Isoperimetric behavior of generalized Stallings-Bieri groups

Noel Brady, Pratit Goswami, Rob Merrell · 2025

Read Isoperimetric behavior of generalized Stallings-Bieri groups by Noel Brady, Pratit Goswami, Rob Merrell (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-2509-09939.txt.

Introduction

Noel Brady and 2 co-authors measure how hard it is to fill loops inside a broad family of groups in a 2025 geometric group theory paper. The Dehn function of a group counts the cheapest way to fill a closed word, a yardstick for the difficulty of deciding when two expressions name the same element. The groups in question sit inside direct products and project onto a free abelian group, a pattern that includes the classic Stallings-Bieri examples and far more.

A new bookkeeping device, the n-split of such a projection, organises the estimate. With it the authors bound Dehn functions for these coabelian subgroups in terms of the pieces they are built from. The framework widens older results considerably, giving one lens for a whole landscape of product subgroups.

Catalog record

CREATORNoel Brady, Pratit Goswami, Rob Merrell
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE172 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-09939

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Isoperimetric behavior of generalized Stallings-Bieri groups Noel Brady, Pratit Goswami, Rob Merrell Abstract . We introduce the notion of n n -split for an epimorphism from a group to a finite rank free abelian group. This is used to provide bounds for the Dehn functions of certain coabelian subgroups of direct products of finitely presented groups. Such subgroups include and significantly generalize the Stallings-Bieri groups. MnLargeSymbols’164 MnLargeSymbols’171 1. Introduction The Dehn function is a fundamental quasi-isometry invariant of finitely presented groups. Although the geometric intuition behind the Dehn function can be found in the work of Dehn [ Deh11 ] , it was not until the early 1990s ( [ Gro93 ] and [ Ger91 ] ) that it was defined as a quasi-isometry invariant. Since then, there has been an active research enterprise estimating bounds for Dehn functions of finitely presented groups. Estimates for Dehn functions of finitely presented subgroups of various groups have been computed as a way of demonstrating the extent to which the large scale geometry of the subgroup differs from that of the parent group. Examples include finitely presented subgroups of the following classes of groups: automatic groups ( [ BBMS97 ] ); CAT(0) groups ( [ BF17 ] , [ BT22 ] ); virtually special cubical groups ( [ BS19 ] , [ Bri02 ] ); hyperbolic groups ( [ Bra99 ] , [ GS02 ] , [ Lod18 ] , [ Kro21 ] , [ IMM23 ] , [ IMP24 ] , [ LIP24 ] ); and direct product of free groups ( [ ABI+25a ] , [ ABI+25b ] ). In other situations, it has been shown that the Dehn function of a finitely presented subgroup necessarily agrees with that of its parent group; examples of such situations are studied in [ DERY09 ] , [ CF17 ] , and [ KL23 ] . This last work is the most relevant to our efforts, as it is the techniques of that article that we investigate. In many of the above examples, the finitely presented subgroups appear as kernels of maps from the ambient group to \mathbb{Z} , giving a short exact sequence {\lx@inpgf@ignorespaces 1} {\lx@inpgf@ignorespaces K} {\lx@inpgf@ignorespaces G} {\lx@inpgf@ignorespaces\mathbb{Z}} {\lx@inpgf@ignorespaces 1\,.} Dison showed in [ Dis08 ] that in the case where G is a right-angled Artin group, the Dehn function of K is bounded above by a quartic function. Dison, Elder, Riley and Young showed in [ DERY09 ] that Stallings’ group has a quadratic Dehn function (in this case G is a direct product of three copies of F_{2} and the homomorphism takes all 6 generators of G to a generator of \mathbb{Z} ). Carter and Forester showed in [ CF17 ] that the Stallings-Bieri groups (where G=(F_{2})^{n} for n\geq 3 ) have quadratic Dehn functions. In [ KL23 ] , Kropholler and Llosa Isenrich abstracted and generalized the techniques of [ CF17 ] to include the case of sequences with \mathbb{Z}^{m} as quotient. Our main result fits into the framework of this last article, and uses a condition called being n -split, which is an analogue of their P -split condition. Theorem 7.3 . Let n>2 be an integer. For each 1\leq i\leq 2n , let G_{i} be a finitely presented group such that {\lx@inpgf@ignorespaces 1} {\lx@inpgf@ignorespaces N_{i}} {\lx@inpgf@ignorespaces G_{i}} {\lx@inpgf@ignorespaces\mathbb{Z}^{m}} {\lx@inpgf@ignorespaces 