Operator Analysis Hilbert Space Methods in Complex Analysis Jim Agler Full Chapter PDF

Operator Analysis: Hilbert Space
Methods in Complex Analysis Jim Agler

Visit to download the full and correct content document:
https://ebookmass.com/product/operator-analysis-hilbert-space-methods-in-complex-
analysis-jim-agler/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Modern Control: State-Space Analysis and Design Methods

Arie Nakhmani
https://ebookmass.com/product/modern-control-state-space-
analysis-and-design-methods-arie-nakhmani/
Gaussian Measures in Hilbert Space: Construction and

Properties Kukush
https://ebookmass.com/product/gaussian-measures-in-hilbert-space-
construction-and-properties-kukush/
Integration with Complex Numbers : A Primer on Complex

Analysis Aisling Mccluskey
https://ebookmass.com/product/integration-with-complex-numbers-a-
primer-on-complex-analysis-aisling-mccluskey/
Visual Complex Analysis: 25th Anniversary Edition

Tristan Needham
https://ebookmass.com/product/visual-complex-analysis-25th-
anniversary-edition-tristan-needham-2/
Visual Complex Analysis: 25th Anniversary Edition
Tristan Needham
https://ebookmass.com/product/visual-complex-analysis-25th-
anniversary-edition-tristan-needham/
Numerical Methods in Environmental Data Analysis Moses

Eterigho Emetere
https://ebookmass.com/product/numerical-methods-in-environmental-
data-analysis-moses-eterigho-emetere/
Genetic Analysis of Complex Disease 3rd Edition William

K. Scott
https://ebookmass.com/product/genetic-analysis-of-complex-
disease-3rd-edition-william-k-scott/
Introduction to Research Methods and Data Analysis in

Psychology Darren Langdridge
https://ebookmass.com/product/introduction-to-research-methods-
and-data-analysis-in-psychology-darren-langdridge/
Research Methods, Design, and Analysis – Ebook PDF

Version
https://ebookmass.com/product/research-methods-design-and-
analysis-ebook-pdf-version/
C A M B R I D G E T R AC T S I N M AT H E M AT I C S
General Editors
B. BOLLOB ÁS, W. FULTON, F. KIRWAN,

P. SARNAK, B. SIMON, B. TOTARO
219 Operator Analysis

CAMBRIDGE TRACTS IN MATHEMATICS
General Editors
B. BOLLOBÁS, W. FULTON, F. KIRWAN,
P. SARNAK, B. SIMON, B. TOTARO
A complete list of books in the series can be found at www.cambridge.org/mathematics.
Recent titles include the following:

183. Period Domains over Finite and p-adic Fields. By J.-F. Dat, S. Orlik, and M. Rapoport
184. Algebraic Theories. By J. Adámek, J. Rosický, and E. M. Vitale
185. Rigidity in Higher Rank Abelian Group Actions I: Introduction and Cocycle Problem. By A. Katok and
V. Niţică
186. Dimensions, Embeddings, and Attractors. By J. C. Robinson
187. Convexity: An Analytic Viewpoint. By B. Simon
188. Modern Approaches to the Invariant Subspace Problem. By I. Chalendar and J. R. Partington
189. Nonlinear Perron–Frobenius Theory. By B. Lemmens and R. Nussbaum
190. Jordan Structures in Geometry and Analysis. By C.-H. Chu
191. Malliavin Calculus for Lévy Processes and Infinite-Dimensional Brownian Motion. By H. Osswald
192. Normal Approximations with Malliavin Calculus. By I. Nourdin and G. Peccati
193. Distribution Modulo One and Diophantine Approximation. By Y. Bugeaud
194. Mathematics of Two-Dimensional Turbulence. By S. Kuksin and A. Shirikyan
195. A Universal Construction for Groups Acting Freely on Real Trees. By I. Chiswell and T. Müller
196. The Theory of Hardy’s Z-Function. By A. Ivić
197. Induced Representations of Locally Compact Groups. By E. Kaniuth and K. F. Taylor
198. Topics in Critical Point Theory. By K. Perera and M. Schechter
199. Combinatorics of Minuscule Representations. By R. M. Green
200. Singularities of the Minimal Model Program. By J. Kollár
201. Coherence in Three-Dimensional Category Theory. By N. Gurski
202. Canonical Ramsey Theory on Polish Spaces. By V. Kanovei, M. Sabok, and J. Zapletal
203. A Primer on the Dirichlet Space. By O. El-Fallah, K. Kellay, J. Mashreghi, and T. Ransford
204. Group Cohomology and Algebraic Cycles. By B. Totaro
205. Ridge Functions. By A. Pinkus
206. Probability on Real Lie Algebras. By U. Franz and N. Privault
207. Auxiliary Polynomials in Number Theory. By D. Masser
208. Representations of Elementary Abelian p-Groups and Vector Bundles. By D. J. Benson
209. Non-homogeneous Random Walks. By M. Menshikov, S. Popov and A. Wade
210. Fourier Integrals in Classical Analysis (Second Edition). By C. D. Sogge
211. Eigenvalues, Multiplicities and Graphs. By C. R. Johnson and C. M. Saiago
212. Applications of Diophantine Approximation to Integral Points and Transcendence. By P. Corvaja and
U. Zannier
213. Variations on a Theme of Borel. By S. Weinberger
214. The Mathieu Groups. By A. A. Ivanov
215. Slenderness I: Foundations. By R. Dimitric
216. Justification Logic. By S. Artemov and M. Fitting
217. Defocusing Nonlinear Schrödinger Equations. By B. Dodson
218. The Random Matrix Theory of the Classical Compact Groups. By E. S. Meckes
Operator Analysis
Hilbert Space Methods in Complex Analysis
J I M AG L E R
University of California, San Diego
J O H N E DWA R D McC A RT H Y
Washington University in St. Louis
N I C H O L A S YO U N G
Leeds University and Newcastle University
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi 110025, India
79 Anson Road, #06–04/06, Singapore 079906
Cambridge University Press is part of the University of Cambridge.

It furthers the Universitys mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781108485449
DOI: 10.1017/9781108751292
© Jim Agler, John Edward McCarthy, and Nicholas Young 2020
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2020
Printed in the United Kingdom by TJ International Ltd, Padstow Cornwall
A catalogue record for this publication is available from the British Library.
ISBN 978-1-108-48544-9 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
To
Sarah, Suzanne, and
Contents
Preface page xiii

