Contraction properties for holomorphic functions via isoperimetric stability on the Bergman ball

David Kalaj, Jian-Feng Zhu · 2026

Read Contraction properties for holomorphic functions via isoperimetric stability on the Bergman ball by David Kalaj, Jian-Feng Zhu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-22524.txt.

Introduction

Kalaj and Zhu prove a local contraction property for holomorphic functions that are nearly constant. The setting is weighted Bergman spaces on the Bergman ball. Their route turns shape facts about weighted level sets into analytic deficit bounds. A key input is a sharp stability result for the isoperimetric inequality in that ball. Along one special line of exponents, they show the target deficit is controlled by the source deficit. The same bound also measures how far the level sets sit from perfect spheres.

Catalog record

CREATORDavid Kalaj, Jian-Feng Zhu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE56587
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.22524

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Contraction properties for holomorphic functions via isoperimetric stability on the Bergman ball David Kalaj, Jian-Feng Zhu Abstract . We prove a local contraction property for holomorphic functions that are nearly constant, relating weighted Bergman spaces A α p ​ ( 𝔹 n ) A^{p}_{\alpha}(\mathbb{B}_{n}) and A β q ​ ( 𝔹 n ) A^{q}_{\beta}(\mathbb{B}_{n}) . Our approach converts geometric information on weighted superlevel sets into analytic deficit inequalities and rests crucially on a quantitative stability result (of Fuglede type) for the isoperimetric inequality in the Bergman ball. As an application, along the contractive line q / p = β / α q/p=\beta/\alpha , we obtain a deficit contraction near the extremizer f ≡ 1 f\equiv 1 : if f = 1 + ϕ f=1+\phi with ϕ \phi small and its weighted level sets are nearly spherical (after recentering), then the A β q A^{q}_{\beta} -deficit is controlled by the A α p A^{p}_{\alpha} -deficit, and the same deficit quantitatively controls the deviation of the level sets from spheres. 1. Introduction 1.1. Introduction to the local theory A central theme of this work is the relationship between analytic norms of holomorphic functions on the unit ball and the geometry of their level sets with respect to the invariant (Bergman) measure. In particular, extremal inequalities for weighted Bergman and Hardy norms are closely tied to sharp isoperimetric properties of sublevel and superlevel sets of nonlinear functionals of the form z\longmapsto|f(z)|^{p}(1-|z|^{2})^{\alpha}. A guiding principle in geometric–analytic inequalities is that extremizers are highly symmetric and that near-extremizers inherit quantitative symmetry. Our approach links contraction phenomena for holomorphic function spaces to a geometric analysis of weighted superlevel sets: a monotonicity principle for their hyperbolic distribution functions, combined with quantitative isoperimetric stability (Fuglede-type estimates) in the Bergman ball. In the one-dimensional setting, recent work of Kulikov shows that broad classes of Hardy and Bergman functionals are maximized at normalized reproducing kernels, leading to sharp contractive embeddings. Related contraction results in higher dimensions appear in several forms; see, for instance, [ 5 , 9 , 11 ] . The novelty here is that isoperimetric stability can be leveraged to produce local contraction estimates for perturbations of the constant extremizer f\equiv 1 , i.e. f=1+\phi with \phi small, under a quantitative near-sphericity assumption on the relevant weighted level sets. The constant function f\equiv 1 plays a distinguished role: it is an extremizer for a wide class of sharp inequalities, and its level sets are exactly geodesic balls for the Bergman metric. A natural question is whether these extremal properties are stable and whether they persist, at least locally, under small perturbations of the extremizer. The first main results, Theorem 3.2 and Theorem 3.4 , establish a local contraction principle along the contractive line \frac{q}{p}=\frac{\beta}{\alpha}, for holomorphic functions that are sufficiently close to 1 in a C^{1} sense on compact subsets of \mathbb{B}_{n} . Under mild geometric assumptions on the weighted superlevel sets—namely, a barycenter normalization, a volume cap, and a small radial-graph condition—these theorems yield (i) distribution dominance of the superlevel volumes on a centered interval and (ii) a quantitative stability estimate in which the deficit in a stronger Bergman norm controls a coercive shape functional measuring deviation from spherical symmetry. In particular, dropping the nonnegative remainder term in Theorem 3.4 yields the (local) Bergman–Bergman contraction stated in Corollary 3.3 . The proof is given in Subsection 3.4 . Building on Proposition 3.7 , Theorem 3.5 connects the normalized weighted Bergman scale to the Hardy endpoint for holomorphic functions near f\equiv 1 . The critical-limit proposition is proved by combining the monotonicity of invariant integral means with the polar disintegration of the invariant measure: as \alpha\downarrow n , the normalized weight (1-|z|^{2})^{\alpha}dv_{g} concentrates in an outer annulus, forcing the Bergman average to approach the boundary H^{rn} average. Under the barycenter and small radial-graph hypotheses on the level sets, the local isoperimetric/Fuglede mechanism then yields the norm chain \|f\|_{A_{\beta}^{q}(\mathbb{B}_{n})}\leq\|f\|_{A_{\alpha}^{p}(\mathbb{B}_{n})}\leq\|f\|_{H^{nr}(\mathbb{B}_{n})},\qquad\frac{q}{p}=\frac{\beta}{\alpha},\ \ r=\frac{p}{\alpha}, and the Hardy endpoint agrees with the critical limit \gamma\downarrow n via the Poisson–Szegő majorant, Minkowski’s inequality, and a Forelli–Rudin estimate for the Poisson kernel. These results provide a local, quantitative counterpart to a conjecture of Nicola–Riccardi–Tilli [ 14 , Conjecture 5.1] . That conjecture asserts that, for every convex function \Phi:[0,1]\to\mathbb{R} , the extremizer f\equiv 1 maximizes the functional \int_{\mathbb{B}_{n}}\Phi\!\left(|f(z)|^{p}(1-|z|^{2})^{\alpha}\right)\,dv_{g}(z) among all holomorphic functions normalized in A_{\alpha}^{p}(\mathbb{B}_{n}) . This may be interpreted as a convex-order domination principle for the random variable U_{f}(z):=|f(z)|^{p}(1-|z|^{2})^{\alpha} with respect to the invariant measure. We now recall the following conjecture of Nicola–Riccardi–Tilli [ 14 , Conjecture 5.1] , and later in this section we prove its local validity near f\equiv 1 . Conjecture 1.1 . [ 14 , Conjecture 5.1] Let \Phi:[0,1]\to\mathbb{R} be a convex function. For every p\in(0,\infty) , \alpha>n , and every holomorphic function f\in A^{p}_{\alpha}(\mathbb{B}_{n}) normalized by \|f\|_{A^{p}_{\alpha}}=1, one has \int_{\mathbb{B}_{n}}\Phi\!\left(|f(z)|^{p}(1-|z|^{2})^{\alpha}\right)\,dv_{g}(z)\ \leq\ \int_{\mathbb{B}_{n}}\Phi\!