Skip to the Main Content

Note:These pages make extensive use of the latest XHTML and CSS Standards. They ought to look great in any standards-compliant modern browser. Unfortunately, they will probably look horrible in older browsers, like Netscape 4.x and IE 4.x. Moreover, many posts use MathML, which is, currently only supported in Mozilla. My best suggestion (and you will thank me when surfing an ever-increasing number of sites on the web which have been crafted to use the new standards) is to upgrade to the latest version of your browser. If that's not possible, consider moving to the Standards-compliant and open-source Mozilla browser.

June 10, 2008

Bagger-Lambert Again

I haven’t attempted to post more about Bagger-Lambert Theory, since my earlier post. Every time I think it might be worthwhile to pause and take stock of developments, two or three new papers on the subject appear on the arXivs, and I drop that silly idea.

Still, one thread which got a fair amount of attention was the proposal by three different groups of a whole new class of “Bagger-Lambert” algebras, obtained by relaxing the condition that the bilinear form (the “trace”) on the algebra be positive-definite.

With an indefinite bilinear form, it sure looks like the theory has ghosts. Which, to put it mildly, would not be good. A way around this difficulty was proposed by two other groups: by gauging a certain shift symmetry, one can remove the negative-norm states.

Well, along come Ezhuthachan, Mukhi and Papageorgakis, who point out that the resulting theory is on-shell equivalent to the standard D=3D=3, 𝒩=8\mathcal{N}=8 SYM — that is (for one of the classical gauge groups) to the theory on the world volume of a stack of D2 branes. In this dictionary, there is a scalar field, whose VEV is the Yang-Mills gauge coupling. For any finite value of the VEV of that scalar, the would-be SO(8)SO(8) R-symmetry is broken to SO(7)SO(7) (as is the superconformal symmetry).

The computation involves a nice application of a nonabelian duality transformation, due to de Wit, Nicolai and Samtleben. Consider

(1)d 3x12Tr(ϵ μνλB μF νλ(D μϕg YMB μ) 2)\int d^3 x \tfrac{1}{2} Tr\left(\epsilon^{\mu\nu\lambda} B_\mu F_{\nu\lambda} - {(D_\mu\phi - g_{\text{YM}}B_\mu)}^2\right)

In addition to the usual Yang-Mills gauge transformations, A μgA μg 1+ μgg 1 ϕgϕg 1,B μgB μg 1 \begin{gathered} A_\mu \to g A_\mu g^{-1} + \partial_\mu g g^{-1}\\ \phi \to g \phi g^{-1},\quad B_\mu\to g B_\mu g^{-1} \end{gathered} (1) is invariant under a local shift symmetry

(2)ϕϕ+g YMχ,B μB μ+D μχ\phi \to \phi + g_{\text{YM}}\chi,\quad B_\mu\to B_\mu + D_\mu \chi

Using (2) to gauge away ϕ\phi, B μB_\mu becomes an ordinary auxiliary field, and (1) is on-shell equivalent to d 3x14g YM 2Tr(F μν 2) \int d^3 x\, -\tfrac{1}{4 g_{\text{YM}}^2} Tr(F_{\mu\nu}^2)

Now, the observation of Ezhuthachan et al is that ϕ\phi looks like an eighth adjoint-valued scalar, to complement the seven already present in 𝒩=8\mathcal{N}=8 SYM. Its coupling to B μB_\mu, however, breaks that symmetry. To restore the symmetry, treat g YMg_{\text{YM}} as the VEV of another (SO(8)SO(8) vector-valued) scalar, Y=(0,0,0,0,0,0,0,g YM)\langle Y\rangle = (0, 0, 0, 0, 0, 0, 0, g_{\text{YM}}). To ensure that Y IY^I is spatially-constant, they introduce a vector field, C μ IC^I_\mu (and another scalar, Z IZ^I), with action d 3x(C μ I μZ I) μY I \int d^3 x\, (C^I_\mu - \partial_\mu Z^I) \partial^\mu Y^I and local shift symmetry Z IZ I+η I,C μ IC μ I+ μη I Z^I \to Z^I + \eta^I,\quad C^I_\mu \to C^I_\mu + \partial_\mu \eta^I

Putting all the pieces together, they show that this construction yields the gauged version of the “new” Bagger-Lambert actions. Thus, with some particular choice of VEV for Y IY^I, the latter is just on-shell equivalent to standard 𝒩=8\mathcal{N}=8 SYM.

