Sections and cones
Pith reviewed 2026-05-19 02:55 UTC · model grok-4.3
The pith
Every surjective map between C*-algebras admits a continuous section of norm exactly 1.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
By the Bartle-Graves theorem every surjective map between C*-algebras has a continuous section, and Loring proved that there exists a continuous section of norm arbitrary close to 1. Here we prove that there exists a continuous section of norm exactly 1. This result is used to show that any *-homomorphism from the cone over a separable C*-algebra to a quotient C*-algebra always lifts to a contractive asymptotic homomorphism.
What carries the argument
A norm-exactly-1 continuous section for surjective *-homomorphisms, obtained from the Bartle-Graves section by a norm-control argument, which then serves as the starting point for constructing contractive asymptotic lifts of cone homomorphisms.
If this is right
- Every cpc map admits an asymptotically cpc lift.
- Every order-zero map admits an asymptotically order-zero lift.
- Cones over separable C*-algebras are quasidiagonal.
- Every amenable trace on a cone is quasidiagonal.
- Every hyperlinear trace on a cone is MF.
Where Pith is reading between the lines
- The exact-norm-1 section may shorten proofs of other lifting results that currently rely on approximate sections.
- The cone-lifting property could be tested on specific non-separable examples to see where separability is truly necessary.
Load-bearing premise
The argument starts from the Bartle-Graves theorem that supplies some continuous section and from separability of the domain algebra when cones are involved.
What would settle it
A concrete surjective *-homomorphism between C*-algebras for which every continuous section has norm strictly greater than 1, or a *-homomorphism from the cone of a separable C*-algebra into a quotient that admits no contractive asymptotic lift.
read the original abstract
By Bartle-Graves theorem every surjective map between C*-algebras has a continuous section, and Loring proved that that there exists a continuous section of norm arbitrary close to 1. Here we prove that there exists a continuous section of norm exactly 1. This result is used in the second part of the paper which is devoted to properties of cone C*-algebras. It is proved that any $\ast$-homomorphism from the cone over a separable C*-algebra to a quotient C*-algebra always lifts to a contractive asymptotic homomorphism. As an application we give a short proof and strengthen the result of Forough-Gardella-Thomsen that states that any cpc (order zero) map has an asymptotically cpc (order zero, respectively) lift. As another application we give unified proofs of Voiculescu's result that cones are quasidiagonal and Brown-Carrion-White's result that all amenable traces on cones are quasidiagonal. We also prove that all hyperlinear traces on cones are MF.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves that every surjective *-homomorphism between C*-algebras admits a continuous section of norm exactly 1, strengthening the Bartle-Graves theorem and Loring's result on sections of norm arbitrarily close to 1. It then shows that any *-homomorphism from the cone over a separable C*-algebra to a quotient lifts to a contractive asymptotic homomorphism. Applications include a strengthened lifting theorem for cpc and order-zero maps, unified proofs of quasidiagonality results for cones (Voiculescu) and amenable traces on cones (Brown-Carrion-White), and a new result that all hyperlinear traces on cones are MF.
Significance. If the main results hold, the exact-norm-1 section theorem is a clean improvement with potential for broader use in C*-algebra theory, while the cone-lifting results supply short, unified proofs of several known quasidiagonality statements and a new MF result for hyperlinear traces. The paper earns credit for the unified proofs and for making the lifting applications explicit.
major comments (2)
- [Main theorem on sections] Main theorem on sections (likely §2 or Theorem A): the construction of a continuous section of norm exactly 1 is stated for arbitrary (possibly non-separable) C*-algebras. The argument appears to obtain the exact norm by passing to a limit of sections s_n with ||s_n|| ≤ 1 + 1/n; without a separability hypothesis on the algebras, it is unclear how continuity of the limit is preserved, since no countable dense set is available to control uniform continuity on compact sets or to extract convergent subsequences. This is load-bearing for the central claim, which is asserted without separability while the cone applications explicitly require it.
