuous project. Of course one can now argue how much of string theory does really go into predictions for colliders or the such, or whether any such connection would be unique and what one can learn from that at all. These are all questions that one has to think about, that one can argue about, and to address them is part of the scientific endeavor. E.g. you might want to check the paper I found on the arxiv today:

LHC String Phenomenology
Authors: Gordon L. Kane, Piyush Kumar, Jing Shao
hep-ph/0610038

which addresses the LHC inverse problem, that is: when we see some new physics at the LHC, can we uniquely find out what was its cause? Surprising for me, the authors state that this problem has received little attention until lately. This surprises me because it was my favourite question to ask in whatever talk: if you see these signatures, can you be sure it's what you have predicted and not something else? A question that is widely applicable in almost every talk btw, very handy. And the answer is usually: no.

In the last weeks I have read so many nasty things about 'the string community' that I suddenly feel like I have to state that is a very unbalanced polarization taking place in a public debate. A debate about what theoretical physics means in the 21st century, which in my opinion should not have been lead in public in such an unscientific way. Unfortunately, this has already happened and I can only hope, that it calms down to a level where we can discuss raised concerns without being personal. In my experience, string theorists are not more or less scientifically blinded by their own believes than those working out other believes about how nature fools us. Though I admit that some of them behave kind of strange when they are clustered to groups.

To quote another paragraph from the above mentioned article by Geoff Brumfield:

The books leave string theorists such as Susskind wondering how to approach such strong public criticism. "I don't know if the right thing is to worry about the public image or keep quiet," he says. He fears the argument may "fuel the discrediting of scientific expertise".

This, I'd say depends on whether the arguments are discussed with scientific expertise. Or personal insults.


Note added: Lubos has something to say about the Nature articles as well.

25 comments:

paul valletta said...

B, a great overview of what is becoming an issue of immense consequesnces for the scientific community.

I must state that having first encountered:
http://particle.physics.ucdavis.edu/hefti/lecture.html

I was really taken in by his personal viewpoints, and this period made me look at a number of fascinating "pathways", if it was not for Susskind, I would not have contemplated a large number of fruitful Questions, so my admiration for him persononally will always be steadfast.

But the fact remains that I see the problem, as you highlighted, a Mathematical consequence, which can be stated thus:
I am of the opinion that Anthropic reasoning has matured, out of the Mathematical arena, due to the inability of mathematical solutions to certain physical question.

Fail at Maths, and there is no other arena but an Anthropic course to wander in, save your status!

I mean, any mathematical genius that is defeated by string theory, would really be lowering his/her genius status, if it was "NOT" for Anthropic Landscapes.

Mathematics paint's a beautiful picture, but if when the paint dries, the plethora of Colours which is your artwork, cannot explain the finished article, then re-tracing the "wet-paint" backwards, your only left with a Black or White initial condition (the foundation canvas that accepts the paint), thus, an artist can explain the genius of his brushwork, but not the canvas that supports the paint?

Of course, there is the counter example, and anyone NOT of mathematical genius status, may be opposed to the Anthropic conotations, as a matter of default!

As a Non-mathematical genius ( kidding ;), I myself do not fear any sort of physical/experimental questions, and thus, any sort of physical Answers.

If there were no problems, then the "Inverse Problem" would not arise?

I hope I made a little sense! if not then I will NOT give an Anthropic exemption ticket as an excuse, I know my limitations.

All the Best,paul.

Anonymous said...

Dear Bee,

Thanks for being a voice of reason. I have sympathized very much with Lee Smolins appeals, but I simply don't like the way he uses the media to draw attention to himself. The book is superfluous, and mainly an annoyance for those of us who try to work out the funding issue in a calm and civilized way.

Anonymous said...

Bee, just to let you know the link to the paper detailing the inverse problem, does not connect?

Bee said...

Hi Anonymous,

thanks, I have fixed the link. Regarding the media, I am actually not sure who uses whom in this case. Best, B.

fh said...

I have to say I have become mostly bored with this stuff. But I find Susskinds reply slightly disinginious after all String THeory has had a lot of public praise and acclaim over the past decades, which is of course one of the reasons why the critizism is, unfortunately being debated so publicly.

stefan said...

Dear Bee,

thank you for that sensible update of the debate, and your opinon how it should be run...

Now that I have read the Nature pieces, I agree that the editorial is somewhat strange, since in the end, it seems to completly ignore all points and ideas raised in the Ellis review and the Brumfiel report.

