Pith. sign in

REVIEW 2 major objections 4 minor 13 references

Smoothing topological pseudo-isotopies of 4-manifolds

T0 review · 2 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper proves that the Kirby–Siebenmann invariant completely controls when a topological pseudo-isotopy of a 4-manifold can be smoothed, gives broad classes of fundamental groups where it always can, and constructs the first examples…

desk verdict Theorem F is the first real counterexample for smoothability of topological pseudo-isotopies in dimension 4, but the proof of the key square (19) has an unchecked compatibility issue that needs referee attention. read the letter →

arxiv 2507.16984 v1 pith:OAJSEDYD submitted 2025-07-22 math.GT

Patrick Orson , Mark Powell , Oscar Randal-Williams This is my paper · ORCID
classification math.GT MSC 57K4057N3757R52
keywords 4-manifoldspseudo-isotopyKirby–Siebenmanninvariantsmoothingtheorysurgerydiffeomorphismgroupsstableisotopyblockautomorphisms
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 asks when a topological pseudo-isotopy between diffeomorphisms of a closed 4-manifold can be smoothed, meaning replaced, up to topological pseudo-isotopy, by a smooth pseudo-isotopy. It proves that the entire obstruction is a single cohomology class, the Kirby–Siebenmann invariant $KS(F)\in H^2(X;\mathbb{Z}/2)$, and that this class vanishes exactly when the pseudo-isotopy is topologically isotopic, relative to the boundary, to a smooth pseudo-isotopy. For manifolds whose fundamental group lies in a broad class — including trivial, free, finite odd order, and knot groups — the paper shows every topological pseudo-isotopy to the identity is smooth. In contrast, it constructs the first closed smooth 4-manifolds carrying a diffeomorphism that is topologically pseudo-isotopic to the identity but not smoothly so, and hence not smoothly stably isotopic. The counterexample is built from an aspherical 4-manifold together with enough $S^2\times S^2$ summands, and the proof turns on a delicate comparison of PL normal invariants under Postnikov truncation.

What carries the argument

The central object is the relative Kirby–Siebenmann invariant $KS(F)\in H^2(X;\mathbb{Z}/2)$, obtained by putting a smooth structure on the boundary of the 6-manifold $X\times I\times I$ using $F$ and taking the relative Kirby–Siebenmann class; it is a homomorphism from the group $Q(X)$ of topological pseudo-isotopies from the identity to diffeomorphisms, modulo topological pseudo-isotopy. The proof machinery combines block automorphism groups, the surgery exact sequence, the algebraic L-theory assembly map, and the 2-local splitting of L-spectra. For the counterexample, a load-bearing diagram compares PL normal invariants under the Postnikov truncation $t\colon X\to A$; its commutativity is proved using stable framings, an $S$-theory fundamental class, umkehr maps, and a four-dimensional argument about infinite loop space structures on truncations of $\Omega^\infty S^0$.

What would settle it

Take $A = \Sigma \times S^1$ with $\Sigma$ a hyperbolic 3-manifold having 2-torsion in $H_1(\Sigma;\mathbb{Z})$, and $X = A\#g(S^2\times S^2)$ with $g$ as in Definition 7.4. Evaluate square (19) on the loop of homotopy automorphisms of $A$ generated by the circle action: if the PL normal invariant obtained via the umkehr map $t_!$ differs from the one obtained by first truncating the loop to $A$, the diagram does not commute and the proof of Theorem F collapses.

Watch

Extended reading notes

Core claim

The central discovery is that the smoothing problem for topological pseudo-isotopies of 4-manifolds is governed by the Kirby–Siebenmann invariant: the homomorphism $KS\colon Q(X)\to H^2(X;\mathbb{Z}/2)$ is injective, and it detects exactly whether a pseudo-isotopy is topologically isotopic rel boundary to a smooth pseudo-isotopy. On the positive side, when the fundamental group $\pi$ satisfies the algebraic hypotheses that the assembly map $I_2\colon H_2(\pi;\mathbb{Z}_{(2)})\to L_6(\mathbb{Z}[\pi])_{(2)}$ is trivial and $H_1(\pi;\mathbb{Z}_{(2)})$ is torsion-free, the invariant vanishes on all topological pseudo-isotopies to the identity, so topological pseudo-isotopy implies smooth pseudo-isotopy; adding $\mathrm{Wh}_2(\pi)=0$ upgrades this to stable smooth isotopy. On the negative side, after stabilising by connected sums with enough copies of $S^2\times S^2$ every class in $H^2$ arises as $KS(F)$, and for $X = A\#g(S^2\times S^2)$ with $A$ aspherical, stably framed, and with 2-torsion in $H_1$, this yields a diffeomorphism whose every topological pseudo-isotopy has nonzero Kirby–Siebenmann invariant. This gives the first known examples of diffeomorphisms of closed smooth 4-manifolds that are topologically pseudo-isotopic to the identity but not smoothly pseudo-isotopic.

