The Selfless Dichotomy

Miles Gould

REVIEW 3 major objections 3 minor 1 cited by

The Selfless Dichotomy

T0 review · 3 major / 3 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read Nonfaithful selfless C*-algebras are purely infinite, simple.

desk verdict Closes the nonfaithful selfless gap with a credible new excision argument; the Section 4 isomorphism examples overreach but the core result is solid. read the letter →

arxiv 2606.09654 v2 pith:RUS7YIZX submitted 2026-06-08 math.OA

Miles Gould This is my paper · ORCID

classification math.OA MSC 46L0546L3546L54

keywords selflessC*-algebrasC*-probabilityspacespurelyinfinitesimplereducedfreeproductsCuntzsemigroupstablerankonepureKirchbergalgebras

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

This paper closes the last gap in the dichotomy for selfless C*-probability spaces: when the distinguished state is nonfaithful, the underlying C*-algebra must be purely infinite and simple. The proof works by analyzing infinite reduced free products of nonfaithful yet GNS-faithful states, showing that such products are often purely infinite and simple when a computable sum of factor-level constants exceeds one. With this, every selfless C*-algebra is either stably finite with stable rank one (tracial case) or purely infinite and simple (nontracial case). A direct consequence is that every selfless C*-algebra is pure, and the result yields new permanence properties, progress on a conjecture of Choda and Dykema, and new Kirchberg algebras satisfying the UCT.

What carries the argument

The 'tail excision' mechanism on the free-product Hilbert space. Each factor supplies a positive element a_j in the kernel of its state that annihilates the vacuum vector and all centered words whose first letter comes from a different factor; a finite tail a_F built from such elements cuts out arbitrary positive elements, making them comparable to multiples of a_j. Lemma 2.3 turns this into Cuntz-order inequalities that yield a minimal essential ideal and prove it is purely infinite and simple.

What would settle it

A concrete counterexample to the excision step: exhibit factors (A_j, φ_j) with nonfaithful GNS-faithful states, a finite F disjoint from the support of an approximant d, and a nonzero word w with a_F d_0 a_F w ≠ 0, or otherwise compute a tail product a_F c a_F that does not reduce to ω(c) a_F^2. Alternatively, find a selfless (A, φ) with φ nonfaithful whose infinite self-free product is not purely infinite simple—this would directly refute Theorem 1's route.

Watch

Extended reading notes

Core claim

The paper proves Theorem 1: if (A, φ) is a selfless C*-probability space and φ is nonfaithful, then A is purely infinite and simple. The engine is Theorem 6, which states that for an infinite family of C*-probability spaces with nonfaithful states inducing faithful GNS representations, the reduced free product contains an essential purely infinite simple ideal, and if the sum of δ_j (the norms of the states restricted to the ideals generated by the kernels) exceeds 1, the entire product is purely infinite and simple. This condition is computable factor-by-factor and requires no unitaries in the kernels. Applying the theorem to an infinite self-free product of (A, φ) and using the existential

Load-bearing premise

Lemma 2.3(ii) assumes that positive elements in the kernels of the factors annihilate all centered words from other factors in the free-product Hilbert space; if that excision property fails, Theorem 6—and with it Theorem 1—does not follow.

Editorial extensions

If this is right

Every selfless C*-algebra is pure (Corollary 3), removing the faithful-state assumption from earlier results.
Permanence properties improve: reduced free products with a selfless factor and a GNS-faithful factor are selfless even without separability (Prop 3.1), and matrix amplifications and UHF tensor products of selfless algebras are selfless (Prop 3.2).
The Choda–Dykema conjecture holds when simplicity is strengthened to one factor being selfless (Corollary 5).
Infinite reduced free products of separable nuclear UCT algebras with pure states are Kirchberg algebras satisfying the UCT, and their isomorphisms are governed by K-theory via the Kirchberg–Phillips theorem (Theorem 4.1).
New concrete isomorphisms arise, such as O_∞ ≅ (Z, τ)∗(Z, φ) and (O_∞, ω) ≅ ∗_{j=1}^∞ (Z, φ) for pure states φ, ω and trace τ.