Acknowledgments xv
PART ONE COMMUTATIVE THEORY 1
1 The Origins of Operator-Theoretic Approaches
to Function Theory 3
1.1 Operators 3
1.2 Functional Calculi 4
1.3 Operators on Hilbert Space 9
1.4 The Spectral Theorem 12
1.5 Hardy Space and the Unilateral Shift 13
1.6 Invariant Subspaces of the Unilateral Shift 15
1.7 Von Neumann’s Theory of Spectral Sets 16
1.8 The Schur Class and Spectral Domains 19
1.9 The Sz.-Nagy Dilation Theorem 21
1.10 Andô’s Dilation Theorem 22
1.11 The Sz.-Nagy–Foias Model Theory 23
1.12 The Sarason Interpolation Theorem 25
1.13 Historical Notes 30
2 Operator Analysis on D: Model Formulas, Lurking
Isometries, and Positivity Arguments 33
2.1 Overview 33
2.2 A Model Formula for S (D) 33
2.3 Reproducing Kernel Hilbert Spaces 37
2.4 Lurking Isometries 39
2.5 The Network Realization Formula (Scalar Case)
via Model Theory 44
2.6 Interpolation via Model Theory 52
vii
viii Contents
2.7 The Müntz–Szász Interpolation Theorem 56
2.8 Positivity Arguments 61
3 Further Development of Models on the Disc 71
3.1 A Model Formula for SB(H,K) (D) 71
3.2 Lurking Isometries Revisited 74
3.3 The Network Realization Formula 75
3.4 Tensor Products of Hilbert Spaces 78
3.5 Tensor Products of Operators 79
3.6 Realization of Rational Matrix Functions and
the McMillan Degree 81
3.7 Pick Interpolation Revisited 83
3.8 The Corona Problem 85
4 Operator Analysis on D2 93
4.1 The Space of Hereditary Functions on D2 93
4.2 The Hereditary Functional Calculus on D2 95
4.3 Models on D2 99
4.4 Models on D2 via the Duality Construction 104
4.5 The Network Realization Formula for D2 106
4.6 Nevanlinna–Pick Interpolation on D2 109
4.7 Toeplitz Corona for the Bidisc 112
4.8 Operator-Valued Functions on D2 113
4.9 Models of Operator-Valued Functions on D2 116
5 Carathéodory–Julia Theory on the Disc and the Bidisc 131
5.1 The One-Variable Results 131
5.2 The Model Approach to Regularity on D: B-points and C-points 133
5.3 A Proof of the Carathéodory–Julia Theorem on D via Models 137
5.4 Pick Interpolation on the Boundary 141
5.5 Regularity, B-points and C-points on the Bidisc 143
5.6 The Missing Link 146
6 Herglotz and Nevanlinna Representations in Several
Variables 148
6.1 Overview 148
6.2 The Herglotz Representation on D2 149
6.3 Nevanlinna Representations on H via Operator Theory 153
Contents ix
6.4 The Nevanlinna Representations on H2 156
6.5 A Classification Scheme for Nevanlinna Representations
in Two Variables 160
6.6 The Type of a Function 165
7 Model Theory on the Symmetrized Bidisc 169
7.1 Adding Symmetry to the Fundamental Theorem for D2 170
7.2 How to Define Models on the Symmetrized Bidisc 173
7.3 The Network Realization Formula for G 176
7.4 The Hereditary Functional Calculus on G 177
7.5 When Is G a Spectral Domain? 182
7.6 G Spectral Implies G Complete Spectral 185
7.7 The Spectral Nevanlinna–Pick Problem 185
8 Spectral Sets: Three Case Studies 189
8.1 Von Neumann’s Inequality and the Pseudo-Hyperbolic
Metric on D 189
8.2 Spectral Domains and the Carathéodory Metric 192
8.3 Background Material 194
8.4 Lempert’s Theorem 195
8.5 The Carathéodory Distance on G 203
8.6 Von Neumann’s Inequality on Subvarieties of the Bidisc 205
9 Calcular Norms 213
9.1 The Taylor Spectrum and Functional Calculus 213
9.2 Calcular Norms and Algebras 215
9.3 Halmos’s Conjecture and Paulsen’s Theorem 223
9.4 The Douglas–Paulsen Norm 226
9.5 The B. and F. Delyon Norm and Crouzeix’s Theorem 233
9.6 The Badea–Beckermann–Crouzeix Norm 235
9.7 The Polydisc Norm 243
9.8 The Oka Extension Theorem and Calcular Norms 245
10 Operator Monotone Functions 254
10.1 Löwner’s Theorems 254
10.2 An Interlude on Linear Programming 262
10.3 Locally Matrix Monotone Functions in d Variables 269
10.4 The Löwner Class in d Variables 278
x Contents
10.5 Globally Monotone Rational Functions in Two Variables 279
PART TWO NON-COMMUTATIVE THEORY 283
11 Motivation for Non-Commutative Functions 285
11.1 Non-Commutative Polynomials 285
11.2 Sums of Squares 287
11.3 Nullstellensatz 287
11.4 Linear Matrix Inequalities 288
11.5 The Implicit Function Theorem 289
11.6 Matrix Monotone Functions 289
11.7 The Functional Calculus 290
12 Basic Properties of Non-Commutative Functions 291
12.1 Definition of an nc-Function 291
12.2 Locally Bounded nc-Functions Are Holomorphic 295
12.3 Nc Topologies 298
13 Montel Theorems 302
13.1 Overview 302
13.2 Wandering Isometries 303
13.3 A Graded Montel Theorem 305
13.4 An nc Montel Theorem 309
13.5 Closed Cones 310
14 Free Holomorphic Functions 313
14.1 The Range of a Free Holomorphic Function 313
14.2 Nc-Models and Free Realizations 315
14.3 Free Pick Interpolation 319
14.4 Free Realizations Redux 324
14.5 Extending off Varieties 326
14.6 The nc Oka–Weil Theorem 327
14.7 Additional Results 328
15 The Implicit Function Theorem 330
15.1 The Fine Inverse Function Theorem 330
15.2 The Fat Inverse Function Theorem 332
15.3 The Implicit Function Theorem 337
15.4 The Range of an nc-Function 339
Contents xi
15.5 Applications of the Implicit Function Theorem in
Non-Commutative Algebraic Geometry 341
15.6 Additional Results 343
16 Non-Commutative Functional Calculus 344
16.1 Nc Operator Functions 344
16.2 Polynomial Approximation and Power Series 349
16.3 Extending Free Functions to Operators 354
16.4 Nc Functional Calculus in Banach Algebras 356
Notation 358
Bibliography 361
Index 372
Preface
The philosophy of this book is that Hilbert space geometry binds function
theory and operator theory together, not only allowing each to aid the other but
creating a rich structure that can be used to discover new phenomena. There is
a “three-way street” between operator theory and function theory: sometimes
one uses function theory to prove operator theorems, sometimes one uses
operator theory to prove function theorems, and sometimes the theories are so
interwoven that one cannot even state the theorem without using the language
of both operator theory and function theory.
The main thrust of the book is to discover and prove theorems about
holomorphic functions and complex geometry with the aid of Hilbert space
geometry and operator theory. The holomorphic functional calculus permits us
to substitute commuting operators for the variables of a holomorphic function,
a step that reveals hidden properties of the function—something every function
theorist should want to do.
It is remarkable how little operator theory one needs in order to prove
significant facts in function theory. There will be no call here for the detailed
and subtle theories of particular classes of operators, but we shall make heavy
use of the functional calculus for operators. The theories of operator dilations
and spectral sets also play an important role. We explain carefully what is
required from these theories. The reader may either take this material on trust
or consult the references that we give.
Part I of the book is devoted to holomorphic functions on domains in
complex Euclidean space, beginning with scalar-valued functions on the unit
disc in the complex plane, where intuition is most easily developed. Here
the central notion of a Hilbert space model of a function is introduced, as
are several types of arguments that will recur throughout the book. Gradually
we build up to various domains in higher dimensions and to operator-valued
functions, the latter being important for engineering applications. We do not
xiii
xiv Preface
aim for maximum generality, which can come at the cost of a sacrifice of
elegance and impact.
Part II concerns non-commutative functions, that is, functions of non-
commuting variables. This is a topic of relatively recent study, and one that
is currently undergoing rapid development. It transpires that the Hilbert space
methods of Part I are well suited to this new context, and we derive many
analogs of classical theorems.
Our intended audience is graduate students and mathematicians interested
in complex function theory and/or operator theory. We do not require famil-
iarity with several complex variables, but we do assume that the reader has a
basic knowledge of complex analysis and functional analysis.
The moment you buy into the functional calculus, you’re ready to roll!
Acknowledgments
The catalyst for this book was two series of lectures, given by the first and
second authors at Newcastle University in an LMS lecture series in April
2017. The authors would like to thank the London Mathematical Society for
supporting this event. In addition, the first author was partially supported by
National Science Foundation Grant DMS 1665260, the second author was
partially supported by National Science Foundation Grant DMS 1565243,
and the third author was partially supported by UK Engineering and Physical
Sciences Research Council grants EP/K50340X/1 and EP/N03242X/1, and
London Mathematical Society grants 41219 and 41730. We would like to thank
the National Science Foundation, the UK Engineering and Physical Sciences
Research Council, and the London Mathematical Society for their support.
Parts of the book were read in draft form by several people, who caught
some mistakes. We are specially grateful to Alberto Dayan, Chris Felder,
Michael Hartz, Mark Mancuso, James Pascoe, and Jeet Sampat. All remain-
ing mistakes and typographical errors are, of course, the responsibility of
the authors.
xv
P AR T ONE
Commutative Theory
1
The Origins of Operator-Theoretic Approaches
to Function Theory
In this chapter we present some of the highlights of operator theory over the
last century and briefly describe their connections with function theory. It is
intended to establish notation and provide orientation for the reader, before we
go on to a detailed and systematic development in Chapter 2 (which can be
understood independently of this scene-setting chapter). The reader should not
attempt to master all the contents of this chapter in detail before progressing–
some of them are substantial. Our choice, inevitably, is partial; we mention
some other major contributions in historical notes at the end of the chapter, but
we do not attempt a history of the subject.
1.1 Operators
We adopt the standard Bourbaki symbols R and C for the sets of real numbers
and complex numbers, respectively, and also the notations
T = {z ∈ C: |z| = 1} and D = {z ∈ C: |z| < 1}.
For any Banach space X , we shall denote by ball X the closed unit ball of X ,
that is,
ball X = {u ∈ X : u ≤ 1}.
If X , Y are Banach spaces, we say that a linear transformation T : X → Y is
bounded if there exists c ∈ [0, ∞) such that
T u ≤ cu for all u ∈ X . (1.1)
For a bounded linear transformation T : X → Y we define T , the norm of T ,
by the formula
T = sup T u, (1.2)
u∈ball X
3
4 Commutative Theory
or equivalently, by the formula
T = inf {c: T u ≤ cu for all u ∈ X }.
When X , Y are Banach spaces, we denote by B(X ) the set of bounded
linear transformations from X to X , and by B(X , Y) the set of bounded linear
transformations from X to Y. We refer to the elements of B(X ) or B(X , Y)
as operators. B(X )and B(X , Y) are Banach spaces under the norm defined by
equation (1.2). Furthermore, B(X ) (with composition as multiplication) is a
Banach algebra with identity, as it satisfies the axioms1 1 = 1 and
T1 T2 ≤ T1 T2
for all T1 , T2 ∈ B(X ).
Every element of B(X ) has a spectrum. For T ∈ B(X ) we say that T is
invertible if there exists S ∈ B(X ) such that ST = T S = 1. The spectrum of T ,
denoted by σ (T ), is defined by
σ (T ) = {z ∈ C: z − T is not invertible}.
The set σ (T ) is a nonempty compact subset of C provided that X = {0}.
1.2 Functional Calculi

Informally, the term functional calculus refers to a rule that enables a function
f to act on an operator T ∈ B(X ) to produce an operator f (T ) ∈ B(X ). There
are numerous functional calculi that have been defined in various settings.
However, in these notes we shall restrict ourselves principally to the case that
the functional calculus is holomorphic, that is, the function f is assumed to be
holomorphic on a neighborhood of σ (T ). To be an honest functional calculus,
the rule should satisfy the following natural conditions.
Properties of a holomorphic functional calculus
Let X be a Banach space and let T ∈ B(X ).
(1) If f (z) = 1 for all z in a neighborhood of σ (T ), then f (T ) = 1, the identity
operator on X .
(2) If f (z) = z for all z in a neighborhood of σ (T ), then f (T ) = T .
(3) If f , g are holomorphic in a neighborhood of σ (T ) then
(f + g)(T ) = f (T ) + g(T ) and (f g)(T ) = f (T )g(T ).
(4) If f is holomorphic in a neighborhood of σ (T ), λ ∈ C and g = λf then
g(T ) = λf (T ).
1 Here the first 1 denotes the identity of B(X ), which is the identity operator on X .
The Origins of Operator-Theoretic Approaches to Function Theory 5
(5) The functional calculus has an appropriate continuity property.
For any open set U in C, the algebra of scalar-valued holomorphic functions
on U will be denoted by Hol(U ). Observe that, for any T ∈ B(X ), if U is a
neighborhood of σ (T ) and τ : Hol(U ) → B(X ) is defined by τ (f ) = f (T ),
then τ is a homomorphism of algebras that satisfies τ (1) = 1 and τ (z) = T .
The simplest example of a functional calculus is the polynomial calculus.
Here, for f ∈ C[z], the algebra of polynomials with complex coefficients, if

n
f (z) = ak zk ,
k=1
then f (T ) is defined by

n
f (T ) = ak T k .
k=1
The polynomial calculus can easily be extended to the power series

calculus. If
∞

f (z) = ak (z − a)k
k=1
defines an analytic function on {z ∈ C: |z − a| < r} for some r > 0, then

under the assumption that σ (T ) ⊆ {z ∈ C: |z − a| < r}, the operator f (T )
is defined by
∞

f (T ) = ak (T − a)k .
k=1
That this series converges in the operator norm of B(X ) is a simple exercise on
the spectral radius formula and Cauchy’s formula for the radius of convergence
of a power series.
Another simple modification of the polynomial calculus is the rational cal-
culus. For K, a compact subset of C, let Rat(K) denote the algebra of rational
functions with poles off K. If T ∈ B(X ) and f ∈ Rat(σ (T )), then we may write
f (z) = p(z)q(z)−1 ,
where p and q are polynomials and q(z) = 0 for all z ∈ σ (T ). By the spectral
mapping theorem, which asserts that q(σ (T )) = σ (q(T )), the spectrum of
q(T ) does not contain 0, so that q(T ) is invertible. Therefore, f (T ) can be
defined by the formula
f (T ) = p(T )q(T )−1 . (1.3)
A sweet spot in the theory of functional calculi is the Riesz–Dunford
functional calculus. Treatments of this topic, complete with proofs, are in
[60, VII.4; 140, section 17.2], and many other introductory texts on functional
analysis. We shall just state the important results.
Fix an operator T ∈ B(X ). For any open set U in C such that σ (T ) ⊆ U , let
be a finite collection of closed rectifiable curves in U \ σ (T ) that winds once
around each point in the spectrum of σ (T ) and no times around each point in
the complement of U . If f ∈ Hol(U ) then, by the Cauchy integral formula,

1 f (w)
f (z) = dw,
2π i w − z
whenever z ∈ σ (T ). The Riesz–Dunford functional calculus defines f (T ) for
f ∈ Hol(U ) by the substitution z = T in this formula, that is to say,

1
f (T ) = f (w)(w − T )−1 dw. (1.4)
2π i
Note that this formula makes sense, since the assumption ⊆ C\σ (T ) implies
that w − T is invertible for all w ∈ . Moreover, the defining equation (1.4)
depends neither on the choice of contour nor on the choice of U , since by
Cauchy’s theorem, for any vectors u, v in X and X ∗ ,

1
f (T )u, v = f (w) (w − T )−1 u, v dw
2π i
is independent of .
Properties of the Riesz–Dunford functional calculus
(1) If f (z) = 1, then f (T ) = 1 and if f (z) = z, then f (T ) = T .

