Pith. sign in

REVIEW 4 major objections 6 minor 20 references

Scalability and asymptotic adjunction

T0 review · 4 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read For scalable pairs of metric spaces, the continuous-function functor and the relative uniform Roe functor are asymptotically adjoint, giving suspension-free models of E-theory and K-homology.

desk verdict Main theorem is solid and new, but all applications ride on an unproved imported theorem from the author's own preprint, so the corollaries are conditional until that's supplied. read the letter →

arxiv 2510.07883 v2 pith:ANF4W3LR submitted 2025-10-09 math.OA

Georgii S. Makeev This is my paper · ORCID
classification math.OA MSC 46L8519K3546L80
keywords asymptoticadjunctionrelativeRoefunctoruniformalgebraE-theoryscalablemetricspaceK-homologygoodendofunctorscoarsegeometry
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's central claim is that, for any scalable pair of proper metric spaces, tensoring with C0(X,X0) admits a right 'asymptotic adjoint': the relative uniform Roe functor built from a discretization of the space. An asymptotic adjunction is a weakened adjunction in the category of good endofunctors of C*-algebras, where the usual triangular identities need only commute up to homotopy after a stabilization shift. The paper argues this weak notion is still enough: it induces an isomorphism of monoids of generalized morphisms, and from that the author derives an unsuspended description of Connes–Higson E-theory, an E-theoretic analog of the KK1-versus-Ext correspondence, and a Roe-algebra formula for the K-homology of compact metric spaces via metric cones. The reason to care is that suspensions and colimits, analytic conveniences that obscure the underlying geometry, can be replaced by operator-algebraic objects attached directly to the metric space.

What carries the argument

The load-bearing object is the relative uniform Roe functor N^u_{X,X0} := M^u_X / M^u_{X⊃X0}: the norm closure of finite-propagation matrices indexed by the discrete space X, modulo the ideal of matrices supported in neighborhoods of X0. It is a good labeled endofunctor, so it inherits the homotopy and stabilization calculus of good endofunctors. The unit η quantizes a square partition of unity into matrix coefficients; the counit ε feeds the scaling family sc_t into a diagonal matrix and passes to the asymptotic algebra. The asymptotic-adjunction diagrams are the mechanism that converts these two explicit natural transformations into monoid isomorphisms: if they commute up to homotopy, then

What would settle it

Check the asymptotic-adjunction diagrams on the simplest nontrivial scalable pair, X=R, X0=∅, with the standard shrinking scaling sc_t(x)=t^{-1}x: take B=C and a smooth compactly supported f∈C0(R), and compute the norm of the difference between the two composite maps in Proposition 3.19 as t→∞. The theorem predicts this norm tends to zero; an explicit estimate exhibiting a positive lower bound would refute the adjunction, while a verification would confirm the mechanism that all the applications rely on.

Watch

Extended reading notes

Core claim

Introducing relative uniform Roe functors as quotients N^u_{X,X0} := M^u_X / M^u_{X⊃X0}, the author proves that for a scalable pair X=(X,X0), any Δ-discretization X=(X,X0), and any square partition of unity subordinate to the Δ-ball cover, there is an asymptotic adjunction C_{X,X0} ⊣_as N^u_X, with explicit unit η: Id ⇒ N^u_X C_{X,X0} and counit ε: C_{X,X0} N^u_X ⇒ AK. The unit encodes a partition of unity as matrix coefficients; the counit evaluates continuous functions along the scaling maps and passes to the asymptotic algebra. By the isomorphism theorem for asymptotic adjunctions, this yields natural isomorphisms of monoids colim_k [C_{X,X0}A, A^kK, B] ≅ colim_k [A, N^u_{X,X0}A^kK, B]. F

Load-bearing premise

The applications all pass through the previously established theorem, stated here without proof, that an asymptotic adjunction induces isomorphisms of generalized-morphism monoids; if that framework has a gap, the advertised corollaries of Section 3.4 do not follow, even if the main adjunction theorem itself stands.

