arXiv subjects
Publications and source records attributed to Max Karoubi.
The purpose of this short paper is to investigate relations between various real K-theories. In particular, we show how a real projective bundle theorem implies an unexpected relation between Atiyah's KR-theory and the usual equivariant K-theory of real vector bundles. This relation has been used recently in a new computation of the Witt group of real curves by Schlichting, Weibel and the author. We also interpret Atiyah's theory as a special case of twisted K-theory.
We establish some structural results for the Witt and Grothendieck-Witt groups of schemes over $\mathbb{Z}[1/2]$, including homotopy invariance for Witt groups and a formula for the Witt and Grothendieck-Witt groups of punctured affine spaces over a scheme. All these results hold for singular schemes and at the level of spectra.
Let $V$ be an algebraic variety defined over $\mathbb R$, and $V_{top}$ the space of its complex points. We compare the algebraic Witt group $W(V)$ of symmetric bilinear forms on vector bundles over $V$, with the topological Witt group $WR(V_{top})$ of symmetric forms on Real vector bundles over $V_{top}$ in the sense of Atiyah, especially when $V$ is 2-dimensional. To do so, we develop topological tools to calculate $WR(V_{top})$, and to measure the difference between $W(V)$ and $WR(V_{top})$.
We introduce a version of the Brauer--Wall group for Real vector bundles of algebras (in the sense of Atiyah), and compare it to the topological analogue of the Witt group. For varieties over the reals, these invariants capture the topological parts of the Brauer--Wall and Witt groups.
We introduce the notion of the "covering type" of a space, which is more subtle that the notion of Lusternik Schnirelman category. It measures the complexity of a space which arises from coverings by contractible subspaces whose non-empty intersections are also contractible.
Let $V$ be an algebraic variety over $\mathbb R$. The purpose of this paper is to compare its algebraic Witt group $W(V)$ with a new topological invariant $WR(V_{\mathbb C})$, based on symmetric forms on Real vector bundles (in the sense of Atiyah) on the space of complex points of $V$, This invariant lies between $W(V)$ and the group $KO(V_{\mathbb R})$ of $\mathbb R$-linear topological vector bundles on $V_{\mathbb R}$, the set of real points of $V$. We show that the comparison maps $W(V)\to WR(V_{\mathbb C})$ and $WR(V_{\mathbb C})\to KO(V_{\mathbb R})$ that we define are isomorphisms modulo bounded 2-primary torsion. We give precise bounds for the exponent of the kernel and cokernel of these maps, depending upon the dimension of $V.$ These results improve theorems of Knebusch, Brumfiel and Mahé. Along the way, we prove a comparison theorem between algebraic and topological Hermitian $K$-theory, and homotopy fixed point theorems for the latter. We also give a new proof (and a generalization) of a theorem of Brumfiel.
We introduce a general framework to unify several variants of twisted topological $K$-theory. We focus on the role of finite dimensional real simple algebras with a product-preserving involution, showing that Grothendieck-Witt groups provide interesting examples of twisted $K$-theory. These groups are linked with the classification of algebraic vector bundles on real algebraic varieties.
The main purpose of the present article is to establish the real case of "Karoubi's conjecture" in algebraic K-theory. The complex case was proved in 1990-91 by the second author and Andrei Suslin. Compared to the case of complex algebras, the real case poses additional difficulties. This is due to the fact that topological K-theory of real Banach algebras has period 8 instead of 2. The method we employ to overcome these difficulties can be used for complex algebras, and provides some simplifications to the original proofs. We also establish a natural analog of "Karoubi's conjecture" in Hermitian K-theory.
Let A be a commutative ring with 1/2 in A. In this paper, we define new characteristic classes for finitely generated projective A-modules V provided with a non degenerate quadratic form. These classes belong to the usual K-theory of A. They generalize in some sense the classical "cannibalistic" Bott classes in topological K-theory, when A is the ring of continuous functions on a compact space X. To define these classes, we replace the topological Thom isomorphism by a Morita equivalence between A-modules and C(V)-modules, where C(V) denotes the Clifford algebra of V, assuming that the class of C(V) in the graded Brauer group of A is trivial. We then essentially use ideas going back to Atiyah, Bott and Shapiro together with an alternative definition of the Adams operations due to Atiyah. When C(V) is not trivial in the graded Brauer group, the characteristic classes take their values in an algebraic analog of twisted K-theory. Finally, we also make use of a letter written by J.-P. Serre to the author, in order to interpret these classes as defined on the Witt group W(A) of the ring A. One aspect of this letter is summarized in our Lemma 3.5 where it is shown that in our situation the Bott class has a canonical square root in the K-theory of A.
