On Matricial Order Operator Spaces
Roy Araiza and Timur Oikhberg · 2026
Read On Matricial Order Operator Spaces by Roy Araiza and Timur Oikhberg (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-2605-21982.txt.
Introduction
Roy Araiza and Timur Oikhberg study matricial order operator spaces, a generalization of operator systems carrying matricial norms plus matricial order. For these objects they develop duality theory, naming two properties for the interplay of order and norm, called normality and generation, and proving the two are dual to each other. Operator systems, including C-star algebras, and Schatten spaces serve as examples. The authors also lay out the minimal and maximal matricial order structures, again in duality, and show that Banach lattices can carry such structures.
Catalog record
| CREATOR | Roy Araiza and Timur Oikhberg |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 110696 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.21982
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On Matricial Order Operator Spaces Thanks: During this project the first author was supported as a J.L. Doob Research Assistant Professor at the University of Illinois at Urbana-Champaign, and as a Trond Mohn Stiftelse Guest Professor at Universitetet I Oslo. The authors wish to express their gratitude to E. Bilokopytov and E. Pernecka for many helpful comments. Roy Araiza and Timur Oikhberg Abstract . We investigate the category of “matricial order operator spaces,” which generalize operator systems, being equipped with both matricial norms and matricial order. For these objects, we develop duality theory. Taking a cue from the theory of ordered normed spaces, we introduce two important properties describing the interplay between order and norm – “normality” and “generation,” and show that they are dual to each other. As examples, we consider operator systems (in particular, C ∗ -algebras), and Schatten spaces. We also describe the minimal and maximal matricial order structures (which, again, turn out to be in duality), and show how Banach lattices can be equipped with such structures. 1. Introduction The present paper is devoted to investigating matricial order operator spaces (defined below). Our goal is to unify two streams of Functional Analysis: (i) the theory of ordered normed spaces, and (ii) operator space theory. Ordered spaces (such as function spaces) have been studied since the dawn of functional analysis. For a general approach, we refer the reader to e.g. [ 6 ] or [ 4 ] . Below, we recall the facts and definitions relevant to this work. Operator spaces arose in the late 1980s, as subspaces of C^{*} -algebras. Alternatively, one equips a given normed space with a matricial structure, satisfying the so called Ruan’s Axioms. Accessible monographs on the topic include [ 26 ] , [ 28 ] , and [ 14 ] ; again, the most pertinent results are recalled below. From the beginning of operator space theory, people were interested in ( unital ) operator systems – that is, involution-invariant subspaces of B(H) which contain the identity, and hence naturally carry “matricial order” (see e.g. [ 26 ] ). Instead of concrete embeddings into B(H) , one can give “abstract” descriptions of such objects, involving interactions of the unit, the matricial norms, and the matricial order, see [ 11 ] . Recently, general (not necessarily unital) ordered operator spaces have come into focus, due to their connections to non-commutative convexity (see e.g. [ 21 ] ). Here, the order is required to satisfy certain conditions (which we make explicit below), but doesn’t need to arise from a specific embedding into B(H) . In fact, [ 20 ] already notes the difficulty of obtaining such embeddings of duals of C^{*} -algebras. All this points to the necessity of developing a general theory of “matricial order” on operator spaces, which is the purpose of the present work. We begin this paper by laying the necessary foundation. In Section 2 , we briefly introduce the notion of an ordered Banach space, and review “order” versions of local reflexivity. Apart from being of independent interest, local reflexivity will also be used to describe the duality between minimal and maximal matricial order structures in Section 7 . Section 3 