gpt 6 astra pro attempt on Math-117

Positive Scalar Curvature on Aspherical Manifolds

Definitions and mathematical statement

A closed manifold is compact without boundary, and asphericity means its universal cover is contractible. The scalar curvature of a Riemannian metric is the trace of its Ricci curvature. The target is \[ M\text{ closed and aspherical}\quad\Longrightarrow\quad \nexists g\text{ on }M\text{ with }\operatorname{Scal}_g(x)>0\text{ for every }x. \] Strict pointwise positivity is intended. A prohibition only on positive sectional curvature is weaker, and additional spin or fundamental-group assumptions would restrict the conjecture.

Formulation and concept references: [S1].

Short English statement

Does asphericity rule out every everywhere-positive scalar-curvature metric on a closed manifold?

Research attempt

Editorial source note.

The original source entry is retained unchanged. Its URL identifies a paper by Shuli Chen, Jianchun Chu, and Jintian Zhu, not by Simone Cecchini, Jianchun Chu, and Jintian Zhu. The corrected bibliographic entry is added below. [E1]

Current frontier.

The December 2025 version of that paper records the closed aspherical obstruction through dimension five, and proves stronger connected-sum and noncompact-domination results in dimensions three, four, and five. It explicitly distinguishes these from the general higher-dimensional problem. Its introduction also explains why asphericity alone need not supply the homology classes used in elementary dimension descent. [E1], introduction and Theorems 1.2 and 1.8. The spinorial curvature obstruction used below is standard; Listing's formulation makes the operator-norm dependence of the twisted Bochner formula explicit. [E2]

Attack routes.

The index route needs a nonzero index that survives twisting by bundles of arbitrarily small curvature. Contractibility of the universal cover does not directly construct those bundles. The minimal-hypersurface route needs essential smooth hypersurfaces and an inductively preserved topological obstruction. A counterexample route would construct a positive-scalar-curvature metric without accidentally losing asphericity through surgery. We pursue the first two estimates and then identify the topology that neither estimate provides. Spin is an additional hypothesis of one route, not of the original problem.

Attempt I: the exact analytic input for a spin obstruction.

Let $M^n$, $n\geq2$, be closed, spin, and even-dimensional, with a hypothetical metric $g$ of scalar curvature at least $\kappa>0$. The positive constant exists because $M$ is compact and the curvature is strictly positive. For a Hermitian bundle $E$ with unitary connection, the twisted spin Dirac operator satisfies \begin{equation}\label{wsref-0-33-30-30-30-35-38-37-30-3a-64-69-72-61-63} D_E^2=\nabla^*\nabla+\frac{R_g}{4}+\mathcal R_E, \qquad \mathcal R_E=\sum_{i<j}c(e_i)c(e_j)F_E(e_i,e_j). \end{equation} Here $c$ is Clifford multiplication and $F_E$ is curvature. This is the twisted Bochner--Lichnerowicz identity used in [E2], Sections 2 and 3.3. Define $\|F_E\|$ by the supremum of the operator norm on orthonormal pairs of tangent vectors. Since unit Clifford multiplications have norm one, \[ \|\mathcal R_E\|\leq\binom n2\|F_E\|. \] In particular this bound does not acquire an extra factor equal to $\operatorname{rank}E$.

If $\|F_E\|<\kappa/(8\binom n2)$, integration of (30005870:dirac) for a smooth spinor $s$ gives \[ \|D_Es\|_2^2\geq\|\nabla s\|_2^2+\frac\kappa8\|s\|_2^2. \] Both chiral kernels vanish, hence $\operatorname{index}D_E^+=0$. Therefore the following extra geometric condition rules out $g$: \begin{equation}\label{wsref-0-33-30-30-30-35-38-37-30-3a-62-75-6e-64-6c-65} \text{for every $\eta>0$ there is $E$ with } \|F_E\|<\eta\text{ and }\operatorname{index}D_E^+\ne0. \end{equation} This implication is completely proved. The issue is whether the appropriate topological construction supplies (30005870:bundle) for the manifold under consideration; the implication does not prove its hypothesis. Nontriviality of $E$ alone is not enough, since it need not pair nontrivially with the Dirac index.

For an odd-dimensional spin manifold, the same conditional obstruction can be applied to $M\times S^1$: a product with a flat circle preserves strictly positive scalar curvature and gives even dimension. It still requires the corresponding nonzero twisted indices on the product. For a nonspin manifold, passing to the contractible universal cover makes that cover spin but noncompact. The closed Fredholm-index argument just given then does not apply. Existence of a finite spin cover has not been assumed or proved. An orientable double cover does not in general remove the spin obstruction.

Attempt II: what stable minimal hypersurfaces do provide.