That was fun while it lasted …

Update (6/11/2008):

As Chethan points out in the comments, it’s never a good time to try to write about this stuff.

In Monday’s listings, a paper by Aharony et al appeared. They couple the standard 𝒩=3\mathcal{N}=3 supersymmetric Chern-Simons theory1 to matter. For a particular choice of gauge group — U(N)×U(N)U(N)\times U(N) or SU(N)×SU(N)SU(N)\times SU(N), with Chern-Simons levels (k,k)(k,-k) — and matter representations — chiral multiplets A i(N,N¯)A_i\in(N,\overline{N}) and B i(N¯,N)B_i\in (\overline{N},N), i=1,2i=1,2 — they show that the resulting theory has an enhanced 𝒩=6\mathcal{N}=6 supersymmetry. The SO(3) RSO(3)_R symmetry is enhanced to and SO(6) RSO(6)_R, under which the scalars (A 1,A 2,B¯ 1,B¯ 2)(A_1,A_2,\overline{B}_1,\overline{B}_2) transform as a 4\mathbf{4}. In the particular case when the gauge group is SU(2)×SU(2)SU(2)\times SU(2), the (2,2)(2,2) is a real representation, and the SO(6) RSO(6)_R is enhanced to an SO(8) RSO(8)_R.

For higher NN, they argue that the theory (with 𝒩=6\mathcal{N}=6 superconformal invariance) is the correct description of NN M2-branes transverse to a 4/ k\mathbb{C}^4/\mathbb{Z}_k orbifold. There is much here that bears further discussion. Perhaps fodder for another post …

1 The 𝒩=2\mathcal{N}=2 supersymmetric Chern-Simons action is

(3)S CS 𝒩=2 =k2πd 3xd 4θ 0 1dτTr(VD¯ α(e τVD αe τV)) =k4πAdA+23A 3χ¯χ+2σD\begin{aligned} S_{\text{CS}}^{\mathcal{N}=2} &= \tfrac{k}{2\pi}\int d^3x \int d^4\theta \int_0^1 d\tau Tr\left(V \overline{D}^\alpha \left(e^{-\tau V} D_\alpha e^{\tau V}\right)\right)\\ &= \tfrac{k}{4\pi} \int A d A +\tfrac{2}{3} A^3 - \overline{\chi}\chi + 2 \sigma D \end{aligned}

in Wess-Zumino gauge. Here DD and σ\sigma are scalar fields, and χ\chi is a Dirac fermion, all in the adjoint. The 𝒩=3\mathcal{N}=3 Chern-Simons action contains an additional chiral multiplet, Φ\Phi, in the adjoint

(4)S CS 𝒩=3=S CS 𝒩=2k4πd 3x(d 2θTr(Φ 2)+c.c.)S_{\text{CS}}^{\mathcal{N}=3} = S_{\text{CS}}^{\mathcal{N}=2} - \tfrac{k}{4\pi}\int d^3 x \left(\int d^2\theta Tr(\Phi^2) + \text{c.c.}\right)

We can couple matter chiral multiplet(s), in representation RR, to (3) S Q =d 3xd 4θQ¯e VQ =d 3xD μQ¯D μQ+iψ¯γ μD μψ +Q¯(Dσ 2)Qψ¯σψ+iQ¯χ¯ψiψ¯χQ \begin{aligned} S_{Q} &=\int d^3x \int d^4\theta \overline{Q}e^V Q\\ &= \int d^3x \overline{D_\mu Q} D^\mu Q + i \overline{\psi} \gamma^\mu D_\mu \psi\\ & +\overline{Q}(D-\sigma^2)Q - \overline{\psi}\sigma\psi + i\overline{Q}\overline{\chi}\psi - i\overline{\psi}\chi Q \end{aligned} To get 𝒩=3\mathcal{N}=3 supersymmetry, the matter (Q,Q˜)RR¯(Q,\tilde{Q})\in R\oplus \overline{R}, and the action S matter 𝒩=3=S Q+S Q˜+d 3x(d 2θQ˜ΦQ+c.c.) S_{\text{matter}}^{\mathcal{N}=3} = S_{Q} + S_{\tilde{Q}} + \int d^3 x \left(\int d^2\theta \tilde{Q} \Phi Q + \text{c.c.}\right)

Posted by distler at June 10, 2008 11:30 PM

TrackBack URL for this Entry:

6 Comments & 1 Trackback

Re: Bagger-Lambert Again

That was fun while it lasted…

So what is the lesson to be learned here? Just a weird reformulation of something known, or a reformulation of intrinsic value?

