In a public letter addressed to Peter Woit, Lee Smolin clarifies some issues and possible misunderstandings of his essay Why no new Einstein?
Lee's point was to emphasize that young ambitious theorists should be specifically supported to pursue their own research programs - independently of the "big" research directions. I have argued that the natural equilibrium - the invisible hand of the free market of ideas, if you wish - defines a very good balance between originality and reliability, and every bias such as Lee's proposal is bound to be counter-productive.
For example, the most likely reason why no young (or old, for that matter) person has convinced others about her alternative to string theory is that there probably exists no alternative to string theory.
If a field or subfield is overstudied, eventually the people realize that the amount of interesting problems in that field or subfield is rather small and there is too much overlap of their work with the work of others. If a field or subfield is understudied, it naturally attracts people because there are chances that significant progress can be done quite easily. The same comments apply to the situations in which too many people believe that something is true or something is false - as long as the community's goal is to search for the truth instead of finding evidence for some political goals outside of science.
If there is no progress or very little progress in a field, it usually means that no one knows how to make the next big step, and no amount of social engineering can change this basic fact because science can't be done by bureaucrats and politicians.
Lee says "String theory is criticized in the essay mainly because it is currently sociologically dominant". It's hard for me to understand this kind of thinking. I imagine a scientist as someone who judges the value of scientific claims and theories by their content, instead of their sociological dominance. Evolution is also dominant as a description of the origin of species which does not mean that I must immediately start to criticize it.
There are people who believe that there is a universal key to make progress in science and technology - to worship theories that have become sociologically dominant; this is the "scientific-consensus" type of people. Lee is apparently in the opposite, but not exactly competing group - according to which the progress is made by criticizing theories that are sociologically dominant.
I say that these opposite approaches are not really competing because their essence is more or less identical. The essence is to replace science by superficial sociological observations. The essence is to determine the answers to the big questions in advance. The essence is to decide how many talks should be positive and how many talks should be negative before the actual research is done. The essence is to guarantee a certain "share of the market of ideas" for a particular group - either the global warming alarmists or the loop quantum gravity proponents. I guess that both of these groups know that their importance will decrease if the scientific research is going to be done properly, so some of them try to emphasize non-scientific criteria.
Neither of these two "opposite" prescriptions can lead to any progress in the long term. Whether or not either of these two prescriptions "wins" a particular battle is a matter of chance. And the war is always lost because neither of these two approaches is a scientific approach.
Lee repeats some mysterious comments about "problems with string theory". As far as the detailed analyses of his statements by The Reference Frame can say, Lee's hypothetical "problems" don't exist. The last example of this sort was about Lee's conjecture that string theory may become inconsistent or divergent even at a finite order in perturbation theory. We've explained why we know that this conjecture is unjustifiable, and as far as we can say, demonstrably untrue. Lee's "sociological" justifications can't change the fact that what he says is not true.
They also can't change the fact that an even more rigorous proof of the perturbative finiteness of superstring theory is not one of the important topics in current theoretical physics simply because the question has been settled for quite some time. Everyone is free to investigate it; it's a very interesting piece of math. But no one has the power to force others to think that this is the golden problem of 2005.
Lee also mentions that the reason why Ted Jacobson spoke at a conference about loop quantum gravity is that that he became critical of loop quantum gravity. That's a pretty strange reason for a person to become a speaker at a scientific conference. At other scientific conferences, people usually speak because they have something interesting (and usually constructive) to say about the scientific questions that define the meeting themselves; not because they just become critical of something.
Ted Jacobson is a great guy who may be critical of loop quantum gravity - which means, of course, that he is correct in one particular "Yes/No" question - but he currently believes, among many other things, that black holes have a much higher entropy than the Bekenstein-Hawking entropy and that it's not a problem because this additional entropy is "hidden" below the horizon. Needless to say, this contradicts most of the detailed tests of black hole thermodynamics that have been done, especially the microscopic counting of the entropy in string theory.
Being critical of loop quantum gravity does not guarantee that the person has the right answers about other questions, beyond the single binary question "is LQG correct?". And the opinion about any question is only interesting for a scientific conference is there is some calculation, argument, observation, or experiment that supports it - not just because the sign of the answer fits an organizer's political agenda.
Lee even suggests that this bizarre policy to allow "critics" to speak just because they are critics should be adopted by other fields such as string theory. I could not disagree more. People speak at a conference about some topic because they study related questions - in a way that is sufficiently interesting and convincing for their peers - and because they have found some answers or evidence supporting particular answers, whether or not the evidence has a "positive" or "negative" flavor and whether or not they can be viewed as optimists or skeptics.
