Several people wrote in this morning to tell me about Phil Anderson’s comments about string theory that appeared in the New York Times today. These originally come from John Brockman’s “Edge” web-site where he has gathered responses from more than a hundred scientists and others to the question “What do you believe is true even though you cannot prove it?”.

Beside’s Anderson’s answer, also interesting is Paul Steinhardt’s. Steinhardt refers to the currently fashionable use of the anthropic principle as “an act of desperation” and “millennial madness”, notes that the Weinberg anthropic “prediction” of the cosmological constant gives the wrong value, and even acknowledges that string theory may just be wrong. For sheer weirdness, as usual these days, Lenny Susskind is hard to beat. Brockman doesn’t seem to have located any string theorists who believe string theory but can’t prove it. Since it can’t be proved, I guess even they don’t believe it anymore.

Phil Anderson has always been somewhat of an intellectual hero for me. He’s really the person who discovered the Higgs mechanism, among many other things. Despite a reputation for being a curmudgeon, at one point he was quite kind to me. At some sort of social event at Princeton to mark students passing their generals, he came up to me and told me that he had graded my solid state physics exam. He complimented me on one problem in particular, one I had got wrong. I had realized something was wrong with my solution of that problem, noting on my exam that the result I was getting couldn’t be right and explaining why. He told me that this had impressed him, that one should always know what the result of a calculation should look like before attempting it.

Ah yes, I remember this: T & V rishons, or whatever they’re called. The sub-structure of quarks and leptons. Maybe when someone constructs a satisfactory relativistic theory of bound states we’ll be able to see whether we can test Harari’s ideas beyond the mere counting of quantum numbers.

See Haim Harari’s response.

I guess if we are to take you Peter at your word, about where we should spend our money, then such avenues of exploration are then unsuitable?

A lot of people are picking up on Steinhardt, but lets say we discover something in regards to Planck, how would this effect areas of quantum geometry?

I don’t think mathematical physicists doing string theory have gotten any funding that would otherwise go to molecular biologists.

There’s also no evidence that the overhyping of string theory, M-theory, branes, etc. in recent years has lead to any funding for new accelerators. The only thing that has gotten extra funding because of this are the grants going directly to those theorists doing the overhyped work.

Freeman Dyson’s answer was the coolest.

Curious layman has two questions

1. How can you attract large scale funding from either public or private sources without all the nonsensical hype(eg. wormholes, etc). In otherwords, you came clean and stated the impact on everyday life would likely be much smaller than even GR plus you have to pick up a huge bill from the cost of a collider. How do you sell that?

2. RE: needless complexity of the mathematics. I read where the mathematician Rota stated that developments in knot theory and braids would probably be necessary to solve a lot of problems in *both* physics and biology (eg. protein folding). It has always seemed to me that mathematicians have played a minor role in molecular biology. Could it be that the umbrella of ‘String Theory’ is a clever backdoor for mathematical physicists to secure significant funds in an attempt to elbow into the turf of molecular biologists?

Reading the developement of someone else’s thesis is interesting approach to encourage further thought developement?

That we know full well model comprehension can stimulate other areas, and is always worth looking at for execising the brain’s further possibilties:)

*It is vital that such developement understand the evolution of geometrical design(roads to to higher dimensions) to include topological considerations, as part of the developing framework?*

The one common thread relating the twentieth century higher-dimensional models of the previous section is Klein’s compactification paradigm. Standard Kaluza-Klein, supergravity, and superstring theories all explain our seemingly (3+1)-dimensional world via compact topologies for extra dimensions. However, one the sidelines there have been concurrent models that do not make use of Klein’s idea; models that either make no assumptions about the size of extra dimensions or hypothesize that they have macroscopic extent. Up until very recently, such models were outside of the mainstream because of a fairly strong prejudice in favor of compactification.

But there has been a notable change of heart in the community at large, and there is currently a huge amount of interest in models that involve large extra dimensions, which we will now proceed to describeI know what people think in general about these proposals, but I highlight the question here* above.

