Pith. sign in

REVIEW 2 major objections 3 minor 28 references

On the ${\Ext}^2$-problem for Hilbert spaces

T0 review · 2 major / 3 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 The paper proves that ℓ2 has a nontrivial two-step extension, built by splicing two Kalton–Peck twisted Hilbert spaces.

desk verdict The main result is real: an explicit nontrivial Ext^2(ℓ2,ℓ2), built from a spliced Kalton–Peck sequence, with checkable estimates that should withstand refereeing. read the letter →

arxiv 1908.06529 v2 pith:CQGK3BLQ submitted 2019-08-18 math.FA

Félix Cabello Sánchez , Jesús M.F. Castillo , Willian H. G. Corrêa , Valentin Ferenczi , Ricardo García This is my paper · ORCID
classification math.FA MSC 46B2546M18
keywords Ext^2problemHilbertspaceKalton-Peckcentralizerquasilinearmaptwistedsumfour-spacelocalconvexity
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 proves that the second extension group $\operatorname{Ext}^2(\ell_2,\ell_2)$ is nonzero in the category of Banach spaces, settling a problem that earlier work on twisted Hilbert spaces had left open. The counterexample is explicit: two short exact sequences, one the classical Kalton–Peck twisted Hilbert space $Z_2$ and the other a 'chunked' relative $\widetilde{Z}_2$ built from a partition of the integers, are spliced into a four-term sequence $0\to \ell_2\to \widetilde{Z}_2\to Z_2\to \ell_2\to 0$. The splice is shown to be nontrivial by proving that the concatenation of the corresponding centralizers, $\widetilde{\Omega}\Omega$, cannot be trivial, through a dyadic-partition estimate that contradicts any constant bound. The same methods give $\operatorname{Ext}^2(\ell_1,\mathbb{K})\neq 0$ in the category of quasi Banach spaces, a negative solution to the four-space problem for local convexity. A reader should care because $\ell_2$ is the most symmetric of Banach spaces, and even there the extension problem does not collapse at the second step.

What carries the argument

The machinery is the homological algebra of quasi Banach spaces expressed through quasilinear maps and centralizers. A short exact sequence is encoded by a homogeneous quasilinear map $\Phi:X\to Y$; splicing two such sequences $Y\hookrightarrow E_1\twoheadrightarrow E$ and $E\hookrightarrow E_2\twoheadrightarrow X$ gives a four-term sequence, and the criterion (Lemma 3.3) says it is trivial in $\operatorname{Ext}^2(X,Y)$ exactly when there is a witness $H:X\to Y$ satisfying the estimate involving $\Phi(\Psi(x+y)-\Psi(x)-\Psi(y))$. The key identity is the Kalton–Peck centralizer $\Omega(x)=x\log(\|x\|/|x|)$ together with its chunked version $\widetilde{\Omega}$; the specific mechanism that makes the contradiction work is the averaging of $H$ over the compact group of real unitaries, which makes $H$ commute with that group and therefore preserve supports. With support preservation, the proof reduces to an estimate over nodes of a dyadic tree (Lemma 4.4) that directly contradicts the explicit value of $\sum_\alpha(\widetilde{\Omega}(x_\alpha)-\widetilde{\Omega}(y_\alpha))$ for two lexicographic adequate partitions (Lemma 4.5).

What would settle it

Compute the norm in Lemma 4.5 for the two lexicographic adequate partitions with parameters r=s (so k=2r): the paper obtains $\lVert\sum_{\alpha\in T_k}(\widetilde{\Omega}(x_\alpha)-\widetilde{\Omega}(y_\alpha))\rVert_2 = \log(2)\,r^2\,\sqrt{2^k}$, while Lemma 4.4 would force this norm to be at most $2Kk\sqrt{2^{k-1}}$ for a constant $K$ independent of $k$. After factoring out $\sqrt{2^k}$, the left-hand side grows like $r^2$ while the bound grows like $k=2r$; the ratio grows linearly, so no finite $K$ can hold for all $r$. Numerically checking this ratio for $r=1,2,3$ would confirm the contradiction, and an explicit witness $H$ providing a $K$ that does not grow with $r$ would refute the theorem.