Thank you also for pointing to the paper about the inverse problem for new physics at the LHC. I can imagine that this may be a harder problem than one first thinks. Since heavy ion collisions have been mentioned again, I may say this inverse problem reminds me of the story of the J/Psi:

In 1986, there was a proposal by Matsui and Satz that the "suppression" of the J/Psi would be an unambiguous signal of the formation of a quark-gluon plasma in heavy ion collisions. In the plasma, the string constant (or the slope) of the linear confining potential between quarks vanishes, and moreover, the Coulomb part of the colour-electric potential is damped by Debye screening. As a result, there should be no more bound states of charm anti-charm qarks, hence no J/Psi. Unfortunately, the signal was extremely murky, and difficult to measure. In the end, the CERN people could find it for central collisions of lead nuclei.

However, there have been many different explanations besides the original one that also predicted a suppression of the J/Psi signal. Some of them used hadronic models, creating a suppression mainly by what they called comover absorbtion or rescattering - reactions with the dense, hadronic gas in the fireball, which all resulted in the destruction of the J/Psi. These models could describe the observations as well. Moreover, taking an incoherent sperposition of all the different models, not a single J/Psi would have been left to be detected any more ;-)

The point is that there were many different plausible explanations for the observed lowered yields of J/Psi, but based on very distinct physical assumptions. In the end, the suppression of the J/Psi as such could not be used as a "smoking gun" signal for the formation of the QGP, as was hoped initially. So, if you have a complicated experiment with many factors which can come into play, as in heavy ion collisions, it may be extremely difficult to draw conclusions, and very different models can, in fact, fit the data.

Fortunately, if I understand that correctly, the proton-proton collsions at the LHC will produce much cleaner data than those obtained in heavy ion collisions.


Best, stefan

Bee said...

Dear fh,

I totally agree with you. I can't understand all the fuss there is. Yes, the current funding distribution is suboptimal, and yes, we should try to improve it. And yes, if I say it this way it sounds pretty boring, and it's certainly nothing new, so who'd want to write an editorial about this?

So I'll just repeat what I have said several times before (see e.g. Science and Democracy I and II)

a) In my opinion the central point is not 'the string community', but the question how support researcher to optimize progress, and what means progress. Esp. in the absence of evidential support, as George Ellis points out, we are in a situation that hasn't happened before and we should think about what 'theoretical physics' means to us.

b) If this discussion is lead in public by pointing at people to punch in representation for whole fields, it's most likely that funding agencies will conclude that we (= theoretical physicists) should first sort out our problems before they decide what to do. As a result, more funding will go into applications or PHENOMENOLOGY (in capital letters). The latter you can already see happening. It's not what I want, and I doubt that many of us want that.

c) Oh yes, and Lee Smolin has written a book, so what?

Best, B.

Thomas D said...

I don't think the editors at Nature know very much about theoretical physics. That is natural for a journal whose focus is on quite different areas. So it is not really surprising to have an editorial about string theory which makes little sense but eventually comes out for the status quo.

Does Ellis explain, or know, what he means by:

"an approach to theory validation that is adequate where insubstantial evidential support has to be supplemented by other principles of inference."

which to me sounds ... well, insubstantial. If I can try to translate:

"String theorists have to find a way of judging their theories without paying attention to experiment because experiments probably won't tell us anything."

which is both crazy and defeatist, particularly if in the first place he criticizes string theory for being too divorced from experiment.

Well, we know that theories can be invalidated without doing experiments, e.g. if they make no mathematical sense. In fact huge efforts have already been devoted to finding out if string theory does make sufficient mathematical sense, and that may still lead to many models being ruled out.

What other 'approach to theory validation' might Ellis have in mind?

And how many scientific advances has 'serious effort to engage with the philosophy of science' ever led to?

I thought that strings were being criticized for having too much pure thought and too little science...

Bee said...

Hi Stefan,

thank you for that sensible update of the debate, and your opinon how it should be run...

Well, you know that I always think the world would be much less complicated if just everybody would listen to me ;-) I'd like to kick some people, and tell them to go back to work, instead of proclaiming there is a 'war' going on. Gee, what a crap! I mean, we should remember that despite all the differences, we all share the same interests. We all want to understand how nature works.

See, this reminds me of the constant disagreements between the maths department and the physics department back in Frankfurt. There wasn't really any reason for this, except that each of them found it enormously funny to make cynical remarks about the research of the others, and sooner or later things got personal. There were several occasions in which millions of funding were lost because some committees did not get along with each other. Eventually everybody had to suffer. Though in this case the situation was (and is) much more serious for the maths department. Temporarily there were even rumours that it would have to be closed.

So. I'll go back to work now before someone kicks me ;-)

- B.

Bee said...

Hi Thomas,

I deleted one of your comments, it was submitted twice.

Does Ellis explain, or know, what he means by:

"an approach to theory validation that is adequate where insubstantial evidential support has to be supplemented by other principles of inference."


He doesn't explain. That's the last sentence. I hope he knows.

which to me sounds ... well, insubstantial. If I can try to translate:

"String theorists have to find a way of judging their theories without paying attention to experiment because experiments probably won't tell us anything."


