Range strongly exposing operators between Banach spaces
Geunsu Choi, Helena del Río, Audrey Fovelle, Mingu Jung, Miguel Martín · 2025
Read Range strongly exposing operators between Banach spaces by Geunsu Choi, Helena del Río, Audrey Fovelle, Mingu Jung, Miguel Martín (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-2503-18581.txt.
Introduction
Geunsu Choi and four co-authors define a new class of operators between Banach spaces, spaces of functions or vectors with a size rule. Some operators reach their largest value on some input; these are called norm-attaining, and they anchor a rich theory. Their 2025 paper sets out range strongly exposing operators, a class that sits between a strong form due to Bourgain and the classic attaining kind. Basic results from the older theory are carried over to the new class.
One main result shows that for any infinite-dimensional target space, some source space makes these operators fail to be dense. Another links a classic property of the target space to good behavior of the class. The authors offer the class as a sharper tool for a long-running line of analysis.
Catalog record
| CREATOR | Geunsu Choi, Helena del Río, Audrey Fovelle, Mingu Jung, Miguel Martín |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 106 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-18581
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Range strongly exposing operators between Banach spaces Geunsu Choi, Helena del Río, Audrey Fovelle, Mingu Jung, Miguel Martín Abstract . We introduce a new class of bounded linear operators, called range strongly exposing ( RSE , in short) operators, which form a natural intermediate class: weaker than absolutely strongly exposing operators, yet stronger than both uniquely quasi norm-attaining and classical norm-attaining operators. Several foundational results on norm-attaining operators are extended to the RSE setting. Among our main contributions, we establish that for every infinite-dimensional Banach space Y Y , there exists a Banach space X X such that the RSE operators from X X to Y Y are not dense—an RSE analogue of a result by Acosta (1999) which applies only when Y Y is strictly convex. We also show that the Radon-Nikodým property of Y Y is sufficient to obtain that RSE operators from L 1 ( μ ) L_{1}(\mu) to Y Y are dense and that this is also necessary if μ \mu is not purely atomic. This extends and sharpens classical results by Uhl (1976). As a consequence, we prove that the set of RSE operators between L 1 ( μ ) L_{1}(\mu) and L 1 ( ν ) L_{1}(\nu) is dense if and only if at least one of the measures μ \mu or ν \nu is purely atomic, in contrast with the classical result by Iwanik (1979) which guarantees the denseness of norm-attaining operators for all measures μ \mu and ν \nu . We also prove that weakly compact operators from any C ( K ) C(K) space can always be approximated by (weakly compact) RSE operators, thereby strengthening a result of Schachermayer (1983). Additionally, we present several improvements of more recent results concerning finite-rank operators and Γ \Gamma -flat operators which give, in particular, RSE versions of classical results on compact operators by Johnson–-Wolfe (1979). Finally, we discuss RSE counterparts of results by Zizler and Lindenstrauss on the denseness of operators whose adjoints attain their norm. 2020 Mathematics Subject Classification Primary 46B04; Secondary 46B20, 46B22, 46B25, 47B01, 47B07 1. Introduction Let X and Y be Banach spaces over the field \mathbb{K} (which will be the field \mathbb{R} of real numbers or the field \mathbb{C} of complex numbers). Our notation is standard, as in most textbooks on Banach space theory; we refer, e.g., to [ 18 , 36 ] . For example, we denote by B_{X} and S_{X} the closed unit ball and the unit sphere of X , respectively, and \mathbb{T}:=S_{\mathbb{K}} is the set of modulus one scalars, \mathcal{L}(X,Y) is the space of all bounded linear operators from X into Y , and X^{*}=\mathcal{L}(X,\mathbb{K}) denotes the (topological) dual space of X . Most of the standard concepts from Banach space theory used in the paper (such as rotundity, smoothness, and various operator ideals) can