Editorial extensions

If this is right

  • For separable A, the colimit in the definition of generalized morphisms is unnecessary: [C_X A, AK, B] ≅ [A, N^u_X AK, B].
  • Connes–Higson E-theory gains suspension-free models: E^0(A,B) ≅ [[A, M^u_{Z^2}, B]] and E^1(A,B) ≅ [[A, M^u_Z, B]].
  • E^1(A,B) is isomorphic to a monoid of homotopy classes of extensions with asymptotic coefficients, via the relative uniform Roe functor of the half-line discretization; this parallels the classical KK^1 ≅ Ext^{-1} correspondence.
  • The K-homology of any compact metrizable space X is expressible as [[C, N^u_{(OX)discr,{0}}, K]], where OX is the metric cone on X.
  • The adjunction is insensitive, up to the isomorphism, to the choice of Δ-discretization and square partition of unity, so the analytic construction is geometrically robust in that sense.

Reading between the lines

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

  • Editorial: The paper itself notes that naturality of the K-homology formula in X is left open; if a natural refinement can be constructed, the cone description would become a full functorial model of K-homology rather than an object-level invariant.
  • Editorial: The same pattern plausibly extends to all degrees: the pair (R^n,∅) suggests identifications E^n(A,B) ≅ [[A, M^u_{Z^n}, B]], and the boundary-relative cases suggest E^n(A,B) ≅ [[A, N^u_{Z^n_+,{0}}, B]], for every n.
  • Editorial: The extension correspondence points toward a direct comparison with Kasparov theory; testing whether [[A, N^u_{Z+,{0}}, B]] is invertible as a group for nuclear A would sharpen the analogy with KK^1 ≅ Ext^{-1}.
  • Editorial: The proof uses only the scaling axioms, not a group action, so versions of the adjunction may survive for spaces with very weak self-similarity; subjecting the construction to spaces that are not locally finite would clarify where bounded geometry is truly essential.
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

4 major / 6 minor

Summary. The paper introduces relative uniform Roe functors N^u_{X,X0} for a discrete bounded-geometry space X and subspace X0, and shows (Theorem 3.15) that for a scalable pair of proper metric spaces (X,X0), with a ∆-discretization X=(X,X0), the functor C_X := C_0(X,X0; ·) is asymptotically adjoint to N^u_X in the author's framework of good endofunctors (Definition 1.50). The unit and counit are given explicitly and the two asymptotic triangle identities are treated in Propositions 3.19 and 3.20, using the scaling axioms, partition-of-unity estimates, and a technical reparametrization lemma. Theorem 3.21 then uses the imported Theorem 1.51 to deduce a monoid isomorphism between colim_k [C_X A, A^k K, B] and colim_k [A, N^u_X A^k K, B], yielding three applications: unsuspended descriptions of E^0 and E^1, an extension/E_1 correspondence, and a K-homology formula via metric cones. Section 4 proves a technical theorem (Theorem 4.1) showing that for separable A the colimit over A^n K collapses to a single copy of A K.

Significance. The analytic core of the paper is substantial and mostly coherent. The explicit construction of the unit and counit, and the norm estimates in Proposition 3.19 (in particular inequality (3.5)) and Proposition 3.20 (Claims 1–4), are genuine contributions and appear to be checked correctly within the stated framework. The relative uniform Roe functor is a useful new object, and the technical theorem of Section 4 is a self-contained stabilization result that is valuable independently. If the implied applications hold, they would provide attractive descriptions of unsuspended E-theory, E_1/extension theory, and K-homology. However, every advertised application passes through Theorem 1.51, which is imported from the companion paper [12] and stated without proof in §1.11. The manuscript is commendably explicit about this dependence, but as a refereed standalone paper this is a load-bearing gap. The K-homology application has a second, independent gap, because the coarse homotopy invariance needed for relative Roe functors with nonempty subspace is not covered by the cited reference. These issues are fixable in principle but require either including the missing proofs or giving precise po