Watch

Extended reading notes

Core claim

The central claim, proved as Theorem 4.6, is that the spliced four-term exact sequence $$0 \to \ell_2 \to \widetilde{Z}_2 \to Z_2 \to \ell_2 \to 0$$ is not trivial in $\operatorname{Ext}^2(\ell_2,\ell_2)$: no finite chain of commutative diagrams of Banach spaces and operators connects it to the zero sequence. Here $Z_2$ is the Kalton–Peck twisted Hilbert space, the middle space of the short exact sequence induced by the centralizer $\Omega(x)=x\log(\|x\|/|x|)$, and $\widetilde{Z}_2$ is the space induced by the chunked centralizer $\widetilde{\Omega}(x)=\sum_i x_i\log(\|x\|/\|x_i\|)$ associated to a partition of the integers. Nontriviality is established by showing that the concatenation $\widetilde{\Omega}\Omega$ admits no quasilinear extension; the proof forces a witness $H$ to commute with the real unitary group and hence preserve supports, then compares sums of $H$-differences over nodes of two adequately chosen dyadic partitions with an explicit computation of $\sum_\alpha(\widetilde{\Omega}(x_\alpha)-\widetilde{\Omega}(y_\alpha))$, and the two estimates are incompatible. The paper also proves $\operatorname{Ext}^2(\ell_1,\mathbb{K})\neq 0$ in the quasi-Banach category, via the Ribe space, solving the four-space problem for local convexity.

Load-bearing premise

The argument relies on the assertion, stated without proof in Section 4, that every centralizer on a sequence space is strongly equivalent to one commuting with the real unitary group; if that equivalence failed for the Kalton–Peck or chunked centralizers, the support-preserving witness needed for the partition estimates could not be guaranteed.

Editorial extensions

If this is right

  • $\operatorname{Ext}^2_{\mathcal{B}}(\ell_2,\ell_2)\neq 0$, so the nontrivial sequence lives in the Banach category, not merely the quasi Banach category (Corollary 4.7).
  • If $X$ and $Y$ are Banach spaces containing $\ell_2^n$ uniformly complemented, then $\operatorname{Ext}^2_{\mathcal{B}}(X,Y)\neq 0$ (Corollary 5.1).
  • In the category of quasi Banach spaces, $\operatorname{Ext}^2(\ell_1,\mathbb{K})\neq 0$, giving a negative answer to the four-space problem for local convexity (Proposition 5.9).
  • There exists an embedding $u:\ell_2\to C[0,1]$ and an operator $v:\ell_2\to C[0,1]/u[\ell_2]$ that cannot be extended to the Kalton–Peck space $Z_2$ (Corollary 5.2).
  • For $0<p<\infty$, $\operatorname{Ext}^2_{\ell_\infty}(\ell_p,\ell_p)\neq 0$ in quasi Banach $\ell_\infty$-modules, and for $1\le p<\infty$ the Banach-space group is also nonzero (Section 5.6).

Reading between the lines

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

  • The reflection machinery of Section 5.6 suggests that the same construction, transported from $\ell_2$ to $\ell_p$ for $1<p<\infty$, yields $\operatorname{Ext}^2_{\mathcal{B}}(\ell_p,\ell_p)\neq 0$ for every $p$ in that range, even though splicing a centralizer with itself is always trivial.
  • The averaging trick over the real unitary group is a general principle: any witness inequality preserved under a compact group of module symmetries can be made equivariant, so similar contradictions might be built for other homogeneous spaces with large symmetry groups.
  • The finite-dimensional partition sums in Lemma 4.5 give a family of explicit inequalities that could be checked numerically for small $k$; a test that ever produced a constant $K$ not growing with $k$ would localize a flaw in the support-preservation step rather than in the partition arithmetic.
  • The paper's method decouples the two difficulties of constructing a nontrivial four-term sequence: the construction is purely from centralizer splices, while the nontriviality is certified by a combinatorial estimate that does not require incomparability of subspaces or duals, suggesting higher $\operatorname{Ext}^n$ problems may be approachable by similar finite estimates.
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.

Referee Report

2 major / 3 minor

Summary. The paper proves that Ext^2(ℓ2, ℓ2) is nonzero in the category of Banach spaces, solving a sharpened version of Palamodov's problem and a second-order version of Palais's problem. The counterexample is a four-term exact sequence 0 → ℓ2 → ~Z2 → Z2 → ℓ2 → 0 obtained by splicing the Kalton–Peck sequence with a vector-valued, 'chunked' version of it. Section 3 develops criteria, using quasilinear maps and a witness function H, for deciding when a spliced sequence is trivial in Ext^2. Section 4 gives the core argument: under an adequate dyadic partition, a nontriviality estimate for the chunked centralizer contradicts an upper bound that would follow from triviality of the spliced sequence. Section 5 draws several applications, including nonvanishing of Ext^2(X,Y) for Banach spaces containing ℓ2^n uniformly complemented, and a proof that Ext^2(ℓ1,K) ≠ 0 in the category of quasi Banach spaces, solving the four-space problem for local convexity.