Reading between the lines

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

The tail-excision mechanism suggests a general principle: in infinite free products, nonfaithfulness becomes a resource rather than an obstruction, potentially useful for proving pure infinity in amalgamated or graph products where kernel positives behave similarly.
The δ_j > 1 condition is likely far from sharp; the mechanism may work under weaker divergence conditions, e.g., infinitely many δ_j > 0, since the theorem already covers the case of a divergent sum.
The transfer of pure infinity from an infinite free product to a factor via an existential embedding hints at a broader model-theoretic transfer principle: existential embeddings of nonfaithful states may preserve purely infinite simplicity in other contexts.
The new Kirchberg-algebra isomorphisms indicate that infinite reduced free products with pure states are classified by K-theory alone, and the distinction between state-preserving and non-state-preserving isomorphisms is controlled by purity of the factor states.

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.

Desk Editor's Note

read the letter

Let me give you the short version. This paper closes a real gap in the selfless dichotomy: it shows that a selfless C*-probability space with nonfaithful state is purely infinite simple. That was explicitly open in Robert's work, and it completes the dichotomy so that every selfless algebra is either stably finite with stable rank one or purely infinite simple. The central mechanism, Theorem 6, is a genuinely new tail-excision criterion for infinite reduced free products. It uses nonfaithfulness positively: positive elements in the kernels of infinitely many factors let you excise any positive element. The proof of Theorem 6 checks out on the free-product Hilbert space; I rechecked the key identity a_F d_0 a_F = 0 and the containment argument, and they hold. The model-theoretic transfer from the infinite free product to the original algebra via existential embeddings is standard and fine.

What the paper does well beyond the main theorem: it removes faithful assumptions from several permanence properties, sharpens the Choda-Dykema conjecture statement to the selfless case, and the exposition is refreshingly direct. The author also labels the gap honestly.

Soft spots. Section 4 is where I get uncomfortable. Several isomorphism examples use pure states on C*(Z), and pure states need not induce faithful GNS representations. The theorems being invoked, especially Corollary 5, require GNS-faithful states. So the examples as stated are not consequences of the paper's own results. This is an overclaim in the applications, not in the core dichotomy. It should be fixed by either proving the examples under correct hypotheses or stating them as conditional. There's also a compressed step in Lemma 2.3(iv) where 'standard Cuntz semigroup analysis' covers the fact that 2[b] <= [b]; the referee should ask for that to be spelled out. The reliance on [17, 2.6(vi)] and [6, 3.13] for existential embedding and axiomatization is normal, but worth flagging for the referee to verify that the nonfaithful case is covered.

The central argument holds up. I think this deserves a serious referee. The main theorem is important, the proof is credible, and the flaws are localized to a clearly identifiable section. I'd send it out.

For a reading group, it's maybe: the core proof is worth working through, but the applications section might mislead. I'd cite it once the Section 4 issues are resolved, or with a caveat.

Referee Report

3 major / 3 minor

Summary. The paper claims to complete the selfless dichotomy for C*-probability spaces by proving Theorem 1: if (A, φ) is selfless and φ is nonfaithful, then A is purely infinite and simple. The proof passes through Theorem 6, which asserts that certain infinite reduced free products of nonfaithful GNS-faithful states are purely infinite and simple whenever a local δ_j condition holds. The announced consequences are that every selfless C*-algebra is pure, that several permanence properties can be strengthened, and that new Kirchberg-algebra isomorphisms follow from reduced free products. The main technical content is in Section 2: Lemma 2.3 is supposed to establish Cuntz comparison and the existence of a minimal essential purely infinite ideal in the free product, and Theorem 1 then transfers pure infiniteness back to a factor using existential embeddings.

