Threads like this remind me that some CV’ers not only know a lot of science, they also know a lot about science.

If I weren’t spending my time commenting here, I might get around to writing up an easily digestible summary of some of the subsequent work.

Jacques is also well-aware that there’s good reason not to take such arguments seriously. When the first of these “swampland” (swampland = theories that may or may not be low energy limits of string theory) papers came out, he wrote a blog entry praising it and giving as the best example of this the prediction that you couldn’t get a single generation of fermions as a low energy limit of string theory. Unfortunately for him, someone wrote into his comment section to explain how to do this. Just about all claims that “you can’t get X out of string theory” have turned out to be wrong when someone seriously sat down and tried to find a way to do it.

So, I’ve looked very carefully for a plausible idea about how this is going to lead to a scientific prediction in the scientific papers in this area, and in Susskind’s book, and now neither Sean nor Jacques can come up with one either. There really isn’t such an idea, and having such an idea is what science is about. This stuff really is no better than astrology. Like astrology, it is driven not by rational effort to understand how the world works, but by people believing stuff because they just want to, not because there’s any evidence for it, or any hope of ever getting any.

I think it’s obviously possible that it could be done, if for example someone shows that the only compactifications with odd numbers of families have gauge-mediated susy breaking rather than gravitationally-mediated breaking.

I don’t think you’re going to see something quite of that ilk.

But, as several recent papers (and more in preparation) have shown, requiring a UV completion including gravity (by which the authors invariably mean “can be embedded in string theory”) imposes surprising contraints on effective field theories. “Surprising,” in the sense that these constraints are not explicable in terms of standard low-energy effective field theory considerations.

This discussion is going in circles, branching into other complicated topics, and continually evading the point I keep asking you about, in response to your disagreement with me about whether what Susskind is promoting is science or not.

By the way, I notice Jacques is evading the same point, while at the same time he seems to be claiming that Susskind’s book is no different than other popular books on science, a claim expressed with his trademark sneering at the supposed ignorance of anyone who disagrees with him (”Is Susskind’s the first book of this genre that you’ve read?”).

You say that Susskind is “recalibrating our notions of what constitutes a natural low-energy action”.
Sure, he says that the landscape means you can stop worrying about hierarchy and fine-tuning problems and use anthropic explanations. That’s a valid scientific point. What I’m objecting to is not that but his claim that what theorists should be doing is studying the details of the landscape, despite being unable to give a plausible account of how this is going to lead to a testable prediction of anything. Can you give me a plausible example of how studying the string theory landscape is going to lead to a testable scientific prediction? If you can’t, why do you think studying the string theory landscape is science?

As long a no theory of quantum gravity exists, judgement of what is the most promising approach is purely subjective. For decades, young people have been indoctrinated to believe that string theory, whatever it may be, is the most promising route to QG. In recent years this seems to have changed. Smart kids in high school and collage, who have no stakes in string theory, will undoubtedly see the anthropic principle for what it is: evidence that string theory has reached the end of the road.

If the smart kids a decade from now want to do LQG, because they see (or have been indoctrinated to see) background independence as the only important aspect of QG, then there will be ways for them to do LQG, and LQG will be considered as the most promising program by practioners of the field. Not that this is a development that I particularly welcome, but the trend in recent years has been very clear and it is presumably impossible to revert.

Most people shape their worldview when they are young and malleable. It seems to me like many of us who are skeptical about string theory entered physics in the early 1980s, between the analytic S-matrix and the first string revolution, when QFT ruled supreme. The next generation was brought up on string theory and finds it difficult to give it up, whereas those who are young today will undoubtedly have their views shaped by the string theory Landscape.

It was clear to me, e.g., in Hawking’s book, what was primarily speculation.

To you, perhaps. But not to the average reader.

(I’m sure you were equally capable of discerning what, in Susskind’s book, was speculation.)

Since we’re asking questions, let me ask you: Do you believe that Susskind is doing a service to science in relegating testability to a secondary or optional role?

I don’t think that’s what he’s doing.

Which is not to say that I agree with his assessment about what “the Landscape” may (or may not) mean for physics. But “relegating testability to a secondary or optional role” is a gross mischaracterization of his position.

It’s ironic that as important scientific distinctions are becoming more subtle the public’s ability to deal with subtlety is declining; social scientists seem to feel that TV people tend to be less able to think inferentially than are printed word consumers.