Significance. If the proof is sound after the corrections noted below, this is a substantial advance in the homological theory of Banach spaces: it answers a longstanding open question and provides the first nontrivial element in Ext^2(ℓ2, ℓ2). The paper is largely self-contained, the counterexample is explicit, and the main estimates are checkable. The criteria in Section 3 and the applications in Section 5 give the result genuine mathematical reach beyond the single counterexample. No parameters are fitted to force the conclusion, and the central construction is reproduced in the text.

major comments (2)
  1. [Section 4, Lemma 4.5 and the paragraph preceding it] The identity 'if x has exactly q nonzero chunks and they all have the same norm, then ~Ω_{m,n}(x)=log(q)x' is false by a factor of 1/2. From (5) with p=2 and φ(t)=t, if q chunks have equal norm a, then ||x||=√q·a, so ~Ω(x)=∑ x_i log(||x||/||x_i||) = log(√q) x = (1/2)log(q)x. Consequently the four displayed formulas in Lemma 4.5 carry an extra factor of 2, and the final norm should be (log2/2) rs √(2^k), not log2 rs √(2^k). The contradiction in Theorem 4.6 still works after this correction, because the lower bound remains of order k^2 2^k against an upper bound of order k 2^k, but the lemma and the displayed equality in the proof of Theorem 4.6 must be corrected.
  2. [Section 4, Corollary 4.2] The proof of Corollary 4.2 does not follow directly from Lemma 3.3(b) by homogeneity alone. For disjoint equal-norm x and y, Lemma 3.3(b) gives ||H(x+y)-H(x)-H(y) - (log2/2)Φ(x+y)|| ≤ 2K||x||. To obtain the stated inequality with Φ(x+y) in place of (log2/2)Φ(x+y), one must replace H by (2/log2)H and adjust K accordingly. This rescaling is harmless, and the estimate is correct after the change, but it should be stated explicitly rather than left implicit.
minor comments (3)
  1. [Section 4, before Lemma 4.1] The assertion 'Every centralizer defined on a sequence space is strongly equivalent to one that commutes with U' is stated without proof or reference. It is not needed for Theorem 4.6, since the centralizers Ω and ~Ω already commute with U by formulas (4) and (5); please either supply a short proof (e.g., by averaging over U) or clearly mark the statement as a remark not used in the main argument.
  2. [Section 3, Lemma 3.2] The proof of Lemma 3.2 is left 'to the patient reader.' Since this lemma is the operative criterion used in the rest of the paper, please include a brief proof or at least a precise reference to the completed arguments of Lemma 3.1.
  3. [Throughout (typesetting and notation)] The displayed commutative diagrams contain many OCR-like artifacts (e.g., '/d47' and '/d38' glyphs) that make them difficult to read, and the display in Lemma 4.5 showing division by log 2 is ambiguous. These should be cleaned up in the final version. In particular, clarify in Lemma 4.5 that the computation divides the sum by log 2 before reinserting the factor.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central Ext^2(ell_2, ell_2) counterexample is constructed from explicit centralizers and proved by a direct estimate, with no fitted input or self-referential reduction.

full rationale

The main theorem (Theorem 4.6, Ext^2(ell_2, ell_2) != 0) is self-contained. The two centralizers Omega and ~Omega are given by explicit formulas (4) and (5); Lemma 4.5 computes the norm of the difference sum directly from the definition of the chunked centralizer as log(2) r s sqrt(2^k), while Lemma 4.4 derives the opposite bound 2 K k sqrt(2^(k-1)) from the witness H supplied by Lemmas 3.3 and 4.1. The contradiction between the two estimates is a genuine mathematical argument, not a restatement of the desired conclusion. The Kalton-Peck map is an external input, cited to [21] and reproduced in formula (4), so the external content is explicit and not smuggled in by citation. The unproved assertion before Lemma 4.1 that every centralizer on a sequence space is strongly equivalent to one commuting with U is not load-bearing for Theorem 4.6, because the centralizers used there, Omega and ~Omega, commute with U by their explicit formulas; Lemma 4.1 only averages over UR for maps that already commute with UR. Self-citations such as [1], [2], [4], [5], and [9] appear in applications (Sections 5.1, 5.3, 5.6, and 5.7), not in the main counterexample; they are prior published theorems whose assumptions do not include the target results Ext^2(ell_2, ell_2) != 0 or Ext^2(ell_1, K) != 0. They therefore function as independent supporting results rather than as a circular chain. No parameter is fitted and then renamed as a prediction, and the equivalence criteria of Lemmas 3.1 and 3.3 are derived from the standard zigzag definition of Ext^2 rather than being imposed by definition. I find no circular step.