Significance. If Theorem 1 were correct, it would settle the last open case of Robert's dichotomy and imply that every selfless C*-algebra is pure. The strategy of using infinitely many nonfaithful factors as 'tails' to excise arbitrary positive elements is attractive, and the model-theoretic transfer from an infinite free product to a factor is a plausible use of the existing literature. A correct proof of the dichotomy would be a substantial contribution. However, the central free-product lemma contains a false Hilbert-space calculation, and this invalidates the main theorem as written. The paper also overstates the applicability of the free-product results to pure states in Section 4.

major comments (3)

The proof asserts 'As the free product representation is faithful, a_i a_j = 0 whenever i ≠ j.' This is false. Faithfulness of the reduced free-product representation does not imply that elements from different factors annihilate each other. Concretely, take nonzero a_i ∈ (A_i)_+ and a_j ∈ (A_j)_+ with φ_i(a_i)=φ_j(a_j)=0, and choose k ∉ {i,j} and a nonzero vector h_k in the centered Hilbert space of the k-th factor. Then a_j h_k = a_j ⊗ h_k and a_i(a_j ⊗ h_k) = a_i ⊗ a_j ⊗ h_k, a nonzero reduced word in H. Hence a_i a_j ≠ 0. The same erroneous annihilation claim is used in the computation a_F d_0 a_F = ω(d) a_F^2 in (ii): if d_0 contains a centered word from A_j, then a_i d_0 a_i is generally nonzero, not zero. Since (ii) is the source of the comparison mα ≤ [b], Lemma 2.3(ii) and all results depending on it (Theorem 6, and hence Theorem 1) are unsupported.
Theorem 4.1 invokes Theorem 6 and Corollary 5 for pure states φ_j, j ≥ 2. But Theorem 6 and Corollary 5 require each state to induce a faithful GNS representation. A pure state on a non-simple C*-algebra generally has a nonzero GNS kernel; for example, a pure state on C*(Z) ≅ C(S^1) has a one-dimensional, non-faithful GNS representation. Since A_j is only assumed separable, nuclear, and UCT (not simple), the hypotheses are not met. The proof of δ_j = 1 also uses ∥x∥ = ∥π_φ(x)∥, which is exactly faithfulness of π_φ and can fail for pure states on non-simple algebras. Consequently, the Kirchberg-algebra and isomorphism claims in Section 4 do not follow from the preceding theorems as written.
Even setting aside the false excision claim, the step from 'mα ≤ [b] for every m' to '2[b] ≤ [b]' is asserted rather than proved: 'some standard Cuntz semigroup analysis grants 2[b] ≤ [b].' This is a load-bearing implication in the proof that I is purely infinite and simple. A precise proof or a stated theorem reference is needed; without it, Theorem 6(ii) is incomplete even if the earlier comparison were repaired.

minor comments (3)

In Definition 2.1, the symbol ψ is used both for a state and for a quantifier-free formula; this is confusing and should be changed (e.g., use σ(¯x,¯y)).
The abstract says the paper completes the 'stably finite/purely infinite dichotomy,' while Theorem 2 states the tracial case as stable rank one. The terminology is not always consistent; please unify.
The approximation of c by d in a finite free-product subalgebra is plausible, but the phrase 'sufficiently large so that there is self-adjoint d ... with ∥c−d∥<t/3' should justify that c is in the closure of the union of finite free products; since C is the reduced free product, this is true but should be stated.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central dichotomy proof transfers pure infiniteness through an existential embedding and does not fit or define its conclusion into its hypotheses.