major comments (4)
  1. [§1.11, Theorem 1.51] Theorem 1.51 is the sole mechanism that converts the asymptotic adjunction of Theorem 3.15 into the monoid isomorphism of Theorem 3.21, and hence into all three applications of §3.4. It is stated without proof, with only the remark 'one can easily obtain'. The formulas (1.9) define Φ and Ψ, but well-definedness requires showing they pass to F-homotopy classes, that they are independent of the colimit index, that the compositions are homotopic to identities using the asymptotic triangle identities and stability homotopy (Lemma 1.41), and that the maps are monoid homomorphisms. If [12] contains a proof, a precise pointer is needed; otherwise the proof should be included. As written, the central applications do not follow from the analytic results of Section 3 on the basis of the present manuscript.
  2. [§3.4, K-homology (3.19)] The assertion that the right-hand side of (3.19) is well-defined uses two further imported facts: coarse homotopy invariance of relative Roe functors N^u_{X,X0}K for pairs with X0 nonempty, and independence of the cone from the embedding. The text notes that [11] proves coarse homotopy invariance only for X0=Y0=∅. In (3.19) the second entry is {0}, so this is exactly the case not covered by the cited reference. The construction of K^1(X) therefore requires a proof or a precise reference for the relative case. This issue is independent of Theorem 1.51.
  3. [§3.4, E-theory applications] The unsuspended E-theory statements E^0(A,B) ≅ [[A,M^u_{Z^2},B]] and E^1(A,B) ≅ [[A,M^u_Z,B]] are derived from (3.18), which is quoted from [4] and needs to be checked in the present generalized-morphism framework. More importantly, the step 'Applying Theorem 3.21 to the right-hand sides of (3.18)' requires identifying C_{R^2} with the suspension functor S in the presence of the left-hand multiplication in Theorem 3.21. These tensor/identification steps should be written out, since Theorem 3.21 involves C_X acting on the left on A, while the relevant functor in (3.18) is S acting on A.
  4. [§3.2, Theorem 3.15] The theorem is stated for a fixed discretization scale ∆ and a particular square partition of unity {α_x}. The proof shows that for that datum the maps η and ε are well-defined and satisfy the asymptotic adjunction. However, the applications in §3.4 implicitly treat the right-hand functor N^u_X, which depends on the chosen discretization X and the partition {α_x}, as associated to the pair (X,X0). The manuscript does not prove independence of the discrete model, the partition, or the scale ∆, nor does it identify the resulting Roe functor up to homotopy. This is load-bearing for the isomorphism statements in Theorem 3.21 and the corollaries; a brief argument or reference is needed.
minor comments (6)
  1. [Abstract/Introduction] The abstract switches between 'scalable locally compact metric spaces' and 'scalable proper metric spaces'; the body consistently uses proper metric spaces. Please harmonize the terminology.
  2. [§1.11, Definition 1.50] The triangle diagrams in Definition 1.50 would benefit from being displayed with more separation; as typeset they are hard to parse. Also, the second diagram's lower horizontal arrow should be labeled consistently as Nαι00.
  3. [§2.2, Definition 2.7] The set in (2.3) is presented as a ∗-subalgebra; a sentence clarifying that finite-propagation matrices form a ∗-subalgebra under multiplication would help the reader, since the propagation bound for products is not immediate.
  4. [§3.3, Proposition 3.20] The notation in the proof of Proposition 3.20 uses both φ1/φ2 as natural transformations and then introduces 'φ01' and 'φ12' as homotopies; the naming is confusing because the subscripts are not consistent. Consider renaming the homotopies (for instance, H^{01} and H^{12}).
  5. [§4.1, Lemma 4.6] Lemma 4.6 is used crucially in Proposition 4.7, but its proof is omitted with 'left as a simple exercise'. A one-sentence justification would improve readability and make the paper more self-contained.
  6. [General] There are several typos: 'assiosiated' in Definition 4.3, 'remarametrization' in Lemma 4.12, and a missing article/phrase in the sentence 'Using Lemma 4.9, we conclude that ψ is also a ∗-homomorphism' (it should say why). The dependence on the unpublished companion [12] should also be flagged explicitly in the introduction.

Circularity Check

1 steps flagged · score 4.0 of 10

Applications are bridged by the author's own unproved Theorem 1.51 from [12], not by Theorem 3.15 alone.

