The Trouble With … Smolin?


A Caveat…

I should note up front that I don’t have any `trouble with [Lee] Smolin,’ as my title might suggest. I couldn’t resist using the above title after I saw a Heffers book display featuring a new book, The Dawkins Delusion; a pun on Richard Dawkins’ book The God Delusion. I actually found Dr. Smolin’s talk rather fun and this is not meant to be a critique of his book or his talk. In fact, most of this post is only loosely motivated by Smolin’s talk.

A couple of days ago I joined a small group of Part III theoretical physics students to attend Lee Smolin’s public talk promoting his book, The Trouble with Physics: The rise of string theory, the fall of a science and what comes next. The talk was held in Cavendish (the Physics department); even though all of the formal theorists are situated in DAMTP (Maths). This led to half-serious speculation that a prominent DAMTP string theorist barred Smolin from speaking at the nicer CMS facilities. Much more likely, however, was the fact that the talk was jointly arranged by the IOP and the Physics Society, making the Cavendish a much more natural location.

Background (In)dependence

The question of the day was how to quantize gravity. This was among five `big questions’ in theoretical physics that Smolin identified; the others include dark matter, dark energy, and some other things that I don’t remember. (Rest assured that you’ve probably heard them before.) He started off historically and philosophically, noting his undergraduate background in philosophy of science.

Borrowing from the terminology of Loop Quantum Gravity, Smolin described the history of 20th century physics in terms of `background independent’ theories (e.g. general relativity) and `background dependent’ theories (e.g. quantum mechanics). He used this division as an ongoing theme of his talk, though unfortunately I didn’t readily understand how he extended this line. He claimed that the early founders of quantum mechanics who thought about `deep questions’ in quantum theory approached the subject from a `background independent’ school of thought. Meanwhile, the developers of the Standard Model (those we might call phenomenologists today) were more practical and `background dependent’ in their approach. That is to say, `background dependence’ stems from the “shut up and calculate” school of quantum mechanics.

Don’t worry too much about the meaning of the terms background dependent and background independent. The gist of the idea (in the physics context) is that in GR spacetime itself is a dynamical quantity explained by the theory, whereas quantum mechanics `lives’ on a static previously specified background. Smolin extends this definition to describe approaches to physics, where (if I understood correctly) the background dependent school is phenomenological while the background independent school is fundamental/formal.

I think Smolin meant to present these two `schools of thought’ as somehow dual to one another, but his main point was that the untestable nature of String Theory requres that we take a step back and return to `background independence’ — meaning reasserting what is meant by science.

Unfortuantely, I think I’ve largely misunderstood (or disagree with) this line of thought in his talk. The division between `deep thinkers’ and `practitioners’ of theories is arbitrary. The fact that Einstein, Heisenberg, and Dirac thought about questions of a different nature than Feynman and Gell-Mann is more of a reflection of the historical context of the field. The `formal theorists’ of the 1920s were developing the formalism for new physics without many new experiments guiding the way. As his popular image would have it, Einstein largely developed GR based on gendanken experiments. The `phenomenologists’ of the post-WWII generation had a wealth of experimental data in the form of hadron resonances coming from shiny new particle accelerators. The relevant question was to construct a quantum theory that explained these particles.

And yet Feynman’s thesis on the path integral formalism is also a `deep’ statement about physics. And Gell-Mann led the charge in making symmetry a fundamental part of quantum theories. String theory was, itself, developed in the 70s out of attempts to create a phenomenological model of the strong force. Since then it has become a leading candidate for a fundamental theory. The point is, I’m not sure if this divide between `background independent’ and `background dependent’ schools really exists. To be sure there’s a difference between a phenomenologist and a formal theorist, but the difference isn’t much deeper than the particular questions that they choose to study.

But, again, it’s altogether probable that I’ve misunderstood this part of the talk as it was directed primarily for the general public and I had let myself zone out. (My German friend sitting next to me periodically whispered to object to his definition of `background dependence’ or to remark that, “He can’t even spell ‘Schrodinger’!”)


What did perk up my attention was the end of his talk, where after going through a laundry list of `alternate theories,’ Smolin made the claim that at the end of the day, physicists should be doing experiments and that there are experiments on the near horizon.

Two things rattled me about this:

  1. He had spent a a large part of the talk explaining that there was a philosophical (or moral/ethical) issue at hand regarding how one can approach science without experiments, with an eye on the lack of predictions from string theory. It seems a bit evasive to brush this all aside and claim that the debate is neither here nor there because, oh look, there are experiments to be done. But…
  2. There are no experiments now or in the forseeable future that can accurately probe the Planck scale. (Unless the fundamental Planck scale is actually much lower than we think, such as in some extra dimensional scenarios.)

