23. Eye on ArXiv: 21 Apr 2008 – SUSY/non-SUSY duality
ASTRO, HEP-TH/PH — By Dmitry Podolsky on April 21, 2008 at 1:18 pm
Save This Post as PDF
And so the new week began… Would be fun to finish another paper till Friday (the subject is “IR divergencies in de Sitter space”). In the mean time, let us take a closer look on some papers released to arxives during the previous week.
Although the idea of SUSY/non-SUSY duality may sound a bit contradictory for an unprepared reader 
, there are situations where it seems to exist. The idea of the paper is the following.
Let us consider N=1 SYM with adjoint superfield \Phi. We will choose the superpotential to be
where 
is the field strength superfield.
(In string theory, such superpotentials appear in the situation when D5 branes are wrapped on vanishing cycles of local CY 3-folds and background B-field is added depending holomorphically on one complex coordinate of the 3-fold. This gives 
in the formula above.)
As it turns out, if the function 
is chosen appropriately, there exist vacua in the theory where SUSY is broken. The idea is that 
can be rendered negative for some factors of the gauge group by introducing high-dimensional operators in the Lagrangian. Negative means that gauge theory becomes unstable (w.r.t. Dyson instability) and wants to choose another vacuum. This vacuum turns out to be non-SUSY.
Duality between SUSY and non-SUSY works as follows. We start with effective IR (
) N=1 SYM. The theory depends on couplings t in the formulae above. For each t the authors find two sets of vacua – one is SUSY and one is not. However, only one beween the two is physical for any t (“physical” here means that VEV of chiral field is below 
), so as one changes t, the SUSY solution may leave allowed region of the field space, and SUSY gets (spontaneously) broken.
2. Ruth Gregory, “Braneworld black holes”
Lecture notes on braneworlds and black hole solutions in brane worlds (I strongly recommend them since Ruth is known expert in the field). Let me remind you what is the essence of the problem.
The brane world is 5 dimensional, and matter is localized on the brane. Since we are living on the brane, we expect gravitational potential between two bodies to fall as 
with distance. However, if we construct a 4 dim Schwarzchild like solution (with desirable 
behavior of the grav. potential at large distances), it will actually look like a string in 5 dim Universe. The latter is of course dynamically unstable and the characteristic time scale of the development of this instability is of course much shorter than the evaporation time for 4-dim black hole. Moreove, this instability appears on the purely classical level. If we instead want to construct 5-dim blackhole with spherical event horizon etc., we find that its gravitational potential falls as 
. Please see the lectures to learn what is the present point of view on this issue.
A very critical point of view on Boyle-Steinhardt-Turok PRL paper is introduced. Let me remind you that BST argued that in natural theories of inflation 
is bounded by 0.98 and 
for 
. The latter constraint is very important since one might than expect detection of gravitational waves from the B-mode polarization measurements of CMB (even in current experiments). The authors of the present paper argue that in natural models 
should not be restricted so much. The issue of course is how both groups define the word “natural” itself
The Boyle et al. define this word as follows:
1. The energy density or scalar perturbations generated during inflation have amplitude
on the scales that left the horizon 60 efolds before the end of inflation.
2. The universe undergoes at least 60 efolds of inflation.
3. After inflation, inflaton evolves smoothly to an analytic minimum with 
.
4. If the minimum is metastable, it must be long-lived and 
must be bounded from below.
5. The universe has to be reheated without spoiling large scale homogeneity and isotropy.
Hotchkiss et al. drop the last 4 requirements out. I think, dropping out (3) is reasonable (take old inflation for example, where 
evolves from small to large fields), (5) is not a really strong constraint (coupling of inflaton to matter fields is small in order not to spoil the flatness of the inflaton potential by radiative corrections) and (4) is not very much constraining as well.
4. Ted Jacobson, Aron C. Wall, ” Black Hole Thermodynamics and Lorentz Symmetry”
Lubos Motl was so much faster than me and already has a great discussion of this paper on his blog featuring one of the authors of the paper. The picture emerging is quite clear, and I am not sure I am able to add anything to the discussion. Maybe just this:
1) I think, it is indeed wrong to say that string theory predicts Lorentz violation at high energies. Lorentz invaiance is built into the theory; at high energies gravity is described in terms of dual theory – weakly coupled YM, which is of course Lorentz invariant. It is meaningless to talk in this regime about effective disersion relation for the graviton, because there is simply no such fundamental degree of freedom there (let us be consistent and, say, also discuss what happens with its width as quantum state in high energy regime, in other words, how well the graviton can be described as a particle).
