Unifying derived deformation theories (Q973926)

At the time, there are two main approaches to derived moduli theory. There is the local approach using differential graded Lie algebras (DGLAs) and strong homotopy Lie algebras (SHLAs). These yield derived deformation functors for a very wide range of problems, but are restricted to characteristic zero, with their construction seldom extending to global problems. The derived moduli stacks give a global formulation valid in all characteristics, but are only constructed for a comparatively narrow class of examples. The author connects these two approaches: Take a geometric \(D^-\)-stack \(F\) over \(k\) with a \(k\)-valued point \(x\), let \(\Omega_x F\) be the loop space of \(F\) at \(x\), \(L_x(F)\) its tangent space at \(x\), and let \(F_x\) be the formal neighborhood functor. The author then proves that \(L_x(F)\) has the natural structure of an \(L_\infty\)-algebra, and the functors \(F_x\) and \(\Sigma(\mathcal L_q C_x(F))\) (Hinich's simplicial nerve) are weakly equivalent, where \(C_x(F)\) is the dg coalgebra generated by \(L_x(F)\). It is proved that for all classical deformation problems there can be assigned functors defined on Artinian simplicial rings. The classical deformation groupoid will then be the fundamental groupoid of these functors, restricted to rings. The author gives the proper definitions and explains the properties of these functors in general. The crucial new ingredient is a property of functors \(F\) called quasi-smoothness meaning that \(F\) maps small extensions to fibrations, and acyclic small extensions to trivial fibrations. For such functors the author defines cohomology groups \(H^i(F)\), for \(i\in\mathbb Z\), and there are long exact sequences in which these groups simultaneously plays the role of tangent and obstruction spaces. This gives the \(\infty\)-theory without the inductive construction. The author shows how to put a model structure on the category of all left-exact functors from Artinian simplicial rings to simplicial sets. In this model structure, the fibrations are precisely the quasi-smooth maps, so each equivalence class has a quasi-smooth representative. There are analogues of Eilenberg-Maclane spaces for representing cohomology groups, and every weak equivalence class has a unique minimal model. The homotopy category satisfies a Brown-type representability property analogous to Schlessinger's theorem. The author gives a summary of existing approaches to derived deformations: Manetti's extended functors, Hinich's formal stacks, and the derived stacks of Toën-Vezzosi and Lurie. He then proves that his geometric stacks may be regarded as germs of geometric \(D^-\)-stacks. The model structure defined by the author is compared with established homotopy categories used to study derived deformations in characteristic zero. It is shown to be equivalent to the category of deformation functors defined by Manetti, and then this category, in turn, is equivalent to Kontsevich's category of SHLAs modulo tangent quasi-isomorphisms, and to the homotopy categories of DG coealgebras and DGLAs considered by Hinich. In fact, this shows that all existing approaches to derived deformations are equivalent. The author establishes an Adams-type spectral sequence, enabling the definition of a graded Lie algebra structure on the cohomology groups \(H^\ast(F)\) of any deformation functor. These are all the operations in characteristic \(0\), together with many more in general, and the model structure is applied to outline the operations common to all deformation cohomologies. The article is not self-contained, and builds on a lot of advanced results. With the necessary background given, however, it is properly written and is a very important contribution to the field of derived deformation theory, unifying this theory to a great extent. It also makes the focus on the possibility of unifying deformation theories in general, with its focus on simplicial categories as the basis for all such.