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

The central theorem has no fitted free parameters; the proof is a constructive counterexample using Kalton-Peck's centralizer and a new chunked version. The main non-standard assumptions are the unitary-commuting equivalence (unproved in this paper) and deep external theorems used in applications. No new physical entities are postulated.

assumptions (5)
  • domain assumption Every centralizer on a sequence space is strongly equivalent to one that commutes with the unitary group U.
    Stated without proof in Section 4; used to make ~Ω and the witness H support-preserving, which is essential for the disjoint-support estimate in Lemma 4.4.
  • standard math Kalton-Roberts theorem: c0 is a K-space with an absolute constant (C ≤ 200 for real scalars).
    Deep external result used in Lemma 5.10 to construct a centralizer with values in ℓ1.
  • standard math Ext_{ℓ∞}(ℓr, ℓq) = 0 unless r = q, from [1, main result].
    Used in Proposition 5.8 to prove exactness of Hom(ℓr, -) applied to the splitting diagram.
  • standard math Hom(ℓr, Z(Φ_q)) = Z(Φ_p) for reflected centralizers, from [1, Corollary 2].
    Used in Proposition 5.8 to transfer triviality between ℓp and ℓq.
  • standard math Standard homological algebra facts: five-lemma, open mapping theorem, projectivity of ℓ1, Hahn-Banach.
    Background used throughout for Ext and Ext^2 equivalence relations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of On the ${\Ext}^2$-problem for Hilbert spaces." pith.science (2026). https://pith.science/paper/CQGK3BLQ

@misc{pith2026190806529,
  author       = {Pith},
  title        = {Pith review of: On the $\Ext^2$-problem for Hilbert spaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CQGK3BLQ}},
  note         = {Machine review of arXiv:1908.06529}
}
abstract

We show that $\Ext^2(\ell_2, \ell_2)\neq 0$ in the category of Banach spaces. This solves a sharpened version of Palamodov's problem and provides a solution to the second order version of Palais problem. We also show that $\Ext^2(\ell_1, \K)\neq 0$ in the category of quasi Banach spaces which solves the four-space problem for local convexity.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    Cabello S´ anchez, Nonlinear centralizers in homology , Math

    F. Cabello S´ anchez, Nonlinear centralizers in homology , Math. Ann. 358 (2014) 779–798

    work page 2014

  2. [2]

    Cabello S´ anchez, Pointwise tensor products of function spaces , J

    F. Cabello S´ anchez, Pointwise tensor products of function spaces , J. Math. Anal. Appl. 418 (2014) 317–335

    work page 2014

  3. [3]

    Cabello S´ anchez, J.M.F

    F. Cabello S´ anchez, J.M.F. Castillo, Duality and twisted sums of Banach spaces , J. Funct. Anal. 175 (2000) 1–16

    work page 2000

  4. [4]

    Cabello S´ anchez, J.M.F

    F. Cabello S´ anchez, J.M.F. Castillo, Uniform boundedness and twisted sums of Banach spaces , Hous- ton J. Math. 30 (2004) 523–536

    work page 2004

  5. [5]

    Cabello S´ anchez, J.M.F

    F. Cabello S´ anchez, J.M.F. Castillo, N.J. Kalton. Complex interpolation and twisted twisted Hilbert spaces, Pacific J. Math. (2015) 276 (2015) 287–307

    work page 2015

  6. [6]

    Cabello S´ anchez, J.M.F

    F. Cabello S´ anchez, J.M.F. Castillo, N.J. Kalton, D.T. Yost, Twisted sums with C(K) spaces, Trans. Amer. Math. Soc. 355 (2003) 4523–4541

    work page 2003

  7. [7]

    Banach spaces and their applications in Analysis

    J.M.F. Castillo, The story of Kalton’s last unpublished paper , plenary conference at “Banach spaces and their applications in Analysis” held in the CIRM, Luminy, 12- 16 January 2015. https://library.cirm-math.fr/Record.htm?idlist=11&record=19277570124910957529

    work page 2015

  8. [8]

    Castillo, W

    J.M.F. Castillo, W. Cu´ ellar, V. Ferenczi, Y. Moreno, Complex structures on twisted Hilbert spaces , Israel J. Math. 222 (2017) 787–814

    work page 2017

Show all 28 references
  1. [9]

    Castillo, R

    J.M.F. Castillo, R. Garc ´ ıa,Bilinear forms and the Ext2 problem in Banach spaces , Linear Algebra and its Applications 566 (2019) 199–211

    2019

  2. [10]

    Castillo, D

    J.M.F. Castillo, D. Morales, J. Su´ arez de la Fuente, Derivation of vector valued complex interpolation scales, J. Math. Anal. Appl. 468 (2018) 461–472