1} \scriptstyle{\lx@inpgf@ignorespaces\varphi_{i}} is an n -split short exact sequence. Let G=\prod_{i=1}^{2n}G_{i} , and let \Phi:G\to\mathbb{Z}^{m} be defined by \Phi(g)=\sum_{i=1}^{2n}\varphi_{i}({\rm pr}^{\rm G}_{i}(g)) . Then the kernel \Ker(\Phi) in the short exact sequence {\lx@inpgf@ignorespaces 1} {\lx@inpgf@ignorespaces\Ker(\Phi)} {\lx@inpgf@ignorespaces G} {\lx@inpgf@ignorespaces\mathbb{Z}^{m}} {\lx@inpgf@ignorespaces 1} \scriptstyle{\lx@inpgf@ignorespaces\Phi} is finitely presented and its Dehn function satisfies \delta_{G}(n)\dehnleq\delta_{\Ker(\Phi)}(n)\dehnleq\overline{\delta_{G}}(n)\cdot\log(n) . If, moreover, \frac{\delta_{G}(n)}{n} is superadditive, then \delta_{G}(n)\simeq\delta_{\Ker(\Phi)}(n) . Here \overline{\delta_{G}}(n) denotes the superadditive closure of the Dehn function \delta_{G}(n) of the group G , defined in Definition 3.3 . The following table summarizes the differences and similarities between the analogous versions of Theorem 7.3 in this paper and in [ KL23 ] . Note that although Theorem 4.2 in [ KL23 ] is only stated for a direct product of four groups, grouping some of the subproducts into pairs (the structure of which is forced by their P -splitting) permits the result to hold for arbitrary products of an even number of groups, and so we include this more general setup in the table. {longtblr} colspec = —c—c—c—c—, rowhead = 1, rows=abovesep=2pt,belowsep=2pt, n -split \Ker(\Phi) Geometry of filling Location n=1 \Phi:\displaystyle\prod_{j=1}^{3}G_{j}\to\mathbb{Z}^{m} triangular [ KL23 , Section 3] n=2 \Phi:\displaystyle\prod_{j=1}^{2n}G_{j}\to\mathbb{Z}^{m} square [ KL23 , Section 4] n\geq 3 \Phi:\displaystyle\prod_{j=1}^{2n}G_{j}\to\mathbb{Z}^{m} triangular Section 7 We highlight several features of our construction. • The n -split condition is a generalization of the P -split condition of [ KL23 ] (which corresponds to being 2 -split in our sense). For n\geq 3 , the n -split condition exploits different features of the lattice of subgroups of \Ker(\Phi) than [ KL23 ] . This enables us to construct algebraic triangles which are used to build van Kampen diagrams. These diagrams have different fillings than the algebraic square fillings of [ KL23 ] in the cases where both techniques apply (for example, when n\geq 4 is even). • In our construction of the algebraic triangle, the regions of the triangle closely model the types of subgroups in the lattice of subgroups of \Ker(\Phi) . In particular, quadrilateral regions correspond to subgroups which have a direct product decomposition. Furthermore, triangular and hexagonal regions correspond to subgroups which admit the following 3 -fold “symmetry”; they appear as kernels \Ker(A^{3}\to A) where A is free-abelian and the homomorphism is (a_{1},a_{2},a_{3})\mapsto a_{1}+a_{2}+a_{3} . • Finally, the level of detail provided by the 25 region decomposition of the algebraic triangle in our construction, allows us to explicitly describe the \Ker(\Phi) group elements associated to every vertex (the description in [ KL23 ] is more implicit than our description) and the edge paths associated to every edge of the algebraic triangle. In particular, we can distinguish between purely linear algebra operations and other group operations inside \Ker(\Phi) . See proofs of Lemma 6.10 and Lemma 6.14 for details. The proofs of the main results of [ KL23 ] and the current paper suggest investigating the combinatorics of the lattice of (particular families of) subgroups of a given finitely presented group group as an approach to understanding the Dehn function of the ambient group. In particular, it would be interesting to discover other situations where the combinatorics of a lattice of subgroups of a given group can be used to provide upper bounds for the Dehn function of the group. It is also tempting to think about lattices of subgroups with higher dimensional geometric realizations and possible applications to higher dimensional filling functions of groups. 2. Acknowledgements The first two authors acknowledge travel support from Simons Foundation collaboration grant #430097. The second author acknowledges support from the 2025 Summer Research Fellowship at the University of Oklahoma. The third author also extends thanks to the University of Oklahoma for support via the 2024 Summer Research Fellowship. 3. Background The main theorem of this paper concerns Dehn functions of certain finitely presented subgroups of direct products of groups. This section provides background on equivalence of functions, Dehn functions of groups, and the notion

Read the complete text