Load-bearing premise

The counterexample rests on assuming there is a building block A whose homotopy symmetries are all realised by diffeomorphisms and that a certain large diagram comparing normal invariants really commutes; if either fails, the construction collapses.

Editorial extensions

If this is right

  • For any 4-manifold with fundamental group in the listed class, topological pseudo-isotopy to the identity implies smooth pseudo-isotopy to the identity.
  • When the fundamental group additionally has vanishing $\mathrm{Wh}_2$, such diffeomorphisms are smoothly stably isotopic to the identity, extending the previously known positive cases beyond free groups.
  • The diffeomorphisms produced in Theorem F are not smoothly stably isotopic to the identity, so they are genuine counterexamples to the smoothing principle for pseudo-isotopy.
  • After stabilising by enough connected sums with $S^2\times S^2$, every class in $H^2(X;\mathbb{Z}/2)$ is realised as the Kirby–Siebenmann invariant of some topological pseudo-isotopy, so the invariant is the only obstruction at the stable level.

Reading between the lines

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

  • If a future theory upgrades topological pseudo-isotopies of non-simply-connected 4-manifolds to topological isotopies, the examples here would likely become diffeomorphisms that are topologically isotopic to the identity but not smoothly stably isotopic, answering the paper's motivating question negatively.
  • The paper's split between smoothability and non-smoothability tracks 2-torsion in $H_1$: Theorem C excludes it, while the counterexample requires it, suggesting that the Bockstein kernel is the natural place to look for further obstructions.
  • The proof's reliance on the vanishing of the stable stems $\pi_4^s$ and $\pi_5^s$ suggests that analogous smoothing failures might be detectable in other dimensions by the same combination of stable framings, umkehr maps, and truncated loop spaces, though the paper does not claim this.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper studies when a topological pseudo-isotopy of a closed smooth 4-manifold can be smoothed, introducing a Kirby–Siebenmann invariant KS(F) in H2(X; Z/2) and proving that its vanishing characterizes when F is topologically isotopic, rel. boundary, to a smooth pseudo-isotopy (Theorem A). Under hypotheses on π1(X) involving the vanishing of an L-theoretic assembly map I2 and torsion-freeness of H1(π; Z(2)), the authors prove that topological pseudo-isotopy implies smooth pseudo-isotopy (Theorem C), with consequences for stable smooth isotopy (Corollary D). They also prove a stabilization surjectivity statement (Theorem E) and then construct the first examples of diffeomorphisms of closed smooth 4-manifolds that are topologically pseudo-isotopic to the identity but not smoothly pseudo-isotopic (Theorem F). The proofs use smoothing theory, the surgery exact sequence, L-spectra and Ranicki–Sullivan duality, and are organized through a block automorphism reformulation.

Significance. If the main results are correct, Theorem F answers a natural and previously open question in 4-manifold theory, and Theorem A gives a clean criterion for smoothing topological pseudo-isotopies. The paper is notable for its explicit and detailed use of surgery-theoretic machinery rather than heuristic arguments: there are no fitted parameters, the counterexample is built from an explicit aspherical 4-manifold Σ × S1, and several auxiliary claims are proved directly (e.g., Lemma 5.3 and Lemma 7.17) instead of being quoted. The central Theorem F, however, depends on a delicate commutativity statement, Proposition 7.20, and the current proof of that statement has a gap; this is the main reason the manuscript needs revision before acceptance.

