Quillen's theorem on buildings and the loops on a symmetric space (Q917074): Difference between revisions

In 1975 Garland and Raghunathan and, independently, Quillen [unpublished] proved that the space \(\Omega\) G of based loops in a compact Lie group G is homotopy equivalent to an infinite dimensional flag variety and that Bott's cell decomposition of \(\Omega\) G can be obtained as a Bruhat or Schubert cell decomposition. The proof has two main ingredients. The first one is to exhibit a Tits system for the group \(L_{alg}G_{{\mathbb{C}}}\) of ``algebraic loops'', i.e. regular maps of \({\mathbb{C}}^*\) to the complexified Lie group \(G_{{\mathbb{C}}}\). Then \(L_{alg}G_{{\mathbb{C}}}\) can be regarded as the group of points \(\tilde G_{{\mathbb{C}}}\) of the algebraic group \(G_{{\mathbb{C}}}\) over \({\mathbb{C}}[z,z^{-1}]\). The group P of regular maps \({\mathbb{C}}\to G_{{\mathbb{C}}}\) is a maximal parabolic subgroup and \(\Omega_{alg}G=\{f\in L_{alg}G_{{\mathbb{C}}}\); \(f(S^ 1)\subseteq G\) and \(f(1)=1\}\) can be identified with \(\tilde G_{{\mathbb{C}}}/P\). The axioms of a Tits system then yield a Bruhat or Schubert decomposition of \(\tilde G_{{\mathbb{C}}}/P=\Omega_{alg}G\). The second ingredient of the proof is then the result that \(\Omega_{alg}G\to \Omega G\) is a homotopy equivalence. The proof works more generally for a compact symmetric space M instead of G. The author rederives the Bott-Samelson theorems on \(\Omega\) M and the real and complex Bott periodicity theorems. He also develops a theory of topological Tits systems and their associated buildings.