The experiments that he cited were GLAST and Auger (pronounced ‘ow-shAYe,’ apparently), which measure gamma/cosmic rays. Indeed, these experiments may shed a bit light on high energy physics well beyond the capacities of accelerators: cosmic rays have been observed at up to 108 TeV. However, these cosmic rays have relatively low flux (~1 highest-energy observation per decade is a good rate) and are only observed after scattering in the high atmosphere. The resulting data is sparse and it is (naively) intractably difficult to reconstruct the initial event. And even then, the energy scales are still orders of magnitude below the Planck scale where fundamental questions of quantum gravity can be probed.

I raised this concern during the questions, asking if Smolin could quantify the extent to which these experiments could span the parameter space of a quantum gravity theory. He explained that there were phenomenological models of gravity that had predictions at the scale of these cosmic/gamma-ray experiments, such as `doubly special relativity.’ But these are still only phenomenological models. While these are valuable in themselves, what Smolin is really pitching is a `fundamental theory’ of gravity that, presumably, lives at the Planck scale. Thus one would have to map the parameter space of fundamental theories at the Planck scale to phenomenological predictions at some lower scale. (The lower the scale the more testable the phenomenology.)

However, now we’re still in a bit of a rut. How do we know that the `fundamental theory space’ can be mapped, via predictions, to the ‘phenomenological parameter space’ in an invertible or near-invertible way? That is to say, how do we know that the observations we make at the low scale can tell us anything meaningful about the high scale that might distinguish between high scale `fundamental’ theories? We don’t even know how large the `fundamental theory space’ is, let alone whether or not we can think reasonably about an “Auger-inverse problem.”

My point (#1): Unless I’ve misunderstood something, it is misleading to suggest to the public that these experiments can probe quantum gravity. And while I’m being diplomatic — if I have misinterpreted something, please let me know so I don’t go on spreading myths.

Experimental Inaccessibility of Quantum Gravity

A fact about the nature of quantum gravity: it’s damn hard to test. Gravity couples very, very weakly at energy scales that might reasonably be accessible in our lifetimes.

Compare this to traditional particle physics in the heyday of the Standard Model, which was based on the phenomenological interplay between experiments and model building. New resonances gave hints of an organizing principle, such as the organization of hadrons into SU(3) representations. This, in turn, led to models which predicted particles at the next generation of collider experiments. And, at least while funding for new colliders was readily available, these experiments would dig deeper towards the smaller and smaller scales.

String theory was born in this context, originally an attempt at a phenomenological model of mesons and only later recast as a `fundamental theory’ of quantum gravity. To some degree, this realization was extremely lucky. This isn’t meant to take away from the tremendous insights of the founders of string theory, but only that they had discovered a theory that would have otherwise been generations (or even tens or hundreds of generations) away if it was to have been discovered in the traditional experimental-phenomenological model of progress. In some sense, string theory was a more sophisticated model than physicists were ready for — like one of these science fiction stories where the spacemen from the future go back in time and beat everone up because they have laser weapons and spaceships.

So like the cowboy who has to learn how to fly a spaceship, 21st century physicists have been trying to develop string theory. One can stretch the analogy a little: what the cowboy learns from tinkering with the spaceship engine might eventually teach him how to improve traditional cowboy technologies (see: Wild Wild West), whether or not he was able to get the spaceship to fly. In the same way, string theory has had an impact on physics outside of the realm of quantum gravity, such as in the AdS-CFT correspondence. This is a powerful tool whether or not string theory is a `fundamental’ theory.
But here’s my point (#2): In some sense, we were lucky to have discovered string theory and to be able to speak about theories at the Planck scale. The trade-off for this luck is that we cannot do experimental science at the Planck scale. As I understand it, this is not the same thing as saying that string theory is `untestable’ or `makes no predictions.’ If one were able to build a “galaxy sized Planck-scale collider,” then one could search for stringy effects and test string models.

This is different from saying something inherently untestable. I could alternately propose that little invisible fairies fly around and break supersymmetry, cause the universe to expand, and poop bits of dark matter around the galaxy. These fairies disappear whenever we try to observe them directly or indirectly. This is ridiculous—not necessarily because it’s terrifically implausible, but because it’s manifestly untestable as a hypothesis. To be fair the above scenario may be true, but it would be a truth inaccessible to science even in principle.

String theory is scientifically accessible in principle. Right now and in the forseeable future, it’s just not practically possible. So experimentally we just have to shelve it as something that is a hypothesis, at least in the strictest sense. This isn’t meant to be diminutive towards string theory, only to use the proper word for what it is in a strict scientific sense. String theory is still very important as our currently-most-viable idea for quantum gravity, for its contribution to AdS-CFT, etc; one just has to take this progress with the understanding that in some sense it’s a mathematical framework with assumptions and conclusions that haven’t been checked against nature.

(I understand that this is something of a sensitive subject, so I reiterate my extreme falliability and that I’m only reflecting upon my own limited understanding of the state of the field.)

