The impact of Schur multipliers in

harmonic analysis and operator algebras · 2025

Read The impact of Schur multipliers in by harmonic analysis and operator algebras (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-2510-17732.txt.

Introduction

Javier Parcet surveys fifteen years of work on Schur multipliers. These are basic linear maps that act on matrix algebras. They connect closely with Fourier multipliers, a core tool in harmonic analysis. That link bridges harmonic analysis and operator algebras, two major fields.

Parcet highlights new bounds for Schatten classes, which measure operator size. Those bounds feed results on group algebras and operator rigidity. Rigidity means some algebraic structures cannot bend into different shapes. He also covers non-Toeplitz multipliers and links to singular operators. The survey began as his invited talk for a major congress.

Catalog record

CREATORharmonic analysis and operator algebras
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE80 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-17732

Browse the full library · About the Tower of Babel

Full text

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

The impact of Schur multipliers in harmonic analysis and operator algebras Abstract . Schur multipliers are basic linear maps on matrix algebras. Their close albeit still intriguing connection with Fourier multipliers establishes a powerful bridge between harmonic analysis and operator algebras. In this paper, we survey their growing impact over the past 15 years. Particular attention will be drawn to recent bounds on Schatten p p -classes, with far-reaching applications in harmonic analysis on group von Neumann algebras and operator rigidity phenomena for higher-rank Lie groups and lattices. Key novelties arise from new insights into nonToeplitz Schur multipliers and unprecedented connections with highly singular operators from Euclidean harmonic analysis. Abstract Introduction Inspired by Heisenberg’s matrix mechanics and connections with ergodic theory and dynamical systems, the theory of von Neumann algebras emerged in the 1930s as a noncommutative form of measure theory. Here functions are replaced by linear operators on Hilbert spaces, which lack a commutative product. Harmonic analysis over von Neumann algebras extends beyond traditional “noncommutative harmonic analysis” which prioritizes the (commutative) measure spaces formed by a nonabelian group and its Haar measure. While noncommutative L_{p} -spaces have been widely studied from a functional analytic viewpoint, they have certainly been underexploited in harmonic analysis, since it demands a genuine interaction across several fields. A key turning point in the late 20th century —notably under Pisier’s influence— led to new noncommutative L_{p} -techniques, thanks to the development of operator space theory as well as quantum probability. Sums of independent or free random variables, square and maximal functions, martingale inequalities, singular integral operators and more were investigated thereafter. A strong motivation for this form of harmonic analysis comes from structural properties of group von Neumann algebras, with major implications in geometric group theory and the classification of nonamenable factors. Haagerup’s pioneering work on free groups and rank-one lattices [ 12 , 27 , 37 ] encoded deep geometric aspects of these groups in terms of Fourier approximation properties. In 2011 Lafforgue and de la Salle’s theorem [ 57 ] sparked a powerful revival of Haagerup’s methods, with a remarkable analysis of the failure of L_{p} -approximations in higher ranks and a smart use of Schur multipliers. These are linear maps on matrix algebras —first investigated by Issai Schur [ 95 ] in 1911— whose definition is rather simple for finite matrices S_{M}(A)=\Big(M(j,k)A_{j,k}\Big)_{1\leq j,k\leq N}\hskip 10.00002pt\mbox{for}\hskip 10.00002ptM\colon\{1,2,\ldots,N\}\times\{1,2,\ldots,N\}\to\mathbf{C}. In this paper, we survey a tight relation between Fourier and Schur multipliers along with some new fundamental inequalities for them, and a variety of far-reaching applications in matrix and group von Neumann algebras. The extra flexibility to manipulate Schur multipliers compared to their Fourier peers is a crucial aspect of these achievements. Finally, we briefly review some implications in the context of Lafforgue/de la Salle’s theorem and how new connections with Euclidean harmonic analysis could be relevant towards Connes’ rigidity conjecture. Along the way, we shall also explore related topics and open problems. Overall, we aim to expose why Schur multipliers have become key tools in addressing different challenges in harmonic analysis, operator algebras and geometric group theory. 