(2) For any neighborhood U of σ (T ) the map
f → f (T )
is a unital algebra-homomorphism from Hol(U ) to B(X ).

(3) The calculus is consistent with the polynomial and rational calculi, that is,
if U is a neighborhood of σ (T ), if r is a rational function with poles off
U , and f ∈ Hol(U ) is defined by f (λ) = r(λ) for all λ ∈ U , then
f (T ) = r(T ).
(4) The calculus is consistent with the power series calculus, that is, if U is a
disc, σ (T ) ⊆ U , and g(z) is the power series expansion of f (z) valid in U ,
then f (T ) = g(T ), where g(T ) is defined by the power series calculus.
(5) The calculus is continuous in the following sense. If U is a neighborhood
of σ (T ), if {fn } is a sequence in Hol(U ), f ∈ Hol(U ), and fn → f
uniformly on compact subsets of U , then fn (T ) → f (T ) in the operator
norm of B(X ).
Another way of thinking of the Riesz–Dunford functional calculus is by
way of Runge’s theorem, which implies that, for any open set U in C, every
function f ∈ Hol(U ) is the limit, uniformly on compact subsets of U , of a
sequence of rational functions rn with poles off U . As a consequence, property
(5) can be used to define f (T ) as the norm limit of rn (T ).
The holomorphic functional calculus is one of the most important tools
in operator theory and is central to this book. We shall frequently want to
apply a holomorphic function of d complex variables to a d-tuple of pairwise
commuting operators; for this we need a generalization of the Riesz–Dunford
functional calculus. Such a generalization does indeed exist; the elaboration
of this theory is a heroic chapter in the history of operator theory. It is more
intricate than the one-variable theory and requires a significant input from the
theory of several complex variables. It was developed over several decades by
many mathematicians and, while “multivariable spectral theory” remains an
active research topic, the foundations of the functional calculus now appear to
be in a definitive form.
For a holomorphic function f of d variables and a pairwise commuting
d-tuple T = (T 1 , . . . , T d ) of operators on a Banach space X , to define what
we mean by f (T ) one strategy is again to require f to be holomorphic on
the spectrum of T and to invoke an integral representation formula. To do
this one must overcome some technical difficulties, the first of which is to
find the appropriate notion of spectrum of T . For a start, the spectrum should
contain the joint eigenvalues of T . We say that a non-zero vector x ∈ X is
a joint eigenvector of T and that λ = (λ1 , . . . , λd ) is a corresponding joint
eigenvalue of T if T j x = λj x for j = 1, . . . , d. If dim X < ∞ (and X = {0})
then T has at least one joint eigenvalue. For let λ1 be an eigenvalue of T 1 and
let E1 be the corresponding eigenspace {x ∈ X : T 1 x = λ1 x}. Since the T j
commute with T 1 , the space E1 is invariant under each T j and the operators
T j |E1 commute pairwise. Let λ2 be an eigenvalue of T 2 |E1 and let E2 be the
corresponding eigenspace {x ∈ E1 : T 2 x = λ2 x}. Continuing in this way we
arrive at a point λ = (λ1 , . . . , λd ) ∈ Cd and a non-zero subspace Ed of X such
that T j x = λj x for j = 1, . . . , d and all x ∈ Ed . This point λ ∈ Cd is a joint
eigenvalue of T . However, as we already know from the case d = 1, when X
is infinite-dimensional there need not be any joint eigenvalues of T .
In 1953 Georgii Evgen’evich Shilov [187] constructed a functional calculus
for several elements of a commutative Banach algebra. If T is a commuting
d-tuple of elements of B(X ), one can choose any commutative Banach
subalgebra A of B(X ) containing the elements of T and then define the
spectrum of T with respect to A by
σA (T ) = Cd \ ρA (T ), (1.5)
where

ρA (T ) = λ ∈ Cd : there exist R 1 , . . . , R d ∈ A such that

(T 1 − λ1 )R 1 + · · · + (T d − λd )R d = 1 . (1.6)
Any joint eigenvalue of T does belong to σA (T ), for any commutative

subalgebra A of B(X ) containing the T j (otherwise, simply apply the defining
equation (1.6) of ρA (T ) to the corresponding joint eigenvector x to get a
contradiction).
Shilov’s theory yields a functional calculus for a commuting tuple T , a
calculus that was further developed by numerous authors; see for example
[212]. A drawback of this approach is that the spectrum of a tuple is defined
relative to the chosen commutative subalgebra of B(X ) and is typically differ-
ent for different subalgebras. The simplest choice is to take the subalgebra to
be the one generated in B(X ) by the operators in question; we shall call the
corresponding spectrum the algebraic spectrum of T and denote it by σalg (T ).
However, this choice often results in an unnecessarily large spectrum.
This imperfection of the Banach algebra approach was remedied in the
groundbreaking papers [198, 197] by Joseph Taylor. He introduced a more
geometric notion of spectrum, now called the Taylor spectrum of T , denoted
in this book by σ (T ).
The Taylor spectrum is a subset, sometimes proper, of σA (T ) for any
commutative algebra A containing T (indeed, Taylor showed that it is a
subset of the spectrum with respect to any algebra A with T in its center
[198, lemma 1.1]). This result is important, as a smaller spectrum gives a larger
class of holomorphic functions for which f (T ) is defined. It is now accepted
that the Taylor spectrum is the right notion for the definition of a functional
calculus. We define it in Section 9.1. Let us remark that, for commuting normal
tuples, there is a simpler description of the Taylor spectrum: see Section 8.3.1.
Note also that σ (T ) contains all the joint eigenvalues of T [67, page 22].
A definitive discussion of the various notions of spectrum for d-tuples and
development of the Taylor functional calculus is given in [89], while a more
accessible account can be found in [67]. The present book, on the other hand
is about the use of the functional calculus as a tool for function theory. What
the reader will need is a knowledge of the properties of the functional calculus,
which we shall describe as we need them.
For most of the book (all except Chapter 9) the domains that we study will
be polynomially convex.2 For such a domain U , the Taylor spectrum of T is
contained in U if and only if the algebraic spectrum is. Furthermore, every
holomorphic function f on U can be approximated uniformly on compact
2 See Definition 1.43.
subsets by a sequence of polynomials pn (this result is called the Oka-Weil
theorem). The Taylor functional calculus in this case can be defined concretely
by f (T ) = lim pn (T ).
1.3 Operators on Hilbert Space

Operator analysis in the sense of this book concerns the special case that X is
a complex Hilbert space H, that is, a complete inner product space over the
field of complex numbers. We assume familiarity with the basic properties of
Hilbert space (see [214]). In this case, B(X ) has additional structure beyond
that of a Banach algebra, a structure that plays a critical role in the development
of operator-analytic methods.
Definition 1.7. For Hilbert spaces H and K and for T ∈ B(H, K), the adjoint
of T is the operator T ∗ in B(K, H) given by
∗
T v, u H = v, T u K for all u ∈ H, v ∈ K.
The adjoint satisfies the C ∗ -axiom
T ∗ T = T 2 . (1.8)
In the C ∗ -axiom, since T : H → K and T ∗ : K → H, the operator T ∗ T is
well defined and maps H to H. The techniques that we shall present from
operator analysis rely essentially on the meaningfulness of such algebraic
expressions in T and T ∗ (especially on the statement (1.13)). For a general
Banach space, if T : X → Y, then T ∗ : Y ∗ → X ∗ , so one cannot make sense
of T ∗ T . For this reason, in the remainder of the book, the term operator will
always refer to a bounded linear operator acting on a complex Hilbert space or
from one complex Hilbert space to another.
The techniques of operator analysis depend essentially on the notion of
positivity. Classically, an n × n matrix [aij ] is said to be positive semi-definite
if, for all c1 , . . . , cn ∈ C,

n
aij cj c̄i ≥ 0.
i,j =1
We extend this notion to operators on Hilbert space in the following way.

Definition 1.9. An operator T ∈ B(H) is said to be positive, written T ≥ 0, if,
for all u ∈ H,
T u, u H ≥ 0.
There are other revealing characterizations of the positivity of an operator.
See [60, theorem VIII.3.6] for a proof of the following statement.
Proposition 1.10. For T ∈ B(H), the following statements are equivalent.
(i) T ≥ 0;
(ii) T ∗ = T and
σ (T ) ⊆ {x ∈ R: x ≥ 0};
(iii) T = X ∗ X for some Hilbert space K and some X ∈ B(H, K);

(iv) T = X2 for some X ∈ B(H) such that X = X∗ .
There is a close connection between positivity and some other important
classes of operators.
Definition 1.11. Let H, K be Hilbert spaces. An operator T ∈ B(H, K) is
(i) a contractive operator if T ≤ 1;

(ii) an isometric operator if T x = x for every x ∈ H.
A simple but, in this book, all-pervading principle is the following.
Proposition 1.12. (The fundamental fact of operator analysis) If H, K are
Hilbert spaces and T ∈ B(H, K), then
T is contractive if and only if 1 − T ∗ T ≥ 0. (1.13)

Proof.
T ≤ 1 ⇔ x2 ≥ T x2 for all x ∈ H

⇔ x, x ≥ T x, T x for all x ∈ H
∗
⇔ x, x − T T x, x ≥ 0 for all x ∈ H

⇔ (1 − T ∗ T )x, x ≥ 0 for all x ∈ H
⇔ 1 − T ∗ T ≥ 0.
The importance of the equivalence (1.13) in Proposition 1.12 cannot be

overstated. It enables the expression of the analytic concept of size in terms
of the algebraic concept of positivity. The norm of T is bounded by 1 if and
only if 1 − T ∗ T can be represented as X ∗ X or X2 , as in Proposition 1.10, a
condition that can often be resolved by algebra.
In Proposition 1.12 the extremal case of the inequality 1 − T ∗ T ≥ 0 is that
it hold with equality, that is 1 − T ∗ T = 0.
Proposition 1.14. If H, K are Hilbert spaces and T ∈ B(H, K), then
T is isometric if and only if 1 − T ∗ T = 0. (1.15)

A slight modification of the proof of Proposition 1.12 yields the fact that
T is isometric if and only if (1 − T ∗ T )x, x = 0 for all vectors x. Proposition
1.14 is therefore a consequence of the following statement.
Lemma 1.16. If Z is an operator on a Hilbert space H then Z = 0 if and only
if Zx, x = 0 for all x ∈ H.
Proof. Necessity is trivial. Conversely, suppose Zx, x = 0 for all x ∈ H.
Consider any x, y ∈ H and λ ∈ C. We have
0 = Z(x + λy), x + λy
= Zx, x + λ Zy, x + λ̄ Zx, y + |λ|2 Zy, y
= λ Zy, x + λ̄ Zx, y .
By choosing in succession λ = 1 and λ = i we deduce that Zx, y = 0 for all

x, y ∈ H, and therefore Z = 0.
In Lemma 1.16 we make use of our standing assumption that Hilbert spaces
are vector spaces over C. The statement would be false in a real Hilbert space
(consider the operation of rotation through a right angle on R2 ).
We defined isometries to be operators that preserve lengths; an interesting
consequence of Proposition 1.14 is that they also preserve inner products. If
T ∗ T = 1 then, for any vectors x, y,

T x, T y = T ∗ T x, y = x, y .
For Hilbert spaces H and K, there is a useful notation that allows the
identification of elements of the algebraic tensor product H ⊗ K with operators
from K into H. If u ∈ H and v ∈ K we define u ⊗ v ∈ B(K, H) by the formula
(u ⊗ v)(x) = x, v K u for all x ∈ K. (1.17)
In the context of a vector space V there is an algebraic notion of projection.

