Irreducible Nonsurjective Endomorphisms of $F_n$ are Hyperbolic

Jean Pierre Mutanguha

REVIEW 1 major objections 5 minor 22 references

Irreducible Nonsurjective Endomorphisms of $F_n$ are Hyperbolic

T0 review · 1 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Every irreducible nonsurjective endomorphism of a finitely generated free group has a word-hyperbolic mapping torus, because such endomorphisms admit a unique clean immersion representative and cannot carry periodic conjugacy classes.

desk verdict This paper proves a long-standing conjecture for nonsurjective free group endomorphisms with clean, careful arguments; the one external dependency is explicit and not a hidden flaw. read the letter →

arxiv 1908.08214 v3 pith:JM5AH6GF submitted 2019-08-22 math.GR

Jean Pierre Mutanguha This is my paper · ORCID

classification math.GR MSC 20F6520E0520F6720E36

keywords freegroupsirreducibleendomorphismsfullymappingtoriword-hyperbolictraintrackmapsouterspaceWhiteheadgraphs

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves that every irreducible nonsurjective endomorphism of a finitely generated free group has a word-hyperbolic mapping torus—the ascending HNN extension $F*_\varphi = \langle F,t \mid t^{-1}xt=\varphi(x)\rangle$ has a hyperbolic Cayley graph. More generally, the conclusion holds for every endomorphism represented by a clean immersion of a finite core graph. The proof shows that such an endomorphism has a unique irreducible immersion representative, that irreducibility implies full irreducibility in the nonsurjective case, and that any finitely generated invariant subgroup carrying an expanding conjugacy class has finite index in some iterated image. A prior hyperbolicity criterion then reduces the task to excluding periodic conjugacy classes, which the invariant-subgroup result accomplishes. If correct, the theorem turns a broad family of non-invertible free-group maps into a source of hyperbolic groups.

What carries the argument

The carrying objects are clean immersions and the hyperbolicity criterion from the author's earlier work. A clean immersion is a graph map that is locally injective, whose transition matrix is primitive, and whose Whitehead graphs at every vertex are connected; it represents the endomorphism on the fundamental group. The criterion (Theorem 6.2) states that if an endomorphism is represented by an immersion, then the mapping torus is word-hyperbolic exactly when no nontrivial element $a$ has $\varphi^n(a)$ conjugate to $a^d$ for some $d,n\ge 1$. The new work supplies the clean immersion representative and proves that such a periodic conjugacy class would violate Proposition 5.4, making the criterion's obstruction impossible. Bounded cancellation estimates and the clean train-track structure drive the length and fold arguments that produce the representative and control iterated images.

What would settle it

Exhibit a clean immersion representing an irreducible nonsurjective endomorphism for which some nontrivial $a\in F$ and integers $d,n\ge 1$ satisfy $\varphi^n(a)=g a^d g^{-1}$; by the paper's own criterion (Theorem 6.2) the mapping torus would then fail to be word-hyperbolic, disproving Theorem 6.3. In a concrete example this is a finite check: if the word length of $\varphi^n(a)$ grows linearly for some nontrivial $a$, such a periodic class exists, and the predicted hyperbolicity fails.

Watch

Extended reading notes

Core claim

The central claim is Theorem 6.3: if an injective endomorphism $\varphi:F\to F$ is represented by a clean immersion, then its mapping torus $F*_\varphi$ is word-hyperbolic; in particular, every irreducible nonsurjective endomorphism of a finitely generated free group is hyperbolic in this sense. Supporting results include Theorem 4.5, which produces a unique irreducible immersion representative with connected Whitehead graphs for every irreducible nonsurjective endomorphism; Theorem 4.6, a partial converse using Whitehead graphs with no cut vertices; and Theorem 5.5, which characterizes fully irreducible injective endomorphisms as those with no periodic cyclic free factor, a clean representative, and image not contained in a proper free factor. Corollary 5.6 then identifies irreducible and fully irreducible for nonsurjective endomorphisms. The decisive technical step is Proposition 5.4: if a finitely generated subgroup $H$ satisfies $\varphi(H)\le H$ and contains a $\varphi$-expanding conjugacy class, then $[\varphi^k(F):\varphi^k(F)\cap H]<\infty$ for some $k\ge 0$, which rules out the cyclic periodic conjugacy class that would obstruct hyperbolicity.

Load-bearing premise

The load-bearing premise is the previously proved hyperbolicity criterion (Theorem 6.2), that for an immersion representative the mapping torus is hyperbolic if and only if no nontrivial element satisfies $\varphi^n(a)$ conjugate to $a^d$; the present paper does not reprove that criterion, so an error or unstated hypothesis in it would collapse the main conclusion.

Editorial extensions

If this is right

Every irreducible nonsurjective endomorphism of a finitely generated free group has a word-hyperbolic mapping torus, so the group $\langle a,b,t \mid t^{-1}at=ab,\ t^{-1}bt=ba\rangle$ is word-hyperbolic.
Irreducibility and full irreducibility coincide for nonsurjective endomorphisms, so iterates of an irreducible nonsurjective endomorphism remain irreducible.
The characterization in Theorem 5.5 gives a checkable criterion for full irreducibility of an injective endomorphism: no periodic cyclic free factor, a clean representative, and an image not contained in a proper free factor.
An irreducible nonsurjective endomorphism acts on outer space with a unique attracting fixed point, represented by the unique irreducible immersion, giving a canonical graph model for the endomorphism.
The partial converse Theorem 4.6 shows that a clean immersion with connected Whitehead graphs and no cut vertices represents a nonsurjective fully irreducible endomorphism.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

An implication the author leaves implicit is that the mapping tori constructed here form a class of hyperbolic groups that are not free-by-cyclic in the usual sense, so the proof gives a template for showing hyperbolicity of more general ascending HNN extensions of free groups.
The uniqueness of the immersion representative suggests that the expanding graph together with its leading eigenvalue is an invariant of the endomorphism, potentially allowing distinct irreducible nonsurjective endomorphisms to be distinguished by their clean immersion data.
A natural testable extension is to ask whether the conclusion holds for endomorphisms some iterate of which is represented by a clean immersion; the current proof does not immediately cover that 'eventually clean' class.
Since the obstruction to hyperbolicity is exactly a periodic conjugacy class, the theorem predicts that in all explicit examples, the length of $\varphi^n(a)$ grows exponentially for every nontrivial $a$; this can be checked directly in small-rank examples.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The headline result here is genuine: every irreducible nonsurjective endomorphism of a free group has a word-hyperbolic mapping torus. The proof goes through a clean structure theorem—such endomorphisms have a unique clean immersion representative—and gives a nice corollary that irreducibility implies full irreducibility. The paper also supplies a new proof of Reynolds' immersion theorem, which was previously unpublished, and a useful characterization of fully irreducible injective endomorphisms in Theorem 5.5. I was not expecting all of this to be packaged in one paper, and the main results are new and significant.

The exposition is careful and honest. The action on outer space is treated properly, the limiting tree argument in Theorem 4.5 is spelled out, and the subgroup rigidity result in Section 5 is a real contribution. The examples, especially the Sapir group, are well chosen and actually illustrate the scope of the theorems.

The real soft spots are the dense folding arguments in Theorems 4.5 and 4.6. I could not find a gap, but these are hard to check in a quick read. The other structural dependency is Theorem 6.2 from the author's own previous paper [15]: the main hyperbolicity theorem relies on an "if and only if" criterion for all immersions that is not reproved here. That is a legitimate reliance on prior work, not a circularity, and the reduction from a clean immersion to the absence of periodic powers is sound. If [15] has a hidden hypothesis, the main theorem would fail, but there is no evidence of that from this paper; it is a normal conditional result.

One weaker point: Theorem 4.6 gives a criterion for full irreducibility that the author immediately shows is not necessary, which is fine, but it means the title theorem is not derived from that criterion and the paper is slightly less unified than the introduction suggests.

Who should read this? Geometric group theorists working on free group endomorphisms, outer space, or mapping tori of free groups. The paper deserves a serious referee; I would send it to a strong journal and recommend the referee focus on the folding arguments and on verifying the applicability of [15, Thm 6.3]. I would cite this if I worked in the area.

Referee Report

1 major / 5 minor

Summary. The paper proves several results about irreducible nonsurjective endomorphisms of finitely generated free groups. It gives a new proof that such an endomorphism can be represented by a unique clean immersion (Theorem 4.5), proves a partial converse for clean immersions whose Whitehead graphs have no cut vertices (Theorem 4.6), characterizes invariant finitely generated subgroups (Propositions 5.3 and 5.4), characterizes fully irreducible injective endomorphisms (Theorem 5.5), and derives that irreducible nonsurjective endomorphisms are fully irreducible (Corollary 5.6). The main hyperbolicity result (Theorem 6.3) states that if an endomorphism is represented by a clean immersion then its mapping torus is word-hyperbolic; in particular, the mapping torus of an irreducible nonsurjective endomorphism is word-hyperbolic. The proof relies on a previously stated hyperbolicity criterion for mapping tori of immersions (Theorem 6.2, from the author's earlier work).

Significance. If the results are correct, the paper makes a substantial contribution to the study of nonsurjective endomorphisms of free groups. It extends train-track and outer-space methods from automorphisms to irreducible nonsurjective endomorphisms, establishes that irreducibility implies full irreducibility in this setting, and proves hyperbolicity of the corresponding mapping tori, generalizing the free-by-cyclic case. The proof of Theorem 4.5 is a genuine new proof of a result previously announced by Reynolds, and the treatment of invariant subgroups in Section 5 is careful and appears to be new. The paper also gives a quick proof that the Sapir group is word-hyperbolic. The technical machinery is standard and the arguments are mostly detailed. The main caveat is that the final hyperbolicity theorem depends on the author's earlier criterion in Theorem 6.2, which is invoked without a full statement or proof.

major comments (1)

[Section 6, Theorem 6.3] The proof of Theorem 6.3 invokes Theorem 6.2 of [15] as a black box. Since this external criterion supplies exactly the implication 'absence of roots implies word-hyperbolicity' that the argument needs, the central conclusion of the paper is conditional on Theorem 6.2. The manuscript should include a complete statement of Theorem 6.2, clarify its publication status, and explicitly verify that all hypotheses of that theorem are satisfied by clean immersions. If Theorem 6.2 has any hidden assumptions, they must be stated so that the reader can check them.

minor comments (5)

[Abstract and Introduction] The abstract states several results without theorem numbers; please add references to Theorems 4.5, 4.6, 5.5, and 6.3 to help the reader navigate.
[Theorem 4.5 proof, Section 4] The sentence 'Since there are finitely many combinatorially distinct ways to fold a forest, there are fixed j > i and k ≥ 1 such that the composition of folds S_j → S_{j+k} is homotopic to a homeomorphism' is terse; the finiteness argument should be expanded, since the number of folds may depend on the number of edges in S_j.
[Theorem 4.6 proof, Section 4] The final contradiction in the proof of Theorem 4.6 asserts that the induced map on Δ∞ would have disconnected Whitehead graphs or Whitehead graphs with cut vertices; this step is not fully justified and would benefit from a more explicit explanation of which identified vertices produce the cut vertices.
[Example 4.7] In Example 4.7, the conclusion that φ_A is fully irreducible is derived from word-hyperbolicity of its mapping torus together with the fact that A has rank two; this implication is not immediate and should be explained explicitly.
[Throughout] There are several typographical issues: 'eventially' should be 'eventually' in Proposition 5.3, the affiliation line contains an extra space in 'F ayetteville', and 'c.f.' should be 'cf.'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main theorem is derived from a transparently cited external hyperbolicity criterion and in-paper dynamical lemmas.

full rationale

The paper's central claim, Theorem 6.3, is not assumed at the start and does not reduce to its inputs by construction. The proof invokes Theorem 6.2, quoted from the author's previous paper [15], as a criterion for word-hyperbolicity of mapping tori of immersions. That criterion is a separate theorem stated for arbitrary immersions and does not itself assume irreducibility, nonsurjectivity, or cleanliness; it is therefore not an instance of the conclusion being proved. The paper then supplies the genuinely new dynamical content: a clean immersion expands every nontrivial conjugacy class, and Proposition 5.4, proved in Section 5, rules out an invariant cyclic subgroup with finite-index intersection. The auxiliary results cited from the literature, such as Proposition 2.6 and Corollary 2.9, are used as standard background rather than as smuggled versions of the target theorem. No fitted parameters are renamed as predictions, no uniqueness theorem from the author's prior work is used to force the main choice, and no displayed equation identifies the conclusion with an assumption. The main dependency, Theorem 6.2 of [15], is a load-bearing citation, but it is transparently identified and is independent support in the sense of a parameter-free theorem with stated assumptions that do not include the present result; a possible gap in that external theorem would be a correctness risk, not circularity. The derivation chain is therefore self-contained relative to its cited external theorem, and no circular step is exhibited.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new objects; it relies on standard tools of the field. There are no empirical free parameters.

assumptions (5)

domain assumption Stallings fold theory and subgroup separability of free groups
Used in Lemma 2.2 and Lemma 2.4 to prove injectivity of irreducible endomorphisms and unboundedness of Stallings graphs.
domain assumption Bestvina-Handel train track theorem (Theorem 2.8)
Provides irreducible train track representatives for irreducible endomorphisms, used throughout.
domain assumption Bestvina-Feighn-Handel lamination convergence (Lemma 4.1)
Gives convergence of the action on outer space to the limit tree T_f, central to Theorem 4.5.
domain assumption The author's prior hyperbolicity criterion (Theorem 6.2 from [15])
Used as a black box to prove Theorem 6.3; it is an external published result.
standard math Perron-Frobenius theory of nonnegative matrices
Used in Proposition 2.6 to show weakly clean implies clean.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Irreducible Nonsurjective Endomorphisms of $F_n$ are Hyperbolic." pith.science (2026). https://pith.science/paper/JM5AH6GF

@misc{pith2026190808214,
  author       = {Pith},
  title        = {Pith review of: Irreducible Nonsurjective Endomorphisms of $F_n$ are Hyperbolic},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JM5AH6GF}},
  note         = {Machine review of arXiv:1908.08214}
}

read the original abstract

Previously, Reynolds showed that any irreducible nonsurjective endomorphism can be represented by an irreducible immersion on a finite graph. We give a new proof of this and also show a partial converse holds when the immersion has connected Whitehead graphs with no cut vertices. The next result is a characterization of finitely generated subgroups of the free group that are invariant under an irreducible nonsurjective endomorphism. Consequently, irreducible nonsurjective endomorphisms are fully irreducible. The characterization and Reynolds' theorem imply that the mapping torus of an irreducible nonsurjective endomorphism is word-hyperbolic.

Figures

Figures reproduced from arXiv: 1908.08214 by the authors.

**Figure 1.** Computation showing σ(R2, ϕ−1 a ) · [φ] = σ(B, β). 4. Irreducible Nonsurjective Endomorphisms are Immersions Given an irreducible train track map f : Γ → Γ representing an injective endomorphism φ : F → F, then the iterates of φ act on the F-tree Γ by taking minimal trees of their ˜ twisted actions. If lΓ˜ : F → R is the length function for Γ, then ˜ lΓ˜·φn = lΓ˜φ n . We define the limit tree Tf to be the tree whose… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 22 canonical work pages

[15]

Hyperbolic Immersions of Free Groups

Jean Pierre Mutanguha. Hyperbolic Immersions of Free Groups. Groups Geom. Dyn. , to appear. arxiv:1809.04761

work page Pith review arXiv
[1]

R. B. Bapat and T. E. S. Raghavan. Nonnegative matrices and applications, volume 64 of Encyclopedia of Mathematics and its Applications . Cambridge University Press, Cambridge, 1997

work page 1997
[2]

Laminations, trees, and irreducible automor- phisms of free groups

M. Bestvina, M. Feighn, and M. Handel. Erratum to: “Laminations, trees, and irreducible automor- phisms of free groups” [Geom. Funct. Anal. 7(2):215–244, 1997]. Geom. Funct. Anal., 7(6):1143, 1997

work page 1997
[3]

Laminations, trees, and irreducible automor- phisms of free groups

Mladen Bestvina, Mark Feighn, and Michael Handel. Laminations, trees, and irreducible automor- phisms of free groups. Geom. Funct. Anal., 7(2):215–244, 1997

work page 1997
[4]

Train tracks and automorphisms of free groups

Mladen Bestvina and Michael Handel. Train tracks and automorphisms of free groups. Ann. of Math. (2), 135(1):1–51, 1992

work page 1992
[5]

Neumann, and Ralph Strebel

Robert Bieri, Walter D. Neumann, and Ralph Strebel. A geometric invariant of discrete groups. Invent. Math., 90(3):451–477, 1987

work page 1987
[6]

Hyperbolic automorphisms of free groups

Peter Brinkmann. Hyperbolic automorphisms of free groups. Geom. Funct. Anal. , 10(5):1071–1089, 2000

work page 2000
[7]

Moduli of graphs and automorphisms of free groups

Marc Culler and Karen Vogtmann. Moduli of graphs and automorphisms of free groups. Invent. Math., 84(1):91–119, 1986

work page 1986

Show all 22 references

[8]

The group ﬁxed by a family of injective endomorphisms of a free group, volume 195 of Contemporary Mathematics

Warren Dicks and Enric Ventura. The group ﬁxed by a family of injective endomorphisms of a free group, volume 195 of Contemporary Mathematics. American Mathematical Society, Providence, RI, 1996

1996
[9]

Leininger

Spencer Dowdall, Ilya Kapovich, and Christopher J. Leininger. Dynamics on free-by-cyclic groups. Geom. Topol., 19(5):2801–2899, 2015

2015
[10]

Leininger

Spencer Dowdall, Ilya Kapovich, and Christopher J. Leininger. Endomorphisms, train track maps, and fully irreducible monodromies. Groups Geom. Dyn., 11(4):1179–1200, 2017

2017
[11]

Algorithmic detectability of iwip automorphisms

Ilya Kapovich. Algorithmic detectability of iwip automorphisms. Bull. Lond. Math. Soc. , 46(2):279– 290, 2014

2014
[12]

Detecting fully irreducible automorphisms: a polynomial time algorithm

Ilya Kapovich. Detecting fully irreducible automorphisms: a polynomial time algorithm. Exp. Math., 28(1):24–38, 2019. With an appendix by Mark C. Bell

2019
[13]

Irreducible automorphisms of Fn have north-south dynamics on compactiﬁed outer space

Gilbert Levitt and Martin Lustig. Irreducible automorphisms of Fn have north-south dynamics on compactiﬁed outer space. J. Inst. Math. Jussieu , 2(1):59–72, 2003

2003
[14]

Irreducibility of a Free Group Endomorphism is a Mapping Torus Invariant

Jean Pierre Mutanguha. Irreducibility of a Free Group Endomorphism is a Mapping Torus Invariant. arXiv e-prints, Oct 2019. arXiv:1910.04285

2019 arXiv
[16]

Ideal Whitehead graphs in Out(Fr) II: the complete graph in each rank

Catherine Pfaﬀ. Ideal Whitehead graphs in Out(Fr) II: the complete graph in each rank. J. Homotopy Relat. Struct., 10(2):275–301, 2015

2015
[17]

Dynamics of Irreducible Endomorphisms of Fn

Patrick Reynolds. Dynamics of Irreducible Endomorphisms of Fn. ArXiv e-prints , August 2010. arxiv:1008.3659

2010 arXiv
[18]

Stallings

John R. Stallings. Topology of ﬁnite graphs. Invent. Math., 71(3):551–565, 1983

1983
[19]

Thurston

William P. Thurston. Three-dimensional manifolds, Kleinian groups and hyperbolic geometry. Bull. Amer. Math. Soc. (N.S.) , 6(3):357–381, 1982. 20 JEAN PIERRE MUTANGUHA

1982
[20]

Thurston

William P. Thurston. On the geometry and dynamics of diﬀeomorphisms of surfaces.Bull. Amer. Math. Soc. (N.S.), 19(2):417–431, 1988

1988
[21]

Automorphisms of free groups and outer space

Karen Vogtmann. Automorphisms of free groups and outer space. In Proceedings of the Conference on Geometric and Combinatorial Group Theory, Part I (Haifa, 2000) , volume 94, pages 1–31, 2002

2000
[22]

Contractibility of outer space: reprise

Karen Vogtmann. Contractibility of outer space: reprise. In Hyperbolic geometry and geometric group theory, volume 73 of Adv. Stud. Pure Math. , pages 265–280. Math. Soc. Japan, Tokyo, 2017. Department of Mathematical Sciences, University of Arkansas, F ayetteville, AR Web add...

2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [15]

Hyperbolic Immersions of Free Groups

Jean Pierre Mutanguha. Hyperbolic Immersions of Free Groups. Groups Geom. Dyn. , to appear. arxiv:1809.04761

work page Pith review arXiv

[2] [1]

R. B. Bapat and T. E. S. Raghavan. Nonnegative matrices and applications, volume 64 of Encyclopedia of Mathematics and its Applications . Cambridge University Press, Cambridge, 1997

work page 1997

[3] [2]

Laminations, trees, and irreducible automor- phisms of free groups

M. Bestvina, M. Feighn, and M. Handel. Erratum to: “Laminations, trees, and irreducible automor- phisms of free groups” [Geom. Funct. Anal. 7(2):215–244, 1997]. Geom. Funct. Anal., 7(6):1143, 1997

work page 1997

[4] [3]

Laminations, trees, and irreducible automor- phisms of free groups

Mladen Bestvina, Mark Feighn, and Michael Handel. Laminations, trees, and irreducible automor- phisms of free groups. Geom. Funct. Anal., 7(2):215–244, 1997

work page 1997

[5] [4]

Train tracks and automorphisms of free groups

Mladen Bestvina and Michael Handel. Train tracks and automorphisms of free groups. Ann. of Math. (2), 135(1):1–51, 1992

work page 1992

[6] [5]

Neumann, and Ralph Strebel

Robert Bieri, Walter D. Neumann, and Ralph Strebel. A geometric invariant of discrete groups. Invent. Math., 90(3):451–477, 1987

work page 1987

[7] [6]

Hyperbolic automorphisms of free groups

Peter Brinkmann. Hyperbolic automorphisms of free groups. Geom. Funct. Anal. , 10(5):1071–1089, 2000

work page 2000

[8] [7]

Moduli of graphs and automorphisms of free groups

Marc Culler and Karen Vogtmann. Moduli of graphs and automorphisms of free groups. Invent. Math., 84(1):91–119, 1986

work page 1986