major comments (2)
  1. [§7.5.4, Proposition 7.20 and diagram (19)/(28)] The commutativity of square (19) is load-bearing for Theorem F, and the proof in Proposition 7.20 does not currently establish it. The authors compare the umkehr maps tS_! and tG_! after replacing Ω∞+1S0 by Ωτ≤3Ω∞_0S0, and they argue that the difference tS_! − tG_! vanishes on the truncated spaces. However, the actual square (19) involves the basepoint translation (T0,1)_*, which the authors themselves note is not an equivalence of infinite loop spaces. To conclude (T0,1)_* ∘ tS_! = tG_! ∘ (T0,1)_*, one must know that the fibre-sequence identification ρ between τ≤3Ω∞_0S0 and τ≤3SG is compatible with the truncation of T0,1. The proof never checks this; uniqueness of infinite loop structures on the base and fibre only shows that the two umkehr maps agree with respect to some identification of the truncated spaces, not that this identification is the one induced by T0,1. The homotopy-group discussion is also internally inconsistent: τ≤1Ω∞_0S0 is said to have π0 = Z, but if Ω∞_0 denotes the identity component then π0 = 0, while if it denotes the full Ω∞S0 then τ≤1 is not equivalent to τ≤1SG. This makes the short exact sequence (28) and the subsequent diagram chase inconclusive as written.
  2. [§2.1 and exact sequence (1)] The block automorphism model is introduced with the sentence 'Eliding some technical details regarding collars in the smooth case' and the paper relies on this model for the exact sequence (1), for the definition of Q(X), and for the reformulation in Theorem 2.3. Since the collar issue is explicitly acknowledged rather than proved, the manuscript should either state the precise collar normalization used or give a specific reference/location in [HLLRW21, Section 2.2] that supplies the missing details. This is likely a local fix, but as written it leaves a technical gap in the foundations of the block model.
minor comments (4)
  1. [§7.3] The phrase 'We use the unkehr construction' should read 'umkehr construction', as elsewhere in the paper.
  2. [References] The reference [BLR75] is listed as 'Burgelea–Lashof–Rothenberg' but the correct spelling is 'Burghelea–Lashof–Rothenberg'.
  3. [§7.5.4, Proposition 7.20] The notation τ[2,3]Ω∞_0S0 is used without explanation; the authors should clarify that it denotes the relevant homotopy fibre, as this is essential for the fibre-sequence argument in (28).
  4. [Diagram (15)] The labels (17) and (19) for regions of diagram (15) duplicate equation numbers used elsewhere in the paper; renaming these regions would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main theorems are derived from smoothing theory, surgery theory, and external uniqueness results; the only self-citations are technical or background.

full rationale

The derivation chain is self-contained. Theorem A follows from Kirby–Siebenmann smoothing theory: the equivalences in Theorem 3.4 are obtained from the existence of a smooth structure on X×I×I and the concordance-implies-isotopy theorem, neither of which assumes the conclusion. Theorem C is proved from the geometric surgery exact sequence, Ranicki–Sullivan duality, and the Taylor–Williams splitting, with I2=0 an input about L-theory assembly rather than about pseudo-isotopies. Theorem E realizes an arbitrary class x by converting it into a smoothing of X×I via the identification H3(X×I,∂;Z/2)≅[X×I/∂,Top/O] and then stabilizing by Quinn's stable s-cobordism theorem; the equality KS(F)=(x,0) is a computation from that construction, not a fitted parameter. Theorem F reduces to commutativity of diagram (15), and the critical square (19) is addressed in Proposition 7.20 using stable homotopy truncation and the external uniqueness theorem of Mathew–Stojanoska; the paper explicitly notes that the basepoint translation T0,1 is not an infinite loop equivalence and works around that difficulty. The only self-citations, [HLLRW21] for a technical correction to smooth block automorphism groups and [GGH+23] as background, are not load-bearing for the main argument. Possible correctness concerns about the compatibility of the identification ρ with the truncation of T0,1, or about the displayed homotopy groups of τ≤1Ω∞0S0, would be gaps in the proof, not circular reductions; no equation in the paper is equivalent to its input by construction.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

No free parameters are fitted. The axioms listed are either standard background from surgery theory and smoothing theory or domain assumptions that the paper explicitly verifies in examples. No new physical or mathematical entities are postulated.