If V is expressed as the direct sum of two subspaces M and N , then one may
define a linear transformation P on V by the relation
P (x + y) = x
for all x ∈ M and y ∈ N . This operator P is called the projection on M along

N ; it is clearly a linear transformation on V with range M and kernel N . In the
context of a Hilbert space H, the word projection is usually reserved for the
special case of a projection on a closed subspace M of H along its orthogonal
complement. We shall denote it by PM .
1.4 The Spectral Theorem
One of the early triumphs of operator theory was the development of the
modern form of the spectral theorem, together with its myriad applications
to analysis and the mathematics of physics. An operator N ∈ B(H) is said to
be normal if it commutes with its adjoint, that is, if
N ∗N = N N ∗.
Two important special cases of normality are when N is self-adjoint, that is,
N ∗ = N, and when N is unitary, that is, N ∗ N = 1 = N N ∗ . If N is normal, then
(1) N is self-adjoint if and only if σ (N) ⊆ R, and
(2) N is unitary if and only if σ (N) ⊆ T.
If H is a Hilbert space and is a σ -algebra of subsets of a set X, then we say
that E is a B(H)-valued spectral measure defined on (X, ) if E: → B(H)
is a mapping satisfying
(1) for all ∈ , E( ) is a projection on H;
(2) E(∅) = 0 and E(X) = 1;
(3) for all 1 , 2 ∈ , E( 1 ∩ 2 ) = E( 1 )E( 2 ), and
(4) if { n } is a sequence of pairwise disjoint sets in , then
∞ ∞

E n = E( n ),
n=1 n=1
where the sum on the right-hand side converges in the strong operator
topology.3
Theorem 1.18. (The spectral theorem for normal operators) If H is a
Hilbert space and N is a normal operator on H, then there exists a B(H)-
valued spectral measure E defined on the σ -algebra of Borel subsets of σ (N )
such that

N= z dE(z). (1.19)
σ (N )
See [140, section 31.6, theorem 15] or [60, theorem IX.2.2] for a proof.
An immediate corollary of the spectral theorem is the following result, a
cornerstone of the theory of C ∗ -algebras. For K a compact subset of C, let
C(K) denote the algebra of continuous functions on K with supremum norm
and involution f ∗ = f¯, the complex conjugate of f . For a normal operator
N on H, let C ∗ (N ) denote the smallest norm-closed subalgebra of B(H)
containing 1 and N and satisfying X ∗ ∈ C ∗ (N ) whenever X ∈ C ∗ (N ).
3 A sequence (T ) of operators converges to T ∈ B(H) in the strong operator topology if, for
n
every x ∈ H, Tn x − T x → 0 as n → ∞.
Theorem 1.20. (The continuous functional calculus for normal operators)
Let H be a Hilbert space, N be a normal operator on H and E be the spectral
measure of N as in Theorem 1.18. For ϕ ∈ C(σ (N)), define ϕ(N) ∈ ϕ ∈
C(σ (N)) by the formula

ϕ(N) = ϕ(z) dE(z). (1.21)
σ (N )
The map ρ: C(σ (N)) → B(H) given by the formula ρ(ϕ) = ϕ(N) is an
isometric ∗-isomorphism of C(σ (N)) onto C ∗ (N ), that is, ρ is an algebra-
isomorphism satisfying ρ(ϕ̄) = ρ(ϕ)∗ and ρ(ϕ) = ϕ for all ϕ ∈ B(H).
This theorem illustrates the fact that not all functional calculi are holo-
morphic functional calculi. Nevertheless, the continuous functional calculus
for normal operators agrees with the Riesz–Dunford functional calculus on
functions that are holomorphic on a neighborhood of the spectrum of a normal
operator. In particular, if N is a normal operator, f ∈ Rat(σ (N )), and f (N) is
defined by equation (1.3), then

f (N ) = f (z) dE(z). (1.22)
σ (N )
1.5 Hardy Space and the Unilateral Shift

The name Hilbert space is a tribute to a seminal paper of 1906 by David
Hilbert [115]. According to Jean Dieudonné,4 this is one of four landmark
papers in the period 1900–1906 which led to the sudden crystallization of
functional analysis as a branch of analysis (the others were a paper by Ivar
Fredholm on integral equations, Henri Lebesgue’s thesis on integration and
Maurice Fréchet’s thesis on metric spaces). In this paper Hilbert made use of
the space now denoted by 2 . It comprises all sequences
a = (a0 , a1 , a2 , . . .)
of complex numbers, indexed by the non-negative integers, such that
∞
1
def 2
a = |an |2
n=0
is finite. It is equipped with the inner product
∞

a, b = an bn .
n=0
4 In [72, chapter V] he wrote “By the depth and novelty of its ideas, it is a turning point in the
history of Functional Analysis, and indeed deserves to be considered as the very first paper
published in that discipline.”
One could say that 2 is the “purest” Hilbert space, in that it exhibits the
geometry of Hilbert space in the presence of minimal mathematical structure.
Other Hilbert spaces typically have elements that are functions or operators
of some kind, which may involve delicate notions such as measurability or
analyticity, and such spaces often occur in applications. However, all separable
infinite-dimensional Hilbert spaces are isomorphic ([172, section 82] or [214,
theorem 4.19]), and so it is natural to try to gain insight into operators on a
general Hilbert space by studying them on the “bare” Hilbert space 2 .
One operator on 2 that has been much studied is the unilateral shift S. This
is the operator defined on 2 by the formula
S (a0 , a1 , a2 , . . .) = (0, a0 , a1 , . . .). (1.23)
This simple definition conceals a treasure trove of analytical subtlety, much of

it understandable via an avatar of S dwelling on a certain Hilbert space H2 of
analytic functions on D. The space H2 , known as Hardy space, is the Hilbert
space of analytic functions on D that have a square-summable power series
about 0. To say that a power series
∞

f (z) = an zn (1.24)
n=0
is square summable means that

∞

|an |2 < ∞. (1.25)
n=0
This condition ensures that the series (1.24) converges locally uniformly on D
and thereby defines f as an analytic function in D. The inner product on H2 is
defined by
∞
∞
∞

an zn , bn zn = an b̄n . (1.26)
n=0 n=0 n=0
Condition (1.25) also ensures that the series on the right-hand side of equation
(1.26) converges, so that the inner product is well defined.
Thus elements of H2 can be regarded either as power series or as analytic
functions on D. It is useful to have a criterion for membership of H2 , and an
expression for the norm, in terms of function values (as distinct from Taylor
coefficients). In fact, for an analytic function f on D to belong to H2 , it is
necessary and sufficient that
2π
lim sup |f (reiθ )|2 dθ < ∞. (1.27)
r→1− 0
When this inequality holds, the lim sup is in fact a limit as r → 1−, and
1
2π 2
dθ
f H2 = lim |f (re )|
iθ 2
. (1.28)
r→1− 0 2π
One consequence of equation (1.28) (together with the maximum principle)

is that, for every λ ∈ D, the map
f → f (λ)
is a bounded linear functional on H2 . It follows from the Riesz–Fréchet
theorem that there exists a unique element sλ of H2 such that, for all f ∈ H2 ,
f (λ) = f , sλ . (1.29)
It is easy to check by means of the formula (1.26) for the inner product that sλ
is given by
1
sλ (z) = for z ∈ D. (1.30)
1 − λ̄z
The function sλ on D is called the Szegő kernel. The property described by
equation (1.29) is called the reproducing property of the Szegő kernel.
If Mz denotes the operator defined on H2 by the formula
(Mz f )(w) = wf (w) for f ∈ H2 and w ∈ D, (1.31)
and L: 2 → H2 is defined by
L(a0 , a1 , a2 , . . .)(z) = a0 + a1 z + a2 z2 + · · · , (1.32)
then it is apparent that L is a Hilbert space isomorphism and Mz L = LS. In
other words, if we identify 2 and H2 by means of L, then the shift operator
S becomes the operation of multiplication by the idependent variable on H2 .
We may therefore recruit the rich theory of analytic functions on D to help us
understand the unilateral shift operator.
A simple account of the basic facts about H2 is given in [214, chapter 13,
pages 157–170]. Much fuller accounts are in [84, 116].
1.6 Invariant Subspaces of the Unilateral Shift

A highly original paper of Arne Beurling [48] in 1949 exploited the relation
between the shift and Mz to answer a major open question in ergodic theory
at the time. The problem was to describe the closed invariant subspaces of the
unilateral shift operator defined in the last section. Here an invariant subspace
of an operator T ∈ B(H) is a subspace N of H such that T N ⊆ N . Much effort
has been devoted to the identification of all the closed invariant subspaces of
particular operators and, more generally, classes of operators. Indeed, it is still
unknown whether every operator on a Hilbert space of dimension at least 2
has a non-trivial closed invariant subspace.
For motivation of the notion of invariant subspace, notice that, in finite
dimensions, an invariant subspace of an operator T is just the span of a set
of eigenvectors of T . In infinite dimensions the eigenvectors of an operator do
still play a role, but since an operator on an infinite-dimensional space need
have no eigenvectors at all, they cannot be so central as in finite dimensions.
The notion of invariant subspace to some extent replaces that of eigenvector.
The use of analytic functions to describe invariant subspaces has been
an extraordinarily fruitful theme in operator theory. Beurling constructed
the general closed invariant subspace for Mz using “inner functions” on D
and thereby described the closed invariant subspaces of S by means of the
isomorphism L of equation (1.32).
We say that a function ϕ: D → C is inner on D if ϕ is bounded and analytic
in D and |ϕ(z)| = 1 for almost every z ∈ T with respect to Lebesgue measure
on T. If ϕ is an inner function on D, then one can associate with ϕ a subspace
of H2 by forming
ϕH2 = {ϕf : f ∈ H2 }.
Theorem 1.33. (Beurling’s theorem) A subspace N of H2 is a closed
invariant subspace for Mz if and only if there exists an inner function ϕ on
D such that N = ϕH2 .
As a corollary, Beurling proved that the cyclic functions for Mz are precisely
the outer functions. His readable paper [48] caused many operator theorists
to become interested in function theory on the disc. Reciprocally, many
function theorists began to work on problems motivated by operator-theoretic
considerations. There have been many extensions and refinements of Theorem
1.33, which have underlain the development of two substantial theories: the
Sz.-Nagy-Foias model theory, which we outline in Section 1.11, and the Lax-
Phillips scattering theory [141].
1.7 Von Neumann’s Theory of Spectral Sets

In 1951 von Neumann initiated the theory of spectral sets in the paper [210].
After observing that for any normal operator N ∈ B(H) and f ∈ Rat(σ (N )),
by virtue of equation (1.22),

f (N ) = f (z) dE(z)
≤ max |f (z)|,
z∈σ (N )
σ (N )
von Neumann introduced the following terminology.

