Stably exotic 4 4 -manifolds
Daniel Kasprowski, Mark Powell · 2025
Read Stably exotic 4 4 -manifolds by Daniel Kasprowski, Mark Powell (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-10499.txt.
Introduction
Daniel Kasprowski and Mark Powell mapped in 2025 which four-dimensional shapes come in stably exotic pairs. Two shapes are stably exotic when they match after gluing on standard pieces yet never match smoothly, even then. For turnable shapes such pairs are ruled out, but unturnable ones abound. The team asks which group and twist data allow such pairs.
Given a mild condition on the group's homology, they give a full account of when pairs exist. New exotic pairs appear along the way, and new settings where none can. The work closes a long-open census question for this corner of four-dimensional topology.
Catalog record
| CREATOR | Daniel Kasprowski, Mark Powell |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 89 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-10499
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Stably exotic 4 4 -manifolds Daniel Kasprowski, Mark Powell Abstract . A pair of closed, smooth 4 4 -manifolds M M and M ′ M^{\prime} are stably exotic if they are stably homeomorphic but not stably diffeomorphic, where stabilisation refers to connected sum with copies of S 2 × S 2 S^{2}\times S^{2} . Orientable stable exotica do not exist by a result of Gompf, but Kreck showed that nonorientable examples are plentiful. We investigate which values of the fundamental group π \pi and the first and second Stiefel-Whitney classes w 1 w_{1} and w 2 w_{2} admit stably exotic pairs, providing a complete description if H 5 ( π , ℤ ) = 0 H_{5}(\pi;\mathbb{Z})=0 . In particular we produce new stable exotica, and new settings in which they do not arise. 1991 Mathematics Subject Classification 57K40. 1. Introduction Two closed, smooth 4 -manifolds M and M^{\prime} are stably diffeomorphic or stably homeomorphic if there exists k\in\mathbb{N} such that M\#k(S^{2}\times S^{2}) and M^{\prime}\#k(S^{2}\times S^{2}) are diffeomorphic or homeomorphic, respectively. Gompf proved that orientable 4 -manifolds are stably diffeomorphic if and only if they are stably homeomorphic. In fact, he proved the following stronger result, generalising Wall’s theorem [ 24 ] . Theorem (Gompf [ 12 ] ) . Let M and M^{\prime} be closed, smooth, stably homeomorphic 4 -manifolds. (i) Then M\#(S^{2}\mathrel{\widetilde{\times}}S^{2}) and M^{\prime}\#(S^{2}\mathrel{\widetilde{\times}}S^{2}) are stably diffeomorphic. (ii) Suppose that M and M^{\prime} are orientable. Then M and M^{\prime} are stably diffeomorphic. Remark 1.1 . If the universal cover \widetilde{M} is non-spin, then M\#(S^{2}\times S^{2}) is diffeomorphic to M\#(S^{2}\mathrel{\widetilde{\times}}S^{2}) [ 23 ] *Theorem 5.2. Hence by Gompf’s theorem 4 -manifolds M with either w_{1}(M)=0 or w_{2}(\widetilde{M})\neq 0 are stably diffeomorphic if and only if they are stably homeomorphic. We investigate the existence of stably exotic 4 -manifolds, i.e. 4 -manifolds that are stably homeomorphic but are not stably diffeomorphic. By Remark 1.1 , these are necessarily nonorientable and the universal cover must be