“In the case of special relativity, the reason why causality implies vanishing commutators at spacelike separation can be found in any textbook”Not really. The best you can find is the notion that currents comprised of bilinear products of fields commuting for spacelike intervals implies commutation or anticommutation of the fields. It is just

assumedthat commuting currents mean causal separation.The problem, AFAIC, is

whatcurrents? Physical states are of positive energy & so the same thing should apply to non-local convolutions of the currents formed by filtering out the positive energy parts. However these donotin general commute for spacelike intervals, even when the correct spin/statistics connection obtains.Causality is in any case about the flow of information. What exactly one means by “information” is hard to define at the quantum level and what one sees in the text books – when they bother to address the matter at all – is (to my mind) far too glib.

There are some interesting things here, but I’ve always thought Edge was one of the (untentionally) funniest sites on the Internet. Something about giving people enough rope to make fools of themselves.

Thomas,

C is independent of causality, micro or macro! That is, homogeneity and isotropy of space and time produce a light cone structure – whether or not the world is causal (how it *uses* that structure) is not determined at that level.

-drl

Aperion wrote:

…and then, in the real future perhaps, trumpet the accuracy of the few that turned out to be correct. Is this not really pretty silly?

Relax, not everything in life needs to fit into some perfectly rigid framework. Sometimes, it’s ok for things to be just fun or entertaining. It’s not like the edge is PRL.

BTW

I can’t stand all this talk about Occam’s razor??? Nature feels no obligation to follow our wishy-washy, sorta-kinda, maybe-sometimes statements, and neither do I.

I propose a new scientific guiding principal. We’ll call it Ksh95’s Razor.

He who is guided by imprecise, unrigorous, folksy wisdom is an idiot.

In the case of special relativity, the reason why causality implies vanishing commutators at spacelike separation can be found in any textbook, which can explain it much better than me. In general-covariant theories, 3+1 decomposition and the very notion of spacelikeness is problematic, as has recently been emphasized by Savvidou. Her (and Isham’s) History Projection Operator formalism is philosophically close to what I suggest. However, she does not explicitly introduce an observer and can hence not formulate observer-dependent anomalies.

However, I tend to understand locality more intuitively. The problem with fundamental, non-local objects is that we cannot transmit information faster than the speed of light (or so I believe). So one endpoint of the string does not know what the other is doing, but only what is was doing a string-length/c ago. This is OK for real-world strings made up of atoms, but it just seems to be wrong for fundamental strings. The same argument applies to non-local observers. One part of a macroscopic, real-world detector does not know what the other is doing in this instant, but that is OK because it is made out of point-like quarks, gluons and electrons, which should be regarded as the irreducible observers.

It might be possible to transmit information faster than light over Planck-scale distances, so microscopic strings might (just might) be consistent with macrocausality. However, the by far simplest way to prevent superluminal signals over macroscopic distances is to forbid superluminal signals over microscopic distances.

Thomas,

I am thinking about this problem too. One could argue that the observed spin-statistics connection is excellent evidence for microcausality but personally I am still a little puzzled as to why (anti-)commuting operators for spacelike intervals should necessarily be the expression of the latter.

Chris W ,

It might also help to ask what could physically motivate the transition (at the classical level) from point particles to 2-dimensional objects tracing out worldsheets in spacetime?There are strong physical motivations for

notransition from point particles to 1D strings (or higher-dimensional branes): locality and causality. These properties, which are fundamental consequences of special relativity, are automatic in point particle QFT by construction. You don’t need to be Einstein to realize that there might be problems with locality in a theory of non-local objects – evidently it’s enough if you are Gross (though apparently it took him 20 years to take this problem seriously).One should distinguish between microcausality and macrocausality, where the border between micro and macro is perhaps at 1 fm. Macrocausality is firmly established and will not yield, but microcausality can of course never be tested – you can only push the demarcation line. However, just because microcausality cannot be experimentally tested does not mean that it is violated. QFT is microcausal, and that suffices for macrocausality. From this perspective, the idea of cosmic strings is really outlandish – how can you apply the cluster decomposition principle to a fundamental cosmic string?