Definition 1.34. Let H be a Hilbert space, let T ∈ B(H) and let K be a
compact subset of C. We say that K is a spectral set for T if σ (T ) ⊆ K and,
for all f ∈ Rat(K),
f (T ) ≤ max |f (z)|. (1.35)

z∈K
Thus von Neumann’s observation is equivalent to the assertion that σ (N ) is

a spectral set for any normal operator N .
In his paper von Neumann went on to promulgate his famous inequality,
whose influence on the subsequent development of operator theory has been
profound.
Theorem 1.36. (Von Neumann’s inequality) If H is a Hilbert space,
T ∈ B(H), and T ≤ 1 then, for any p ∈ C[z],
p(T ) ≤ max |p(z)|.

|z|≤1
We prove the inequality in Section 1.9 by a different method.

Corollary 1.37. If H is a Hilbert space, T ∈ B(H) and T ≤ 1, then D− is a
spectral set for T .
Proof. Fix T ∈ B(H) with T ≤ 1 and let f ∈ Rat(D− ). Since f is
holomorphic on a neighborhood of D− , it follows by a simple power series
argument that there exists r > 1 and a sequence of polynomials {pn } such
that pn converges to f uniformly on rD. But then by the continuity of the
functional calculus,
f (T ) = lim pn (T ) ≤ lim max |pn (z)| = max |f (z)|.

n→∞ n→∞ |z|≤1 |z|≤1
There are numerous ways to generalize the notion of spectral sets to several
variables. One idea is to juice up the inequality (1.35) to all functions f
holomorphic on a neighborhood of K.
Lemma 1.38. Let T ∈ B(H) and assume that K ⊆ C is compact. Then K is
a spectral set for T if and only if σ (T ) ⊆ K and, for all f holomorphic in a
neighborhood of K,
f (T ) ≤ max |f (z)|. (1.39)

z∈K
Proof. The proof of necessity is immediate from the observation that the
elements of Rat(K) are holomorphic on a neighborhood of K.
To prove the converse, assume that σ (T ) ⊆ K and inequality (1.35) holds.
Fix a neighborhood U of K and f ∈ Hol(U ). By Runge’s theorem, there exists
a sequence of rational functions {fn } in Rat(U − ) such that fn converges locally
uniformly to f on U . By the continuity of the functional calculus,
f (T ) = lim fn (T ) ≤ lim max |fn (z)| = max |f (z)|.

n→∞ n→∞ z∈K z∈K
There are several natural candidates for the definition of a holomorphic

function (possibly vector valued) of several variables; fortunately though, they
all turn out to be equivalent. This issue is discussed in detail in [136, sections
0.2 and 1.2]. We shall adopt the following.
Definition 1.40. Let X be a Banach space, let be an open set in Cd for
some positive integer d, and let f : → X be a mapping. We say that f is
differentiable at a point z0 ∈ if there exists a linear map Df (z0 ): Cd → X
such that
f (z) − f (z0 ) = Df (z0 )(z − z0 ) + o(z − z0 ) as z → z0 , z ∈ .
The linear map Df (z0 ), when there is one, is unique and is called the derivative
or Fréchet derivative of f at z0 . It is also denoted by f (z0 ). The mapping f
is holomorphic on if it is differentiable at every point of .
It is clear that if f is holomorphic then the function of one variable
λ → f (λ, z2 , z3 , . . . , zd )
is analytic in λ for any fixed values of z2 , . . . , zd , and likewise for each of the
d co-ordinates. One can summarize this statement by saying that holomorphic
maps are holomorphic in each variable separately. Remarkably enough, the
converse statement is also true: any separately holomorphic map on an open
set in Cd is holomorphic. This is a theorem of Hartogs, which caused a surprise
when it was proved in 1906. Another criterion for a function to be holomorphic
is that it have a convergent power series expansion in a neighborhood of
every point.
Lemma 1.38 suggests the following definition of spectral sets in several
variables. For a compact set K in Cd , let Hol(K) denote the set of functions
holomorphic on a neighborhood of K (the neighborhood can depend on the
function).
Definition 1.41. Let H be a Hilbert space, let T be a pairwise commuting d-
tuple of operators on H and let K be a compact subset of Cd . We say that K
is a spectral set for T if σ (T ) ⊆ K and, for all f ∈ Hol(K),
f (T ) ≤ max |f (z)|. (1.42)

z∈K
We give another version of the definition, valid for polynomially convex
compact sets, which involves only the algebraic spectrum of T , as defined in
Section 1.2.
Definition 1.43. A compact subset K of Cd is said to be polynomially convex
if, for every λ ∈ Cd \ K, there exists a polynomial f in d variables such that
|f | ≤ 1 on K and |f (λ)| > 1. The polynomial hull of a bounded subset A
of Cd is the intersection of all polynomially convex compact sets in Cd that
contain A.
An open set U in Cd is said to be polynomially convex if, for every compact
subset K of U , the polynomial hull of K is contained in U .
Observe that the polynomial hull of a compact set is a compact polynomi-
ally convex set.
Definition 1.44. Let H be a Hilbert space, let T be a pairwise commuting
d-tuple of operators on H and let K be a polynomially convex compact subset
of Cd . We say that K is a spectral set for T if σalg (T ) ⊆ K and, for all
f ∈ Hol(K),
f (T ) ≤ max |f (z)|. (1.45)
z∈K
Just as long as σ (T ) ⊆ K, Lemma 1.38 asserts that there is no difference

between the conditions (1.35) and (1.39) in one variable. However, in several
variables this statement is true if and only if K is rationally convex. With the
exception of Chapter 9, we shall be dealing with functions on polynomially
convex sets, in which case K is a spectral set for T if and only if inequality
(1.39) holds for polynomials.
1.8 The Schur Class and Spectral Domains

For any open set U in Cd , we denote by H∞ (U ) the Banach algebra of bounded
holomorphic functions on U equipped with the norm U defined by
ϕU = sup |ϕ(λ)| for all ϕ ∈ H∞ (U ).
λ∈U
We define the Schur class of U , S (U ), by

S (U ) = ball H∞ (U ).
Schur himself, in [184, 185], studied the class S (D) of analytic functions in
D bounded in modulus by 1.
We shall adopt a slight change of perspective on the notion of spectral
set described in the preceding section. Instead of fixing K and assuming that
σ (T ) ⊆ K and f is holomorphic on a neighborhood of K, let us fix an open
set U and assume that σ (T ) ⊆ U and f is holomorphic on U . This leads to the
following definition.
Definition 1.46. If T is a d-tuple of pairwise commuting operators and U is
an open set in Cd , we say that U is a spectral domain for T if σ (T ) ⊆ U and,
for all ϕ ∈ S (U ),
ϕ(T ) ≤ 1.
Thus K is a spectral set for T if and only if every neighborhood of K is a
spectral domain for T . Lemma 1.38 is equivalent via a simple approximation
argument to the assertion that D is a spectral domain for any contraction with
spectrum in D. Theorem 1.36 can be recast in the following form.
Theorem 1.47. (Schur form of von Neumann’s inequality) If H is a Hilbert
space, T ∈ B(H), T ≤ 1, and σ (T ) ⊆ D, then
ϕ ∈ S (D) ⇒ ϕ(T ) ≤ 1.
We close this section with the formalization of a problem that remains one
of the most fundamental in operator theory.
Problem 1.48. (The first Holy Grail of spectral set theory) Given a compact
set K or a domain U in Cd , find simple necessary and sufficient conditions on
a tuple T for K or U to be a spectral set or spectral domain respectively for T .
How would success in this quest provide value to function theory? In
proving his inequality, von Neumann found the first Holy Grail for D, namely,
T ≤ 1. What does this mean in the simple case

0 1
T = ?
0 0
Clearly, T ≤ 1. Therefore, by the Schur form of von Neumann’s inequality,
if ϕ ∈ S (D), then ϕ(T ) ≤ 1. But if ϕ ∈ S (D) and ϕ(0) = 0, then

0 ϕ (0)
ϕ(T ) = ,
0 0
which obviously has norm equal to |ϕ (0)|. Thus von Neumann’s inequality in
this simple case is the assertion that, for all ϕ ∈ S (D),
ϕ(0) = 0 ⇒ |ϕ (0)| ≤ 1.
This is the classical Schwarz lemma, which appears in every introductory
course in complex analysis. We shall see many, much less trivial examples
in these notes of how finding the first Holy Grail for a domain U leads to
interesting function theory on U .
1.9 The Sz.-Nagy Dilation Theorem
Two years after the appearance of von Neumann’s inequality, in the 1953 paper
[193] Béla Szőkefalvi-Nagy published his remarkable dilation theorem, which
was the starting point for a major subfield of operator theory.
Definition 1.49. If H is a Hilbert space and T ∈ B(H), a power dilation of T
is an operator V on a Hilbert space K ⊇ H such that T n = PH V n |H for all
integers n ≥ 0.
Theorem 1.50. (Sz.-Nagy dilation theorem) If H is a Hilbert space,

T ∈ B(H), and T ≤ 1, then T has a unitary power dilation, that is, it
has a power dilation to a unitary operator on some superspace of K.
Sz.-Nagy’s theorem gave much deeper insight into von Neumann’s inequal-
ity than the original proof, which had used measure theory (which was not
widely understood at the time) and a hefty dose of classical function theory.
1.9.1 Sz.-Nagy’s Proof of von Neumann’s Inequality

Let H be a Hilbert space, T ∈ B(H), and T ≤ 1. By Sz.-Nagy’s theorem,
there exists a unitary power dilation U of T . As noted earlier, the spectral
theorem implies that T is a spectral set for U , that is,
p(U ) ≤ max |p(λ)|
λ∈T
for all polynomials p. But by the maximum principle, if p is a polynomial,

then maxλ∈T |p(λ)| = maxλ∈D− |p(λ)|. Therefore,
p(U ) ≤ max |p(λ)|
λ∈D−
for all polynomials p. By taking linear combinations of the power dilation

conditions T n = PH U n |H for n ≥ 0 one deduces that
p(T ) = PH p(U ) |H
for all polynomials p. Therefore,
p(T ) = PH p(U ) |H ≤ p(U ) ≤ max |p(λ)|
λ∈D−
for all polynomials p. 2

Thus dilations yield spectral sets. William Arveson in his 1969 article [36]
and 1972 article [37] developed a converse to Sz.-Nagy’s dilation theorem,
one that has had numerous applications throughout operator theory and the
theory of C ∗ -algebras. Arveson introduced a powerful variant of the idea of a
spectral set. In order to describe it, we observe that if Aij is a bounded linear
operator on H for i, j = 1, . . . , n, then the operator matrix
n
A = Aij i,j =1
defines a bounded linear operator on Hn in a natural way, and therefore the

operator norm A is well defined.
We introduce the notation ∂X for the topological boundary of a subset X
of Cd .
Definition 1.51. Let T be a d-tuple of pairwise commuting operators acting
on a Hilbert space H and let K be a polynomially convex compact subset of
Cd containing σalg (T ). Then K is said to be a complete spectral set for T if, for
each positive integer n and each n × n matrix of functions f = [fij ] analytic
on a neighborhood of K,
[fij (T )] ≤ max [fij (λ)] .