full rationale

No circular step is present. Theorem 1 assumes (A, φ) selfless and nonfaithful, forms C = ∗∞(A, φ), proves C purely infinite simple via Theorem 6, and then transfers the property back to A using the existentiality of the first-factor embedding [17, 2.6(vi)]. That cited result is external prior work by Robert, not a self-citation, and the transfer is a standard preservation argument rather than an assumption equivalent to the target. Theorem 6 itself is derived from the free-product Hilbert space and Cuntz-semigroup lemmas; the δ_j > 1 condition is computed from the individual factor states and is not fitted to the conclusion. Lemma 2.3's excision identity a_F d_0 a_F = 0 is justified from the reduced-word structure of the free-product Hilbert space, so it is not a definitional identity with the theorem's conclusion. There are no fitted parameters called predictions, no ansatz smuggled in via self-citations, and no renaming of a known result as a new dichotomy. For the record, two proof details are terse—"As δ_j > 0 is constant" (Theorem 1 proof) and "Therefore, ker(φ_j)_+ ⊆ I" (Theorem 6 proof)—but these are potential correctness gaps, not circular reductions, since neither step defines its conclusion in terms of its own target.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

There are no fitted parameters and no invented entities. The paper defines δ_j and I_j as bookkeeping data computed from the factors, not as free parameters. The central claim rests on prior model theory, Cuntz semigroup facts, and the free-product Hilbert-space structure, all of which are explicit or standard.

assumptions (6)

domain assumption Purely infinite, simple C*-algebras are axiomatized by the displayed sup-inf formula ([6,3.13.7]).
Used in the proof of Theorem 1 to transfer pure-infiniteness from the infinite free product C back to A via the existential embedding. This is a nontrivial model-theoretic fact from the Memoirs volume.
domain assumption For a selfless C*-probability space, the first-factor embedding into the infinite reduced free product is existential ([17,2.6(vi)]).
Bridges Theorem 6 (pure infiniteness of the infinite free product) to pure infiniteness of A. The paper does not prove this; it is quoted from Robert.
domain assumption Any state on a purely infinite, simple C*-algebra makes it selfless ([15,3]).
Used in Corollary 5 and Theorem 4.1 to convert purely infinite simple factors into selfless probability spaces.
domain assumption For a pure state φ on a C*-algebra, ∥x+L_φ∥ = ∥x∥_φ ([19, III.2.16]).
Used to show δ_j=1 for pure factors in Theorem 4.1.
standard math Standard Cuntz semigroup facts: a≾a² for 0≤a≤1, orthogonal positive elements add in the Cuntz semigroup, and 2[b]≤[b] implies pure infiniteness.
Used throughout Lemma 2.3; the step 'some standard Cuntz semigroup analysis' in Lemma 2.3(iv) is not written out.
standard math In the reduced free-product Hilbert space, positive elements annihilated by a state act as centered operators that kill words whose first letter is different from the factor's index.
The mechanism behind a_i a_j=0 and a_F d_0 a_F=0 in Lemma 2.3.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The Selfless Dichotomy." pith.science (2026). https://pith.science/paper/RUS7YIZX

@misc{pith2026260609654,
  author       = {Pith},
  title        = {Pith review of: The Selfless Dichotomy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RUS7YIZX}},
  note         = {Machine review of arXiv:2606.09654}
}

abstract

The purpose of this note is to address the gap in the stable rank one/purely infinite dichotomy of selfless $C^*$-probability spaces. In particular, we show that nonfaithful selfless $C^*$-probability spaces are purely infinite, simple. This completes the dichotomy: Every selfless $C^*$-algebra either has stable rank one or is purely infinite. Notably, this shows that every selfless $C^*$-algebra is pure. To accomplish this, we show that infinite reduced free products of $C^*$-probability spaces with nonfaithful states inducing faithful GNS representations are often purely infinite, simple. Having resolved the dichotomy, we improve existing permanence properties of selfless $C^*$-probability spaces, make progress on a conjecture of Choda and Dykema, and produce several new isomorphisms arising from reduced free products.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Selfless C*-correspondences, operator valued C*-probability spaces and completely positive maps
math.OA 2026-07 conditional novelty 8.0 of 10

A unified theory of selfless C*-correspondences is developed and applied to completely positive maps and conditional expectations, yielding new permanence, regularity, and absorption results.