spin. Henceforth we restrict to such 4 -manifolds. Hypothesis 1.2 . By a 4 -manifold we will always mean a closed, smooth, nonorientable 4 -manifold with spin universal cover. Kreck proved that there is an abundance of examples of stable exotica; see also constructions of Cappell-Shaneson [ 5 , 6 ] and Akbulut [ 1 , 2 ] for fundamental groups \mathbb{Z}/2 and \mathbb{Z} respectively. Theorem (Kreck [ 15 ] *Theorem 1) . Let \pi be a finitely presented group and let w\colon\pi\to\mathbb{Z}/2 be a nontrivial homomorphism. Then there exist stably exotic 4 -manifolds M and M^{\prime} with fundamental group \pi and orientation character w , of the form M=D\#K3 and M^{\prime}=D\#11(S^{2}\times S^{2}) , for some 4-manifold D . In the case where the 2-Sylow subgroup is \mathbb{Z}/2 , the stable homeomorphism and diffeomorphism classifications were also comprehensively studied by Debray [ 9 ] . The stably exotic 4 -manifolds appearing in these works satisfy w_{2}(\nu_{M})=w_{1}(\nu_{M})^{2} , where \nu_{M} is the stable normal bundle of M . Our aim in this article is to understand, for a fixed fundamental group and orientation character, which values of the second Stiefel-Whitney class are realised by stably exotic 4 -manifolds. To formalise this, let M be a 4 -manifold as in 1.2 , and let c\colon M\to B\pi be a 2-connected map; in particular \pi_{1}(M)\cong\pi . Then there exist classes w_{1}\neq 0\in H^{1}(\pi;\mathbb{Z}/2) and w_{2}\in H^{2}(\pi;\mathbb{Z}/2) with c^{*}(w_{i})=w_{i}(\nu_{M}) ; to see this for w_{2} consider the exact sequence H^{2}(\pi;\mathbb{Z}/2)\to H^{2}(M;\mathbb{Z}/2)\to H^{2}(\widetilde{M};\mathbb{Z}/2) . The triple (\pi,w_{1},w_{2}) is called the normal 1-type of M . Kreck’s work quoted above shows that stably exotic 4-manifolds exist whenever w_{2}=w_{1}^{2} . Our first main result reads as follows. Theorem A . Let \pi be a finitely presented group and let w_{1}\neq 0\in H^{1}(\pi;\mathbb{Z}/2) and w_{2}\in H^{2}(\pi;\mathbb{Z}/2) . (i) If there exist stably exotic 4 -manifolds with normal 1-type (\pi,w_{1},w_{2}) , then w_{1}^{3}=w_{1}w_{2}\in H^{3}(\pi;\mathbb{Z}/2) . (ii) [Kreck] If w_{2}=w_{1}^{2}\in H^{2}(\pi;\mathbb{Z}/2) , then there exist stably exotic 4 -manifolds with normal 1-type (\pi,w_{1},w_{2}) . (iii) If \pi has cohomological dimension at most 3 and w_{1}^{3}=w_{1}w_{2}\in H^{3}(\pi;\mathbb{Z}/2) , then there exist stably exotic 4 -manifolds with normal 1-type (\pi,w_{1},w_{2}) . Remark 1.3 . Note that w_{2}=w_{1}^{2} implies w_{1}^{3}=w_{1}w_{2} , but the converse does not hold in general. So in particular A ( iii ) produces new stable exotica that did not appear in [ 15 ] or [ 9 ] , and ( ii ) does not imply ( iii ). For a (cohomological dimension two) example where w_{1}^{3}=w_{1}w_{2} but w_{2}\neq w_{1}^{2} , let \pi\cong\mathbb{Z}^{2} . Then H^{*}(B\mathbb{Z}^{2};\mathbb{Z}/2)\cong\wedge^{2}(\mathbb{Z}/2) . Let w_{1}=e_{1} and let w_{2}=e_{1}\wedge e_{2} . Then w_{1}^{3}=e_{1}\wedge e_{1}\wedge e_{1}=0=e_{1}\wedge e_{1}\wedge e_{2}=w_{1}w_{2} but w_{1}^{2}=e_{1}\wedge e_{1}=0\neq e_{1}\wedge e_{2}=w_{2} . By A ( iii ), it follows that there exist stably exotic 4 -manifolds with this normal 1-type. 1.1. The case w_{1}^{3}=w_{1}w_{2} We