  1. self citation load bearing [Introduction and §1.11, Theorem 1.51; used at §3.4, Theorem 3.21]
    "Section 1 recalls the category machinery of good endofunctors developed in [12]. ... Theorem 1.51. Let A and B be C*-algebras, let S, N ∈ hGEFC, and let S ⊣_as N be an asymptotic adjuction witnessed by a unit η and a counit ε. Then, the formulas ... define mutually inverse isomorphisms of monoids natural in A and B. ... Proof. The first statement follows from Theorems 3.15 and 1.51."

    Theorem 1.51 is the exact bridge that converts the analytic asymptotic adjunction of Theorem 3.15 into the monoid isomorphism of Theorem 3.21, and hence into every advertised application in §3.4: unsuspended E-theory, extensions, and K-homology. The theorem is not proved in this paper; the framework is attributed to the same author's preprint [12], and the Introduction says only that 'one can easily obtain the isomorphism.' No independent proof, machine-checked verification, or external source is given. Thus the derived claims rest on a load-bearing same-author citation whose content is precisely the step between the theorem and the corollaries. A gap in [12] would invalidate Theorem 3.21 and all §3.4 consequences even though Theorem 3.15 itself could still be sound.

full rationale

The core analytic result, Theorem 3.15, is not circular: the unit η and counit ε are given by explicit formulas involving the partition of unity and the scaling, and Propositions 3.19–3.20 verify the triangle diagrams with independent norm estimates; scalability is a hypothesis, not a conclusion, and no fitted parameter is renamed as a prediction. The circularity concern is narrower and located at the bridge: Theorem 3.21 is obtained by applying Theorem 1.51, a general statement about asymptotic adjunctions that is quoted without proof from the author's own preprint [12]. Since every advertised corollary passes through Theorem 3.21, the applications are load-bearing on that self-citation. I also note that the K-homology section explicitly concedes that coarse homotopy invariance of relative Roe functors is proved in [11] only for the case X0 = Y0 = ∅, while the application needs the nonempty-subspace case; this is a further unverified bridge, but I treat it as a support gap rather than a separate circular step. Overall, the paper has significant independent analytic content, but its advertised payoff is completed by an unproved same-author theorem.

Assumptions & free parameters 1 free parameters · 7 assumptions · 2 invented entities

The central theorem (3.15) is analytic and largely self-contained, but every application flows through Theorem 1.51 and the framework of [12] (imported without proofs) and through coarse homotopy invariance for pairs (asserted via [11], proved only for X0 = ∅). Scaling data (sc, ρ_t) and the discretization choice are assumed structures, not fitted numbers; no parameter is tuned to data. The novel entities (asymptotic adjunction, relative Roe functor) carry no external falsifiable prediction; their support is internal to the theorems they enable.