Reference graph

Works this paper leans on

22 extracted references · 7 linked inside Pith · cited by 1 Pith paper

[1]

Ramon Antoine, Francesc Perera, Hannes Thiel, and Eduard Vilalta,PureC ∗-algebras (2024), preprint, available atarXiv:2406.11052

arXiv 2024
[2]

5, Cambridge University Press, Cambridge, 1998

Bruce Blackadar,K-Theory for Operator Algebras, 2nd ed., Mathematical Sciences Research Institute Publications, vol. 5, Cambridge University Press, Cambridge, 1998

1998
[3]

Dykema,Purely infinite, simpleC ∗-algebras arising from free product constructions, III, Proc

Marie Choda and Kenneth J. Dykema,Purely infinite, simpleC ∗-algebras arising from free product constructions, III, Proc. Amer. Math. Soc.128(2000), no. 11, 3269–3273

2000
[4]

Dykema,Purely infinite, simpleC ∗-algebras arising from free product construc- tions, II, Math

Kenneth J. Dykema,Purely infinite, simpleC ∗-algebras arising from free product construc- tions, II, Math. Scand.90(2002), 73–86

2002
[5]

Dykema and Mikael Rørdam,Purely infinite, simpleC ∗-algebras arising from free product constructions, Canad

Kenneth J. Dykema and Mikael Rørdam,Purely infinite, simpleC ∗-algebras arising from free product constructions, Canad. J. Math.50(1998), no. 2, 323–341

1998
[6]

271, American Mathematical Society, Providence, RI, 2021

Ilijas Farah, Bradd Hart, Martino Lupini, Leonel Robert, Aaron Tikuisis, Alessandro Vignati, and Wilhelm Winter,Model Theory ofC ∗-Algebras, Memoirs of the American Mathematical Society, vol. 271, American Mathematical Society, Providence, RI, 2021

2021
[7]

Felipe Flores, Mario Klisse, M ´ ıche´ al´O Cobhthaigh, and Matteo Pagliero,Selfless reduced free products and graph products ofC ∗-algebras(2025), preprint, available atarXiv:2510.24675

arXiv 2025
[8]

David Gao, Srivatsav Kunnawalkam Elayavalli, Gregory Patchell, and Lizzy Teryoshin,Self- less reduced amalgamated free products and HNN extensions(2026), preprint, available at arXiv:2604.06982

arXiv 2026

Show all 22 references

[9]

Kei Hasegawa,KK-equivalence for amalgamated free productC ∗-algebras, Int. Math. Res. Not. IMRN24(2016), 7619–7636

2016
[10]

Ben Hayes, Srivatsav Kunnawalkam Elayavalli, and Leonel Robert,Selfless reduced free prod- uctC ∗-algebras(2025), preprint, available atarXiv:2505.13265

2025 arXiv
[11]

Ben Hayes, Srivatsav Kunnawalkam Elayavalli, Gregory Patchell, and Leonel Robert,Selfless inclusions ofC ∗-algebras(2025), preprint, available atarXiv:2510.13398

2025
[12]

Christopher Phillips,An isomorphism theorem for infinite reduced free products(2026), preprint, available atarXiv:2602.10220

Ilan Hirshberg and N. Christopher Phillips,An isomorphism theorem for infinite reduced free products(2026), preprint, available atarXiv:2602.10220

2026 arXiv
[13]

Akitaka Kishimoto, Narutaka Ozawa, and Shˆ oichirˆ o Sakai,Homogeneity of the pure state space of a separableC ∗-algebra, Canad. Math. Bull.46(2003), no. 3, 365–372

2003
[14]

Narutaka Ozawa,Nuclearity of reduced amalgamated free productC ∗-algebras, RIMS Kˆ okyˆ uroku1250(2002), 49–55

2002
[15]