describe a secondary obstruction to the existence of stably exotic 4-manifolds with normal 1 -type (\pi,w_{1},w_{2}) such that w_{1}^{3}=w_{1}w_{2} . For this we consider the fibration sequence F\xrightarrow{u}K(\mathbb{Z}/2,2)\times K(\mathbb{Z}/2,1)\xrightarrow{\iota_{1}^{3}+\iota_{1}\iota_{2}}K(\mathbb{Z}/2,3), where \iota_{k}\in H^{k}(K(\mathbb{Z}/2,k);\mathbb{Z}/2)\cong\mathbb{Z}/2 , for k=1,2 , are the generators, and F is by definition the homotopy fibre of the map \iota_{1}^{3}+\iota_{1}\iota_{2} . We write \Sq^{2}_{w_{1},w_{2}}:=\Sq^{2}(-)+\big(w_{1}\cup\Sq^{1}(-)\big)+\big(w_{2}\cup-\big)\colon H^{2}(\pi,\mathbb{Z}/2)\to H^{4}(\pi;\mathbb{Z}/2). Theorem B . (i) The space F is 3-coconnected with \pi_{1}(F)\cong\mathbb{Z}/2 , \pi_{2}(F)\cong(\mathbb{Z}/2)[\mathbb{Z}/2] , and trivial k -invariant. (ii) Let p\colon\widetilde{F}\to F be the universal cover and let \{x_{1},x_{2}\} be the basis of H^{2}(\widetilde{F};\mathbb{Z}/2)\cong\pi_{2}(F) such that T^{*}x_{1}=x_{2} , where T is the deck transformation. There exists a unique class \mathfrak{o}\in H^{4}(F;\mathbb{Z}/2) such that p^{*}\mathfrak{o}=x_{1}x_{2} and s^{*}\mathfrak{o}=0 for any section s\colon K(\mathbb{Z}/2,1)\to F . (iii) If there exist stably exotic 4 -manifolds with normal 1-type (\pi,w_{1},w_{2}) , then [f^{*}\mathfrak{o}]=0\in H^{4}(\pi;\mathbb{Z}/2)/\operatorname{Im}\Sq^{2}_{w_{1},w_{2}}, for any lift f\colon B\pi\to F of w_{2}\times w_{1}\colon B\pi\to K(\mathbb{Z}/2,2)\times K(\mathbb{Z}/2,1) . (iv) If H_{5}(\pi;\mathbb{Z})=0 , then the converse to ( iii ) holds, i.e. if w_{1}^{3}=w_{1}w_{2} and [f^{*}\mathfrak{o}]=0\in H^{4}(\pi;\mathbb{Z}/2)/\operatorname{Im}\Sq^{2}_{w_{1},w_{2}} for some lift f\colon B\pi\to F of w_{2}\times w_{1}\colon B\pi\to K(\mathbb{Z}/2,2)\times K(\mathbb{Z}/2,1) , then there exist stably exotic 4 -manifolds with this normal 1-type. For B ( iii ) note that the existence of a lift f follows from A ( i ). Remark 1.4 . We will show in Proposition 6.3 , that the class [f^{*}\mathfrak{o}]\in H^{4}(\pi;\mathbb{Z}/2)/\operatorname{Im}\Sq^{2}_{w_{1},w_{2}} is independent of the choice of lift f . Remark 1.5 . In Proposition 6.5 we will use B ( iii ) to give an example of a normal 1 -type (\pi,w_{1},w_{2}) such that w_{1}^{3}=w_{1}w_{2} but [f^{*}\mathfrak{o}]\neq 0 , and no stably exotic 4 -manifolds with normal 1 -type (\pi,w_{1},w_{2}) exist. In particular, this shows that the converse of A ( i ) does not hold. If w_{1}^{3}=w_{1}w_{2} and 0=[f^{*}\mathfrak{o}] , there is one further obstruction to the existence of stably exotic 4 -manifolds with normal 1-type (\pi,w_{1},w_{2}) . To describe it, we have to first give a brief overview of the proofs of A and B . By Proposition 3.10 , the existence of stably exotic 4 -manifolds with normal 1-type (\pi,w_{1},w_{2}) depends on the bordism class of the K3 surface in \Omega_{4}(\xi_{\pi}) as defined in ( 3.1 ). There is a James spectral sequence converging to \Omega_{4}(\xi_{\pi}) and the K3 surface generates the term E_{0,4}^{2}=H_{0}(\pi;\Omega_{4}^{\spin}) . Since the action of \pi on \Omega_{4}^{\spin} is given by the nontrivial w_{1} , E_{0,4}^{2}\cong\mathbb{Z}/2 . Hence 0=[K3]\in\Omega_{4}(\xi_{\pi}) if and only if