Suppose $\Sigma^m\subset M^{m+1}$ is a closed smooth two-sided stable minimal hypersurface, and $R_M\geq\kappa>0$. Stability and the Gauss identity give, for every smooth real $\phi$ on $\Sigma$, \[ \int_\Sigma|\nabla\phi|^2\geq \int_\Sigma(\operatorname{Ric}_M(\nu,\nu)+|II|^2)\phi^2, \quad R_\Sigma=R_M-2\operatorname{Ric}_M(\nu,\nu)-|II|^2. \] Eliminating the normal Ricci term yields the useful lower bound \begin{equation}\label{wsref-0-33-30-30-30-35-38-37-30-3a-73-74-61-62-6c-65} \int_\Sigma\left(2|\nabla\phi|^2+R_\Sigma\phi^2\right) \geq\int_\Sigma(R_M+|II|^2)\phi^2 \geq\kappa\int_\Sigma\phi^2. \end{equation} This reconstructs the scalar-curvature part of the dimension-descent mechanism underlying the source's geometric discussion. [E1], Sections 1.1 and 2.6.

For $m\geq3$, the conformal Laplacian has gradient coefficient $a_m=4(m-1)/(m-2)>2$. Its first eigenvalue is positive by (30005870:stable), and its first eigenfunction $u$ can be chosen strictly positive. The scalar curvature of the conformally changed metric $u^{4/(m-2)}g_\Sigma$ is \[ u^{-(m+2)/(m-2)}(-a_m\Delta_\Sigma u+R_\Sigma u)>0. \] For an orientable surface $m=2$, using $\phi=1$ in (30005870:stable) gives $\int R_\Sigma>0$, hence positive Euler characteristic by Gauss--Bonnet. A connected such surface is a sphere. Thus a smooth stable hypersurface inherits a precise positive-scalar- curvature conclusion, with no unproved sign in the second variation.

The desired induction would now need this $\Sigma$ to carry an obstruction contradicting its positive curvature, for example to be aspherical. That conclusion is absent. An embedded hypersurface of an aspherical manifold need not be aspherical: the boundary of a small coordinate ball is a sphere. This example does not claim that the sphere is stable minimal in a hypothetical positive-scalar-curvature metric; it refutes the purely topological inheritance assertion that an attempted induction would otherwise use. Essentiality, existence, and stability must all be supplied together.

Even constructing essential hypersurfaces by dualizing first cohomology cannot cover all aspherical manifolds. The formulation source notes aspherical homology-sphere examples in dimension four. A strategy requiring nonzero first Betti number has already restricted the target. In higher dimensions a minimizing hypersurface may also have a singular set; the smooth conformal argument cannot simply be iterated through that set without a separate theorem. [E1], introduction.

A counterexample-oriented check.

Attaching handles or taking connected sums is a familiar way to alter a metric while retaining positive scalar curvature under suitable surgery hypotheses. It is not a way to preserve asphericity for free. For example, the connected sum with $S^2\times S^{n-2}$, in dimensions where this factor is simply connected, introduces an explicit higher-homotopy obstruction detectable in the universal cover. No surgery construction is presented here with a verified contractible universal cover and an everywhere-positive scalar-curvature metric. Thus these metric modifications do not furnish a counterexample to the catalogue's exact class.

Stress test.

The index estimate uses curvature in operator norm, not a trace bound that could deteriorate as the twisting rank grows. The inequality is applied on a closed manifold, where the ordinary chiral index is defined; it is not extended to a noncompact cover without Fredholm hypotheses. The minimal-hypersurface estimate assumes smoothness, two-sidedness, and stability explicitly, and establishes scalar, not sectional, curvature. Neither spin nor an abundance of degree-one cohomology has been added to the universal statement.

The companion symbolic checks verify the elimination of the Ricci term and the inequalities between the conformal-Laplacian coefficients. They test algebraic signs only; they do not construct missing bundles, essential hypersurfaces, or singularity-removal arguments.

Outcome.

Outcome: Exploratory approach with unresolved gap.

Both analytic mechanisms have been pushed to precise sufficient hypotheses. Small-curvature nonzero-index twists give a contradiction in the indicated spin setting, and smooth stable hypersurfaces inherit positive scalar curvature in the stated sense. The general aspherical topology has not been shown to supply the objects required by either mechanism, and the nonspin and high-dimensional difficulties remain. No extension of the established dimension range or new counterexample is claimed.

Sources

• S1 Simone Cecchini, Jianchun Chu, and Jintian Zhu, Positive scalar curvature metrics and aspherical summands, arXiv:2312.04698v3 $2025$, introduction. https://arxiv.org/html/2312.04698. Catalogue formulation source.
• E1 Shuli Chen, Jianchun Chu, and Jintian Zhu. Positive scalar curvature metrics and aspherical summands, arXiv:2312.04698v3, 18 December 2025. Introduction, Theorems 1.2 and 1.8, and the geometric preliminaries. Corrects the first author in the retained catalogue reference. https://arxiv.org/html/2312.04698v3. Consulted 27 September 2026 for the indicated result or scope.
• E2 Mario Listing. Scalar curvature and vector bundles, arXiv:1202.4325v1, 20 February 2012. Sections 2 and 3.3, including the twisted Dirac curvature term and its operator-norm estimate. https://arxiv.org/html/1202.4325v1. Consulted 27 September 2026 for the indicated result or scope.