Even if you don’t believe Vafa’s hypotheses, it is still true that one could discover a low-energy field theory that causes huge difficulties for string theorists in trying to get anything resembling it.

You might even say, field theory contains more models (in fact a continuous infinity) than string theory – but people don’t complain that field theory is untestable.

Richter’s remarks were a disappointingly blunt instrument, it sounded from the beginning as if either he doesn’t understand the anthropic principle as Weinberg used it, or he doesn’t want to. And it also sounded as if he hadn’t noticed that the great majority of conference talks were experimental or phenomenological and had nothing to do with landscaping.

Now there are lots of objections to be made to statistical-type arguments – that we don’t understand the prior distribution, and that we can’t calculate reliably the probability that intelligent observers will emerge – but it seems to me a logically valid method.

If you knew the probability distribution of values of some observable X from your underlying theory, the probability of *observing* a certain value of X depends on the probability that observers will exist given that particular value.

P(Observe X) = P(Theory gives you X)*P(Observers exist given X)

There is a problem with the use of ‘atomic principle’ or ’stellar principle’ or ‘weak symmetry breaking principle’ as recently fashionable, since they don’t actually give you the probability P(Observers exist), and they disagree among themselves. Also to ask for ‘Life as we know it’ makes no sense. Observations could be made by anyone or anything, not specifically by carbon-based life forms with ten fingers… so I think some of the current work being done using such arguments is of poor quality, but that doesn’t mean that the approach is necessarily unscientific.

So what has this to do with ‘naturalness’? The usual naturalness or fine-tuning argument implicitly assumes some probability distribution over the continuous space of parameters of field theory and then asks what kind of field theory could give what we observe with non-tiny probability. It is not so clear how this is better than anthropic-type statistics since we still assume that Nature has some sort of roulette wheel to choose a distribution of parameters.

The alternative is just to get on and measure stuff and interpret it within an effective field theory and see what field theory models gets ruled out. That’s what a lot of people at SUSY were talking about and will be for many years to come. And I don’t know any string theorist who is *against* measuring things, or anyone who wants the budget for measuring things (which is much greater than the budget for theorists!) to be cut. The question is whether you go on to think about what a fundamental theory could be. On that you can be pessimistic or optimistic depending on the state of play in experiment and theory, but I think you ought to go on thinking about it until it is absolutely certain that no progress can be made.

T

Just subscribe to above podcast, & hook up your iPod

[Irrelevant long babble about tech stuff deleted, once again.

Chimpanzee, for the umpteenth time, please try and distinguish threads which are about physics from threads which are about tech stuff. I created a thread elsewhere ("SUSY06 Goes HiTech) where you can put discussions about tech stuff. This is your last warning, ok? Last warning. -cvj]

the way I understood the text, one performs the path integral over all histories compatible with what we observe now (basically fixing the end-state rather than an initial state).
In the usual ‘landscape’ approach one would estimate probabilities over all possible histories or vacua. At least this is how I understood this so until now After what you wrote I am not so sure I understood it correctly.
I assume Hawking talked about this in Beijing and perhaps some participant can tell us more about it.

At first glance at the news article to which you link, I don’t see what is different about what they are doing *in practice* from what the standard classical landscape statistics searches aim to do: figure out the features of the landscape to see if one can then apply conditional probabilities to assess the likelihood of some vacuum or class thereof. They’re instead wanting to do a path integral over all vacua. Nice words, but using what action principle? Also, without knowing both the action principle and the classical features of the landscape, I don’t understand how one can safely ignore any contributions to the path integral as insignificant. In other words, what is the measure in the sum over paths? How do you know that a type of configuration that makes only a tiny contribution does not at the same time have a huge number of representatives, thereby making a significant contribution in any case? So it seems that the existing program of classical explorations is essential making this “quantum” approach viable.

But I’m most bothered by the lack of a believable action principle. Doesn’t this just boil down to the usual search for a better understanding of string theory’s mechanism for choosing vacua? If you don’t have it, the proposal seems to be missing something crucial, no?

As described in the article, what they seem to want would be like saying that you can measure the anomalous magnetic moment of the electron, and from that number, deduce the Lagrangian of QED, and then use it to predict other physics like Bhabba scattering, etc. Nice…but it seems like an inverse problem that is…problematic.

Of course, I’ve not read the actual paper, so there could be a lot more going on than all that.

Cheers,

-cvj

Hawking seems to propose some sort of ‘inverse anthropic principle’ as announced e.g. in this press release
http://www.aip.org/pnu/2006/split/781-2.html?source=rsspnu

It seems very reasonable to me. What do you think about it?

So, no great annoucements, just trying to piece it together?

So then of course, how is such measures today, taking our views “so very close” to this beginning? What is “the trigger?” Does no one like the perspective in regards to the “arrow of time” as listing “strings” in the “microsecond realm,” as part of that inflationary universe?

Is what I am saying all about some Voodoo science?