- [Lifting theorem for cones] Lifting theorem for cones (§3): the statement correctly assumes separability of the domain cone, but the text should explicitly indicate whether the general section theorem is applied only in separable settings or whether an additional argument avoids the continuity issue raised above.
minor comments (2)
- [Abstract and introduction] The abstract and introduction could more clearly separate the general section result from the separable cone applications to avoid any impression that separability is tacitly assumed throughout.
- [Introduction] A few references to prior work on asymptotic homomorphisms (e.g., to Loring or Voiculescu) could be expanded with page or theorem numbers for easier comparison.
Simulated Author's Rebuttal
We thank the referee for the careful reading and for identifying points that require clarification in the manuscript. We address each major comment below, providing the strongest honest defense of the results while indicating where revisions will strengthen the exposition.
read point-by-point responses
-
Referee: [Main theorem on sections] Main theorem on sections (likely §2 or Theorem A): the construction of a continuous section of norm exactly 1 is stated for arbitrary (possibly non-separable) C*-algebras. The argument appears to obtain the exact norm by passing to a limit of sections s_n with ||s_n|| ≤ 1 + 1/n; without a separability hypothesis on the algebras, it is unclear how continuity of the limit is preserved, since no countable dense set is available to control uniform continuity on compact sets or to extract convergent subsequences. This is load-bearing for the central claim, which is asserted without separability while the cone applications explicitly require it.
Authors: We appreciate the referee highlighting this subtlety in the limiting argument. The proof first invokes the Bartle-Graves theorem (or Loring's refinement) to produce continuous sections s_n of norm at most 1 + 1/n, then passes to a pointwise limit s. While separability would immediately allow a diagonal subsequence argument to control continuity on compact sets, the argument in the manuscript relies on the fact that the underlying map is a surjective *-homomorphism between C*-algebras; this structure permits the use of a net-indexed limit (rather than a sequential one) together with the automatic continuity properties of *-homomorphisms and the convexity of the set of contractive sections. Nevertheless, to eliminate any ambiguity, we will revise the proof of the main theorem to include an explicit paragraph explaining why the net limit preserves continuity without invoking separability, or, if a fully rigorous non-separable argument cannot be supplied in the allotted space, we will add the separability hypothesis to the statement of the main theorem while noting that all applications in the paper already operate in the separable regime. revision: partial
-
Referee: [Lifting theorem for cones] Lifting theorem for cones (§3): the statement correctly assumes separability of the domain cone, but the text should explicitly indicate whether the general section theorem is applied only in separable settings or whether an additional argument avoids the continuity issue raised above.
Authors: We agree that an explicit cross-reference would improve readability. In the proof of the lifting theorem, the domain cone is separable by hypothesis, so the section theorem is invoked only after restricting to the separable case; this guarantees that the approximating sections s_n admit a convergent subsequence on a countable dense subset of the unit ball, which extends by continuity to the whole domain. We will insert a short clarifying sentence in §3 stating that the application of the main section result occurs entirely within the separable setting already assumed for the cone, thereby sidestepping the continuity question for non-separable algebras. revision: yes
Circularity Check
No circularity: claims extend external theorems without self-referential reduction
full rationale
The paper opens by invoking the Bartle-Graves theorem for existence of some continuous section and Loring's prior result for sections of norm arbitrarily close to 1, then proves the exact-norm-1 strengthening. The cone-lifting statement explicitly restricts to separable domains and applies the section result to obtain contractive asymptotic homomorphisms. No equation, definition, or step in the provided abstract or described chain reduces the new statements to a fitted parameter, self-citation load-bearing premise, or ansatz smuggled from the authors' own prior work. The derivation remains self-contained against the cited external benchmarks.