assumptions (8)
  • standard math Surgery exact sequence and algebraic L-theory (Wall, Ranicki)
    Used throughout Sections 5-7 for structure sets and normal invariants; standard foundational input.
  • standard math Kirby-Siebenmann smoothing theory for manifolds of dimension at least 5
    Theorem 3.1, used to show KS(F)=0 iff a smooth structure extends; the dimension of X x I x I is 6.
  • domain assumption Gabai's theorem that zero Hatcher-Wagoner obstruction implies stably smoothly isotopic
    Invoked in Corollaries B and D; Gabai's result is a recent preprint (arXiv:2212.02004).
  • domain assumption Galvin-Nonino's construction of topological Hatcher-Wagoner invariants in dimension 4
    Cited as [GN25, Theorem 1.1], a recent preprint, used to factor SigmaDiff through SigmaTop.
  • standard math Quinn's stable s-cobordism theorem for 4-manifolds
    Used in proof of Theorem E to realise all H2 classes as KS of pseudo-isotopies after stabilising.
  • standard math Taylor-Williams splitting of 2-local L-spectra and Morgan-Sullivan classes
    Used in Lemmas 5.3, 5.5, 5.6 to compute the Kirby-Siebenmann class in terms of cohomology operations.
  • standard math Mathew-Stojanoska full faithfulness of Omega^infty for spectra with homotopy concentrated in degrees 2 and 3
    Used in Proposition 7.20 to identify infinite loop space structures on tau_[2,3] SG.
  • domain assumption Existence of an aspherical, stably framable closed 4-manifold A with 2-torsion in H1 and with pi1(Diffeo+(A)) -> pi1(hAut+(A)) surjective
    Conditions (i)-(iv) of Definition 7.4; verified for A = Sigma x S^1 in Example 7.3.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Smoothing topological pseudo-isotopies of 4-manifolds." pith.science (2026). https://pith.science/paper/OAJSEDYD

@misc{pith2026250716984,
  author       = {Pith},
  title        = {Pith review of: Smoothing topological pseudo-isotopies of 4-manifolds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OAJSEDYD}},
  note         = {Machine review of arXiv:2507.16984}
}
abstract

Given a closed, smooth 4-manifold $X$ and self-diffeomorphism $f$ that is topologically pseudo-isotopic to the identity, we study the question of whether $f$ is moreover smoothly pseudo-isotopic to the identity. If the fundamental group of $X$ lies in a certain class, which includes trivial, free, and finite groups of odd order, we show the answer is always affirmative. On the other hand, we produce the first examples of manifolds $X$ and diffeomorphisms $f$ where the answer is negative. Our investigation is motivated by the question, which remains open, of whether there exists a self-diffeomorphism of a closed 4-manifold that is topologically isotopic to the identity, but not stably smoothly isotopic to the identity.

Figures

Figures reproduced from arXiv: 2507.16984 by the authors.

Figure 1
Figure 1. Schematic for X×I×I with the X direction suppressed. Maps defining the equivalence relation (f0, F0) ∼ (f1, F1) for Q(X) are depicted. For [F] ∈ Q(X), we explained in the introduction how to produce a smooth manifold ∂(X × I × I)F by pulling back the smooth structure σ on X × I × {1} using the homeomorphism F. Then X × I × I is a topological 6-manifold with a given smooth structure on its boundary so has a Kirby–Sie… view at source ↗
Figure 2
Figure 2. We denote the standard smooth structure on X ×I by σ. Suppressing the X direction, we indicate the stacking proposed, together with the smooth structures on the various copies of X × I that make up the boundaries of the two copies of X × I × I. The additivity of sm with respect to ∪ is due to the Eckmann-Hilton argument explained above. There is an identification of topological manifolds with smoothed boundaries F ×… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 12 canonical work pages

  1. [1]

    [Akb84] Selman Akbulut, A fake 4-manifold, Four-manifold theory (Durham, N.H., 1982), 1984, pp. 75–141. [Akb85] Selman Akbulut, On fake S3 ˜×S1 #S2×S2, Combinatorial methods in topology and algebraic geometry (Rochester, N.Y., 1982), 1985, pp. 281–286. [Ati61] Michael F. Atiyah, Thom complexes, Proc. London Math. Soc. (3) 11 (1961), 291–310. [BFL14] A. Ba...

    work page 1961

  2. [6]

    [GJ09] Paul G

    Preprint, available at arXiv:2311.11196. [GJ09] Paul G. Goerss and John F. Jardine, Simplicial homotopy theory , Modern Birkh¨ auser Classics, Birkh¨ auser Verlag, Basel,

    arXiv

  3. [8]

