The Becker-Gottlieb transfer is not functorial: An example using C infinity-algebras
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.
Ada: We're looking at the paper: "The Becker-Gottlieb transfer is not functorial".
Emmy: Here's the gist of the paper. The Becker–Gottlieb transfer is not functorial, even on rational cohomology The paper proves that there exist two maps of spaces with finitely dominated fibers X g−→ Y f−→ Z such that the maps of spectra (f◦g)!
Ada: and g!
Emmy: ◦ f!
Ada: induce different maps on rational cohomology.
Title and authors: Ada: So we’re starting with the title and authors again for "The Becker-Gottlieb transfer is not functorial: An example using C infinity-algebras". It’s by Florian Naef and Robin Stoll, right? They are setting up the stage to explain what this paper is about.
Emmy: They are tackling a fundamental question in algebraic topology here, which is whether the Becker-Gottlieb transfer is functorial. Functoriality means that if you have two maps with finitely dominated fibers, the composite of their transfers should be the same as applying the transfer to the composite map.
Kolya: So what’s the big takeaway for us? If it's not functorial, it means we can't always expect that consistency when we look at these spaces using rational cohomology.
Ada: Exactly. The paper is showing that this expected consistency breaks down even when we restrict our attention to rational numbers, which is a pretty strong result in itself because it often implies a failure in the more general setting as well.
Emmy: They are specifically constructing a situation with two maps of spaces X and Z, and they show that the maps of spectra induced by composing the transfers do not induce the same map on rational cohomology.
Kolya: What kind of complexity are we talking about here? Is this just some small algebraic hiccup or something more substantial in terms of stable homotopy groups?
Ada: It’s more substantial than a small hiccup because they are using explicit geometric constructions involving an S1-bundle and a specific fibration with fiber S2 or S n, which gives them a very concrete counterexample.
Emmy: They are using these geometric pieces to show that the composite transfer is different from the composition of individual transfers on rational cohomology. That difference means they aren't homotopic.
Kolya: So, what does this mean for us when we think about the general picture of how these maps interact? Does it suggest that the relationship between maps and their induced operations isn't as rigid as we usually assume?
Ada: It suggests that the relationship is not as rigid as we think. This paper shows that even under relatively mild conditions like finitely dominated fibers, you can still find ways for the transfers to behave in a non-functorial way when viewed through the lens of rational cohomology.
The paper's summary: Ada: Now, let's summarize what the paper actually does. They are providing a detailed construction of two maps of spaces X g to Y f to Z, and then they show that the induced maps on rational cohomology from the composite transfers are different.
Emmy: They build this using an oriented S1-bundle g over X, and a fibration f from X to S n plus one with fiber S2 or S n. This setup is what allows them to create the necessary ingredients for the non-triviality they need.
Kolya: So, in simpler terms, they are showing that by setting up these specific geometric objects—the bundle and the fibration—you can engineer a situation where the transfer of their composition doesn't match what you get from composing them separately on rational cohomology.
Ada: That’s right. They are showing that this non-triviality is what they need to prove that the maps are not homotopic, which is a direct statement about their topological relationship.
Emmy: The paper then relies on previous work to model the transfer rationally by describing it as a Gysin homomorphism into lower dimensional cohomology followed by a map induced by some trace calculation involving mu S3.
Kolya: So, what’s the role of that mu S3 term in this rational description? Is it just a technical detail, or does it carry some structural information about the failure?
Ada: It’s not just technical. The paper uses that specific structure to complete the argument using the explicit description of the Hochschild homology transfer obtained in a follow-up paper, and expresses everything in terms of that C infinity-algebra S.
Emmy: So it ties together the geometry, the algebra, and then shows how it all manifests as a non-trivial map on rational cohomology through this specific algebraic structure.
Kolya: It sounds like they are showing that the geometric setup forces a specific algebraic behavior on the transfer that is not functorial.
The paper's improvements: Ada: Now, let’s look at what the paper suggests for improvement or alternative approaches. They go into this part by showing they can use C infinity-algebra models for odd spherical fibrations as an alternative proof that is more conceptual.
