Michael Dine: Is There a String Theory Landscape

[marcus]

Michael Dine just posted this
"Is There a String Theory Landscape: Some Cautionary Notes"
http://arxiv.org/hep-th/0402101

as a sample, here are the concluding remarks:

"Mention of the anthropic principle brings out strong reactions from most physicists, who ask what can be the role of science in such a situation. But the lesson of the KKLT proposal is not so pessimistic. First, the existence of a landscape within string theory is a question we should be able to decide. If we decide that there is such a discretuum, we will probably be forced to contemplate the anthropic principle; if not, we can dismiss it.

But even if we do adopt the anthropic principle, it will at best explain only a few quantities: either we will falsify string theory, or we will uncover principles which explain most of the features of the Standard Model. We will likely make additional predictions for accelerators and cosmology as well. So surrender to the anthropic principle will not be necessary or possible; we won’t have to give up."

Probably the paper should be seen in the context of papers that came out last year by Leonard Susskind, Michael Douglas, and Tom Banks.

L. Susskind, "The Anthropic Landscape of String Theory"
http://arxiv.org/hep-th/0302219
T. Banks, M. Dine, E. Gorbatov, "Is There a String Theory Landscape?"
http://arxiv.org/hep-th/0309170
M. R. Douglas, "The statistics of string/M theory vacua"
http://arxiv.org/hep-th/0303194

For additional context, shedding light on Leonard Susskind's contribution, is the December 15 2003 Edge interview involving Leonard Susskind, Paul Steinhardt, and Lee Smolin.
http://www.edge.org/discourse/landscape.html
The links are provided in case anyone wishes to comment on this nexus of string issues

 

[notevenwrong]

I had trouble figuring out what Dine meant
with those last few sentences.

"Either we will falsify string theory"
How does he envision string theory being
falsified?

"We will likely make additional predictions
for accelerators and cosmology as well"
How does he expect to make these predictions
and what evidence is there he is "likely" to
be able to make them?

This whole literature about the "landscape"
is very weird. Hardly any equations, just
rambling sci-fi kinds of speculations, a huge
amount of really fuzzy thinking. Seems to
me all just a way of avoiding the obvious
point that string theory is now dead as a
unified theory. If the whole picture of the
perturbative string makes sense, it is
completely vacuous and can never
predict anything.

 

[Haelfix]

Predictivity is a very hard thing these days in theoretical physics.

Its not just String Theory that suffers from this malaise, SuSY, inflation, even simple extensions of standard model all are afflicted.

Often theorists output a theory, and for one reason or the other, have a bunch of free parameters that are involved. You can then either

a) Look for a new theory that's more general and predictive (hard)
b) Wait for experiment to catch up and give the answer. This is nice, b/c it can often lead to a legit prediction for another observable, than can then be tested for
c) Invoke the anthropic principle, and give up, waiting for experiments that are probably not even in principle doable.

Then there's a whole new problem of uniqueness of solution, and of course the celebrated finetuning problem.

In my opinion, a lot of this has to do with theoretical physicists being hamstrung by group theory. Its often a cornerstone that such mathematical methods tend to lead to those dilemas. Don't get me wrong, its a tried and true workhorse for physics, but it only can get you so far. Old fashioned guess work and a great intuitive mind is needed for the last critical steps..

String theory is not dead. Far from it! But then again, its not really a theory *yet*, more like a program to understand certain aspects of quantum gravity, that has a few proofs of concept and some tantalizing glimpses of yet to be discovered physics. Theres much work that still needs to be done however. I very much doubt that String Theory in 50 years is going to look much like what we deal with today. Of course, there's a few particuraly arrogant STheorists who will insist that we must accept ST as it is currently formulated, that anything that kinda goes contrary to the currently formulated tenets of ST is bunk.

But in a sense you are right. If no experiment can ever falsify st, and it never outputs a legit testable prediction. Then like it or not, it will die and be remembered only as a supreme intellectual exercise. Techniques that were learned will be used in new theories that will hopefully get us further.

 

[notevenwrong]

Didn't really intend to start arguing about
string theory. I'm still trying to figure out
what Dine actually thinks he is saying. I
often disagree with what string theorists say,
but at least I can figure out what they are
saying and what their argument is. This
paper seems to take things to a whole new level.

 

[selfAdjoint]

I spent much of the afternoon reading the paper. Dine's main point is that the Anthropic Principle, suitably constrained, is not a woozy or philosophical notion but is capable in specific instances of forming predictions and being falsified.

He then shows how this would work out in the course of analyzing the new attempts to derive low energy physics from stringy hypotheses. None of the ideas he treats are rock solid, but he says that IF they are assumed, THEN anthtopic reasoning about them can come up with testable hypotheses.

 

[Mike2]

Originally posted by selfAdjoint
I spent much of the afternoon reading the paper. Dine's main point is that the Anthropic Principle, suitably constrained, is not a woozy or philosophical notion but is capable in specific instances of forming predictions and being falsified.

He then shows how this would work out in the course of analyzing the new attempts to derive low energy physics from stringy hypotheses. None of the ideas he treats are rock solid, but he says that IF they are assumed, THEN anthtopic reasoning about them can come up with testable hypotheses.

Isn't the effort to develop a theory that conforms to observation an anthropic principle in and of itself where the laws of nature are such that they much produce what we observe (and by implication, produce us in the process)? There are other modes of development such as a theory developed from the requirement of logical consistency alone (my personal favorite).

 

[selfAdjoint]

I confess I can't find much difference between the WEAK anthropic principle, and any other falsification pronciple such as experiment or astrophysical observation. The idea that a theory must fail if it requires nature to behave in a way nature obviously doesn't is surely not constroversial.

But Dine describes using the w.a.p. to constrain a google or so of inequivalent theories. Insofar as this amounts to throwing out theories that clearly violate nature, we're still with falsification and thus OK. But if we start fiddling with the w.a.p. in order to further discriminate between candidate theories, I see a slippery slope.

Because the STRONG anthropic principle states that the world is the way it is just because we are in it - that the causality runs from human existence to cosmological fine-tuning. And that is superstition, or so I believe.

 

[Haelfix]

With the anthropic principle and say the cosmological principles (we are in no favored place in the universe) you can run sensemble statistics to see if indeed we live in a special place. Sean Carrol does this in some of his recent papers.

But yea, its metaphysics and frankly uninteresting physics

 

[Mike2]

Originally posted by selfAdjoint

Because the STRONG anthropic principle states that the world is the way it is just because we are in it - that the causality runs from human existence to cosmological fine-tuning. And that is superstition, or so I believe.

The anthropic principle is just a valid as absolute determinism. If everything is consistent with everything else, it doesn't matter where you start your consistency requirements. But the assumption in the process is that everything is indeed consistent with everything else, and that logical consistency requires the universe to be no other way. So the anthropic principle is equal to determinism. But it still doesn't answer why some variables turn out the way they do.