be found in Subsection 1.1 at the end of this introduction. It is said that T\in\mathcal{L}(X,Y) attains its norm (or that T is norm-attaining ) if the operator norm \|T\| coincides with \max\{\|Tx\|\colon x\in B_{X}\} , that is, if T(B_{X})\cap\|T\|S_{Y}\neq\emptyset . We denote by \NA(X,Y) the set of norm-attaining operators from X into Y , and write \NA(X) when Y=\mathbb{K} for simplicity. The study of norm-attaining operators goes back to the seminal paper by Lindenstrauss [ 33 ] , who proved, among other results, that there are Banach spaces X and Y for which \NA(X,Y) fails to be dense in \mathcal{L}(X,Y) , and that \NA(X,Y) is dense if either X is reflexive or if Y has a geometrical property called Lindenstrauss’ property \beta (which is satisfied by finite-dimensional polyhedral spaces and by closed subspaces of \ell_{\infty} containing the canonical copy of c_{0} , among many other examples). To simplify the notation, Lindenstrauss introduced the following terminology. A Banach space X has property A if \NA(X,Y) is dense in \mathcal{L}(X,Y) for every target space Y . Analogously, a Banach space Y has property B if \NA(X,Y) is dense in \mathcal{L}(X,Y) for every domain space X . In 1977, Bourgain linked in [ 7 ] the study of norm-attaining operators with the Radon-Nikodým property (for short, RNP) by showing that the RNP implies property A and presented a sort of reciprocal result. Huff improved the result further in [ 24 ] by showing that X must have the RNP if any renorming of X has property A. We refer the reader to the expository papers [ 5 , 14 , 35 ] and the recent papers [ 6 , 13 , 20 , 28 , 29 ] and references therein for more information and background. Let us mention that both properties A and B are of isometric nature, that is, depend on the concrete norm we are using in the involved spaces, and that every separable Banach space can be renormed to have property A and every Banach space can be renormed to have property B. To relate Bourgain’s work with the concept we are introducing in this paper, let us describe it more in detail. What was actually proved in [ 7 ] is that the set of all absolutely strongly exposing operators (denoted by \ASE(X,Y) ) is dense in \mathcal{L}(X,Y) when X has the RNP. Recall that T\in\ASE(X,Y) if there exists x_{0}\in S_{X} such that whenever a sequence (x_{n})\subseteq B_{X} satisfies that \|Tx_{n}\|\to\|T\| , one can find a subsequence (x_{\sigma(n)})\subseteq(x_{n}) and \theta\in\mathbb{T} such that x_{\sigma(n)}\to\theta x_{0} . That is, T attains the norm at a unique element (up to rotation) in the strong way that all sequences in B_{X} approximating the norm have to converge (up to rotation) to this element. This is, of course, a very strong concept. For instance, the point at which T attains its norm has to be a strongly exposed point of B_{X} (immediately) and the denseness of \ASE(X,Y) for a non-zero Y implies the denseness of the set of strongly exposing functionals in X^{*} [ 28 , Proposition 1.5] (see Subsection 1.1 for the definitions of strongly exposed points and strongly exposing functionals). We refer the reader to the recent paper [ 28 ] for extensions of Bourgain’s work. Although Bourgain’s positive result is both strong and useful when X has the RNP, the concept of absolutely strongly exposing operator turns out to be highly restrictive when the space X fails the RNP and is entirely useless when B_{X} has no strongly exposed points. This occurs in classical Banach spaces such as infinite-dimensional C(K) spaces and L_{1}(\mu) spaces when \mu is atomless. Let us now turn to property B, the counterpart to property A for range spaces. This property is also connected to the RNP: If every equivalent renorming of Y has property B, then Y has the RNP [ 7 , 24 ] . However, unlike property A, the converse is false; as observed by Gowers [ 22 ] , the Banach spaces \ell_{p} fail to have property B for 1<p<\infty . Actually, the same happens for \ell_{1} [ 4 ] , for all strictly convex infinite-dimensional Banach spaces [ 3 ] , and for \ell_{p} -sums of finite-dimensional Banach spaces for 1<p<\infty [ 20 ] . Nevertheless, there remains a way to connect the RNP of a target