λ∈K
T is said to dilate to the boundary of K if there exists a Hilbert space K ⊇ H

and a pairwise commuting d-tuple of normal operators N acting on K with
σ (N ) ⊆ ∂K and such that
f (T ) = PH f (N)|H,
whenever f is holomorphic on a neighborhood of K.

By mimicking Sz.-Nagy’s proof of von Neumann’s inequality, one can
easily show that, if T dilates to the boundary of K, then K is a complete
spectral set for T . Arveson proved the converse of this statement [37].
Theorem 1.52. (Arveson’s dilation theorem) K is a complete spectral set
for T if and only if T has a dilation to the boundary of K.
Problem 1.53. (The second Holy Grail of spectral set theory) For which
compact sets K ⊆ Cd is it true that K is a complete spectral set for T whenever
K is a spectral set for T ?
1.10 Andô’s Dilation Theorem

In 1963 Tsuyoshi Andô [33] extended Sz.-Nagy’s dilation theorem to pairs of
commuting contractions.
Theorem 1.54. (Andô’s dilation theorem) If H is a Hilbert space and
T = (T 1 , T 2 ) is a commuting pair of contractions acting on H, then T has
a unitary power dilation, that is, there exists a Hilbert space K ⊇ H and a pair
of commuting unitary operators U = (U 1 , U 2 ) acting on K such that

(T 1 )n1 (T 2 )n2 = PH (U 1 )n1 (U 2 )n2 |H
for all integers n1 , n2 ≥ 0.
For a proof, see, for example [195, theorem I.6.4] or [11, theorem 10.26].
Clearly, by the method of Sz.-Nagy’s proof of von Neumann’s inequality,
Andô’s dilation theorem has the following corollary.
Theorem 1.55. (Andô’s inequality) If H is a Hilbert space, T = (T 1 , T 2 ) is
a commuting pair of contractions acting on H and σalg (T ) ⊆ D2 , then
ϕ ∈ S (D2 ) ⇒ ϕ(T ) ≤ 1.
This beautiful theorem obtained both of the Holy Grails of spectral set
theory simultaneously. One shows, as in Sz.-Nagy’s proof of von Neumann’s
inequality, that (D2 )− is a spectral set for a commuting pair T if and only if
T 1 , T 2 ≤ 1 (Holy Grail 1). Andô’s dilation theorem also enables us to
prove a result (Theorem 4.93) from which it is easy to deduce that (D2 )− is
a complete spectral set for T whenever (D2 )− is a spectral set for T (Holy
Grail 2) (see Corollary 4.108).
1.11 The Sz.-Nagy–Foias Model Theory

After the discovery of his dilation theorem, Sz.-Nagy worked with Ciprian
Foias to exploit the theorem and thereby develop a model theory for contrac-
tions on Hilbert space through the use of function theory on the unit disc.
In 1967 Sz.-Nagy and Foias published Analyse harmonique des opérateurs
de l’espace de Hilbert, a self-contained treatment of dozens of their papers.
This is certainly one of the most influential works ever published in operator
theory. The first English-language edition was [194]; an updated version by H.
Bercovici and L. Kérchy is [195].
The idea of the Sz.-Nagy–Foias model is to represent a general contraction
on Hilbert space as (up to unitary equivalence) a “piece” of a concrete operator
on a Hilbert space of functions on T, to which classical analyse harmonique
could be applied.
One of the highlights of their theory was the introduction of the character-
istic operator function of a contraction. For T a contraction acting on a Hilbert
space H, let
1 − 1 −
DT = ran(1 − T ∗ T ) 2 and DT ∗ = ran(1 − T T ∗ ) 2 .
The characteristic operator function of T is the B(DT , DT ∗ )-valued function

T defined on D by
1 1
T (λ) = −T + λ(1 − T T ∗ ) 2 (1 − λT ∗ )−1 (1 − T ∗ T ) 2 . (1.56)
T is used to define the space of functions on T on which the “model” of T
acts. The model itself is closely related to the shift operator.
Sz.-Nagy and Foias initiated a substantial program aimed at solving the
invariant subspace problem by extending the classical factorization theory
for scalar-valued Schur functions on the disc to operator-valued functions.
Although this program has had some successes, the invariant subspace problem
itself remains stubbornly open.
In 1974 J. William Helton [101] pointed out that the characteristic operator
function of a contraction was a special case (albeit with a more refined
structure) of the network realization formula widely used by electrical engi-
neers to represent rational functions, typically transfer functions or impedance
functions of linear systems or circuits. The formalism became even more
important in the 1980s in the then newly developing branch of control theory
now known as H∞ -control. The realization formula provides a representation
of operator-valued Schur functions. For Hilbert spaces H, K, let H∞ B(H,K) (D)
denote the Banach space of bounded analytic B(H, K)-valued functions on D
equipped with the norm
ϕD = sup ϕ(z), (1.57)
z∈D
and then define SB(H,K) (D), the B(H, K)-valued Schur class on D, by
SB(H,K) (D) = ball H∞
B(H,K) (D).
Of course, when K = H, we abbreviate H∞ ∞
B(H,K) (D) to HB(H) (D). We give a
brief explanation of the idea of a network realization in Section 2.5.
Theorem 1.58. (Network realization formula) A mapping ϕ: D → B(H, K)
belongs to SB(H,K) (D) if and only if there exist a Hilbert space M and an
isometry

A B
V = : H ⊕ M → K ⊕ M,
C D
such that, for all λ ∈ D,
ϕ(λ) = A + Bλ(1 − Dλ)−1 C. (1.59)
Helton’s seminal observation created new synergies between the operator-
theoretic function theory and mathematical control theory communities that
continue to the present.
Remarks 1.60.
(i) If H = K and
1 1
A = −T , B = (1 − T T ∗ ) 2 , C = (1 − T ∗ T ) 2 , and D = T ∗ ,
then equation (1.59) becomes equation (1.56), and V is not just isomet-
ric, it is unitary.
(ii) In the special case when dim H = dim K < ∞, dim M < ∞, and ϕ is
defined by equation (1.59), the condition that V be isometric implies
that ϕ is a rational inner function, that is, ϕ(τ ) is unitary for all τ ∈ T
and det ϕ is a scalar rational inner function (a finite Blaschke product).
Conversely, if ϕ ∈ H∞ B(H,K) is a rational inner function, then there exists
M with dim M < ∞ and an isometric V such that equation (1.59) holds.
(iii) In the case when H = C, we may assume that

A B a 1⊗β
V = = ,
C D γ ⊗1 D
where a ∈ C, β ∈ M, γ ∈ M, and D ∈ B(M). In this case the realization
formula (1.59) becomes

ϕ(λ) = a + λ (1 − Dλ)−1 γ , β .
M
1.12 The Sarason Interpolation Theorem

In 1967 Donald Sarason added another ingredient to the mix with the
influential paper “Generalized Interpolation in H∞ ” [180]. In it he exploited
yet another way to think about Schur-class functions from an operator-theoretic
perspective.
For ϕ ∈ H∞ (D), we denote by Mϕ the multiplication operator, defined on
2
H by the formula
Mϕ f = ϕf for f ∈ H2 . (1.61)
The following lemma asserts that this formula does indeed define a bounded
operator.
Lemma 1.62. If ϕ ∈ H∞ (D) and Mϕ is defined by equation (1.61), then
Mϕ f ∈ H2 whenever f ∈ H2 . Furthermore, Mϕ (as an operator acting on
H2 ) has norm less than or equal to ϕD .
Proof. By equation (1.28), if f ∈ H2 , then

dt
Mϕ f 2 = lim |ϕ(reit )f (reit )|2
r→1− 2π

dt
≤ ϕD lim
2
|f (reit )|2
r→1− 2π
= ϕ2D f 2H2 .
Therefore, Mϕ ≤ ϕD .