Emmy: This is interesting because it suggests a different path. Instead of relying solely on the Gysin homomorphism description, they build explicit C infinity-models for these fibrations and use those to provide an alternative proof of Proposition two point one <ref:2610.18718#pg1>.
Kolya: So, what’s the advantage there? Is this alternative proof easier to understand, or is it more powerful in terms of what it can prove about the structure?
Ada: It’s more conceptual because it involves constructing explicit C infinity-models for these fibrations, including circle bundles. This approach is really useful for understanding how these geometric objects behave algebraically.
Emmy: They use this to give an alternative proof of Proposition two point one, which means they are showing that the non-functoriality result can be established in a way that doesn't rely only on the initial rational models <ref:2610.18718#pg1>.
Kolya: So it’s about providing a more robust foundation for the argument, suggesting that using these C infinity-algebra models is a key way to make this kind of statement clear.
Conclusion: Ada: So we’re wrapping up. The Becker-Gottlieb transfer is not functorial even on rational cohomology, proven by constructing a specific fibration and bundle structure that forces the composite transfers to induce different maps on rational cohomology.
Emmy: And they provided an alternative proof using C infinity-algebra models for odd spherical fibrations, which is more conceptual and gives a way to see this failure through the algebraic structure of these models.
Kolya: I just want to say that seeing these explicit constructions using C infinity-algebras helps us understand the mechanism behind why this non-functoriality is happening, which is valuable for future research in stable homotopy theory.
Ada: That’s right. The paper, "The Becker-Gottlieb transfer is not functorial: An example using C infinity-algebras," provides a concrete recipe for constructing these counterexamples that show the failure of functoriality on rational cohomology.
Emmy: It’s a very powerful result because it shows that the relationship between geometric operations and algebraic structure is not always perfectly preserved, even in the rational setting.
Kolya: And we have this paper, and it sets up a clear direction for thinking about how we might look at these transfers from different angles going forward.
math.AT, math.KT
Submitted: 2026-10-13
Updated: 2026-11-06
Comments: 10 pages + 12 pages of appendix and references. v2: Improved introduction and fixed inaccuracy
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 88/100
The gist: The Becker–Gottlieb transfer is not functorial, even on rational cohomology The paper proves that there exist two maps of spaces with finitely dominated fibers X g−→ Y f−→ Z such that the…
Key concepts
- Becker–Gottlieb transfer
- This is a map between cohomology groups associated with certain fiber bundles. The paper investigates whether this transfer behaves consistently when you compose maps of spaces, which is the definition of functoriality. The core finding is that it fails this consistency check.
- Odd spherical fibrations
- These are specific types of fiber bundles where the fiber has a particular topological structure related to spheres. The authors use explicit C∞-algebra models to describe these complex geometric situations, which allows them to construct concrete counterexamples showing the transfer's non-functorial nature.
- C∞-algebra models
- These are mathematical structures used to model continuous geometric spaces and their associated bundles. The paper constructs specific C∞-algebras, like Rα, that capture the structure of these fibrations over a base space. These models are crucial for providing explicit algebraic descriptions that prove the transfer's non-functoriality.
- Functoriality on rational cohomology
- Functoriality means that if you have two maps between spaces, applying a certain operation (like the Becker–Gottlieb transfer) to the composed map should yield the same result as applying it to the individual maps separately. The paper proves this property breaks down even when only considering rational cohomology.
Terminology
Summary
The Becker–Gottlieb transfer is not functorial, even on rational cohomology The paper proves that there exist two maps of spaces with finitely dominated fibers X g−→ Y f−→ Z such that the maps of spectra (f◦g)! p. 1 and g! ◦ f! induce different maps on rational cohomology. p. 0 p. 0
The Core Counterexample
The central finding is the construction of a counterexample to functoriality using specific fibrations and bundles The authors construct a fibration f: X → S n+1 with fiber S2 ∨S n, and an oriented S1-bundle g: Y → X such that the transfer of f◦g is non-trivial on rational cohomology This non-triviality implies that the transfer of f◦g and the composite of the transfers of f and g induce different maps on rational cohomology and are thus not homotopic.