,Proximality and selflessness for groupC ∗-algebras(2025), preprint, available at arXiv:2508.07938

2025 arXiv
[16]

Sven Raum, Hannes Thiel, and Eduard Vilalta,Strict comparison for twisted groupC ∗- algebras(2025), preprint, available atarXiv:2505.18569

2025 arXiv
[17]

Math.478(2025), Paper No

Leonel Robert,SelflessC ∗-algebras, Adv. Math.478(2025), Paper No. 110409, 28 pp

2025
[18]

Entropy in Operator Algebras, 2002, pp

Mikael Rørdam,Classification of nuclear, simpleC ∗-algebras, Classification of NuclearC ∗- Algebras. Entropy in Operator Algebras, 2002, pp. 1–145

2002
[19]

124, Springer, Berlin, Heidelberg, 2001

Masamichi Takesaki,Theory of Operator Algebras I, Encyclopaedia of Mathematical Sciences, vol. 124, Springer, Berlin, Heidelberg, 2001

2001
[20]

Klaus Thomsen,On the KK-theory and the E-theory of amalgamated free products ofC ∗- algebras, J. Funct. Anal.201(2003), no. 1, 30–56

2003
[21]

Itamar Vigdorovich,Selfless reducedC ∗-algebras of linear groups(2026), preprint, available atarXiv:2602.10616

2026 arXiv
[22]

Voiculescu, Kenneth J

Dan V. Voiculescu, Kenneth J. Dykema, and Alexandru Nica,Free Random Variables, CRM Monograph Series, vol. 1, American Mathematical Society, Providence, RI, 1992

1992

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.

[1] [1]

Ramon Antoine, Francesc Perera, Hannes Thiel, and Eduard Vilalta,PureC ∗-algebras (2024), preprint, available atarXiv:2406.11052

arXiv 2024

[2] [2]

5, Cambridge University Press, Cambridge, 1998

Bruce Blackadar,K-Theory for Operator Algebras, 2nd ed., Mathematical Sciences Research Institute Publications, vol. 5, Cambridge University Press, Cambridge, 1998

1998

[3] [3]

Dykema,Purely infinite, simpleC ∗-algebras arising from free product constructions, III, Proc

Marie Choda and Kenneth J. Dykema,Purely infinite, simpleC ∗-algebras arising from free product constructions, III, Proc. Amer. Math. Soc.128(2000), no. 11, 3269–3273

2000

[4] [4]

Dykema,Purely infinite, simpleC ∗-algebras arising from free product construc- tions, II, Math

Kenneth J. Dykema,Purely infinite, simpleC ∗-algebras arising from free product construc- tions, II, Math. Scand.90(2002), 73–86

2002

[5] [5]

Dykema and Mikael Rørdam,Purely infinite, simpleC ∗-algebras arising from free product constructions, Canad

Kenneth J. Dykema and Mikael Rørdam,Purely infinite, simpleC ∗-algebras arising from free product constructions, Canad. J. Math.50(1998), no. 2, 323–341

1998

[6] [6]

271, American Mathematical Society, Providence, RI, 2021

Ilijas Farah, Bradd Hart, Martino Lupini, Leonel Robert, Aaron Tikuisis, Alessandro Vignati, and Wilhelm Winter,Model Theory ofC ∗-Algebras, Memoirs of the American Mathematical Society, vol. 271, American Mathematical Society, Providence, RI, 2021

2021

[7] [7]

Felipe Flores, Mario Klisse, M ´ ıche´ al´O Cobhthaigh, and Matteo Pagliero,Selfless reduced free products and graph products ofC ∗-algebras(2025), preprint, available atarXiv:2510.24675

arXiv 2025

[8] [8]

David Gao, Srivatsav Kunnawalkam Elayavalli, Gregory Patchell, and Lizzy Teryoshin,Self- less reduced amalgamated free products and HNN extensions(2026), preprint, available at arXiv:2604.06982

arXiv 2026