Lemma 1.62 suggests the following natural question: given T ∈ B(H2 ),
when is T = Mϕ for some ϕ ∈ H∞ (D)?
Theorem 1.63. Let T : B(H) → B(H) be a function. Then
T = Mϕ for some ϕ ∈ H∞ (D) (1.64)
if and only if T ∈ B(H2 ) and T commutes with Mz . Furthermore, when
equation (1.64) holds,
T = ϕD .
In particular, S (D) corresponds naturally to the set of contractions on H2 that
commute with Mz .
An illuminating proof of Theorem 1.63 relies on the fact that H2 is a
reproducing kernel Hilbert space and exploits a property of the reproducing
kernel functions sλ : they are eigenvectors of all adjoints of multiplication
operators.
Lemma 1.65. If ϕ ∈ H∞ (D) and λ ∈ D, then
Mϕ∗ sλ = ϕ(λ) sλ . (1.66)
Proof. Fix ϕ ∈ H∞ (D) and λ ∈ D. If f ∈ H2 , then

f , Mϕ∗ sλ = Mϕ f , sλ
= ϕf , sλ
= ϕ(λ)f (λ)
= ϕ(λ) f , sλ

= f , ϕ(λ) sλ .
Since f is arbitrary, it follows that equation (1.66) holds.
Furthermore, the reproducing kernel functions sλ are the only eigenvectors
for all adjoints of multiplication operators.
Lemma 1.67. If λ ∈ D, then
ker(λ̄ − Mz∗ ) = Csλ .
Proof. By Lemma 1.65, Mλ∗ sλ = λ̄ sλ . Therefore, ker(λ̄ − Mz∗ ) ⊇ Csλ . To
prove the reverse inclusion, we show that if f ∈ H2 and f ⊥ C sλ , then
f ⊥ ker(λ̄ − Mz∗ ). But if f ∈ H2 and f ⊥ C sλ , then f (λ) = 0. Hence there
exists a holomorphic function g on D such that f (z) = (λ − z)g(z). By the
criterion (1.27), g ∈ H2 . This implies that f ∈ ran(λ − Mz ), or equivalently,
f ⊥ ker(λ̄ − Mz∗ ).
Armed with these lemmas, one can easily prove Theorem 1.63.
Proof. First assume that ϕ ∈ H∞ (D) and T = Mϕ . Lemma 1.62 implies that
T ∈ B(H2 ). Also, if f ∈ H2 and z ∈ D, then
((T Mz )f )(z) = ϕ(z)zf (z) = zϕ(z)f (z) = ((Mz T )f )(z),
that is, T commutes with Mz .
Conversely, assume that T ∈ B(H2 ) and T Mz = Mz T . If λ ∈ D, then Lemma
1.65 implies that
Mz∗ (T ∗ sλ ) = T ∗ (Mz∗ sλ ) = λ̄ T ∗ sλ .
Consequently, Lemma 1.67 implies that for each λ ∈ D, there exists ϕ(λ) ∈ C
such that
T ∗ sλ = ϕ(λ)sλ . (1.68)
We claim that ϕ ∈ H∞ (D). First observe that if f ∈ H2 and λ ∈ D,

(Tf )(λ) = Tf , sλ = f , T ∗ sλ = f , ϕ(λ)sλ = ϕ(λ) f , sλ
= ϕ(λ)f (λ). (1.69)

In particular, if we let f be the constant function 1, we see that ϕ = T 1 ∈ H2 ,
so that ϕ is analytic in D. To see that ϕ is bounded, note that equation (1.68)
implies that if λ ∈ C, then |ϕ(λ)| ≤ T ∗ = T . Hence ϕ ∈ H∞ and
ϕD ≤ T . (1.70)
There remains to show that T = ϕD . This follows from the inequality
(1.70) and the observation that Lemma 1.62 implies that T ≤ ϕD . This
concludes the proof of Theorem 1.63.
In [180], Sarason proved a generalization of Theorem 1.63, Theorem 1.75.
If H is a Hilbert space and E is a subset of H, we denote by H E the
orthogonal complement of E in H:
def
H E = {x ∈ H: x, y = 0 for all y ∈ E}.
When the space H is understood, it is customary to write E ⊥ instead of H E.
Sarason’s work involved the case that H = H2 and E = ϕH2 for some inner
function ϕ.5
The following lemma gives a concrete description of H2 ϕH2 in the case
that ϕ is a finite Blaschke product.
5 So that, by Beurling’s theorem, E is the general M -invariant subspace of H2 .

z
Lemma 1.71. If B is the rational inner function
n
z − λj
B(z) = , (1.72)
j =1
1 − λ¯j z
where λ1 , . . . , λn are distinct points in D, then
H2 BH2 = span {sλ1 , sλ2 , . . . , sλn }. (1.73)

Proof. For each j , by the reproducing property of the Szegő kernel, for any
f ∈ H2 ,

Bf , sλj H2 = B(λj )f (λj ) = 0.
Thus sλj ∈ H2 BH2 , and so
span {sλ1 , sλ2 , . . . , sλn } ⊆ H2 BH2 . (1.74)
f ∈ H orthogonal to the
To prove the reverse inclusion, consider any function 2
left-hand side of the relation (1.74). Then f , sλj = 0 for each j , which is to
say that f vanishes at all the points λ1 , . . . , λn . It follows that the function
def f (z) (1 − λ¯1 z) . . . (1 − λ¯n z)f (z)

g(z) = =
B(z) (z − λ1 ) . . . (z − λn )
also belongs to H2 . Thus f = Bg ∈ BH2 . Hence
span {sλ1 , sλ2 , . . . , sλn }⊥ ⊆ BH2 ,
and therefore equation (1.73) holds.

Theorem 1.75. (Sarason’s theorem) If ϕ is an inner function,M = H2 ϕH2 ,
and T is a bounded operator on M, then T commutes with PM Mz |M if and
only if there exists ψ ∈ H∞ such that
T = PM Mψ |M.
Furthermore, ψ may be chosen with ψ∞ = T .

Sarason showed that his theorem encodes, unifies, and extends classical
interpolation and moment theorems on the disc. We illustrate Sarason’s idea
by giving a proof in his style of the classical Pick interpolation theorem. To
this end we need two lemmas; the first one is proved just like Lemma 1.65.
Lemma 1.76. If B is as in Lemma 1.71, M = H2 BH2 and ψ ∈ H∞ , then
∗
PM Mψ |M sλi = ψ(λi ) sλi for i = 1, . . . , n. (1.77)
Another random document with
no related content on Scribd:
PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK
To protect the Project Gutenberg™ mission of promoting the free

distribution of electronic works, by using or distributing this work (or
any other work associated in any way with the phrase “Project
Gutenberg”), you agree to comply with all the terms of the Full
Project Gutenberg™ License available with this file or online at
www.gutenberg.org/license.
Section 1. General Terms of Use and

Redistributing Project Gutenberg™
electronic works
1.A. By reading or using any part of this Project Gutenberg™
electronic work, you indicate that you have read, understand, agree
to and accept all the terms of this license and intellectual property
(trademark/copyright) agreement. If you do not agree to abide by all
the terms of this agreement, you must cease using and return or
destroy all copies of Project Gutenberg™ electronic works in your
possession. If you paid a fee for obtaining a copy of or access to a
Project Gutenberg™ electronic work and you do not agree to be
bound by the terms of this agreement, you may obtain a refund from
the person or entity to whom you paid the fee as set forth in
paragraph 1.E.8.
1.B. “Project Gutenberg” is a registered trademark. It may only be

used on or associated in any way with an electronic work by people
who agree to be bound by the terms of this agreement. There are a
few things that you can do with most Project Gutenberg™ electronic
works even without complying with the full terms of this agreement.
See paragraph 1.C below. There are a lot of things you can do with
Project Gutenberg™ electronic works if you follow the terms of this
agreement and help preserve free future access to Project
Gutenberg™ electronic works. See paragraph 1.E below.
1.C. The Project Gutenberg Literary Archive Foundation (“the
Foundation” or PGLAF), owns a compilation copyright in the
collection of Project Gutenberg™ electronic works. Nearly all the
individual works in the collection are in the public domain in the
United States. If an individual work is unprotected by copyright law in
the United States and you are located in the United States, we do
not claim a right to prevent you from copying, distributing,
performing, displaying or creating derivative works based on the
work as long as all references to Project Gutenberg are removed. Of
course, we hope that you will support the Project Gutenberg™
mission of promoting free access to electronic works by freely
sharing Project Gutenberg™ works in compliance with the terms of
this agreement for keeping the Project Gutenberg™ name
associated with the work. You can easily comply with the terms of
this agreement by keeping this work in the same format with its
attached full Project Gutenberg™ License when you share it without
charge with others.
1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside the
United States, check the laws of your country in addition to the terms
of this agreement before downloading, copying, displaying,
performing, distributing or creating derivative works based on this
work or any other Project Gutenberg™ work. The Foundation makes
no representations concerning the copyright status of any work in
any country other than the United States.
1.E. Unless you have removed all references to Project Gutenberg:
1.E.1. The following sentence, with active links to, or other

immediate access to, the full Project Gutenberg™ License must
appear prominently whenever any copy of a Project Gutenberg™
work (any work on which the phrase “Project Gutenberg” appears, or
with which the phrase “Project Gutenberg” is associated) is
accessed, displayed, performed, viewed, copied or distributed:
This eBook is for the use of anyone anywhere in the United
States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it away
or re-use it under the terms of the Project Gutenberg License
included with this eBook or online at www.gutenberg.org. If you
are not located in the United States, you will have to check the
laws of the country where you are located before using this
eBook.
1.E.2. If an individual Project Gutenberg™ electronic work is derived

from texts not protected by U.S. copyright law (does not contain a
notice indicating that it is posted with permission of the copyright
holder), the work can be copied and distributed to anyone in the
United States without paying any fees or charges. If you are
redistributing or providing access to a work with the phrase “Project
Gutenberg” associated with or appearing on the work, you must
comply either with the requirements of paragraphs 1.E.1 through
1.E.7 or obtain permission for the use of the work and the Project
Gutenberg™ trademark as set forth in paragraphs 1.E.8 or 1.E.9.
1.E.3. If an individual Project Gutenberg™ electronic work is posted

with the permission of the copyright holder, your use and distribution
must comply with both paragraphs 1.E.1 through 1.E.7 and any
additional terms imposed by the copyright holder. Additional terms
will be linked to the Project Gutenberg™ License for all works posted
with the permission of the copyright holder found at the beginning of
this work.
1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files containing a
part of this work or any other work associated with Project
Gutenberg™.
1.E.5. Do not copy, display, perform, distribute or redistribute this

electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1 with
active links or immediate access to the full terms of the Project
Gutenberg™ License.
1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if you
provide access to or distribute copies of a Project Gutenberg™ work
in a format other than “Plain Vanilla ASCII” or other format used in
the official version posted on the official Project Gutenberg™ website
(www.gutenberg.org), you must, at no additional cost, fee or expense
to the user, provide a copy, a means of exporting a copy, or a means
of obtaining a copy upon request, of the work in its original “Plain
Vanilla ASCII” or other form. Any alternate format must include the
full Project Gutenberg™ License as specified in paragraph 1.E.1.
1.E.7. Do not charge a fee for access to, viewing, displaying,

performing, copying or distributing any Project Gutenberg™ works
unless you comply with paragraph 1.E.8 or 1.E.9.
1.E.8. You may charge a reasonable fee for copies of or providing

access to or distributing Project Gutenberg™ electronic works
provided that:
• You pay a royalty fee of 20% of the gross profits you derive from
the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
about donations to the Project Gutenberg Literary Archive
Foundation.”
• You provide a full refund of any money paid by a user who

notifies you in writing (or by e-mail) within 30 days of receipt that
s/he does not agree to the terms of the full Project Gutenberg™
License. You must require such a user to return or destroy all
copies of the works possessed in a physical medium and
discontinue all use of and all access to other copies of Project
Gutenberg™ works.
• You provide, in accordance with paragraph 1.F.3, a full refund of

any money paid for a work or a replacement copy, if a defect in
the electronic work is discovered and reported to you within 90
days of receipt of the work.
• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.
1.E.9. If you wish to charge a fee or distribute a Project Gutenberg™

electronic work or group of works on different terms than are set
forth in this agreement, you must obtain permission in writing from
the Project Gutenberg Literary Archive Foundation, the manager of
the Project Gutenberg™ trademark. Contact the Foundation as set
forth in Section 3 below.
1.F.
1.F.1. Project Gutenberg volunteers and employees expend