free parameters (1)
  • Discretization scale ∆ and square partition of unity {α_x}
    Theorem 3.15 fixes a ∆-discretization (X,X0) and a square partition of unity; the theorem is meant to be insensitive to the choices, but coarse invariance for pairs is imported from [11] rather than proved here.
assumptions (7)
  • domain assumption Theorem 1.51: an asymptotic adjunction S ⊣_as N induces natural isomorphisms of monoids [[SA,Id,B]] ≅ [[A,N,B]]
    Stated in §1.11 without proof as part of the review of [12]; it is the bridge from Theorem 3.15 to every application in §3.4. If it fails, the E-theory/E1/K-homology corollaries do not follow.
  • domain assumption Scalability of the pair (X,X0): existence of sc: X×R+→X and control maps ρ_t satisfying (S1)–(S5) of Definition 3.4
    The main theorem only covers scalable pairs; Propositions 3.19 and 3.20 use (S2)–(S5) in the estimates. Examples are inherited from [6] (locally finite trees, non-positively curved manifolds) and metric cones.
  • domain assumption Coarse homotopy invariance of relative Roe functors for pairs: N^u_{X,X0}K and N^u_{Y,Y0}K homotopy equivalent for coarsely equivalent pairs
    Needed in §3.4 so the K-homology expression (3.19) is independent of the cone embedding/discretization; [11] proves only the X0 = ∅ case. The passage to pairs is asserted without proof.
  • domain assumption Good-endofunctor framework of [12]: decent/labeled endofunctors, monoidal structure, homotopies of natural transformations, generalized morphisms, Lemma 1.41 (K^2 ≅ K)
    Section 1 is a review; Theorems 1.43, 1.45, 1.51 and Lemma 1.41 are used without proof in the body, importing the categorical machinery wholesale.
  • standard math [5, Theorem 2.16] (colimit collapses for separable A) and stability absorption
    Used in Theorem 3.21 and Theorem 4.1 to remove colimits and to match E-theory's suspension/stable conventions with the framework's A^kK colimits.
  • standard math E-theory ≅ KK for separable nuclear algebras (Blackadar [2, Thm 25.6.3]); K^i(X) ≅ E^i(C0(X),C) for compact metrizable X
    Launches the K-homology chain (3.19) in §3.4.
  • standard math Existence of ∆-discretizations of proper metric spaces ([20, Prop. A.3.11]) and of square partitions of unity subordinate to {B_∆(x)}
    Defines X, X0, {α_x} used in Theorem 3.15 and Lemmas 3.10–3.11.
invented entities (2)
  • Asymptotic adjunction (S ⊣_as N)
    purpose: Weakened adjunction between C*-endofunctors that still yields isomorphisms of generalized-morphism monoids; replaces the nonexistent right adjoint of C_X in hAsy.
    A purpose-built categorical relation (Definition 1.50); its only evidence is the theorem that it induces monoid isomorphisms (Theorem 1.51, imported from [12]) — no external falsifiable prediction.
  • Relative uniform Roe functor N^u_{X,X0} = M^u_X / M^u_{X⊃X0}
    purpose: Concrete right asymptotic adjoint to C_{X,X0} for scalable pairs; the vehicle for Roe-algebra descriptions of E-theory and K-homology.
    Newly constructed object (Definition 2.16); internally defined from M^u_X and an ideal. Its support is Theorem 3.15 itself; no external data are predicted.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Scalability and asymptotic adjunction." pith.science (2026). https://pith.science/paper/ANF4W3LR

@misc{pith2026251007883,
  author       = {Pith},
  title        = {Pith review of: Scalability and asymptotic adjunction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ANF4W3LR}},
  note         = {Machine review of arXiv:2510.07883}
}
abstract

In this paper, we introduce relative Roe functors and show that for every pair of scalable locally compact metric spaces with bounded coarse geometry, the functor of continuous functions and the relative Roe functor, both associated with this pair, are asymptotically adjoint. While this asymptotic adjunction is weaker than the genuine one, it retains sufficient categorical properties to be intuitive and useful in applications. These results can be used to provide an unsuspended description of the Connes-Higson $E$-theory, establish connections between $E_{1}$-theory and extension theory, and express $K$-homology of compact metric spaces in terms the corresponding metric cones.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 10 canonical work pages

  1. [12]

    G. S. Makeev. On generalized morphisms associated to endofunctors ofC∗-algebras, Preprint, 2025.https://arxiv.org/abs/2509.02001

    arXiv 2025

  2. [11]

    G. S. Makeev. Roe functors preserve homotopies.Proc. Amer. Math. Soc., 152(8):3467–3474, 2024.https://doi.org/10.1090/proc/16824

    work page doi:10.1090/proc/16824 2024

  3. [4]

    Dadarlat, T

    M. Dadarlat, T. A. Loring.K-homology, asymptotic representations, and unsuspendedE-theory. J. Func. Anal., 126(2):367–383, 1994.https://doi.org/10.1006/jfan.1994.1151

    arXiv 1994

  4. [1]

    C. A. Akemann, G. K. Pedersen, J. Tomiyama. Multipliers ofC ∗-algebras.J. Func. Anal., 13(3):277–301, July 1973.https://doi.org/10.1016/0022-1236(73)90036-0

    work page doi:10.1016/0022-1236(73)90036-0 1973

  5. [2]

    Blackadar.K-theory for Operator Algebras, 2nd ed.Cambridge University Press, September 1998.https://doi.org/10.1017/9781009701907