Well, I don't think he meant to say that it's solely the string theorists who should judge...

which is both crazy and defeatist, particularly if in the first place he criticizes string theory for being too divorced from experiment.

Well, we know that theories can be invalidated without doing experiments, e.g. if they make no mathematical sense. In fact huge efforts have already been devoted to finding out if string theory does make sufficient mathematical sense, and that may still lead to many models being ruled out.


Yes, this is one aspect, but this is a quality control that certainly is there.

I don't know what Ellis had in mind, but to me it's a good question: if you have a theory that has no experimental verification, how do you decide whether it's mathematics or physics? And how highly rated should be closeness to experiment? On which timescale?

And how many scientific advances has 'serious effort to engage with the philosophy of science' ever led to?

Has it ever been necessary? If such effort has never lead to anything, does that mean it can't lead to anything?

Best,

B.

QUASAR9 said...

Hi Bee, I guess it is the Nature of things that there is more than one possibility until the others are cancelled out anhihilated, transformed or assimilated:

Do other dimensions exist, how many are there, how relevant are they to our four dimensions.
In biology we can see further and further into a cell, with practical 'physical' applications
With nanotechnology we can 'make' things smaller and smaller

But what exactly do we do with string theory. It seems we are concerned with whether string theory is or is not right, and yet little concerned with what the real implications or applications are in the real 'physical' world.

The same seems to be so with LQG, what does LQG mean to us in our everyday lives. We are trying to explain things as they are, does it serve any purpose other than philosophical whether the universe disappears up its own beginning every several billion years, or that it does so cyclically. What does LQG tell us about particles and what can it tell us about the Cosmos are legitimate answers, and therefore legitimate questions.

What analogy shall I use for the string theory & LQG debate - that you haven't aleady used.

Certain physicists would argue that learning Latin or Hebrew, is pointless because they are dead languages and not 'modern' languages. Diversity means there is a worldwide resurgence and interest in native or local tongues.
Certain physicists probably think it is pointless to read history, hmmm I could be tempted to agree considering some of the history that was and is 'being' written. But culturally and educationally history is taught even with contradictory facts and including certain significant gaps or dark ages (and not just prior to the middle ages), because it has an educational value.
Some physicists are downright luddites when it comes to art or literature "burn the books" and yet hollywood, bollywood, music, video games and all entertainment are all very 'real' parts of our daily life.

As someone said: It seems the less progress made in either theory, the more determined both are in conclusively disproving the other, as if that would be their proof.

JoAnne said...

B, thanks for your clear minded take on the brouhaha in the press about string theory.

Regarding the LHC inverse problem, you say:

"when we see some new physics at the LHC, can we uniquely find out what was its cause? Surprising for me, the authors state that this problem has received little attention until lately."

Contrary to this statement by the authors on that paper, the LHC inverse problem is very well-known and studied, for years now. In fact, it is in essence the backbone of the physics case for the International Linear Collider. In particular, there is a 472 page report by 140 authors, entitled "Physics Interplay of the LHC and the ILC" hep-ph/0410364. This tome studies a plethora of physics scenarios and how we can use ILC data to identify the theory which gives rise to the new effects observed at the LHC.

Bee said...

Hi JoAnne,

yes, that was also my perception, esp. after all that olympics...

But then. I am also guilty of beginning most of my papers by stating that whatever-I-am-working on hasn't received enough attention. Might just be part of the marketing :-)

Hi Quasar,

It seems we are concerned with whether string theory is or is not right, and yet little concerned with what the real implications or applications are in the real 'physical' world.

Depends on what you mean with 'right'? That's the problem: if the theory has no applications in the real world, it can still be 'right' in a mathematical sense, but is this the 'right' description of nature?

From a practical point of view, one can at least say that the theory is useless unless it predicts how nature behaves. That however gets more complicated if you consider how closely progress in mathematics and theoretical physics are tied together. And mathematics by itself has it's own purpose of existence, beyond being the language of physics. Just exactly with whom do we want to fight for DOE grants?

Best,

B.

Christine said...

Gee, what a crap! I mean, we should remember that despite all the differences, we all share the same interests. We all want to understand how nature works.

I once read a book that said that people can be divided into 3 groups (this is supposed to be valid for any human activity):

1) those motivated by money, the subject matter is not important (they look out for their own welfare);

2) those motivated by recognition from others, the subject matter is not important (socially motivated);

3) those motivated by the subject they work.

(And mixing of these in different weights of importance of course arise).

I came to understand that, in science, #1 people are almost absent (which is an obvious fact that we learn at early age), #3 are the romantics and idealists ("we all want to understand how nature works"), and #2 are those that gradually show to the #3 people how dim their careers will end up if they insist on such a pure and naive motivation. However, some lucky #2 people do survive. Some do probe Nature in a genuine disposition. But perhaps that was easier in the past. Today, the system favors more and more the #2 ones, which usually are more aggresive than the others, and end up, in many instances, in the control of the system itself.

