The Daugavet property is equivalent to the polynomial Daugavet property

Sheldon Dantas, Miguel Martín, Yoël Perreau · 2025

Read The Daugavet property is equivalent to the polynomial Daugavet property by Sheldon Dantas, Miguel Martín, Yoël Perreau (2025) 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-2507-06143.txt.

Introduction

Sheldon Dantas and co-authors prove that the Daugavet property and its polynomial form are the same for Banach spaces. This property governs how operators nearly reach their norm. A long-standing open problem asked if the polynomial version is stronger.

Their proof leans on a geometric picture of the property. That picture was first stated with the weak topology. It stays valid when the weaker polynomial topology replaces it. The same trick shows linear centers are polynomial centers too. A weak operator version also implies its polynomial twin. New examples follow for tensor products of spaces. Those products now gain the property in fresh cases. The note closes a chapter in Banach space theory.

Catalog record

CREATORSheldon Dantas, Miguel Martín, Yoël Perreau
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE41 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-06143

Browse the full library · About the Tower of Babel

Full text

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

The Daugavet property is equivalent to the polynomial Daugavet property Sheldon Dantas, Miguel Martín, Yoël Perreau Abstract . In this note, we prove that the Daugavet property implies the polynomial Daugavet property, solving a longstanding open problem in the field. Our approach is based on showing that a geometric characterization of the Daugavet property due to Shvidkoy, originally formulated in terms of the weak topology, remains valid when the weak topology is replaced by the weak polynomial topology. Using similar techniques, we further establish that every linear Daugavet center is also a polynomial Daugavet center, and that the weak operator Daugavet property implies its polynomial counterpart. As an application of the latter result, we present new examples of Banach spaces whose N N -fold symmetric tensor products satisfy the Daugavet property. 2020 Mathematics Subject Classification 46B04 (primary), and 46B20; 46B25; 46B28; 46G25 (secondary) 1. Introduction Given a Banach space X over a field \mathbb{K} ( \mathbb{K}=\mathbb{R} or \mathbb{K}=\mathbb{C} ), we denote by B_{X} its closed unit ball and by S_{X} its closed unit sphere. We also denote by X^{*} the topological dual of X and by X^{**} its bidual. For Banach spaces X and Y , we write \mathcal{L}(X,Y) to denote the Banach space of all bounded linear operators from X to Y endowed with the operator norm. The symbol {\mathrm{Id}} stands for the identity operator. We will write \re z to denote the real part of z\in\mathbb{K} ; of course, when \mathbb{K}=\mathbb{R} , this simply means \re z=z . A Banach space X has the Daugavet property ( DPr , for short), if the Daugavet equation (DE) \left\|{\mathrm{Id}}+T\right\|=1+\left\|T\right\| holds for every rank-one T\in\mathcal{L}(X,X) . In this case, we have that every weakly compact operator T\in\mathcal{L}(X,X) also satisfies ( DE ) (in fact, many others classes of linear operators satisfy the DPr, see the survey [ Wer01 ] for more on the topic). This property takes its name after I. K. Daugavet’s paper from 1963, in which it is proved that compact linear operators on C[0,1] satisfy ( DE ). Examples of Banach spaces with the DPr include C(K) -spaces when K is perfect [ Dau63 ] , L_{1}(\mu) -spaces when \mu has no atoms [ Lo66 ] , Lipschitz-free spaces \mathcal{F}({M}) and Lipschitz spaces \mathop{\mathrm{}}{Lip}(M) when M is a length metric space [ IKW07 , GLPRZ18 ] , non-atomic C^{*} -algebras and preduals of diffuse von Neumann algebras [ BM05 , Oik02 ] , and some Banach algebras of holomorphic functions on an arbitrary Banach space [ Jun23 , Wer97 , Woj92 ] , among many others. It is immediate that the DPr passes from the dual to the space itself, while the other implication is false, as the above list of examples shows. We refer the reader to the recent monograph [ KMRZW25 ] for a detailed account and background on the Daugavet property. Considerable attention has been devoted to this isometric property of Banach