1. New inequalities for Schur multipliers Let (\Omega,\mu) be a \sigma -finite measure space. Given p\geq 1 , the Schatten class S_{p}(\Omega) is the Banach space of all bounded linear maps A\colon L_{2}(\Omega)\to L_{2}(\Omega) with finite norm \|A\|_{S_{p}}=(\mathrm{tr}|A|^{p})^{1/p} , where |A|^{p} arises from A by functional calculus. The Schatten class S_{2}(\Omega) is the space of Hilbert-Schmidt operators on L_{2}(\Omega) and coincides with L_{2}(\Omega\times\Omega) by identifying A\in S_{2}(\Omega) with its kernel K_{A} Af(x)=\int_{\Omega}K_{A}(x,y)f(y)d\mu(y). Given M\colon\Omega\times\Omega\to\mathbf{C} measurable, its Schur S_{p} -multiplier is defined (when it exists) as the unique bounded operator S_{M} on S_{p}(\Omega) assigning A\in S_{2}(\Omega)\cap S_{p}(\Omega) to the operator S_{M}(A) with kernel M(x,y)K_{A}(x,y) . In what follows, we shall formally identify K_{A}(x,y) with a matrix (A_{x,y}) . Next, S_{M} is completely S_{p} -bounded when additionally S_{M}\otimes\mathrm{id}_{S_{p}(\Gamma)} extends to a bounded map on S_{p}(\Omega\times\Gamma) for any countable index set \Gamma . Given 1<p\neq 2<\infty , Pisier conjectured in 1998 the existence of Schur S_{p} -multipliers failing complete boundedness, but no examples are known. We refer to [ 57 , Section 1] to see why no such examples exist when \Omega has no atoms and for a rather complete presentation of basic properties of Schur multipliers. 1.1. Fourier-Schur transference Fourier multiplier theory is a cornerstone in harmonic analysis. Given m\colon\mathbf{Z}\to\mathbf{C} bounded, the Fourier multiplier T_{m} is densely defined on square-integrable functions f\colon\mathbf{T}\to\mathbf{C} by pointwise multiplication on their Fourier coefficients \widehat{T_{m}}f(k)=m(k)\widehat{f}(k) . It is completely L_{p} -bounded when T_{m}\otimes\mathrm{id}_{S_{p}(\Gamma)} extends to a bounded map on L_{p}(\mathbf{T};S_{p}(\Gamma)) , a space of matrix-valued functions. We set \big\|T_{m}\colon L_{p}(\mathbf{T})\to L_{p}(\mathbf{T})\big\|_{\mathrm{cb}}:=\sup_{\Gamma}\big\|T_{m}\otimes\mathrm{id}_{S_{p}(\Gamma)}\colon L_{p}(\mathbf{T};S_{p}(\Gamma))\to L_{p}(\mathbf{T};S_{p}(\Gamma))\big\|. A similar definition applies for symbols m\colon\mathrm{G}\to\mathbf{C} on any locally compact abelian group \mathrm{G} and Fourier multipliers on L_{2} -functions over its dual group. Complete boundedness is a strengthening in the category of operator spaces of Banach space boundedness [ 79 , 81 ] . A great portion of Euclidean Fourier L_{p} -multiplier theory holds verbatim in the cb-setting. The Marcinkiewicz, Hörmander-Mikhlin and Carleson-Sjölin multiplier theorems or de Leeuw’s transference theorems are some illustrations. The cb-validity of the Littlewood-Paley-Rubio de Francia theorem [ 92 ] in its full form is a notable open problem. The main known Fourier L_{p} -multipliers which fail cb-boundedness are due to Pisier and arise from \Lambda_{p} -sets whose Hankel extensions fail unconditionality in the matrix unit system [ 79 , Chapter 8] . Pisier’s counterexample above relies on a tight relation between the trigonometric system of characters \{\chi_{g}\colon\widehat{\mathrm{G}}\to\mathbf{T}\ \mathrm{s.t.}\ g\in\mathrm{G}\} and the system of matrix units \{e_{g,h}\colon g,h\in\mathrm{G}\} . To be more precise, the * -homomorphism \int_{\mathrm{G}}\widehat{f}(g)\chi_{g}(\cdot)\,d\mu(g)\longmapsto\Big(\widehat{f}(gh^{-1})\Big) is well-defined for \widehat{f}\in\mathcal{C}_{c}(\mathrm{G}) and extends to a map \Lambda\colon L_{\infty}(\widehat{\mathrm{G}})\to\mathcal{B}(L_{2}(\mathrm{G})) by weak- * density. Here, \mu denotes the Haar measure of \mathrm{G} and this suggests a relation between Fourier multipliers in the dual group (\widehat{\mathrm{G}},\widehat{\mu}) and certain class of Schur multipliers over (\mathrm{G},\mu) . This is indeed confirmed by the following result. Theorem 1.1 . (Fourier-Schur transference I) . Let 1\leq p\leq\infty and consider a locally compact abelian group \mathrm{G} . Assume that m\colon\mathrm{G}\to\mathbf{C} defines a completely L_{p} -bounded Fourier multiplier and set M(g,h)=m(gh^{-1}) . Then \big\|S_{M}\colon S_{p}(\mathrm{G})\to S_{p}(\mathrm{G})\big\|_{\mathrm{cb}}=\big\|T_{m}\colon L_{p}(\widehat{\mathrm{G}})\to L_{p}(\widehat{\mathrm{G}})\big\|_{\mathrm{cb}}. Schur multipliers defined with symbols of the form M(g,h)=m(gh^{-1}) in some topological group are known as Herz-Schur or Toeplitz multipliers. Theorem 1.1 for p=\infty follows from the elementary identity \Lambda\circ T_{m}=S_{M}\circ\Lambda and was originally noted by Bożejko/Fendler [ 6 ] in 1984. The L_{p} -case requires to suitably modify the map \Lambda , which relies in turn on certain limiting/averaging process. It holds for nonabelian amenable groups as well (see below). The first result in this direction is due to Neuwirth/Ricard [ 68 ] , who proved it for amenable discrete groups. The general case was finally settled by Caspers/de la Salle in [ 18 ] . This important transference theorem —along with its nonabelian form in Theorem 2.2 — allows to encode Fourier multipliers as Schur multipliers with Toeplitz symbols, which will be particularly useful in the rest of this paper. NonToeplitz multipliers are no longer tied to Fourier multipliers and Theorem 1.1 shows that any result valid for arbitrary Schur multipliers can be understood as a nontrigonometric extension of a Fourier multiplier theorem —the one

Read the complete text