    2018

  3. [11]

    P. Enflo, J. Lindenstrauss, G. Pisier, On the three space problem , Math. Scand. 36 (1975) 199–210

    1975

  4. [12]

    Frerick, D

    L. Frerick, D. Sieg, Exact categories in functional analysis , Script 2010. Available at: www.mathematik.uni-trier.de:8080/abteilung/analysis/H omAlg.pdf

    2010

  5. [13]

    Greenleaf, Invariant means on topological groups and their applicatio ns

    F. Greenleaf, Invariant means on topological groups and their applicatio ns. Van Nostrand Math. Studies, No. 16 New York: Van Nostrand Reinhold Company, 1969

    1969

  6. [14]

    Hilton, U

    P.J. Hilton, U. Stammbach, A course in homological algebra , Second Edition. Graduate Texts in Math. 4, Springer-Verlag, 1996

    1996

  7. [15]

    Kalton, The three-space problem for locally bounded F-spaces , Compositio Math

    N.J. Kalton, The three-space problem for locally bounded F-spaces , Compositio Math. 37 (1978) 243–276

    1978

  8. [16]

    Kalton, Nonlinear commutators in interpolation theory , Memoirs of the AMS, vol

    N.J. Kalton, Nonlinear commutators in interpolation theory , Memoirs of the AMS, vol. 385, 1988

    1988

  9. [17]

    Kalton, Differentials of complex interpolation processes for K¨ othe function spaces , Trans

    N.J. Kalton, Differentials of complex interpolation processes for K¨ othe function spaces , Trans. Amer. Math. Soc. 333 (1992) 479–529

    1992

  10. [18]

    Kalton, Quasi-Banach spaces, in Handbook of the geometry of Banach spaces, Vol

    N.J. Kalton, Quasi-Banach spaces, in Handbook of the geometry of Banach spaces, Vol. 2, Edited by W.B. Johnson and J. Lindenstrauss. Elsevier (2003), 1099–113 0

    2003

  11. [19]

    Kalton, Extension of linear operators and Lipschitz maps into C(K)-spaces, New York J

    N.J. Kalton, Extension of linear operators and Lipschitz maps into C(K)-spaces, New York J. Math. 13 (2007) 317–381

    2007

  12. [20]

    Kalton, S

    N.J. Kalton, S. Montgomery-Smith, Interpolation of Banach spaces , in Handbook of the geometry of Banach spaces, Vol. 2, Edited by W.B. Johnson and J. Lindenstra uss. Elsevier (2003), 1131–1175

    2003

  13. [21]

    Kalton, N.T

    N.J. Kalton, N.T. Peck, Twisted sums of sequence spaces and the three space problem , Trans. Amer. Math. Soc. 255 (1979) 1–30

    1979

  14. [22]

    Kalton, W

    N.J. Kalton, W. Roberts, Uniformly exhaustive submeasures and nearly additive set f unctions, Trans. Amer. Math. Soc. 278 (1983) 803–816

    1983

  15. [23]

    MacLane, Homology, Grund

    S. MacLane, Homology, Grund. der math. Wissenschaften 114, Springer 1975. ON THE Ext 2-PROBLEM FOR HILBERT SPACES 29

    1975

  16. [24]

    Moreno Salguero, Theory of z-linear maps , Ph.D

    Y. Moreno Salguero, Theory of z-linear maps , Ph.D. Thesis, Universidad de Extremadura, 2003

    2003

  17. [25]

    Palamodov, Homological methods in the theory of locally convex spaces

    V. Palamodov, Homological methods in the theory of locally convex spaces . (Russian) Uspehi Mat. Nauk 26 (1971) 3-65. (English Transl.) Rusiian Math. Surveys 26 (19 71) 1-64

    1971

  18. [26]

    Ribe, Examples for the nonlocally convex three space problem , Proc

    M. Ribe, Examples for the nonlocally convex three space problem , Proc. Amer. Math. Soc. 73 (1979) 351–355

    1979

  19. [27]

    Roberts, A non-locally convex F -space with the Hahn-Banach extension property , LNM 604, Springer (1977)

    J.W. Roberts, A non-locally convex F -space with the Hahn-Banach extension property , LNM 604, Springer (1977)

    1977

  20. [28]

    J. Wengenroth, Palamodov’s questions from homological methods in the theo ry of locally convex spaces, in Banach space methods, Proceedings of the V Conference in Ban ach spaces, C´ aceres, 2004, (J.M.F. Castillo and W.B. Johnson eds.), LN London Math. Society 337, Cambridge U...

    2004

Pith tools

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