I am convinced that the answer lies elsewhere. In QM, observation is a complicated, non-local process, and as such it is in conflict with the spirit of locality and (special) relativity. To make observation into a local process, one needs to assign it to a definite event in spacetime, and to treat the observer’s trajectory as a material object to be quantized together with the fields. This is clearly a very small fix to the logical structure of QFT, and it does not really involve any new physics, but it has a profound consequence: new types of gauge and diff anomalies arise, which cannot be seen in conventional QFT because the cocycles are functionals of the observer’s trajectory.

This does not rule out any new physics, of course, although observations do not seem to indicate anything beyond the SM and GR. But no matter what happens in future experiments, the insight that we need to localize the process of observation and that this leads to new anomalies can not be undone.

Excuse me for pointing out the obvious, but if one applied the Principle of Parsimony, aka Occam’s Razor, consistently, one would never have arrived at String theory in the first place.

btw – bad spelling intentional

I’m leaving this post – as admittedly a low investment observer.

Observations:

-Typical Question –

Peter:- String theroy is so terrible – Lubos Motl is an idiot!

Lubos:- Peter is not even qualified to understand why my superior knowledge is vodka-stle-eastern-european-vodka correct.

Chris O,

Given my characterization of string/M-theory as a theory generator, Occam’s razor is more properly applied to the theories (models*) it generates. The question then becomes, what physical principles can we invoke to select among the generated theories?

Consider general relativity. Its formulation in terms of Riemannian geometry inspired the consideration of numerous alternatives using more or less the same mathematical framework. The latter could be looked upon as a meta-theory from which one could draw many candidate physical theories for describing gravitation, some claiming to incorporate other physical fields in some kind of unification. However, the physical ideas underlying GR result in a remarkably restrictive sieve for choosing among alternatives; indeed, in the end it is hard to come up with viable alternatives.

At least, this was the case before quantum gravity came on the scene, with GR regarded as a low energy effective theory. The question now becomes (in the QG context), what physical principles can help us select among the mess of alternatives offered by the string/M-theory meta-theoretical framework? After two decades of investigation we still don’t know. It might help to at least state the problem in this way.

It might also help to ask what could

physicallymotivate the transition (at the classical level) from point particles to 2-dimensional objects tracing out worldsheets in spacetime? (Taking “physics as geometry” to the next level doesn’t cut it, in my opinion; more on that later.) The original motivation in hadron physics seems completely irrelevant now. What should (or could) replace it? The mathematical discoveries that enthralled Witten and others in the early 80s strike me as things calling for an explanation—for deep physical insight—not just a mathematical treasure hunt.** Of course this is especially difficult when the original physical motivation is so obscure, and the mathematical territory is so vast and intriguing in its own right.—

* BTW, I’ve always had some discomfort with the term

model, regarding it as something of a weasel word.(** Peter’s recollection of his encounter with Philip Anderson is relevant here.)

Sorry Danny, but I’m a big believer in minimizing the amount of trouble software brings into your life by sticking to the defaults when they are something you can live with, and this default is something I can live with.

Oak –

How right you are! In fact KK theory is the ultimate theory generator. If you can make bowling balls generally covariant, you can strike up a theory of unified bowling ball gravity in your spare time. Then you can refer your theory to a moving frame. Of course, in the end you’ll still never make a 7-10 split.

XXXXXXXXX/

Peter – aside – could you please change the font color to “black”? For some reason the MT people think charcoal grey is a nice default color for text.

Here’s how:

http://www.elise.com/mt/archives/000677colors_background_banner_headline_and_font.php

-drl

I read the entire Edge comment group. Nearly all the postings involve saying something like:

“I personally believe that X is true, and I think that in the future, X will be universally believed to be true by people like me.”

This is really quite a meaningless statement, and in fact may involve a serious misunderstanding of the nature of truth or potentially the conception of future and past.

I suspect the point of the exercise is to make unsubstantiated claims and then, in the real future perhaps, trumpet the accuracy of the few that turned out to be correct. Is this not really pretty silly?

Chris W,

The paradigm shift therefore is away from Occam’s Razor towards something much more bizarre. I would never go there, as I view the activity as opening a Pandora’s box. How many other “theory generators” besides M-theory could one dream up, I wonder? In any case, what satisfaction is there to be derived from explaining simple things in a complex way?