We offer here a more direct approach to twisted K-theory, based on the notion of twisted vector bundles (of finite or infinite dimension) and of twisted principal bundles. This is closeely related to the classical notion ot torsors and bundles of modules over an algebra bundle. Twisted K-theory is simply defined as the Grothendieck group of twisted vector bundles (with a given twist). The usual operations on vector bundles (exterior powers, Adams operations) are easily extended to this twisted framework. Graded twisted K-theory, which is important for the definition and proof of the Thom isomorphism in this framework, is defined with the same formalism. We also define a twisted Chern character with target "twisted cohomology". Although there are many definitions of this character in the literature, our method is more elementary and is based on the classical definitions of the Chern-Weil theory, via connections and curvatures. In this way, we get explicit formulas very much in the spirit of Steenrod's coordinate bundles.
Bott periodicity plays an important role in topological K-theory. The purpose of this paper is to extend the periodicity theorem in a discrete context, where all classical groups are involved and not just the general linear group. The present paper generalizes previous results of the author [K1] and [K2], where 2 was assumed to be invertible in the rings involved. For the proof, two important ideas have to be mentioned : the first one is due to Ranicki [R] who introduced a kind of "enlarged" orthogonal group ; the second one is a genuine cup-product between quadratic forms due to Clauwens [C]. As an example of results obtained, we prove that the higher Witt groups of a finite field of characteristic 2 are all isomorphic to Z/2. They generalize in some sense the Dickson and Arf invariants.
The K-theory of a functor may be viewed as a relative version of the K-theory of a ring. In the case of a Galois extension of a number field F/L with rings of integers A/B respectively, this K-theory of the "norm functor" is an extension of a subgroup of the ideal class group Cl(A) by the 0-Tate cohomology group with coefficients in A*. The Mayer-Vietoris exact sequence enables us to describe quite explicitly this extension which is related to the coinvariants of Cl(A) under the action of the Galois group. We apply these ideas to find results in Number Theory, which are known for some of them with different methods.
The purpose of this shord paper is to make the link between the fundamental work of Atiyah, Bott and Shapiro (MR0167985/29/5250) and twisted K-theory (MR0282363/43/8075). This link was implicit for a long time in the literature (for the description of the real K-theory of spheres as an example) but was not explicitly defined before.
Twisted K-theory has its origins in the author's PhD thesis [27] : http://www.numdam.org/item?id=ASENS_1968_4_1_2_161_0 and in the paper with P. Donovan http://www.numdam.org/item?id=PMIHES_1970__38__5_0 The objective of this paper is to revisit the subject in the light of generalizations and new developments inspired by Mathematical Physics. See for instance E. Witten (hep-th/9810188), J. Rosenberg http://anziamj.austms.org.au/JAMSA/V47/Part3/Rosenberg.html, C. Laurent-Gentoux, J.-L. Tu, P. Xu (math/0306138) and M.F. Atiyah, G. Segal (math/0407054), among many authors. The unifiyng theme in our presentation is the notion of K-theory of graded Banach algebras,implicit in [27], from which most of the classical theorems in twisted K-theory are derived. We also prove some new results in the subject : a Thom isomorphism in this setting, explicit computations in the equivariant case and new cohomology operations (in the graded and ungraded cases).
We give a necessary and sufficient condition for lagrangians in a symplectic vector bundle to be deformed stably into transversal lagrangians. In the case of three lagrangians, we show that the associated Grothendieck group can be identified with a Hermitian K-theory group.
Let G be a finite group acting on a finite dimensional real vector space V. We denote by P(V) the projective space associated to V. In this paper we compute in a very explicit way the rank of the equivariant complex K-theory of V and P(V), using previous results by Atiyah and the author. The interest of this computation comes from explicit formulas given by the Baum-Connes-Slominska Chern character and the basic fact that the equivariant K-theory of V is free. We use these topological computations to prove algebraic results like computing the number of conjugacy classes of G which split in a central extension. Our main example is the case where V = R^n and G = the symmetric group of n letters acting on V by permutation of the coordinates. This example is related to the famous pentagonal identity of Euler and (ironically) the Euler-Poincare characteristic of the equivariant K-theory of V.
This survey paper is an expanded version of lectures given at the Clay Mathematics Academy ; see http://www.claymath.org/programs/outreach/academy/colloquium2005.php These lectures were intended to very young (and motivated) college students with little background. Therefore, they are accessible to a mathematician of any speciality willing to understand the subject. A much more complete introduction to K-theory may be found in the "Handbook of K-theory", recently edited by Springer.
This is a survey paper, starting from the general notion of coordinate bundle taken from Steenrod. Its aim is to provide a motivation for the introduction of cyclic homology (and the closely related noncommutative de Rham cohomology) by Connes, Tsygan and the author. The bridge is made through a generalization of Chern-Weil theory, explained in a very concrete manner through the transition functions of the bundles involved. The final result is a "Chern character" introduced by Connes and the author in the framework of K-theory and functional analysis.