Courant algebroids and strongly homotopy Lie algebras (Q1272373)

Given a manifold \(M\), consider \(TM\) with its standard Lie algebroid structure and \(T^*M\) with zero anchor and bracket. Then \((TM,T^*M)\) is a Lie bialgebroid with the double bracket \([\cdot,\cdot]\) originally introduced by \textit{T. Courant} [Trans. Am. Math. Soc. 319, 631-661 (1990)], to investigate Dirac structures on \(M\) being subbundles \(L \subset TM\oplus T^*M\) that are maximally isotropic under the canonical symmetric bilinear form \(\langle\cdot, \cdot\rangle\) and whose sections are closed under \([\cdot,\cdot]\). We add that general Lie bialgebroids were introduced by \textit{K. Mackenzie} and \textit{P. Xu} [Duke Math. J. 73, 415-452 (1994; Zbl 0844.22005)] as linearization of Poisson groupoids. Next, \textit{Z. Liu}, \textit{A. Weinstein} and \textit{P. Xu} [J. Differ. Geom. 45, 547-574 (1997; Zbl 0885.58030)] have introduced a general notion of a Courant algebroid generalizing the double construction to Lie algebroid: if \((A,A^*)\) is a Lie bialgebroid then \(E=A\oplus A^*\) possesses a structure of a Courant algebroid. A Courant Lie algebroid is a vector bundle \(E\to M\) equipped with a nondegenerate symmetric form \(\langle\cdot, \cdot\rangle\) on the bundle, a skew-symmetric bracket \([\cdot,\cdot]\) on \(\Gamma(E)\), and an anchor \(\rho:E\to TM\) which are all fulfilling a number of properties. The Courant bracket is natural but it has anomalous properties (it does not satisfy the Leibniz rule or the Jacobi identity). The purpose of this letter is to explain the anomalies, which all can be described via differentials of certain expressions. From this, the author noticed that the general Courant algebroids can be viewed as strongly homotopy Lie algebras (known also as SHLA's or \(L_\infty\)-algebras) [\textit{V. Hinich} and \textit{V. Schechtman}, Adv. Sov. Math. 16 (2), 1-28 (1993; Zbl 0823.18004), \textit{T. Lada} and \textit{M. Markl}, Commun. Algebra 23, 2147-2161 (1995)]. A strongly homotopy Lie algebra is a graded vector space \(V\) together with a collection of linear maps \(l_k: \wedge^k V\to V\) of degree \(k-2\), \(k\geq 1\), satisfying the relation \(\sum_{i+j=n+1} (-1)^{i(j-1)} l_jl_i=0\) where \(l_i\) denotes here the extension of \(l_i:\wedge^iV\to V\) to \(\wedge V\) as a coderivation of the coalgebra structure. The interest in \(L_\infty\)-algebras by the authors is for the following reason: the Courant bracket on a Courant Lie algebroid \(E\to M\) can be extended to an \(L_\infty\)-structure on the total space of the following finite homological resolution of the Lie algebra \(H =\text{coker} {\mathcal D}\) \[ 0\to \ker D\hookrightarrow C^\infty (M)@>{\mathcal D}>> \Gamma(E)\to H\to 0 \] where \({\mathcal D}: C^\infty(M) \to\Gamma(E)\) is the map defined by \({\mathcal D}= {1\over 2}\beta^{-1} \rho^* d_0\) \((d_0\) the de Rham differential, \(\beta\) the isomorphism between \(E\) and \(E^*\) given by \(\langle \cdot,\cdot \rangle)\). We add that from the definition of the Courant bracket on \(\Gamma(E)\;\text{Im} {\mathcal D}\) is an ideal in \(\Gamma(E)\) which implies that \(H\) is a Lie algebra.