The speakers must know at least the basics of the field about which their speak, and they must be interested in some particular questions rather than dumb "Yes/No" screams about the whole field. These are the reasons why inviting a general critic - such as Peter Woit - could only be a funny idea to entertain the physicists once, but it certainly can't become a part of the standard organizational process because such steps would strikingly lower the scientific quality of conferences. One can't build a conference about string theory on speakers who have no idea what string theory is.
Incidentally, I don't think that there is not enough diversity of approaches in the string theory community. Sometimes the diversity - and the fragmentation of interest - seems just far too high. The fragmentation of interest is something that distinguishes the more quiet times from the revolutionary ones. Stanford, for example, seems to believe the anthropic principle that about 80% of the community rejects. There are places with a strong focus on topological string theory. There are places where string theorists interact with phenomenologists a lot. There are places where they interact with the loop quantum gravity proponents. There are places that prefer to study questions rooted in field theory and where they worship integrability. There are optimists as well as pessimists within the string theory community. I could continue for a long time. No one is currently able to convince others that everyone should study the same question. Someone may think that this is a great gift of diversity; but for most of us this is just a symptom of a lack of inspiration. Once new great ideas are revealed, diversity will be replaced by focus again.
Lee claims that he did not advocate funding of those who only work on foundations of quantum mechanics, but instead funding of those who are inclined to work on foundations of quantum mechanics. ;-) Big difference, is not it? Lee says that they are deep and independent thinkers. I would say this statement about a very few of them; there are many more crackpots who approach physics in this way. Most of the people who are bothered by the foundational issues in 2005 are those who have not understood the very basic framework of quantum mechanics and its inevitability. Most of them don't want to solve some cutting-edge open questions, but to "undo" even the discoveries made in the late 1920s. The people from the previous sentence are pretty different from those who solve actual problems connected with quantum computing.
Lee argues that the loop quantum gravity researchers don't ignore the well-known fact that infinite-dimensional constraint algebras generically acquire anomalies (a point emphasized to the LQG community by Nicolai et al., among others) and instead they have "rigorous existence and uniqueness theorems". Of course that Nicolai et al. are correct and Lee is wrong. (This is just one example where Lee views the established conclusions in string theory - such as the perturbative finiteness - as uncertain speculations while the statements in loop quantum gravity that are more or less safely known to be false - such as the off-shell closure of the constraint algebra - to be rigorous theorems.)
Nicolai et al. support their statement by a rather detailed analysis why the off-shell closure of the constraint algebra probably does not hold in the versions of loop quantum gravity that they have looked at. Lee only offers words; words that don't seem to be true.
Concerning Lee's comparison of the quantization approaches of string theory and of loop quantum gravity, he seems confused to me. The string-theoretical approach to quantization is not just string-theoretical. It's the correct, universal approach used throughout physics - in all field theories and in all descriptions of string theory that admit a classical limit. The approach of loop quantum gravity gives wrong results even for the harmonic oscillator. Lee says "give me a break, no one wants to quantize harmonic oscillator incorrectly". The reason why only a few want to quantize the harmonic oscillator incorrectly is that it would be very hard for them to convince anyone that it is the right thing to do. However, the loop quantum gravity community wants to apply the very same flawed techniques to quantize a much more complex theory - namely quantum gravity (or to quantize the worldsheet equally incorrectly).
Lee argues that the difference that can justify the application of the methods in gravity but not in the harmonic oscillator is probably that the harmonic oscillator is quadratic (Fock space?) while GR is not. Well, I assure Lee that one gets wrong results even for potentials that are as non-linear as general relativity. If a method fails in a very special and simple case, it is almost guaranteed to fail in a more general and complex case, too - because a more complex case requires one to be careful about many more subtleties (not less). The failure of the loop quantum gravity quantization of the harmonic oscillator shows that even the harmonic oscillator is too subtle for the techniques of loop quantum gravity to be any useful. And if the true essence of loop quantum gravity is in the techniques - which is what Lee always says - then the conclusion is that the failure of the techniques is enough to invalidate the theory.
Finally, I agree with Lee that it is much more likely that even a meaningful work about classical general relativity is written by a string theorist even though this community is fundamentally closer to particle physics, despite the overly self-confident claims of loop quantum gravity proponents that they have inherited the traditions of general relativity.