considerable effort to identify, do copyright research on, transcribe
and proofread works not protected by U.S. copyright law in creating
the Project Gutenberg™ collection. Despite these efforts, Project
Gutenberg™ electronic works, and the medium on which they may
be stored, may contain “Defects,” such as, but not limited to,
incomplete, inaccurate or corrupt data, transcription errors, a
copyright or other intellectual property infringement, a defective or
damaged disk or other medium, a computer virus, or computer
codes that damage or cannot be read by your equipment.
1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except

for the “Right of Replacement or Refund” described in paragraph
1.F.3, the Project Gutenberg Literary Archive Foundation, the owner
of the Project Gutenberg™ trademark, and any other party
distributing a Project Gutenberg™ electronic work under this
agreement, disclaim all liability to you for damages, costs and
expenses, including legal fees. YOU AGREE THAT YOU HAVE NO
REMEDIES FOR NEGLIGENCE, STRICT LIABILITY, BREACH OF
WARRANTY OR BREACH OF CONTRACT EXCEPT THOSE
PROVIDED IN PARAGRAPH 1.F.3. YOU AGREE THAT THE
FOUNDATION, THE TRADEMARK OWNER, AND ANY
DISTRIBUTOR UNDER THIS AGREEMENT WILL NOT BE LIABLE
TO YOU FOR ACTUAL, DIRECT, INDIRECT, CONSEQUENTIAL,
PUNITIVE OR INCIDENTAL DAMAGES EVEN IF YOU GIVE
NOTICE OF THE POSSIBILITY OF SUCH DAMAGE.
1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you

discover a defect in this electronic work within 90 days of receiving it,
you can receive a refund of the money (if any) you paid for it by
sending a written explanation to the person you received the work
from. If you received the work on a physical medium, you must
return the medium with your written explanation. The person or entity
that provided you with the defective work may elect to provide a
replacement copy in lieu of a refund. If you received the work
electronically, the person or entity providing it to you may choose to
give you a second opportunity to receive the work electronically in
lieu of a refund. If the second copy is also defective, you may
demand a refund in writing without further opportunities to fix the
problem.
1.F.4. Except for the limited right of replacement or refund set forth in
paragraph 1.F.3, this work is provided to you ‘AS-IS’, WITH NO
OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.
1.F.5. Some states do not allow disclaimers of certain implied

warranties or the exclusion or limitation of certain types of damages.
If any disclaimer or limitation set forth in this agreement violates the
law of the state applicable to this agreement, the agreement shall be
interpreted to make the maximum disclaimer or limitation permitted
by the applicable state law. The invalidity or unenforceability of any
provision of this agreement shall not void the remaining provisions.
1.F.6. INDEMNITY - You agree to indemnify and hold the
Foundation, the trademark owner, any agent or employee of the
Foundation, anyone providing copies of Project Gutenberg™
electronic works in accordance with this agreement, and any
volunteers associated with the production, promotion and distribution
of Project Gutenberg™ electronic works, harmless from all liability,
costs and expenses, including legal fees, that arise directly or
indirectly from any of the following which you do or cause to occur:
(a) distribution of this or any Project Gutenberg™ work, (b)
alteration, modification, or additions or deletions to any Project
Gutenberg™ work, and (c) any Defect you cause.
Section 2. Information about the Mission of

Project Gutenberg™
Project Gutenberg™ is synonymous with the free distribution of
electronic works in formats readable by the widest variety of
computers including obsolete, old, middle-aged and new computers.
It exists because of the efforts of hundreds of volunteers and
donations from people in all walks of life.
Volunteers and financial support to provide volunteers with the

assistance they need are critical to reaching Project Gutenberg™’s
goals and ensuring that the Project Gutenberg™ collection will
remain freely available for generations to come. In 2001, the Project
Gutenberg Literary Archive Foundation was created to provide a
secure and permanent future for Project Gutenberg™ and future
generations. To learn more about the Project Gutenberg Literary
Archive Foundation and how your efforts and donations can help,
see Sections 3 and 4 and the Foundation information page at
www.gutenberg.org.
Section 3. Information about the Project

Gutenberg Literary Archive Foundation
The Project Gutenberg Literary Archive Foundation is a non-profit
501(c)(3) educational corporation organized under the laws of the
state of Mississippi and granted tax exempt status by the Internal
Revenue Service. The Foundation’s EIN or federal tax identification
number is 64-6221541. Contributions to the Project Gutenberg
Literary Archive Foundation are tax deductible to the full extent
permitted by U.S. federal laws and your state’s laws.
The Foundation’s business office is located at 809 North 1500 West,

Salt Lake City, UT 84116, (801) 596-1887. Email contact links and up
to date contact information can be found at the Foundation’s website
and official page at www.gutenberg.org/contact
Section 4. Information about Donations to

the Project Gutenberg Literary Archive
Foundation
Project Gutenberg™ depends upon and cannot survive without
widespread public support and donations to carry out its mission of
increasing the number of public domain and licensed works that can
be freely distributed in machine-readable form accessible by the
widest array of equipment including outdated equipment. Many small
donations ($1 to $5,000) are particularly important to maintaining tax
exempt status with the IRS.
The Foundation is committed to complying with the laws regulating

charities and charitable donations in all 50 states of the United
States. Compliance requirements are not uniform and it takes a
considerable effort, much paperwork and many fees to meet and
keep up with these requirements. We do not solicit donations in
locations where we have not received written confirmation of
compliance. To SEND DONATIONS or determine the status of
compliance for any particular state visit www.gutenberg.org/donate.
While we cannot and do not solicit contributions from states where

we have not met the solicitation requirements, we know of no
prohibition against accepting unsolicited donations from donors in
such states who approach us with offers to donate.
International donations are gratefully accepted, but we cannot make

any statements concerning tax treatment of donations received from
outside the United States. U.S. laws alone swamp our small staff.
Please check the Project Gutenberg web pages for current donation
methods and addresses. Donations are accepted in a number of
other ways including checks, online payments and credit card
donations. To donate, please visit: www.gutenberg.org/donate.
Section 5. General Information About Project

Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could be
freely shared with anyone. For forty years, he produced and
distributed Project Gutenberg™ eBooks with only a loose network of
volunteer support.
Project Gutenberg™ eBooks are often created from several printed

editions, all of which are confirmed as not protected by copyright in
the U.S. unless a copyright notice is included. Thus, we do not
necessarily keep eBooks in compliance with any particular paper
edition.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.
This website includes information about Project Gutenberg™,

including how to make donations to the Project Gutenberg Literary
Archive Foundation, how to help produce our new eBooks, and how
to subscribe to our email newsletter to hear about new eBooks.

Operator Analysis Hilbert Space Methods in Complex Analysis Jim Agler Full Chapter PDF

Uploaded by

Copyright:

Available Formats

You might also like

Operator Analysis Hilbert Space Methods in Complex Analysis Jim Agler Full Chapter PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Operator Analysis Hilbert Space Methods in Complex Analysis Jim Agler Full Chapter PDF

Uploaded by

Copyright:

Available Formats

Operator Analysis: Hilbert Space

Methods in Complex Analysis Jim Agler

Modern Control: State-Space Analysis and Design Methods

Gaussian Measures in Hilbert Space: Construction and

Integration with Complex Numbers : A Primer on Complex

Visual Complex Analysis: 25th Anniversary Edition

Numerical Methods in Environmental Data Analysis Moses

Genetic Analysis of Complex Disease 3rd Edition William

Introduction to Research Methods and Data Analysis in

Research Methods, Design, and Analysis – Ebook PDF

B. BOLLOB ÁS, W. FULTON, F. KIRWAN,

219 Operator Analysis

Recent titles include the following:

Cambridge University Press is part of the University of Cambridge.

Preface page xiii

1.2 Functional Calculi

The polynomial calculus can easily be extended to the power series

defines an analytic function on {z ∈ C: |z − a| < r} for some r > 0, then

(1) If f (z) = 1, then f (T ) = 1 and if f (z) = z, then f (T ) = T .

is a unital algebra-homomorphism from Hol(U ) to B(X ).

Any joint eigenvalue of T does belong to σA (T ), for any commutative

1.3 Operators on Hilbert Space

We extend this notion to operators on Hilbert space in the following way.

(iii) T = X ∗ X for some Hilbert space K and some X ∈ B(H, K);

(i) a contractive operator if T ≤ 1;

T is contractive if and only if 1 − T ∗ T ≥ 0. (1.13)

T ≤ 1 ⇔ x2 ≥ T x2 for all x ∈ H

The importance of the equivalence (1.13) in Proposition 1.12 cannot be

T is isometric if and only if 1 − T ∗ T = 0. (1.15)

By choosing in succession λ = 1 and λ = i we deduce that Zx, y = 0 for all

(u ⊗ v)(x) = x, v K u for all x ∈ K. (1.17)

In the context of a vector space V there is an algebraic notion of projection.

for all x ∈ M and y ∈ N . This operator P is called the projection on M along

1.5 Hardy Space and the Unilateral Shift

S (a0 , a1 , a2 , . . .) = (0, a0 , a1 , . . .). (1.23)

This simple definition conceals a treasure trove of analytical subtlety, much of

is square summable means that

One consequence of equation (1.28) (together with the maximum principle)

1.6 Invariant Subspaces of the Unilateral Shift

1.7 Von Neumann’s Theory of Spectral Sets

von Neumann introduced the following terminology.

f (T ) ≤ max |f (z)|. (1.35)

Thus von Neumann’s observation is equivalent to the assertion that σ (N ) is

p(T ) ≤ max |p(z)|.

We prove the inequality in Section 1.9 by a different method.

f (T ) = lim pn (T ) ≤ lim max |pn (z)| = max |f (z)|.

f (T ) ≤ max |f (z)|. (1.39)

f (T ) = lim fn (T ) ≤ lim max |fn (z)| = max |f (z)|.

There are several natural candidates for the definition of a holomorphic

f (z) − f (z0 ) = Df (z0 )(z − z0 ) + o(z − z0 ) as z → z0 , z ∈ .

f (T ) ≤ max |f (z)|. (1.42)

Just as long as σ (T ) ⊆ K, Lemma 1.38 asserts that there is no difference

1.8 The Schur Class and Spectral Domains

We define the Schur class of U , S (U ), by

Theorem 1.50. (Sz.-Nagy dilation theorem) If H is a Hilbert space,

1.9.1 Sz.-Nagy’s Proof of von Neumann’s Inequality

for all polynomials p. But by the maximum principle, if p is a polynomial,

for all polynomials p. By taking linear combinations of the power dilation

f (z) − f (z0 ) = Df (z0 )(z − z0 ) + o(z − z0 ) as z → z0 , z ∈ .

H2 BH2 = span {sλ1 , sλ2 , . . . , sλn }. (1.73)

Thus sλj ∈ H2 BH2 , and so

span {sλ1 , sλ2 , . . . , sλn } ⊆ H2 BH2 . (1.74)