Posted by: Urs Schreiber on June 11, 2008 3:33 AM | Permalink | Reply to this

Re: Bagger-Lambert Again

So what is the lesson to be learned here?

I would say: “Naïve attempts to generalize the Bagger-Lambert construction fail.”

  • Relaxing the positivity constraint on the bilinear form allows for new solutions (one for any compact Lie algebra, 𝔤\mathfrak{g}), but these new theories appear not to be unitary.
  • A suitable gauging of the result leads to a unitary theory: the only unitary theory you know about with this amount of supersymmetry, the 𝒩=8\mathcal{N}=8 SYM with gauge group GG.

More broadly, I would say the lesson is: “Physics is hard.” But it’s not clear how one ought to modify one’s behaviour to take that lesson into account.

Posted by: Jacques Distler on June 11, 2008 8:20 AM | Permalink | PGP Sig | Reply to this

Re: Bagger-Lambert Again

> That was fun while it lasted …

Dear Jacques, that sounds depressing, but I am not sure the story is that bleak. I am sure you have seen the paper by Aharony et al. 0806.1218, generalizing the original Bagger-Lambert without getting caught up in the indefinite trace form. Unfortunately, their theory has only 6 SUSYs manifest in the general case. But when the gauge group is SU(2) * SU(2) there is an enhancement in SUSY and you end up with van Raamsdonk’s rewriting of the original Bagger-Lambert. So the 3-algebra structure in this picture is almost incidental.

Perhaps the message is only that 3-algebras might not be the right way to go about constructing/generalizing M2-brane theories.

Warm regards,

Posted by: chethan krishnan on June 11, 2008 10:56 AM | Permalink | Reply to this

Re: Bagger-Lambert Again

the message is only that 3-algebras might not be the right way

Jim Stasheff emphasized that what lately is being called “3-algebra” is precisely a Nambu bracket.

Nambu brackets have been considered in the study of membranes before, but in a context that is at least superficially a bit different. Do any of the recent 3-algebra articles comment on that?

Posted by: Urs Schreiber on June 11, 2008 11:21 AM | Permalink | Reply to this

Re: Bagger-Lambert Again

Matsuo and co. have papers on explicit constructions of 3-algebras using the Nambu-Posisson bracket of an algebra of functions on a manifold. But these things are infinite dimensional and were used by them for generating M5s from the M2s (among other things). In a “normal” Bagger-Lambert theory we are looking for something much more like the gauge structure constants of a Yang-Mills theory, so that we can, for example, relate it to D2s by giving VEVs to some transverse scalars.

I am totally clueless about the pre-Bagger-Lambert work on the relation between M2-branes and 3-algebras, but you should find references in Matsuo et al’s papers. Especially if this older work that you talk about was based on attempts to generalize D-brane noncommutativity to M-branes by messing with function spaces over manifolds or something like that.

Posted by: chethan krishnan on June 11, 2008 12:11 PM | Permalink | Reply to this

Re: Bagger-Lambert Again

Thanks, that’s useful information.

Concerning that other occurence of Nambu brackets:

I think Nambu brackets were at some time hoped to yield a “covariant” matrix model description of the membrane, in that it would allow to write the action (as opposed to just the Hamiltonian) as a square of brackets, but now of trinary brackets.

So it’s a rather different context in which the trinary thing appears, and I don’t know if it every worked out. But I wouldn’t be surprised if there is a relation.

Posted by: Urs Schreiber on June 11, 2008 12:55 PM | Permalink | Reply to this
Read the post Lie 3-Algebras on the Membrane (?)
Weblog: The n-Category Café
Excerpt: Recently a trinary bracket appears in the study of the supermembrane which is sometimes addressed as a homotopy algebraic structure.
Tracked: November 6, 2008 7:22 PM

Post a New Comment