I'm reading the wiki page of qudratic forms. It simply seems curious to me what are some concrete examples of applications of integral quadratic forms in algebraic topology. I've searched a bit but a lot of the readings online are too involved. Hope to see some well-illustrated ones here. Thanks!
What are some examples of applications of integral quadratic forms in $n$ variables in algebraic topology?
155 Views Asked by Bumbble Comm https://math.techqa.club/user/bumbble-comm/detail AtThere are 2 best solutions below
On
The paper
J.H.C. WHITEHEAD: A certain exact sequence. Ann. Math. 52 (1950), 51-110.
introduced a functor $\Gamma$ which is the ``universal quadratic functor" from Abelian groups to Abelian groups. Let $A$ be an Abelian group. Then $\Gamma(A)$ is the Abelian group with generators $\gamma a, a \in A$, and the following relations:
$ \gamma(-a)=\gamma(a), a\in A$
- if $\beta(a,b)=\gamma(a+b)-\gamma a -\gamma b, a,b\in A$, then $\beta : A \times A\to \Gamma(A)$ is biadditive.
This functor also occurs in the paper
R. BROWN and J.-L. LODAY, ``Van Kampen theorems for diagrams of spaces'', — Topology 26 (1987) 311-334.
This paper introduced a nonabelian tensor product $G \otimes H$ of groups $G,H$ which act on each other "compatibly", see this bibliography, so that in particular we have a tensor square $G \otimes G$ and a morphism of groups $\kappa: G \otimes G \to G$ induced by the commutator map $[\;,\;]: G \times G \to G$. The kernel of $\kappa$ is written $J_2(G)$ and is identified in the last paper as $\pi_3(SK(G,1))$. Further there is an exact sequence
$$H_3(G)\to \Gamma(G^{ab}) \to J_2(G) \to H_2(G) \to 0 . $$
I think there are more applications of $\Gamma$ in work of H,-J. Baues.
On a closed $4$-manifold $X$, define the intersection form $q_X:H^2(X)\otimes H^2(X) \to H^4(X) = \mathbb{Z}$ (coefficients are in $\mathbb{Z}$ throughout) by $q(\alpha, \beta) = \alpha \cap \beta$. (Such a form kills torsion, so it can be thought of a quadratic form over $\mathbb{Z}$.) If $X$ is simply-connected, then $q$ completely determines $X$ up to homotopy equivalence. In the opposite direction, any unimodular, symmetric bilinear form over $\mathbb{Z}$ is isomorphic to $q_X$ for some $X$. If $q$ is even, then $X$ is unique up to homeomorphism. If $q$ is odd, then there are two such manifolds up to homeomorphism. For smooth $4$-manifolds, homotopy equivalent manifolds are also homeomorphic. Thus for $q$ odd, at least one of the manifolds with intersection form $q$ must have no smooth structure. Generalizing this result, Donaldson proved that if $q_X$ is definite for $X$ (simply-connected and) smooth, then $q$ must be diagonalizable. In particular, there are quite a few $4$-manifolds that cannot be given a smooth structure. For manifolds that do have a smooth structure, it follows that any two such manifolds $X, X'$ are homeomorphic iff $q_X$ and $q_{X'}$ have the same rank, signature, and parity.