spaces, which has several notable consequences on the isomorphic structure of the underlying Banach space. For instance, a Banach space with the Daugavet property cannot be isomorphically embedded into a Banach space with an unconditional Schauder basis (see [ KSSW00 ] ), extending a classical result by A. Pełczyński for L_{1}[0,1] . Other important isomorphic consequences of the DPr on a Banach space are that it cannot satisfy the Radon-Nikodým property and that it contains (many) isomorphic copies of \ell_{1} (hence, it cannot be Asplund). These latter consequences can be proved by using a celebrated geometric characterization of the DPr in terms of slices of the unit ball which was first obtained in [ KSSW00 ] . This result can be restated in the following way. Proposition 1.1 ( [ KSSW00 , Lemma 2.2] ) . A Banach space X has the DPr if and only if its unit ball B_{X} is equal to the closed convex hull of the set \{y\in B_{X}\colon\left\|x+y\right\|>2-\varepsilon\} for every x\in B_{X} and \varepsilon>0 . In other words, for every x\in S_{X} and every slice S of B_{X} , we have \sup_{y\in S}\left\|x+y\right\|=2 (recall that a slice of a subset C of a Banach space X is a non-empty intersection of C with an open half-space). The above characterization was further refined by R. Shvidkoy in [ Shv00 ] , after he rediscovered a geometric lemma due to J. Bourgain concerning convex combinations of slices. In particular, the following result was established, which we shall refer to as Shvidkoy’s Lemma . Lemma 1.2 ( Shvidkoy’s lemma [ Shv00 , Lemma 3] ) . Let X be a Banach space with the DPr. Then, for every x\in S_{X} and \varepsilon>0 , the set \{y\in B_{X}\colon\left\|x+y\right\|>2-\varepsilon\} is weakly dense in B_{X} . In other words, for every x\in S_{X} and y\in B_{X} , we can find a net (y_{\alpha}) in B_{X} such that y_{\alpha}\to y weakly and \left\|x+y_{\alpha}\right\|\to 2 . Following the terminology of [ KMRZW25 , § 2.6] , two elements x,y\in B_{X} are called quasi-codirected if they satisfy \|x+y\|=2 , and \varepsilon -quasi-codirected if they satisfy \|x+y\|>2-\varepsilon for some \varepsilon>0 . As a consequence of Shvidkoy’s lemma, it can be proved that a Banach space X with the DPr cannot satisfy the CPCP (in fact, every relatively weakly open subset of B_{X} has diameter 2) and that, in the separable case, its bidual contains many L -orthogonal elements (there is a weak-star G_{\delta} -dense subset \mathcal{A} of B_{X^{**}} such that \|x+x^{**}\|=\|x\|+1 for every x\in X and every x^{**}\in\mathcal{A} , see [ KMRZW25 , Corollary 3.1.18] for this version and see [ AMCRZ23 , LRZ2021 ] for further results in this line). Lemma 1.2 can be also used to show that operators not fixing copies of \ell_{1} on a Banach space with the DPr satisfy the Daugavet equation [ Shv00 , Theorem 4] . A natural extension of the DPr to polynomials was first introduced in [ CGMM07 ] , while the name appeared in the later paper [ CGMM08 ] (see Subsection 1.2 below for the necessary background on polynomials on Banach spaces). A Banach space X has the polynomial Daugavet property ( polynomial DPr , for short) if every weakly compact polynomial P\colon X\to X satisfies the equation \left\|{\mathrm{Id}}+P\right\|=1+\left\|P\right\|. As bounded linear operators are, in particular, polynomials, the polynomial DPr implies the DPr. In that same paper [ CGMM07 ] , the following geometric characterization of the polynomial DPr was provided. Proposition 1.3 ( [ CGMM07 , Corollary 2.2] ) . A Banach space X has the polynomial DPr if and only if it satisfies the following condition: given x\in S_{X} , \varepsilon>0 , and a norm-one scalar-valued polynomial p\in\mathcal{P}(X) , we can find y\in B_{X} and a modulus-one scalar \omega\in\mathbb{K} such that \real\omega p(y)>1-\varepsilon and \left\|x+\omega y\right\|>2-\varepsilon . It was proved in [ CGMM07 ] that the polynomial DPr coincides with the DPr in C(K) -spaces and in some spaces of vector-valued functions, and it was observed that a closely related property, the alternative polynomial Daugavet property, differs from its linear version. The question of whether or not the polynomial DPr and the DPr always coincide, while implicit in [ CGMM07 , CGMM08 ] , was first explicitly posed in [ MMP10 , Problem 1.2] .

Read the complete text