What about alternatives?

One of Smolin’s key points was that it is important for the scientific community that young researchers have the freedom and encouragement to think creatively about new directions for quantum gravity. An audience member raised a valid, if interestingly phrased, point, which I paraphrase:

If one studies loop quantum gravity, then one ends up getting stuck in a loop.

By this, I suspect that he meant that young researchers who work on unpopular topics will have a difficult time finding positions in departments that are entrenched in conventional approaches. The Perimiter Institute is certainly a place that has fostered the growth of young researchers interested in diverse directions for `big questions,’ but otherwise I cannot think of any other similar research center in North America, though the Max-Planck-Institut fur Gravitationsphysik in Brandenburg is one in Europe. To be fair, this is probably because string theory has the greatest potential for scientific progress among the different approaches to quantum gravity. It would be difficult for a university physics department to justify `gambling’ on forming a quantum gravity group given the competitive nature of funding and academic publishing. However, perhaps the Perimiter Institute (which is only loosely associated with a university) will be the flagship for broader progress.

Historical note…

String theory rose out of the stagnation of the traditional “experiment – phenomenological model” interplay of particle physics. The first string revolution occured after the Standard Model had more-or-less been written down and discovered. It is not surprising that the second string revolution came following the demise of the Superconducting Super Collider; when a generation of phenomenologists had to shift their research interests. Experimental-phenomenological particle physics hit a stand-still (modulo neutrinos and models of extra dimensions), and the high energy theory community was ripe for new directions and `sexy’ fundamental questions. With the top quark discovery in Fermilab setting the fermion sector of the Standard Model in stone, and it was time for formal theories of quantum gravity (such as string theory) to take center stage. It is possible that novel results from the LHC will bring about Rennaissance in phenomenology, bringing it out of its dark age (with all due respect to the Tevatron and their luminosity records).

Crack pot alert

It’s worth mentioning that the first hand to shoot up during the question and answer session of Smolin’s talk was from a white haired gentleman who quickly went off on a monologue about his personal attempt to modify general relativity. As he was about to explain the `relevant chapter of his MSc. thesis,’ Smolin interjected: “I’m sorry to interrupt you, but I have a very important thing to say.” He went on to explain that he gets plenty of e-mails from people with similar ideas, some of whom may be reasonable, some may not. He emphasized that as a researcher he generally does not have time to discuss such ideas with each person that e-mails him, but that perhaps the gentleman would be interested in speaking to younger researchers. Smolin said that often younger researchers had more time (really?) and could often benefit from a scientific discussion of such ideas.

In less diplomatic terms: Physics professors don’t have time to hear out crack pots. But it might be a `useful exercise’ for graduate students to talk through a crack pot idea once in a while. (I found myself in this position while browsing the Berkeley physics library last summer.)

Last thoughts

Most of the above were my own personal thoughts motivated by Professor Smolin’s talk. The talk was rather enjoyable (though certainly at the level of a popular audience) and I even enjoy the new UK cover for his book. Unfortunately the organisers were unable to arrange for a debate between Smolin and one of the DAMTP string theorists, though private conversations suggest that there was at least one who was champing at the bit. As far as anything I’ve written, this reflects my current understanding of the field and I hope I haven’t said anything terribly inaccurate. I would appreciate any corrections to keep me off the path to crack pottery. 🙂


4 Responses to “The Trouble With … Smolin?”

  1. 1 Richard

    One of the physics society emails about the talk suggested that the reason they arranged the talk is because the Cavendish and DAMTP declined to hold one themselves.

    Someone I spoke to about the talk (who had to leave before the questions to go to a formal) told me that at the other universities Smolin’s been to they haven’t had a talk, just a discussion but for some reason they won’t allowed to do that here (maybe because of the IoP thing).

    I couldn’t go because as it happened I was going to the same formal as the someone but I wasn’t quite as adventurous about getting back to college in time.

  2. 2 Bee

    This is really an interesting post! You raise several good points, especially regarding the experimental situation. Up here at Perimeter, I’ve initiated a discussion group around the topic, you might be interested to check the website

    Quantum Gravity in the Lab

    I also agree with you that the classification into seers and craftsmen, or tree climbers and valley seekers (or swhatever) is not really applicable. Good physicists need to have aspects of both, and most indeed do have. The problem I see is that the current system is in favour of supporting the craftsmen part, and leaves little room for the more creative work that I think many people could and would do, if they weren’t afraid to spoil their career options.

    Regarding wild west, you might like my summary about last week’s debate:

    Good physics is conflict



  3. In contrast with Lee Smolin you are perhaps interested in an architects view on small and large scales in 6 curled up dimensions.

    Best regards,

    Leo Vuyk.

  1. 1 Not Even Wrong » Blog Archive » String Theory Debates

%d bloggers like this: