Discrete spectrum for amenable group actions

Guohua Zhang; Ruifeng Zhang; Tao Yu

REVIEW 1 major objections 4 minor 23 references

Discrete spectrum for amenable group actions

T0 review · 1 major / 4 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 For amenable group actions, discrete spectrum is exactly bounded complexity, and for tempered Følner sequences exactly mean equicontinuity.

desk verdict The paper completes the discrete-spectrum/bounded-complexity/mean-equicontinuity package for countable amenable group actions with proofs that mostly hold up; one arithmetic typo in Lemma 2.5 needs a trivial fix. read the letter →

arxiv 1908.08434 v1 pith:35AMBLE7 submitted 2019-08-22 math.DS

Tao Yu , Guohua Zhang , Ruifeng Zhang This is my paper · ORCID

classification math.DS MSC 37A3537A15

keywords discretespectrumamenablegroupactionsalmostperiodicfunctionsboundedcomplexitymeanequicontinuityinthemeasureHammingdistance

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

The paper establishes a single characterization of discrete spectrum for invariant measures of countable amenable group actions on compact metric spaces. It proves that a measure has discrete spectrum exactly when its complexity with respect to any Følner-averaged metric is bounded, and exactly when the Hamming-distance complexity of partition names is bounded for every finite partition, even every two-element partition. For tempered Følner sequences, the same property is shown to be equivalent to mean equicontinuity and to equicontinuity in the mean. The relevance is that discrete spectrum is thereby not only an operator-theoretic notion but also a combinatorial rate-of-growth property, and this equivalence, previously known for integer actions, now holds for all countable amenable groups.

What carries the argument

The central object is the complexity function $C(w_{F_n},\varepsilon)$, which counts the minimum number of balls of radius $\varepsilon/2$ in the Følner-averaged semimetric $w_{F_n}(x,y)=\frac{1}{|F_n|}\sum_{g\in F_n}w(gx,gy)$ needed to cover $X$ up to $\mu$-measure $1-\varepsilon$; bounded complexity means this count stays uniformly bounded in $n$ for each $\varepsilon$. The companion object is the Hamming distance between partition names, $H_{\alpha,F_n}(x,y)$, measuring the fraction of group elements in $F_n$ on which two points fall into different atoms of $\alpha$. The load-bearing mechanism is a two-way passage: an almost periodic function can be approximated by finitely many finite-dimensional invariant subspaces, and the pointwise ergodic theorem for tempered Følner sequences turns that approximation into a finite uniform covering at every scale $\varepsilon$; conversely, if complexity is bounded but the function is not almost periodic, averaging the Hamming distance over a Følner sequence produces many pairwise separated functions, contradicting the finite-dimensionality of the covering. This transfer is what makes statements (1)-(4) equivalent, and the same covering argument yields the mean-equicontinuity half.

What would settle it

Take a nontrivial Bernoulli shift action of a countable amenable group with its product measure, choose any Følner sequence, and compute the minimal covering number $C(d_{F_n},\varepsilon)$ for the metric averages. The theorem predicts this number is unbounded in $n$ for every small $\varepsilon$, because the measure is not discrete spectrum; finding a Følner sequence along which it stays bounded would refute the implication from bounded complexity to discrete spectrum.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.1: for a countable amenable group acting by homeomorphisms on a compact metric space and an invariant Borel probability measure µ, the following are equivalent for any Følner sequence $F_n$: (1) µ has discrete spectrum; (2) µ has bounded complexity with respect to the Følner-averaged metric averages $d_{F_n}$; (3) µ has bounded complexity with respect to Hamming-distance averages of names of every finite partition; (4) the same for every two-element partition. If the Følner sequence is tempered, these are also equivalent to (5) mean equicontinuity and (6) equicontinuity in the mean. The proof operates through the identification of discrete spectrum with almost periodicity of every square-integrable function, and through a two-way transfer between bounded complexity of a function's oscillation semimetric and almost periodicity. The result lifts the earlier equivalence for integer actions to all countable amenable groups.

Load-bearing premise

The load-bearing premise is that the standard characterization of discrete spectrum as the condition that every square-integrable function has a small, finitely approximable orbit, together with Lindenstrauss's pointwise ergodic theorem for averages over tempered Følner sequences, is available; if either imported result gives way, the argument cannot pass from bounded complexity to discrete spectrum.

Editorial extensions

If this is right

An invariant measure of a countable amenable group action is spectrally pure precisely when its metric-averaged complexity is bounded, so discrete spectrum becomes a checkable rate-of-growth property independent of the group's internal structure.
Checking only two-element partitions suffices: if every measurable set has bounded Hamming-name complexity, then every finite partition does and the measure has discrete spectrum.
Under tempered Følner sequences, mean equicontinuity and equicontinuity in the mean are not merely related to discrete spectrum but identical to it; for such sequences all six viewpoints in Theorem 1.1 describe one property.
Every almost periodic function is individually characterized by bounded complexity of its own oscillation semimetric, so the technique applies function-by-function rather than only to the whole measure.
Because every Følner sequence has a tempered subsequence, the spectral/complexity equivalence holds for arbitrary Følner sequences, not only for the tempered ones where the pointwise ergodic theorem is available.

Reading between the lines

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

The two-element partition criterion suggests a computable diagnostic: in a concrete simulation of a countable amenable group action, estimate the growth of Hamming-name complexity for a small collection of indicator functions; bounded growth across a Følner sequence is a numerical signature of discrete spectrum, while unbounded growth points to weak mixing or entropy.
The paper leaves open whether the mean-equicontinuity equivalence survives for non-tempered Følner sequences; constructing a non-tempered Følner sequence and a mean-equicontinuous measure without discrete spectrum, or proving the converse, would settle the boundary of the tempered hypothesis.
Because the proof only needs continuous semimetrics on compact spaces, the same equivalence plausibly extends to actions on compact Hausdorff spaces by replacing the metric with a separating family of continuous semimetrics, or to locally compact amenable groups with a suitable Følner averaging scheme.

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 is real: for countable amenable group actions, the paper proves the equivalence of discrete spectrum, bounded measure complexity (metric and Hamming), and, along tempered Følner sequences, mean equicontinuity and equicontinuity in the mean. Prior work covered Z-actions and abelian groups, so this is the right next step and the proofs are genuinely new. Section 2 does the heavy lifting: bounded complexity implies almost periodicity via Lindenstrauss's pointwise ergodic theorem, and the converse via finite-dimensional invariant subspaces. I checked the counting arguments in Theorem 2.7 and the measure estimates in Lemma 2.4; they are internally consistent.

Soft spots: one genuine flaw in the written proof of Lemma 2.5. The displayed bound reads ≤ 4C + 4C^2/k + 4C^2/(k-1), which tends to 4C, not 0, so the conclusion “by arbitrariness of k > 1” does not follow as written. The intended correction is immediate: multiplying Lemma 2.4's bound by 2C gives (4C+4C^2)/k + 4C^2/(k-1), which does tend to 0. So this is a typo, not a conceptual hole, but it sits in a lemma used for the converse direction of the main equivalence, and a referee should require the fix. Two smaller issues: the abstract states the equicontinuity equivalences without the tempered hypothesis, which the theorem correctly includes; and the reduction of non-tempered Følner sequences to tempered subsequences in Theorem 3.1 is compressed, though the step is standard. No circularity: the central claims are derived from Zimmer's characterization of discrete spectrum and Lindenstrauss's pointwise theorem, with the first author's Z-action paper cited only for context.

Bottom line: this is a solid, citable contribution for ergodic theorists working on amenable group actions, and it deserves a serious referee. I would send it to peer review and ask for a corrected Lemma 2.5 rather than desk-rejecting.

Referee Report

1 major / 4 minor

Summary. The paper studies invariant probability measures for actions of countable discrete amenable groups on compact metric spaces. It defines measure-theoretic complexity functions with respect to Følner averages of a semimetric and with respect to Hamming distances of partition names, and proves Theorem 1.1: for any invariant measure and any Følner sequence, discrete spectrum is equivalent to bounded complexity with respect to the given metric, to bounded complexity for every finite partition, and to bounded complexity for every two-element partition; when the Følner sequence is tempered, these are also equivalent to mean equicontinuity and equicontinuity in the mean. The technical core, Theorem 2.1, characterizes almost periodic L² functions by bounded complexity with respect to the semimetric H(x,y)=|h(x)-h(y)|. The proofs are written out in full, with explicit constants and with careful use of Lindenstrauss's pointwise ergodic theorem and Zimmer's characterization of discrete spectrum via almost periodic functions.

Significance. If correct, the paper gives a genuine extension to countable amenable group actions of the known equivalences for Z-actions among discrete spectrum, bounded measure complexity, bounded partition complexity, mean equicontinuity, and equicontinuity in the mean. This unifies several existing results (Huang--Wang--Ye, Huang--Li--Thouvenot--Xu--Ye, Ferenczi, Vershik--Zatitskiy--Petrov) in a natural generality. The proof strategy is sound: Section 2 develops a self-contained almost-periodicity criterion via complexity, and Section 3 assembles the full equivalence through semimetric and partition arguments. The paper is also careful about the role of tempered Følner sequences, reserving the mean-equicontinuity statements for the tempered case. The main defect is a single local arithmetic slip in Lemma 2.5, which is load-bearing for the converse direction but is straightforwardly corrected.

major comments (1)

[Lemma 2.5 (Section 2)] The displayed estimate before the final sentence of Lemma 2.5 is incorrect as written. The paper obtains ∫|h∘g_{k,m}-h∘g| dµ ≤ (2+2C)/k + 2C/(k−1) from Lemma 2.4, and then writes ∫|h∘g_{k,m}-h∘g|² dµ ≤ 2C∫|h∘g_{k,m}-h∘g| dµ ≤ 4C + 4C²/k + 4C²/(k−1). The last right-hand side tends to 4C as k→∞, not to 0, so the claim 'by arbitrariness of k>1 we have that h is almost periodic' does not follow from the displayed inequality. The intended inequality is (4C+4C²)/k + 4C²/(k−1), which does tend to 0. Since Lemma 2.5 is used in Theorem 2.7 for the converse direction of Theorem 2.1 (and hence of Theorem 1.1), the proof as printed has a genuine gap in a load-bearing step; however, the correction is immediate and does not affect the rest of the argument.

minor comments (4)

[Abstract] There is a typographical error in the abstract: 'meas ures' should be 'measures'.
[Theorem 2.6, proof of bounded complexity] The conclusion 'In particular, µ has bounded complexity w.r.t. {H_{F_n}}' is slightly too quick: the proof constructs sets on which H_{F_n}(x,y)<ε, whereas the complexity definition uses balls of radius ε/2. This is harmless because one can apply the statement with ε/2, but the sentence should say so.
[Theorem 2.7, beginning of proof] The sentence 'Without loss of generality, we may assume that the Følner sequence {F_n:n∈N} is tempered' is terse. A reader should be told that bounded complexity passes to any tempered subsequence (which exists by [12, Proposition 1.4]), and that the conclusion that h is almost periodic is independent of the choice of Følner sequence.
[Lemma 2.2] The statement says 'if and only if', but the proof only gives the direction '⇒'. The converse is indeed trivial by applying the stated property with ε/2 to obtain sets of H_{F_n}-diameter less than ε/2, but a one-sentence justification would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the main equivalences are derived from external results (Zimmer, Lindenstrauss) and prior non-self work; the sole self-citation [22] is contextual, and the only manuscript-level issue is a non-circular arithmetic gap in Lemma 2.5.

full rationale

The derivation is self-contained relative to external theorems and does not reduce to its own inputs. The paper defines discrete spectrum via the closed subspace H_c generated by finite-dimensional G-invariant subspaces (following Zimmer [23]) and then imports Zimmer's theorem that H_c equals the almost-periodic functions; this is an external characterization, not a self-citation. The central equivalences are proved rather than assumed: Theorem 2.1 characterizes almost periodicity by bounded complexity of H_F_n, Theorem 3.1 transfers this to arbitrary continuous semimetrics, Theorem 3.2 handles Hamming distances of partitions, and Theorem 3.3 derives mean equicontinuity and equicontinuity in the mean. The only self-citation is [22], a prior Z-action paper by the first author mentioned in the introduction as background and as a result being generalized; no proof step invokes [22] as a premise, so it is not load-bearing. The main analytic input is Lindenstrauss's pointwise ergodic theorem for tempered Følner sequences [12], which is external and is used exactly where the tempered hypothesis appears. A non-circular proof-quality concern exists in Lemma 2.5: the displayed inequality before 'By arbitrariness of k>1' appears to omit a factor of 1/k in one term, so the written estimate does not tend to 0; however, the intended bound is evidently (4C+4C^2)/k + 4C^2/(k-1), which does tend to 0, and this is a proof gap or typo, not a circularity.

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

No free parameters and no invented entities. The paper is a theorem paper; the constants ε, δ, r, and the thresholds are universally quantified, not fitted to data. The load-bearing inputs are two published theorems (Lindenstrauss [12] for pointwise ergodic convergence along tempered Følner sequences, Zimmer [23] for the structure of almost periodic functions) plus standard measure-theoretic tools. The domain assumption (countable discrete amenable G acting on compact metric X) is stated in the Introduction. The first author's prior work [22] is cited for the Z-action case but is not used as a proof ingredient.

assumptions (5)

standard math Lindenstrauss's pointwise ergodic theorem [12, Theorem 1.3]: for a tempered Følner sequence {F_n}, (1/|F_n|)Σ_{g∈F_n} φ(gx) converges µ-a.e. for every φ ∈ L¹(X,µ).
Invoked in Lemma 2.2 and Theorem 2.6 to convert bounded complexity into the existence of Borel sets on which the Hamming-distance averages are uniformly small; it holds only along tempered Følner sequences, which is precisely why statements (5)-(6) of Theorem 1.1 carry the tempered hypothesis.
standard math Zimmer's theorem [23, Theorem 7.1]: H_c, the closed subspace generated by all finite-dimensional G-invariant subspaces of L²(X,µ), consists exactly of the almost periodic functions.
This identification is the operative definition of discrete spectrum throughout the paper (Introduction; Theorems 2.6, 3.1, 3.2). The footnote on page 2 notes that for non-abelian G, L² need not be spanned by eigenfunctions, so the almost-periodic formulation is doing real work.
standard math Classical measure-theoretic tools: Egorov's theorem, Lusin's theorem, Fubini's theorem, Chebyshev's inequality, compactness of X, density of C(X) in L²(X,µ).
Used pervasively: Egorov-Lusin in Lemma 2.2 and Theorem 2.6 to choose compact sets of large measure with continuity properties; Fubini and Chebyshev in Lemmas 2.4 and 2.7; compactness in the covering arguments of Theorems 2.6 and 3.3.
domain assumption G is a countable infinite discrete amenable group; Følner sequences exist, every Følner sequence has a tempered subsequence [12, Proposition 1.4], and invariant measures exist.
The stated domain of Theorem 1.1. Countability is used in Lemma 2.3 (a single bound over all g ∈ G) and in the countable decomposition of almost periodic functions in Theorem 2.6. Uncountable acting groups fall outside the paper's scope.
domain assumption All three notions (bounded complexity, mean equicontinuity, equicontinuity in the mean) are defined relative to the fixed Følner sequence {F_n}.
The equivalences are claims about the chosen sequence; the mean-equicontinuity half requires {F_n} tempered. The abstract states this half without the qualification, the overclaim recorded in red_flags.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Discrete spectrum for amenable group actions." pith.science (2026). https://pith.science/paper/35AMBLE7

@misc{pith2026190808434,
  author       = {Pith},
  title        = {Pith review of: Discrete spectrum for amenable group actions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/35AMBLE7}},
  note         = {Machine review of arXiv:1908.08434}
}

read the original abstract

In this paper, we study discrete spectrum of invariant measures for countable discrete amenable group actions. We show that an invariant measure has discrete spectrum if and only if it has bounded measure complexity. We also prove that, discrete spectrum can be characterized via measure-theoretic complexity using names of a partition and the Hamming distance, and it turns out to be equivalent to both mean equicontinuity and equicontinuity in the mean.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 22 canonical work pages

[1]

S´ ebastien Ferenczi,Measure-theoretic complexity of ergodic systems, Israel J. Math. 100 (1997), 189–

work page 1997
[2]

Erling Følner, On groups with full Banach mean value , Math. Scand. 3 (1955), 243–254. MR 0079220

work page 1955
[3]

Gabriel Fuhrmann, Maik Gr¨ oger, and Daniel Lenz, The structure of mean equicontinuous group ac- tions, arXiv:1812.10219v1 (2018)

arXiv 2018
[4]

Systems 37 (2017), no

Felipe Garc´ ıa-Ramos,W eak forms of topological and measure-theoretical equicontinuity: relationships with discrete spectrum and sequence entropy, Ergodic Theory Dynam. Systems 37 (2017), no. 4, 1211–

work page 2017
[5]

Felipe Garc´ ıa-Ramos and Brian Marcus, Mean sensitive, mean equicontinuous and almost periodic functions for dynamical systems , Discrete Contin. Dyn. Syst. 39 (2019), no. 2, 729–746. MR 3918192

work page 2019
[6]

Halmos and John von Neumann, Operator methods in classical mechanics

Paul R. Halmos and John von Neumann, Operator methods in classical mechanics. II , Ann. of Math. (2) 43 (1942), 332–350. MR 0006617

work page 1942
[7]

Bounded complexity, mean equicontinuity and discrete spectrum

Wen Huang, Jian Li, Jean-Paul Thouvenot, Leiye Xu, and Xia ngdong Y e,Bounded complexity, mean equicontinuity and discrete spectrum, Ergodic Theory Dynam. Systems, arXiv:1806.02980 (to appe ar)

work page Pith review arXiv
[8]

Wen Huang, Zhiren Wang, and Xiangdong Y e, Measure complexity and M ¨obius disjointness , Adv. Math. 347 (2019), 827–858. MR 3920840

work page 2019

Show all 23 references

[9]

Katok, Lyapunov exponents, entropy and periodic orbits for diffeo morphisms, Inst

Anatole B. Katok, Lyapunov exponents, entropy and periodic orbits for diffeo morphisms, Inst. Hautes ´Etudes Sci. Publ. Math. (1980), no. 51, 137–173. MR 573822

1980
[10]

Systems, arXiv:1608.05636 (to appear)

Daniel Lenz, An autocorrelation and discrete spectrum for dynamical sys tems on metric spaces , Er- godic Theory Dynam. Systems, arXiv:1608.05636 (to appear)

arXiv
[11]

Daniel Lenz and Nicolae Strungaru, Pure point spectrum for measure dynamical systems on locall y compact abelian groups, J. Math. Pures Appl. (9) 92 (2009), no. 4, 323–341. MR 2569181

2009
[12]

Elon Lindenstrauss, Pointwise theorems for amenable groups , Invent. Math. 146 (2001), no. 2, 259–

2001
[13]

1115, Springer-V erlag, Berlin, 1985

Jean Moulin Ollagnier, Ergodic theory and statistical mechanics , Lecture Notes in Mathematics, vol. 1115, Springer-V erlag, Berlin, 1985. MR 781932 (86h:28013)

1985
[14]

Ornstein and Benjamin Weiss, Entropy and isomorphism theorems for actions of amenable groups, J

Donald S. Ornstein and Benjamin Weiss, Entropy and isomorphism theorems for actions of amenable groups, J. Analyse Math. 48 (1987), 1–141. MR 910005 DISCRETE SPECTRUM FOR AMENABLE GROUP ACTIONS 15

1987
[15]

Koukichi Sakai, On compact transformation semigroups with discrete spectr um, Sci. Rep. Kagoshima Univ. (1985), no. 34, 11–17. MR 824951

1985
[16]

London Math

Bruno Scarpellini, Stability properties of ﬂows with pure point spectrum , J. London Math. Soc. (2) 26 (1982), no. 3, 451–464. MR 684559

1982
[17]

V ershik, Dynamics of metrics in measure spaces and their asymptotic i nvariants, Markov Process

Anatoly M. V ershik, Dynamics of metrics in measure spaces and their asymptotic i nvariants, Markov Process. Related Fields 16 (2010), no. 1, 169–184. MR 2664340

2010
[18]

1, 111–135

, Scaled entropy and automorphisms with a pure point spectrum , Algebra i Analiz 23 (2011), no. 1, 111–135. MR 2760149

2011
[19]

V ershik, Pavel B

Anatoly M. V ershik, Pavel B. Zatitskiy, and Fedor V . Petr ov, Geometry and dynamics of admissible metrics in measure spaces , Cent. Eur. J. Math. 11 (2013), no. 3, 379–400. MR 3016311

2013
[20]

79, Springer- V erlag, New Y ork-Berlin, 1982

Peter Walters, An introduction to ergodic theory , Graduate Texts in Mathematics, vol. 79, Springer- V erlag, New Y ork-Berlin, 1982. MR 648108

1982
[21]

Ward and Qing Zhang, The Abramov-Rokhlin entropy addition formula for amenable group actions, Monatsh

Thomas B. Ward and Qing Zhang, The Abramov-Rokhlin entropy addition formula for amenable group actions, Monatsh. Math. 114 (1992), no. 3-4, 317–329. MR 1203977 (93m:28023)

1992
[22]

Differential Equations, arXiv:1807.05868 (to appear)

Tao Y u, Measure-theoretic mean equicontinuity and bounded comple xity, J. Differential Equations, arXiv:1807.05868 (to appear)

arXiv
[23]

Zimmer, Ergodic actions with generalized discrete spectrum , Illinois J

Robert J. Zimmer, Ergodic actions with generalized discrete spectrum , Illinois J. Math. 20 (1976), no. 4, 555–588. MR 0414832 (T. Y u) D EPARTMENT OF MATHEMATICS , S HANTOU UNIVERSITY , S HANTOU 515063, G UANG - DONG , C HINA – AND – S HANGHAI CENTER FOR MATHEMATICAL SCIENCES...

1976

Pith tools

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

[1] [1]

S´ ebastien Ferenczi,Measure-theoretic complexity of ergodic systems, Israel J. Math. 100 (1997), 189–

work page 1997

[2] [2]

Erling Følner, On groups with full Banach mean value , Math. Scand. 3 (1955), 243–254. MR 0079220

work page 1955

[3] [3]

Gabriel Fuhrmann, Maik Gr¨ oger, and Daniel Lenz, The structure of mean equicontinuous group ac- tions, arXiv:1812.10219v1 (2018)

arXiv 2018

[4] [4]

Systems 37 (2017), no

Felipe Garc´ ıa-Ramos,W eak forms of topological and measure-theoretical equicontinuity: relationships with discrete spectrum and sequence entropy, Ergodic Theory Dynam. Systems 37 (2017), no. 4, 1211–

work page 2017

[5] [5]

Felipe Garc´ ıa-Ramos and Brian Marcus, Mean sensitive, mean equicontinuous and almost periodic functions for dynamical systems , Discrete Contin. Dyn. Syst. 39 (2019), no. 2, 729–746. MR 3918192

work page 2019

[6] [6]

Halmos and John von Neumann, Operator methods in classical mechanics

Paul R. Halmos and John von Neumann, Operator methods in classical mechanics. II , Ann. of Math. (2) 43 (1942), 332–350. MR 0006617

work page 1942

[7] [7]

Bounded complexity, mean equicontinuity and discrete spectrum

Wen Huang, Jian Li, Jean-Paul Thouvenot, Leiye Xu, and Xia ngdong Y e,Bounded complexity, mean equicontinuity and discrete spectrum, Ergodic Theory Dynam. Systems, arXiv:1806.02980 (to appe ar)

work page Pith review arXiv

[8] [8]

Wen Huang, Zhiren Wang, and Xiangdong Y e, Measure complexity and M ¨obius disjointness , Adv. Math. 347 (2019), 827–858. MR 3920840

work page 2019