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.

April 22, 2008

Thoughts (Mostly on Super ∞-Things)

Posted by Urs Schreiber

I am on leave of absence from Hamburg and spending some time travelling before the program in Bonn starts next month. After a very productive week with Jens Fjelstad in Denmark this is now my last evening at Notre Dame, where I spent a very pleasant time with Stephan Stolz.

Time went by quickly, filled with discussions, mostly on nonabelian differential cohomology # and on superQFT, and many thoughts want to be further developed now.

I had planned to collect notes on some such thoughts last Sunday, but Jim Stasheff and Hisham Sati rightly pushed me to work on finalizing our article on Fivebrane structures (section 3), following up the one on L L_\infty-connections #.

One claim is that our L L_\infty-algebraic connection descent objects (section 7) may be integrated by hitting them with the functor DGCAformclassifyingspaceSmoothSpaceformpathωgroupoidSmoothωCatDGCA \stackrel{form classifying space}{\to} SmoothSpace \stackrel{form path \omega-groupoid}{\to} Smooth\omega-Cat to yield nonabelian differential cocycles #, a process reproducing the construction of cocycles by Brylinski-McLaughlin #.

(Fun exercise: read their article and identify how they are secretly integrating L L_\infty-algebras to ω\omega-groups and L L_\infty-connections to differential cocyles.)

But that must wait now until later. With a little luck Hisham will be around at UPenn this week, where I’ll go tomorrow to visit Jim, and we’ll see further.

But Lie nn-tegration has many aspects. Below some comments on super parallel nn-transport (see Florin Dumitrescu’s thesis for a nice discussion of the n=1n=1 case) and integration of super-L L_\infty-algebras (such as our favorite one (page 54)) to smooth super ω\omega-groups.

What’s a super L L_\infty-algebra? (see here for what an ordinary L L_\infty-algebra is.) Let SVectSVect be the category of 2\mathbb{Z}_2-graded vector spaces equipped with the unique non-trivial symmetric braiding. A super L L_\infty-algebra is an L L_\infty-algebra internal to that. So, dually, it’s a \mathbb{N}- 2\mathbb{Z}_2-bigraded quasi-free differential graded commutative algebra.

The archetypical example of such a beast is the super tangent Lie algebroid of any smooth superspace XX: the dual dgc algebra is just that of super-differential forms Ω (X)\Omega^\bullet(X) on XX. For me here, a smooth superspace is a sheaf on the site SS whose objects are superEuclidean spaces and whose morphisms are smooth maps between these Obj(S)=× Obj(S) = \mathbb{N}\times\mathbb{N} S(n|n,m|m)=SmoothSupermanifolds( n|n, m|m). S(n|n',m|m') = SmoothSupermanifolds(\mathbb{R}^{n|n'},\mathbb{R}^{m|m'}) \,.

Given any super L L_\infty-algebra gg we get the smooth superspace S(CE(g))S(CE(g)) classifying gg-valued super-forms using the usual adjunction coming from the ambimorphic object Ω \Omega^\bullet (you are hearing my teacher Todd Trimble speak through me, here) S(CE(G)): n|nHom superDGCA(CE(g),Ω ( n|n)). S(CE(G)) : \mathbb{R}^{n|n'} \mapsto Hom_{superDGCA}(CE(g),\Omega^\bullet(\mathbb{R}^{n|n'})) \,. This classifies (section 6.5) gg-valued forms in that for any smooth super XX SmoothSuperspaces(X,S(CE(g)))Ω (X,g). SmoothSuperspaces(X,S(CE(g))) \simeq \Omega^\bullet(X,g) \,.

A thin homotopy class of a superpath in that classifying space γ: 1|1S(CE(g)) \gamma : \mathbb{R}^{1|1} \to S(CE(g)) is an element in the super ω\omega-group integrating gg. Continuing this way, we form the super path ω\omega-groupoid

BG:=Π ω N(S(CE(g))) \mathbf{B}G := \Pi_\omega^N(S(CE(g)))

of the classifying space, for any NN \in \mathbb{N}, whose kk-morphisms are thin homotopy classes of (k|N)(k|N)-dimensional paths in the space Σ: k|NS(CE(g)) \Sigma : \mathbb{R}^{k|N} \to S(CE(g)) (section 2.3)

Voila, the ω\omega-supergroup GG integrating gg.