    B. Blackadar.K-theory for Operator Algebras, 2nd ed.Cambridge University Press, September 1998.https://doi.org/10.1017/9781009701907

    doi:10.1017/9781009701907 1998

  6. [3]

    Connes, N

    A. Connes, N. Higson. Déformations, morphismes asymptotiques etK-théorie bivariante.CR Acad. Sci. Paris Sér. I Math, 311(2):101–106, 1990

    1990

  7. [5]

    Guentner, N

    E. Guentner, N. Higson, J. Trout. EquivariantE-theory forC∗-algebras.Mem. Amer. Math. Soc., 148(703), Nov. 2000.https://doi.org/10.1090/memo/0703

    work page doi:10.1090/memo/0703 2000

  8. [6]

    Higson, J

    N. Higson, J. Roe. On the coarse Baum–Connes conjecture. InNovikov Conjectures, Index Theory Theorems and Rigidity, S. C. Ferry, A. Ranicki, J. M. Rosenberg, eds., vol. 2 ofLondon Mathematical Society Lecture Note Series, pp. 227–254. Cambridge University Press, 1995.https: //doi.org/10.1017/cbo9780511629365.008

    doi:10.1017/cbo9780511629365.008 1995

Show all 20 references
  1. [7]

    Higson, J

    N. Higson, J. Roe.AnalyticK-homology. Oxford University Press, 2000

    2000

  2. [8]

    K. K. Jensen, K. Thomsen.Elements ofKK-theory. Mathematics: Theory & Applications. Birkhäuser, Boston, Boston, MA, 1991.https://doi.org/10.1007/978-1-4612-0449-7

    1991 doi

  3. [9]

    Cambridge University Press, Cambridge, 1995.https://doi.org/ 10.1017/CBO9780511526206

    E.C.Lance.HilbertC ∗-modules: a toolkit for operator algebraists, vol.210ofLondon Mathematical Society Lecture Note Series. Cambridge University Press, Cambridge, 1995.https://doi.org/ 10.1017/CBO9780511526206

    1995 doi

  4. [10]

    G. S. Makeev. Cycles and extensions inE-theory.Uzbek Math. J., 67(1):79–86, 2023.https: //doi.org/10.29229/uzmj.2023-1-10

    2023 doi

  5. [13]

    V. M. Manuilov. AKK-like picture forE-theory ofC ∗-algebras.Stud. Math., 252:105–128, Mar. 2017.https://doi.org/10.4064/sm170817-12-5

    2017 doi

  6. [14]

    V. M. Manuilov. On theC ∗-algebra of matrix-finite bounded operators.J. Math. Anal. Appl., 478(1):294–303, October 2019.https://doi.org/10.1016/j.jmaa.2019.05.033

    2019 doi

  7. [15]

    P. D. Mitchener, B. Norouzizadeh, T. Schick. Coarse homotopy groups.Math. Nachr., 293(8):1515–1533, 2020.https://doi.org/10.1002/mana.201800523

    2020 doi

  8. [16]

    G. J. Murphy.C ∗-algebras and operator theory. Academic Press Inc., Boston, MA, 1990. 35

    1990

  9. [17]

    G. K. Pedersen. Pullback and pushout constructions inC ∗-algebra theory.J. Func. Anal., 167(2):243–344, October 1999.https://doi.org/10.1006/jfan.1999.3456

    1999

  10. [18]

    Roe.Lectures on coarse geometry

    J. Roe.Lectures on coarse geometry. No. 31 in University Lecture Series. Amer. Math. Soc., Providence, RI, 2003.https://doi.org/10.1090/ulect/031

    2003 doi

  11. [19]

    O. Uuye. Homotopical algebra forC ∗-algebras.J. Noncom. Geom., 7(4):981–1006, December 2013.https://doi.org/10.4171/jncg/141

    2013 doi

  12. [20]

    Willett, G

    R. Willett, G. Yu.Higher index theory. Cambridge University Press, 2020.https://doi.org/ 10.1017/9781108867351 36

    2020 doi

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.