Don't get me wrong, we *all* are motivated in some degree by recognition, the problem is when people get attached so much to their theories and how they can promote themselves with them, that we end up in the realm of pure sociological battlefields, even though that is not supposed to be part of an activity guided by logic, experimentation, creativity, and... wisdom.

Best wishes,
Christine
PS: Clearly, I am an incorrigible #3 person.

Christine said...

Oh, and when someone figures out how to solve the problem, please let me know. :)

For me this is much like a Penelope's web.

Best,
Christine

Bee said...

Hi Christine,

there is undoubtly truth in what you say.

But don't forget, there are only three types of physicists: those who can count to three. And those who can't.

Best,

B.

Anonymous said...

And how many scientific advances has 'serious effort to engage with the philosophy of science' ever led to?

Ever hear of Albert Einstein? Oh, right; he never fully bought into quantum mechanics and quantum field theory, so who cares what he thought.

Of course some will say he didn't so much engage with the philosophy of science as effectively reinvent it by his example and his writings, exerting much influence on a number of well-known professional philosophers in the process. The bottom line is that he was a deep, committed, and critical philosophical thinker, unlike many prominent figures in particle physics, who seem to regard philosophy as at best half-baked physics, and more often, simple bullshit and a waste of time. (God knows that a fair amount of bullshit has flown under the flag of philosophy.) As a consequence they don't do enough philosphical reflection or reading to understand its importance, or appreciate its contributions, including critical examinations of the role and hazards of mathematical formalism in the empirical sciences.

Garrett said...

"The most savage controversies are those about matters as to which there is no good evidence either way."
- Bertrand Russell

But, just to add my favorite bit of savagery against strings... it's stolen from Gertrude Stein:

"There's no there there."

Christine said...

"The personal satisfaction experienced by a discoverer of a scientific truth which used to be the main driving force in the past has given way to more material market-oriented motivations."

-- Bert Schroer
[http://arxiv.org/abs/physics/0603112v4]

QUASAR9 said...

Hi Bee, according to Susskind
"For example, extra dimensions may cause neutrinos to create microscopic black holes, which instantly evaporate and create spectacular showers of particles in the Earth's atmosphere and in the Antarctic ice cap.

Cosmic Matter

Is your interpretation of macroscopic blackholes the same as those who work on string theory
the microstate blackhole is part of the "particle creation" that you meet when you view reductionism and the microstate blackhole as part of the standard model.

QUASAR9 said...

Depends on what you mean with 'right'? That's the problem: if the theory has no applications in the real world, it can still be 'right' in a mathematical sense, but is this the 'right' description of nature?

From a practical point of view, one can at least say that the theory is useless unless it predicts how nature behaves. That however gets more complicated if you consider how closely progress in mathematics and theoretical physics are tied together. And mathematics by itself has it's own purpose of existence, beyond being the language of physics.


Thank you for noting the difference between right in pure maths and right in the physical world. Of course I meant the latter as in applied to particle physics, and predicting(?) or rather defining how Nature behaves.

Bee said...

Hi Quasar,

Is your interpretation of macroscopic blackholes the same as those who work on string theory
the microstate blackhole is part of the "particle creation" that you meet when you view reductionism and the microstate blackhole as part of the standard model.


Yeah. I am not completely sure what you mean with reductionism though. Also, I should point out that nobody really knows how to describe a micro black hole in the context of QFTs in the standard model, and everything that's been done relies on a semiclassical picture. That's a problem string theory might be able to solve.

But yes, I am talking about the same scenario as Susskind in the quotation.

Best,

B.

Count Iblis said...

Perhaps there will be new discoveries at the LHC that will lead to a new revolution in physics, just like in the early 1900s when quantum mechanics was developed.

Supersting theory, or even Supersymmetry, could be just a Fata Morgana of theorists who have been deprived of relevant experimental results for too long.

People in the late 1800s were working on complicated models to make electrodynamics consistent if you take the backreaction of radiation on accelerated charges into acount. But no one had invented quantum mechanics to correctly solve this problem.

Thomas D said...

If Schroer is really saying that string theorists have more

"material market-oriented motivations"

than people in the rest of physics - well, that is absurd and slanderous. (Just try to prove it...)

I don't see what good can come of that sort of trash talk. To claim that one type of physicist is morally better than another is nauseating.

As for (G.) Ellis, he is a smart guy and I ought to be able to figure out what he is trying to say, but no luck yet ;>

Christine said...

Dear Bee,

Finally, I've written a review on Smolin's book and posted it over at my blog.

Best wishes,
Christine