Rational Models for Transfers
The proof relies on a rational description of the transfer (f◦g)! when g is an S1-bundle This description factors as the Gysin homomorphism H∗(Eξ; Q) → H∗−1(E; Q) followed by the map induced on cohomology by S −→ k, x 7−→ − trk µS3(x, −, ξ). p. 1 The argument is completed using the explicit description of the Hochschild homology transfer obtained in the follow-up [NS26b] and expressing the result in terms of the C∞-algebra S. p. 2
Modeling Odd Spherical Fibrations
The paper provides an alternative proof using C∞-algebra models for odd spherical fibrations, which is more conceptual This involves constructing explicit C∞-models for odd spherical fibrations (including circle bundles) and using this to give an alternative proof of Proposition 2.1. The construction in Proposition A. p. 49 equips Rα:= R ⊗ Q⟨1, a⟩ with the structure of a C∞-algebra over k. p. 2
Key Structural Results
Several technical results underpin the main proof, including Proposition 3.4 which shows that for certain maps f: E → S n+1 with fibers S2 ∨ S n and an oriented S1-bundle g: Eξ → E, the Becker–Gottlieb transfer H n+1(Eξ; Q) → H n+1(S n+1; Q) is non-trivial. p. 1 The proof of Proposition 3. p. 34 utilizes the minimal dg Lie models of Sn+1 and S2 ∨ S n and shows that the resulting C∞-algebra structure maps satisfy specific relations. p. 6
Conclusion on Functoriality
The paper concludes by proving Theorem 3.5, stating that the Becker–Gottlieb transfer is not functorial, in fact, it is not even functorial on rational cohomology. p. 9 This is established by showing that there exist maps of animas with compact fibers X g−→ Y f−→ Z such that the maps of spectra g! p. 9◦ f! and (f◦g)! induce different maps on rational cohomology. The failure is detected even on rational cohomology. p. 1
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Better handling of non-functorial transfers in rational cohomology: The AI can now construct explicit counterexamples for functoriality by leveraging models for odd spherical fibrations, specifically using
the explicit Cinfty-algebra models for odd spherical fibrations (including circle bundles)
to show that the transfer of f o g is non-trivial on rational cohomology. -
Detection of failure in torsion: The system can now detect failures of functoriality in the torsion of stable homotopy groups, as demonstrated by the concurrent counterexample found by Lehner–Malkiewich–Naef, which is
of very different flavor to ours.
-
Construction and verification of Cinfty-algebra models for fibrations: The AI can model complex geometric situations using Proposition A.9 to construct a Cinfty-algebra S over k that models the fibration E over B, including the explicit structure maps mu n defined in terms of structure maps mu S n.
-
Rational description of transfers via spectral sequences: The system can use Proposition 2.1 to model the Becker–Gottlieb transfer as a composite involving
the Gysin homomorphism and the map induced on cohomology by s S −→ k, x 7−→ − trk µ S3(x, −, ξ).
-
Verification of algebraic relations in Cinfty-algebras: The AI can verify that structure maps of the constructed Cinfty-algebras satisfy the A infinity algebra relations by showing that
the sum over all terms of the forms (A.12.2) is zero, and thus (A.12.1) holds.
Sources
- The Becker-Gottlieb Transfer Is Functorial
- A rational model for the fiberwise THH transfer I: Sullivan algebras
- A rational model for the fiberwise THH transfer II: $A_\infty$-algebras
Related papers
- An Equivariant Landweber Exact Functor Theorem for Abelian Compact Lie Groups
- Rational homotopy theory of flag manifolds
- K-Theoretic Obstructions to Linearizing QCA Representations
- The sublevel Flood bifiltration: towards scalable 2-parameter persistent homology
- Discrete homotopy groups of cubical sets
- A Note on Topological Hochschild Homology Relative to W(k)[x 0,x 1,,x n]