The response from Steve Giddings is also interesting. He thinks he knows what went wrong with Hawking’s original argument about information loss in black holes. An excerpt:

“We base our argument on a principle we call the locality bound. This is a criterion for when physical degrees of freedom can be independent (in technical language, described by vanishing of commutators of corresponding operators). Roughly, a degree of freedom corresponding to a particle at position x with momentum p and another at y with momentum q will be independent only if the separation x-y is large enough that they are outside of a black hole that would form from their mutual energy. I believe this is the beginning of a general criterion (which will ultimately (be) more precisely formulated) for when locality breaks down in physics.”

I don’t know what to make of this idea.

p.s.: Sorry this comment is a bit off-topic.

At least one respondent, Keith Devlin, started by examining the problematic notion of

proofitself. In practice, proof means (for many people) something like social or cultural acceptance—an idea is proven when it becomes widely discussed and applied, or commonly used as a jumping-off point for further investigations. Here “proven” is a shorthand for “proven to be of value”—interesting and useful at least, and at most, integral to how one understands the world.Personally, I think that in the empirical sciences we should limit our use of the term to the sense in which mathematicians and logicians use it—ie, as showing that a conclusion follows logically from given premises. As Devlin discussed, this notion of proof is problematic enough by itself. I see no need to muddy the waters by talking about “proving” Newtonian mechanics or any other empirical generalization. It is enough to know that a theory is a conjecture that has been rigorously and broadly tested, and has stood up well to all (or most) of the tests. That fact alone makes the theory interesting, and in most cases useful as well.

So, the attention should be focused on tests—how can we know when a theory is wrong? Many people mistrust string theory because it apparently offers so many ways to avoid confronting this question. However, I get the sense that for many string theorists the question is irrelevant. They regard string theory (and M-theory) as a fabulously rich meta-framework from which one can derive a huge variety of interesting models, and modeling is what physics is really about. At least some of the models should be testable, but there is no reason why the meta-framework should be; its purpose is to spawn models.* All one can ask is that some of the models show some promise of producing successful physical predictions. In short, string/M-theory is not a physical theory in the conventional sense; it’s a meta-theory and theory generator. Is a theory generator what we really need?

—-

* Of course, the meta-framework is such that the models are linked by a web of deep mathematical ideas and relations, which makes them interesting even to people who only care about solving mathematical problems.

If Susskind really doesn’t understand why probability theory works [given certain very basic assumptions about the underlying processes involved] then it’s not surprising he thinks that his impressionistic landscape stories and graphics are a real achievement.

Anderson has always had some degree of resentment about the way particle physics often views itself as the most important subfield of physics. He was far from the only one who felt that the cost of the SSC was crowding out funding for other subjects. I remember arguing at the time with various such physicists and mathematicians that if the SSC were canceled they certainly weren’t going to see any of the money, which turned out to be right.

I have mixed emotions when it comes to Phil Anderson.

I agree that his contributions have been notable (actually, Schwinger also had a bit role when it comes to Higgs mechanism).

As an ex-particle theorist, I find it hard to accept that he advised the Congressmen to cancel the SSC and said that “particle physics is no more fundamental than any other branch of physics, \emph{ in any way whasoever}”(I don’t think his testimony was pivotal: budget considerations and lack of political support in the new administration killed it). My opinion is similar to Weinberg as he stated in “Dreams of Final Theory” on the subject: it is about the convergence of arrows of explanation.

But he is correct that reductionism is not everything. I think he may agree that “many-body physics”, be it relativistic as in QFT or NR as in condensed matter physics, is.

Just so people understand this was a selective process (the many names that are presented) by which one will choose what to accept into their world or not?

I think Nicholas Humphrey deserves a lot of credit for thinking of the question,

“WHAT DO YOU BELIEVE IS TRUE EVEN THOUGH YOU CANNOT PROVE IT?”ðŸ™‚If the rational, is to accept matter distinctions as the most relevant, then such thought manifestations have no other basis then the brain:)

Rest easy then:)