Controlled virtual volume representations and efficient virtual dominations

Pierre Derbez, Yi Liu, Hongbin Sun, Shicheng WangarXiv 2016 · arXiv

Abstract

In this paper, it is shown that for any closed orientable $3$-manifold with positive simplicial volume, the growth of the Seifert volume of its finite covers is faster than the linear rate. In particular, each closed orientable $3$-manifold with positive simplicial volume has virtually positive Seifert volume. The result reveals certain fundamental difference between the representation volume of the hyperbolic type and the Seifert type. The proof is based on developments and reactions of the previous work of third author on virtual domination of $3$-manifolds and the previous work of the other authors on virtual representation volumes.

Introduction

In the study of 3-manifolds, mathematicians use various "volume" invariants to understand topological structure. Two major types are hyperbolic volume (associated with hyperbolic geometry) and Seifert volume (associated with Seifert fiber spaces).

While the simplicial volume (introduced by Gromov) is well-behaved—it is proportional to the sum of the volumes of the hyperbolic pieces in a manifold's geometric decomposition—the behavior of Seifert volume has remained "mysterious." Specifically, it was unknown whether a manifold with hyperbolic pieces (positive simplicial volume) would eventually show positive Seifert volume if you looked at its finite covers (a "virtual" property).

The authors tackle two main questions:

  1. Problem 1.5: Is the Seifert volume of a 3-manifold virtually positive if it has positive simplicial volume?
  2. Problem 1.4: How do these volumes grow as we pass to increasingly large finite covers?

Methodology

The authors employ several sophisticated tools from geometric topology:

  • Virtual Domination: They use the concept of one manifold "dominating" another via a non-zero degree map. If manifold $M$ dominates $N$, invariants of $N$ can often be pulled back to $M$.
  • JSJ Decomposition: They analyze manifolds by breaking them into their canonical hyperbolic and Seifert "pieces."
  • CI Completions (Class Inversion): A novel auxiliary construction introduced in this paper. It allows the authors to extend representations from a single geometric piece to the entire manifold by embedding the fundamental group into a larger group that admits "class inversion" (an automorphism that sends elements to their inverses).
  • Universal Realisation of Cycles (URC): Drawing on work by Gaifullin, they use URC manifolds to construct efficient maps between manifolds, providing quantitative bounds on the degrees of these maps.
Commutative diagram illustrating the relationship between various group representations used to define Seifert volume.

Key Insights and Ideas

  • The Hyperbolic-Seifert Divergence: A major "aha moment" is the proof that hyperbolic and Seifert volumes grow differently. Hyperbolic volume is essentially linear (bounded by the simplicial volume), whereas Seifert volume can grow super-linearly. This suggests Seifert volume is far more sensitive to the specific geometric decomposition of the manifold.
  • Virtual Domination as a Bridge: The authors show that a hyperbolic manifold can virtually dominate any other 3-manifold. By picking a target manifold that already has positive Seifert volume, they can "force" the original manifold's covers to inherit that property.
  • Efficiency Matters: The paper doesn't just show that a map exists; Theorem 1.11 provides a "controlled" or "efficient" domination, linking the simplicial volume of the cover to the degree of the map and the simplicial volume of the target.

Key Results

  • Theorem 1.6: Confirms that positive simplicial volume implies virtually positive Seifert volume. If a manifold has a hyperbolic piece, it has a finite cover with non-zero Seifert volume.
  • Theorem 1.8: Demonstrates super-linear growth. For a manifold $M$ with positive simplicial volume, the ratio of the Seifert volume of a cover $M'$ to the degree of the cover $[M':M]$ is unbounded.
  • Corollary 1.7: Establishes a beautiful equivalence. For a closed orientable 3-manifold $N$, the following are equivalent:
  • The set of mapping degrees from any $M$ to $N$ is finite.
  • $N$ has virtually positive Seifert volume.
  • $N$ has at least one prime factor that is Seifert geometric, hyperbolic, or non-geometric (mixed).
  • Theorem 1.11: Provides a quantitative bound for virtual domination, showing that for a hyperbolic $M$, there exists a constant $c(M)$ such that for any $N$, a cover $M'$ can be found that dominates $N$ with $\|M'\| \leq c(M) \cdot |\text{deg}(f)| \cdot (\|N\| + \epsilon)$.

Conclusion and Future Work

The paper effectively bridges the gap between the rigid world of hyperbolic invariants and the more flexible, yet complex, world of Seifert representations. The authors highlight several open problems that remain:

  • Problem 6.3: Does the "Covering Seifert Volume" (a stabilized version of the invariant) exist as a finite value for all non-geometric graph manifolds?
  • Problem 6.4: What are the exact possible growth types (e.g., polynomial, exponential) for virtual Seifert volume?
  • Problem 6.5: Is the stabilized "Covering Hyperbolic Volume" exactly equal to the simplicial volume for every 3-manifold? The authors suspect it is, but a formal proof is missing.

Summary updated

Abstract

In this paper, it is shown that for any close orientable $3$-manifold with positive simplicial volume, the growth of the Seifert volume of its finite covers is faster than the linear rate. In particular, each closed orientable $3$-manifold with positive simplicial volume has virtually positive Seifert volume. The result reveals certain fundamental difference between the representation volume of the hyperbolic type and the Seifert type. The proof is based on developments and reactions of the previous work of third author on virtual domination of $3$-manifolds and the previous work of the other authors on virtual representation volumes.
BibTeX
@misc{derbez2016controlled,
  author = {Pierre Derbez and Yi Liu and Hongbin Sun and Shicheng Wang},
  title = {{Controlled virtual volume representations and efficient virtual dominations}},
  year = {2016},
  eprint = {1603.03558},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1603.03558},
}
Figure