Now take any (flat for brevity) gg-valued superform on XX, represented by a superDGCA morphism Ω (Y)ACE(g) \Omega^\bullet(Y) \stackrel{A}{\leftarrow} CE(g) and hit that entire morphism with our integration functor Π ω NS:superDGCASuperSmoothωCat\Pi_\omega^N \circ S : superDGCA \to SuperSmooth\omega Cat to produce its corresponding super-parallel transport tra A:Π ω N(Y)BG. tra_A : \Pi_\omega^N(Y) \to \mathbf{B}G \,. And so on.

Here is a somewhat more philosophical remark. (Experience shows that it’s of the kind that I tend to find deeply puzzling while you are likely to find it at best pointless.)

I keep thinking about what superification really means. For instance, I keep noticing that superification shares many properties with categorification: categorification introduces \mathbb{N}-gradings, while superification introduces 2\mathbb{Z}_2-gradings.

I have that funny trait that I feel annoyed by the fact that n|n\mathbb{R}^{n|n'} is defined using graded algebras, somewhat messing with the Zen-like beauty of space and quantity, only for Ω ( n|n)\Omega^\bullet(\mathbb{R}^{n|n'}) later to introduce yet another layer of graded algebra.

So I observe this: little to nothing of the above changes if we do a slight repackaging of concepts:

instead of regarding the classifying space S(CE(g))S(CE(g)) of a super L L_\infty-algebra gg as a smooth superspace, I regard it as an ordinary smooth space probed by ordinary Euclidean spaces by setting S(CE(g)): nHom superDGCA(CE(g),Ω N ()), S(CE(g)) : \mathbb{R}^n \mapsto Hom_{superDGCA}(CE(g),\Omega_N^\bullet(\mathbb{R})) \,, where now Ω N :EuclideandomainssuperDGCAs \Omega^\bullet_N : Euclidean domains \to superDGCAs sends n\mathbb{R}^n to the superDGCA of superdifferential forms on n|N\mathbb{R}^{n|N}.

The point being that odd superfunctions are so close in concept to differential forms that we can just repackage them with the forms and assign the full package to any ordinary space.

These are lots of words for the simple fact that using the hom-adjunction we can always think Ω ( n|N) :=SuperSmoothSpaces( n|N,Ω ()) =SuperSmoothSpaces( n|0× 0|N,Ω ()) SuperSmoothSpaces( n,hom( 0|N,Ω ())) =:SuperSmoothSpaces( n,Ω N ()) =:Ω N ( n). \begin{aligned} \Omega^\bullet(\mathbb{R}^{n|N}) &:= SuperSmoothSpaces(\mathbb{R}^{n|N},\Omega^\bullet(-)) \\ &= SuperSmoothSpaces(\mathbb{R}^{n|0}\times \mathbb{R}^{0|N},\Omega^\bullet(-)) \\ & \simeq SuperSmoothSpaces(\mathbb{R}^n,hom(\mathbb{R}^{0|N},\Omega^\bullet(-))) \\ & =: SuperSmoothSpaces(\mathbb{R}^n,\Omega^\bullet_N(-)) \\ & =: \Omega^\bullet_N(\mathbb{R}^n) \end{aligned} \,. But still. It makes me wonder if we shouldn’t rethink.

Posted at April 22, 2008 2:19 AM UTC

TrackBack URL for this Entry:

1 Comment & 0 Trackbacks

Re: Thoughts (mostly on super infinity-things)

identify how they are secretly integrating L ∞-algebras to ω-groups and L ∞-connections to differential cocyles.)

Motivating toy examples and then the case of ordinary Chern-Simons cocycles realizing the first Pontryagin class, obtained from integrating the L L_\infty-connection that obstructs the lift of a GG-connection to a String(G)String(G) 2-connection, is now in section 5 of my talk notes:

On nonabelian differential cohomology (pdf)

Posted by: Urs Schreiber on April 23, 2008 12:57 AM | Permalink | Reply to this

Post a New Comment