Well, as Lubos Motl immediately pointed out, YM is dual to high curvature limit of the AdS string theory, not high energy limit, strictly speaking. So, that was bullshitting from my side 
2) Suppose that string theory has nothing to do with objective reality (doubt that) and Lorentz violation does take place beyond Planckian scale, as, say, in models of emergent gravity Grigory Volovik likes to talk about (so that effective dispersion relation is
). First trivial thing to notice – if the theory is effectively non-interacting there (i.e., we dealt with asymptotically free theory in the IR, and there is no UV-IR phase transition), then one does not have to care whether 2nd law of thermodynamics holds or not. The form of distribution of particles (species etc.) over energy is frozen, nothing thermalizes.
Second thing – in the opposite case, when interaction is important at higher energies (like in QED), troubles begin. 2nd law of thermodynamics is broken down -> theory there is strongly out of equilbrium (take a look on the name of this blog ), thermodynamic unstability develops. It badly messes with everything: even if we say that 2nd law of thermodynamics is effectively valid at low energies and IR degrees of freedom are thermalized quickly, IR dynamics should be influenced by strongly non-equilibrium UV dynamics – phase volume of UV degrees of freedom in Boltzmann integral is large, and they constantly source deviations from thermal equilibrium in the IR.
So, the bottom line, I guess, is that if Lorentz violation may be allowed by a strange flip of Nature, it will be allowed in asymptotically free theories. On the other hand, there are always phase transitions in asymptotically free theories, and IR degrees of freedom are not quite the same as UV ones
The authors calculate the Bayesian posterior probability ratios for three different models that could explain the cold spot: cosmic texture, anomalously large void and Sunyaev-Zeldovich effect caused by a cluster. The texture interpretation is favored, while void and cluster explanations are discarded – temperature decrement produced by them is negligible.
At some point we discussed here the possible explanation of the cold spot as effect caused by a void, so I will outline the statement of the paper regarding void. The radius distribution of voids close to us is known, so the authors take this distribution as prior. It turns out that best fit template corresponds to the radius of void 
MPc and 
. This gives a temperature anisotropy at the level 
, so that the template is completely negligible. Much larger voids do produce a larger temperature decrement but they are strongly penalized by the prior.
6. Diego Chialva, Ulf H. Danielsson, “Chain inflation revisited”
Somehow I forgot to review this paper, although did not to send a trackback to archives Fellow scandinavians somehow forgot to cite us even after I asked them directly by email. I am a friedly guy, though, so will write about their paper.
Diego and Ulf discuss chain inflation as follows from the title. In particular, they determine dynamics of background, construct perturbation theory and introduce a formula for the spectrum of adiabatic perturbations. They also discuss how to embed chain inflation on the landscape – one needs to consider a IIB compactification with fluxes (landscape corresponds to different CY compactifications).
—————————————————–
Also, two new Witten’s papers were released today, and I will have to take a closer look on them to write about them intelligently.
2 Comments
Reply from Lubos:
Dear Dmitry, thanks for your refreshing points but anyway, here’s my comments:
1) YM is the dual of some superselection sectors of string theory only, not a universal “high energy” dual. Moreover, I don’t quite see why you say that YM is “high energy” dual. Weakly coupled YM is the “high curvature” dual of AdS backgrounds of string theory. Also, the Lorentz symmetry of YM is not the same thing as the Lorentz symmetry in the gravity dual. In this particular case, the YM Lorentz symmetry is a subgroup: in this case, the whole isometry of the AdS spacetime – namely the generalized/curved (super)Poincare symmetry – is the (super)conformal symmetry of the YM theory.
2) If you turn off the interactions, indeed, you may make the perpetuum mobiles impossible and the paradox goes away. But turned-off interactions are forbidden in quantum gravity themselves, see e.g. the weak gravity conjecture http://arxiv.org/abs/hep-th/0601001 saying that all forces must be stronger in some proper counting – otherwise you would encounter other paradoxes analogous to this perpetuum mobile paradox here.
If I understand you, I agree that the asymptotically free example could lead to unexpected (and generic) consequences in the IR if you do something dangerous in the UV and even the relevant degrees of freedom change. More concretely, you almost certainly get a high Lorentz violation in the IR if you introduce it in the UV, right?
Best
Lubos
Lubo? Motl |
And my reply to him:
Hi Lubos
1) Thanks for clarification! I was just thinking in too broad Polyakov-like pre-AdS/CFT sense: any gauge theory is dual to some string theory, and the latter necessarily contains gravitation in its low energy effective action. Such a generic statement is most probably wrong.
2) As for quantum gravity, also agree, you are not allowed to turn all the interactions off, of course, if gravity itself survives at high energies. As I remember, Lorentz-violating people like to say that it does not have to (only spin degrees of freedom survive, in the spirit of LQG you love so much
). On the other hand, nobody among them was ever able to introduce a working model providing Einstein-Hilbert in the IR.
Qualitatively, I think, IR influence depends on how rapidly coupling decreases with decreasing energy (Boltzmann integral is proportional to the coupling squared), and in order for UV physics to decouple, this decrease should be very rapid (thermodynamic instability develops exponentially).
Cheers,
Dmitry.
Trackback responses to this post