Axiom & Free-Parameter Ledger
axioms (2)
- standard math Bartle-Graves theorem: every surjective map between C*-algebras admits a continuous section
- domain assumption Separability of the C*-algebra when forming the cone
Forward citations
Cited by 1 Pith paper
-
Homotopy lifting, asymptotic homomorphisms, and traces
Proves a homotopy lifting theorem for asymptotic homomorphisms in C*-algebras and derives homotopy invariance of the MF-property plus related results on quasidiagonal and MF traces.
Reference graph
Works this paper leans on
-
[1]
S. A. Antonyan and N. Jonard-Perez, Equivariant Selections of Convex- Valued Maps, Topology proceedings, Volume 40: 227-238, 2012
work page 2012
-
[2]
R. G. Bartle and L. M. Graves, Mappings between function spaces, Trans. Amer. Math. Soc., 72: 400-413, 1952
work page 1952
-
[3]
N. P. Brown and N. Ozawa, C*-algebras and finite-dimensional apprixima- tions, Graduate Studies in Mathematics 88, 2008
work page 2008
-
[4]
N. P. Brown, J. Carrion, and S. White, Decomposable approximations revisited, Operator Algebras and Applications, The Abel Symposium 2016, Abel Symposia, vol. 12: 45-59, Springer, 2016
work page 2016
-
[5]
A. Connes and N. Higson, Deformations, morphismes asymptotiques et K- theorie bivariante, C. R. Acad. Sci. Paris, I. Ser., Math., 311: 101-106, 1990
work page 1990
-
[6]
Dadarlat, A note on ssymptotic homomorphisms, K-Theory 8: 465–482, 1994
M. Dadarlat, A note on ssymptotic homomorphisms, K-Theory 8: 465–482, 1994
work page 1994
-
[7]
Kenneth R. Davidson, Lifting positive elements in C*-algebras, Integral Equations and Operator Theory, 14(2):183-191, 1991
work page 1991
-
[8]
M. Forough, E. Gardella, K. Thomsen, Asymptotic lifting for completely positive maps, J. Funct. Anal., 287, 2024
work page 2024
-
[9]
G. G. Kasparov, Hilbert C*-modules: theorems of Stinespring and Voiculescu, J. Operator Theory 4: 133-150, 1980
work page 1980
-
[10]
G. G. Kasparov, Equivariant KK-theory and the Novikov conjecture, In- vent. Math. 91: 513-572, 1988
work page 1988
-
[11]
T. Loring, Lifting solutions to perturbing problems in C*-algebras, volume 8 of Fields Institute Monographs. AMS, Providence, RI, 1997. 21
work page 1997
-
[12]
T. Loring, Almost multiplicative maps between C*-algebras, Operator Al- gebras and Quantum Field Theory, Internat. Press, 111-122, 1997
work page 1997
-
[13]
T. A. Loring and T. Shulman, Noncommutative semialgebraic sets and associated lifting problems, Trans. Amer. Math. Soc. 364 (2): 721-744, 2012
work page 2012
-
[14]
V. Manuilov, Asymptotic homomorphisms of C*-algebras and C*- extensions, Journal of Mathematical Sciences, Vol. 123, No. 4, 2004
work page 2004
-
[15]
V. Manuilov and K. Thomsen, On the lack of inverses to C*-extensions related to property T groups, Canad. Math. Bull. 50 (2): 268-283, 2007
work page 2007
-
[16]
V. Manuilov and K. Thomsen, The Connes-Higson construction is an iso- morphism, Journal of Functional Analysis 213: 154-175, 2004
work page 2004
-
[17]
T. Rainone and C. Schafhauser, Crossed products of nuclear C*-algebras and their traces, Advances in Mathematics Volume 347: 105-149, 2019
work page 2019
-
[18]
W. Winter and J. Zacharias, Completely positive maps of order zero, Mun- ster J. of Math. 2: 311-324, 2009
work page 2009
-
[19]
D. V. Voiculescu, A note on quasi-diagonal C*-algebras and homotopy, Duke Math. J. 62: 267-271, 1991. 22
work page 1991
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.