examines matricially normed spaces (operator spaces), and also matricial convexity and duality. We then move on to the core of the present work – namely, the examination of matricial order. Section 4 defines matricial order operator spaces , and introduces their basic properties, such as normality and generation (their classical counterparts are described in Section 2 ). In Section 5 we describe a way to renorm a normal, geneating matricial order space to make it 1 -normal and 1 -generating. It is easy to observe that the dual of a matrical order operator space can be equipped with a matricial order operator space structure. In Section 6 we show that the dual of \mathcal{X} is normal if and only if \mathcal{X} itself is generating, and vice versa. Taking a cue from the theory of operator spaces, in Section 7 we define the maximal and minimal matricial order structures on a given ordered Banach space, and show that they are in duality. We also develop criteria for normality and generation of these “maximal” and “minimal” matricial order operator spaces. Section 8 is devoted to investigating “natural” examples of matricial order operator spaces – namely, C^{*} -algebras (and operator systems more generally) and Schatten spaces. In Section 9 we study a new class of matricial order operator spaces obtained by equipping Banach lattices with their minimal or the maximal structures. For functional analysis background, see [ 2 ] . We mostly adhere to the standard notation; however, some special symbols are used, to wit: \bullet If \mathcal{X} is a vector space, we shall use \mathcal{X}^{\flat} for a space of linear functionals on it. The dual action (linear in both variables) is denoted by \langle\cdot,\cdot\rangle:\mathcal{X}^{\flat}\times\mathcal{X}\to\mathbb{C} . If \mathcal{X} is a topological vector space, then \mathcal{X}^{\flat} stands for the space of continuous functionals. If \mathcal{X} is a normed space, then we equip \mathcal{X}^{\flat} with the usual dual norm. We adopt this unusual notation, to avoid giving double duty to “ * :” when x is an operator on a Hilbert space (or a matrix), x^{*} shall stand for the adjoint of x . \bullet Following the conventions of mathematical physics, we use the “bra” and ”ket” for Hilbert space inner products: the map H\times H\to\mathbb{C}:(\xi,\eta)\mapsto\langle\xi|\eta\rangle is linear in the first variable, and antilinear in the second one. For Hilbert spaces H and K , T\in B(H,K) , and \eta\in H , we write T|\eta\rangle for T\eta . Further, we shall sometimes denote the canonical basis for \ell_{2}^{n} by (|i\rangle)_{i=1}^{n} . \bullet For 1\leq p\leq\infty , S_{p}^{n} will denote the space of n\times n matrices, with the Schatten p -norm (operator norm for p=\infty ), and order determined by positive definiteness of a matrix. More generally, for m\times n matrices we use S_{p}^{m,n} . The Schatten space S_{p} is viewed as the completion of \cup_{n}S_{p}^{n} . \bullet For a normed space (\mathcal{X},\|\cdot\|) , \ball(\mathcal{X}) stands for the closed unit ball \{x\in\mathcal{X}:\|x\|\leq 1\} , and \sphere(\mathcal{X}) – for the unit sphere \{x\in\mathcal{X}:\|x\|=1\} . Further definitions will be introduced in the text as necessary. 2. Ordered vector spaces and local reflexivity In this section, we discuss ordered real Banach spaces, and, in particular, the specifics of local reflexivity therein. For unexplained facts and notation, we refer the reader to [ 4 ] and [ 6 ] . Definition 2.1 . Suppose \mathcal{X} is a real topological vector space. Following [ 4 ] , we say that a closed set \mathcal{X}^{+}\subset\mathcal{X} is a wedge in \mathcal{X} if (i) \mathcal{X}^{+}+\mathcal{X}^{+}\subset\mathcal{X}^{+} , (ii) \mathbb{R}^{+}\mathcal{X}^{+}=\mathcal{X}^{+} . \mathcal{X}^{+} is called a ( positive ) cone if, in addition, it is pointed – that is, \mathcal{X}^{+}\cap(-\mathcal{X}^{+})=\{0\} . Such a wedge turns \mathcal{X} into an ordered space : we say that x\geq y if x-y\in\mathcal{X}^{+} . If \mathcal{X}^{+} is a cone, then x=y iff x\geq y and y\geq x . For A\subset\mathcal{X} , we use a shorthand A^{+} for A\cap\mathcal{X}^{+} . The dual \mathcal{X}^{\flat} of an ordered space \mathcal{X} is turned into an ordered space by defining \mathcal{X}^{\flat+} to be the set