\left((1-|z|^{2})^{\alpha}\right)\,dv_{g}(z). In the perturbative regime f=1+\phi , using the contraction and stability results above, we show that when \|\phi\|_{C^{1}} is sufficiently small, the distribution of U_{f} is dominated by that of the extremizer w(z)=(1-|z|^{2})^{\alpha} in convex order. This yields the conjectured inequality for all convex \Phi near f\equiv 1 , and highlights the geometric mechanism behind the conjecture: monotonicity of level-set measures is associated with quantitative isoperimetric stability in the Bergman ball. 2. Preliminaries 2.1. Hardy and Bergman preliminaries on the disk Let \mathbb{D}=\{z\in\mathbb{C}:|z|<1\} . For \alpha>1 and 0<p<\infty , we say that a function f in \mathbb{D} belongs to the Bergman space A^{p}_{\alpha} if \|f\|_{A^{p}_{\alpha}}^{p}:=\int_{\mathbb{D}}(\alpha-1)|f(z)|^{p}(1-|z|^{2})^{\alpha-2}\,\frac{\,\mathrm{d}x\,\mathrm{d}y}{\pi}<\infty. For 0<r<\infty , we say that a function f analytic in \mathbb{D} belongs to the Hardy space H^{r} if \|f\|_{H^{r}}^{r}:=\sup_{0<\rho<1}\int_{0}^{2\pi}|f(\rho e^{i\theta})|^{r}\frac{\,\mathrm{d}\theta}{2\pi}<\infty. We recall an extremal principle (in the spirit of Kulikov) for convex functionals on Bergman spaces. Theorem 2.1 ( [ 9 ] Extremizers in the disk Bergman space) . Let G:[0,\infty)\to\mathbb{R} be convex. Then the maximum of \int_{\mathbb{D}}G\bigl(|f(z)|^{p}(1-|z|^{2})^{\alpha}\bigr)\,\,\mathrm{d}m(z) over all f\in A^{p}_{\alpha} with \|f\|_{A^{p}_{\alpha}}=1 is attained at constant functions (equivalently, at normalized reproducing kernels via automorphisms). Applying G(t)=t^{s} with s>1 yields the contractive embedding chain: Corollary 2.2 ( [ 9 ] Disk contraction chain) . For 0<p<q<\infty and 1<\alpha<\beta<\infty with \frac{p}{\alpha}=\frac{q}{\beta}=r , every analytic f satisfies \|f\|_{A^{q}_{\beta}}\leq\|f\|_{A^{p}_{\alpha}}\leq\|f\|_{H^{r}}. Equality holds for constants (equivalently, for normalized reproducing kernels). Remark 2.3 . The Möbius group acts on A^{p}_{\alpha} by g(z)=f\!\left(\frac{z-\overline{w}}{1-zw}\right)\frac{(1-|w|^{2})^{\alpha/p}}{(1-zw)^{2\alpha/p}},\qquad w\in\mathbb{D}, preserving both \|\,\cdot\,\|_{A^{p}_{\alpha}} and the distribution of |f(z)|^{p}(1-|z|^{2})^{\alpha} . 2.2. Hardy and weighted Bergman spaces on the unit ball Let \mathbb{B}_{n}:=\{z\in\mathbb{C}^{n}:\ |z|<1\} denote the unit ball and let \mathbb{S}:=\partial\mathbb{B}_{n} be its boundary. We write d\sigma for the normalized surface measure on \mathbb{S} and d\nu for the Lebesgue volume measure on \mathbb{B}_{n} . 2.3. Hardy spaces A holomorphic function f:\mathbb{B}_{n}\to\mathbb{C} belongs to the Hardy space H^{p}(\mathbb{B}_{n}) , 0<p<\infty , if \|f\|_{H^{p}(\mathbb{B}_{n})}^{p}:=\sup_{0<r<1}\int_{\mathbb{S}}|f(r\zeta)|^{p}\,d\sigma(\zeta)<\infty. For f\in H^{p}(\mathbb{B}_{n}) the radial limits f^{*}(\zeta)=\lim_{r\uparrow 1}f(r\zeta) exist for a.e. \zeta\in\mathbb{S} , and one has \|f\|_{H^{p}(\mathbb{B}_{n})}^{p}=\int_{\mathbb{S}}|f^{*}(\zeta)|^{p}\,d\sigma(\zeta). For p\geq 1 this defines a norm, while for 0<p<1 it is a quasi-norm. 2.4. Invariant (Bergman) measure and normalized weighted Bergman spaces Let \mathbb{B}_{n}:=\{z\in\mathbb{C}^{n}:\ |z|<1\} and let d\nu denote Lebesgue volume measure on \mathbb{B}_{n} . We define the invariant (Bergman) volume element by dv_{g}(z):=c_{n}\,(1-|z|^{2})^{-(n+1)}\,d\nu(z), where c_{n}>0 is a fixed normalizing constant (e.g. chosen so that \int_{\mathbb{B}_{n}}dv_{g}=1 ). This measure is invariant under holomorphic automorphisms of the ball: for every \varphi\in\Aut(\mathbb{B}_{n}) and every integrable F ,

Read the complete text