    A note on homotopy and pseudoisotopy of diffeomorphisms of $4$-manifolds

    Preprint, available at arXiv:2402.16167. [Kra22] Manuel Krannich, A homological approach to pseudoisotopy theory. I , Invent. Math. 227 (2022), no. 3, 1093–1167. [Kre79] Matthias Kreck, Isotopy classes of diffeomorphisms of (k − 1)-connected almost-parallelizable 2k- manifolds, Algebraic topology, Aarhus 1978 (Proc. Sympos., Univ. Aarhus), 1979, pp. 643–6...

    work page Pith review arXiv 2022

  4. [1975]

    Brumfiel and I

    [BM76] G. Brumfiel and I. Madsen, Evaluation of the transfer and the universal surgery classes , Invent. Math. 32 (1976), no. 2, 133–169. [Bro72] William Browder, Surgery on simply-connected manifolds, Springer-Verlag, New York-Heidelberg,

    work page 1976

  5. [1979]

    [MS16] Akhil Mathew and Vesna Stojanoska, The Picard group of topological modular forms via descent theory, Geom. Topol. 20 (2016), no. 6, 3133–3217. [MS74] John W. Morgan and Dennis P. Sullivan, The transversality characteristic class and linking cycles in surgery theory, Ann. of Math. (2) 99 (1974), 463–544. [MTW80] Ib Madsen, Laurence R. Taylor, and Br...

    work page 2016

  6. [1982]

    Inst., Univ

    [Qui70] Frank Quinn, A geometric formulation of surgery, Topology of Manifolds (Proc. Inst., Univ. of Georgia, Athens, Ga., 1969), 1970, pp. 500–511. [Qui83] Frank Quinn, The stable topology of 4-manifolds, Topology Appl. 15 (1983), no. 1, 71–77. [Qui86] Frank Quinn, Isotopy of 4-manifolds, J. Differential Geom. 24 (1986), no. 3, 343–372. [Ran79] Andrew R...

    work page 1983

  7. [1995]

    [Ste75] Ronald J

    Corrected reprint of the 1966 original. [Ste75] Ronald J. Stern, On topological and piecewise linear vector fields , Topology 14 (1975), no. 3, 257–269. [Ste77] Elliott Stein, Surgery on products with finite fundamental group , Topology 16 (1977), no. 4, 473–493. [Swi02] Robert M. Switzer, Algebraic topology—homotopy and homology , Classics in Mathematics...

    work page 1975

  8. [1997]

    Conf., Univ

    [TW79] Laurence Taylor and Bruce Williams, Surgery spaces: formulae and structure , Algebraic topology, Waterloo, 1978 (Proc. Conf., Univ. Waterloo, Waterloo, Ont., 1978), 1979, pp. 170–195. [Wal64a] C. Terence C. Wall, Diffeomorphisms of 4-manifolds, J. London Math. Soc. 39 (1964), 131–140. [Wal64b] C. Terence C. Wall, On simply-connected 4-manifolds, J....

    work page 1964

Show all 13 references
  1. [2002]

    [Thu97] William P

    Reprint of the 1975 original [Springer, New York]. [Thu97] William P. Thurston, Three-dimensional geometry and topology. Vol

    1975

  2. [2022]

    [Gal24] Daniel A

    Preprint, available at arXiv:2212.02004. [Gal24] Daniel A. P. Galvin, The Casson-Sullivan invariant for homeomorphisms of 4-manifolds ,

    arXiv

  3. [2023]

    [CS71] Sylvain E

    Preprint, available at arXiv:2303.12857. [CS71] Sylvain E. Cappell and Julius L. Shaneson, On four dimensional surgery and applications , Comment. Math. Helv. 46 (1971), 500–528. [CS76] Sylvain E. Cappell and Julius L. Shaneson, Some new four-manifolds , Ann. Math. (2) 104 (19...

    1971 arXiv

  4. [2024]

    [GGH+23] David Gabai, David T

    Preprint, available at arXiv:2405.07928. [GGH+23] David Gabai, David T. Gay, Daniel Hartman, Vyachslav Krushkal, and Mark Powell, Pseudo-isotopies of simply connected 4-manifolds ,

    arXiv

  5. [2025]

    [Gom84] Robert E

    Preprint, available at arXiv:2506.11905. [Gom84] Robert E. Gompf, Stable diffeomorphism of compact 4-manifolds, Topology Appl. 18 (1984), no. 2-3, 115–120. [Got65] Daniel H. Gottlieb, A certain subgroup of the fundamental group , Amer. J. Math. 87 (1965), 840–856. [HLLR W21] F...

    1984 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.